diff options
Diffstat (limited to 'cddl/contrib/opensolaris/lib/libzfs')
16 files changed, 21449 insertions, 0 deletions
diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h new file mode 100644 index 0000000..ef18b45 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h @@ -0,0 +1,798 @@ +/* + * 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) 2013 by Delphix. All rights reserved. + * Copyright (c) 2012, Joyent, Inc. All rights reserved. + * Copyright (c) 2012 Martin Matuska <mm@FreeBSD.org>. All rights reserved. + * Copyright (c) 2013 Steven Hartland. All rights reserved. + * Copyright 2013 Nexenta Systems, Inc. All rights reserved. + */ + +#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_MAXNAMELEN MAXNAMELEN +#define ZPOOL_MAXNAMELEN MAXNAMELEN +#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_UNKNOWN +} zfs_error_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 *); + +/* + * 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 *); + +/* + * 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; +} splitflags_t; + +/* + * Functions to manipulate pool and vdev state + */ +extern int zpool_scan(zpool_handle_t *, pool_scan_func_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_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_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 *); + +/* + * 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. + */ + 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_IO_FAILURE_WAIT, /* failed I/O, failmode 'wait' */ + ZPOOL_STATUS_IO_FAILURE_CONTINUE, /* failed I/O, failmode 'continue' */ + 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 online */ + 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 **); + +/* + * 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 */ +} importargs_t; + +extern nvlist_t *zpool_search_import(libzfs_handle_t *, importargs_t *); + +/* 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[]; + +extern char *zpool_vdev_name(libzfs_handle_t *, zpool_handle_t *, nvlist_t *, + boolean_t verbose); +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 *); + +/* + * 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 *); + +/* + * 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 *, 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_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 *); +extern int zfs_iter_snapshots_sorted(zfs_handle_t *, zfs_iter_f, void *); +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; + boolean_t cb_verbose; + int (*cb_getone)(zfs_handle_t *, void *); +} get_all_cb_t; + +void libzfs_add_handle(get_all_cb_t *, zfs_handle_t *); +int libzfs_dataset_cmp(const void *, const void *); + +/* + * 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; + + /* WRITE_EMBEDDED records of type DATA are permitted */ + boolean_t embed_data; +} 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, enum lzc_send_flags); + +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; + + /* 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 *, 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); + +/* + * 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_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_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 sun +extern int zmount(const char *, const char *, int, char *, char *, int, char *, + int); +#endif /* !sun */ + +#ifdef __cplusplus +} +#endif + +#endif /* _LIBZFS_H */ diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_changelist.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_changelist.c new file mode 100644 index 0000000..a899965 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_changelist.c @@ -0,0 +1,700 @@ +/* + * 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 2010 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + * + * Portions Copyright 2007 Ramprakash Jelari + * + * Copyright (c) 2011 Pawel Jakub Dawidek <pawel@dawidek.net>. + * All rights reserved. + */ + +#include <libintl.h> +#include <libuutil.h> +#include <stddef.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> +#include <zone.h> + +#include <libzfs.h> + +#include "libzfs_impl.h" + +/* + * Structure to keep track of dataset state. Before changing the 'sharenfs' or + * 'mountpoint' property, we record whether the filesystem was previously + * mounted/shared. This prior state dictates whether we remount/reshare the + * dataset after the property has been changed. + * + * The interface consists of the following sequence of functions: + * + * changelist_gather() + * changelist_prefix() + * < change property > + * changelist_postfix() + * changelist_free() + * + * Other interfaces: + * + * changelist_remove() - remove a node from a gathered list + * changelist_rename() - renames all datasets appropriately when doing a rename + * changelist_unshare() - unshares all the nodes in a given changelist + * changelist_haszonedchild() - check if there is any child exported to + * a local zone + */ +typedef struct prop_changenode { + zfs_handle_t *cn_handle; + int cn_shared; + int cn_mounted; + int cn_zoned; + boolean_t cn_needpost; /* is postfix() needed? */ + uu_list_node_t cn_listnode; +} prop_changenode_t; + +struct prop_changelist { + zfs_prop_t cl_prop; + zfs_prop_t cl_realprop; + zfs_prop_t cl_shareprop; /* used with sharenfs/sharesmb */ + uu_list_pool_t *cl_pool; + uu_list_t *cl_list; + boolean_t cl_waslegacy; + boolean_t cl_allchildren; + boolean_t cl_alldependents; + int cl_mflags; /* Mount flags */ + int cl_gflags; /* Gather request flags */ + boolean_t cl_haszonedchild; + boolean_t cl_sorted; +}; + +/* + * If the property is 'mountpoint', go through and unmount filesystems as + * necessary. We don't do the same for 'sharenfs', because we can just re-share + * with different options without interrupting service. We do handle 'sharesmb' + * since there may be old resource names that need to be removed. + */ +int +changelist_prefix(prop_changelist_t *clp) +{ + prop_changenode_t *cn; + int ret = 0; + + if (clp->cl_prop != ZFS_PROP_MOUNTPOINT && + clp->cl_prop != ZFS_PROP_SHARESMB) + return (0); + + for (cn = uu_list_first(clp->cl_list); cn != NULL; + cn = uu_list_next(clp->cl_list, cn)) { + + /* if a previous loop failed, set the remaining to false */ + if (ret == -1) { + cn->cn_needpost = B_FALSE; + continue; + } + + /* + * If we are in the global zone, but this dataset is exported + * to a local zone, do nothing. + */ + if (getzoneid() == GLOBAL_ZONEID && cn->cn_zoned) + continue; + + if (!ZFS_IS_VOLUME(cn->cn_handle)) { + /* + * Do the property specific processing. + */ + switch (clp->cl_prop) { + case ZFS_PROP_MOUNTPOINT: + if (clp->cl_gflags & CL_GATHER_DONT_UNMOUNT) + break; + if (zfs_unmount(cn->cn_handle, NULL, + clp->cl_mflags) != 0) { + ret = -1; + cn->cn_needpost = B_FALSE; + } + break; + case ZFS_PROP_SHARESMB: + (void) zfs_unshare_smb(cn->cn_handle, NULL); + break; + } + } + } + + if (ret == -1) + (void) changelist_postfix(clp); + + return (ret); +} + +/* + * If the property is 'mountpoint' or 'sharenfs', go through and remount and/or + * reshare the filesystems as necessary. In changelist_gather() we recorded + * whether the filesystem was previously shared or mounted. The action we take + * depends on the previous state, and whether the value was previously 'legacy'. + * For non-legacy properties, we only remount/reshare the filesystem if it was + * previously mounted/shared. Otherwise, we always remount/reshare the + * filesystem. + */ +int +changelist_postfix(prop_changelist_t *clp) +{ + prop_changenode_t *cn; + char shareopts[ZFS_MAXPROPLEN]; + int errors = 0; + libzfs_handle_t *hdl; + + /* + * If we're changing the mountpoint, attempt to destroy the underlying + * mountpoint. All other datasets will have inherited from this dataset + * (in which case their mountpoints exist in the filesystem in the new + * location), or have explicit mountpoints set (in which case they won't + * be in the changelist). + */ + if ((cn = uu_list_last(clp->cl_list)) == NULL) + return (0); + + if (clp->cl_prop == ZFS_PROP_MOUNTPOINT && + !(clp->cl_gflags & CL_GATHER_DONT_UNMOUNT)) { + remove_mountpoint(cn->cn_handle); + } + + /* + * It is possible that the changelist_prefix() used libshare + * to unshare some entries. Since libshare caches data, an + * attempt to reshare during postfix can fail unless libshare + * is uninitialized here so that it will reinitialize later. + */ + if (cn->cn_handle != NULL) { + hdl = cn->cn_handle->zfs_hdl; + assert(hdl != NULL); + zfs_uninit_libshare(hdl); + } + + /* + * We walk the datasets in reverse, because we want to mount any parent + * datasets before mounting the children. We walk all datasets even if + * there are errors. + */ + for (cn = uu_list_last(clp->cl_list); cn != NULL; + cn = uu_list_prev(clp->cl_list, cn)) { + + boolean_t sharenfs; + boolean_t sharesmb; + boolean_t mounted; + + /* + * If we are in the global zone, but this dataset is exported + * to a local zone, do nothing. + */ + if (getzoneid() == GLOBAL_ZONEID && cn->cn_zoned) + continue; + + /* Only do post-processing if it's required */ + if (!cn->cn_needpost) + continue; + cn->cn_needpost = B_FALSE; + + zfs_refresh_properties(cn->cn_handle); + + if (ZFS_IS_VOLUME(cn->cn_handle)) + continue; + + /* + * Remount if previously mounted or mountpoint was legacy, + * or sharenfs or sharesmb property is set. + */ + sharenfs = ((zfs_prop_get(cn->cn_handle, ZFS_PROP_SHARENFS, + shareopts, sizeof (shareopts), NULL, NULL, 0, + B_FALSE) == 0) && (strcmp(shareopts, "off") != 0)); + + sharesmb = ((zfs_prop_get(cn->cn_handle, ZFS_PROP_SHARESMB, + shareopts, sizeof (shareopts), NULL, NULL, 0, + B_FALSE) == 0) && (strcmp(shareopts, "off") != 0)); + + mounted = (clp->cl_gflags & CL_GATHER_DONT_UNMOUNT) || + zfs_is_mounted(cn->cn_handle, NULL); + + if (!mounted && (cn->cn_mounted || + ((sharenfs || sharesmb || clp->cl_waslegacy) && + (zfs_prop_get_int(cn->cn_handle, + ZFS_PROP_CANMOUNT) == ZFS_CANMOUNT_ON)))) { + + if (zfs_mount(cn->cn_handle, NULL, 0) != 0) + errors++; + else + mounted = TRUE; + } + + /* + * If the file system is mounted we always re-share even + * if the filesystem is currently shared, so that we can + * adopt any new options. + */ + if (sharenfs && mounted) + errors += zfs_share_nfs(cn->cn_handle); + else if (cn->cn_shared || clp->cl_waslegacy) + errors += zfs_unshare_nfs(cn->cn_handle, NULL); + if (sharesmb && mounted) + errors += zfs_share_smb(cn->cn_handle); + else if (cn->cn_shared || clp->cl_waslegacy) + errors += zfs_unshare_smb(cn->cn_handle, NULL); + } + + return (errors ? -1 : 0); +} + +/* + * Is this "dataset" a child of "parent"? + */ +boolean_t +isa_child_of(const char *dataset, const char *parent) +{ + int len; + + len = strlen(parent); + + if (strncmp(dataset, parent, len) == 0 && + (dataset[len] == '@' || dataset[len] == '/' || + dataset[len] == '\0')) + return (B_TRUE); + else + return (B_FALSE); + +} + +/* + * If we rename a filesystem, child filesystem handles are no longer valid + * since we identify each dataset by its name in the ZFS namespace. As a + * result, we have to go through and fix up all the names appropriately. We + * could do this automatically if libzfs kept track of all open handles, but + * this is a lot less work. + */ +void +changelist_rename(prop_changelist_t *clp, const char *src, const char *dst) +{ + prop_changenode_t *cn; + char newname[ZFS_MAXNAMELEN]; + + for (cn = uu_list_first(clp->cl_list); cn != NULL; + cn = uu_list_next(clp->cl_list, cn)) { + /* + * Do not rename a clone that's not in the source hierarchy. + */ + if (!isa_child_of(cn->cn_handle->zfs_name, src)) + continue; + + /* + * Destroy the previous mountpoint if needed. + */ + remove_mountpoint(cn->cn_handle); + + (void) strlcpy(newname, dst, sizeof (newname)); + (void) strcat(newname, cn->cn_handle->zfs_name + strlen(src)); + + (void) strlcpy(cn->cn_handle->zfs_name, newname, + sizeof (cn->cn_handle->zfs_name)); + } +} + +/* + * Given a gathered changelist for the 'sharenfs' or 'sharesmb' property, + * unshare all the datasets in the list. + */ +int +changelist_unshare(prop_changelist_t *clp, zfs_share_proto_t *proto) +{ + prop_changenode_t *cn; + int ret = 0; + + if (clp->cl_prop != ZFS_PROP_SHARENFS && + clp->cl_prop != ZFS_PROP_SHARESMB) + return (0); + + for (cn = uu_list_first(clp->cl_list); cn != NULL; + cn = uu_list_next(clp->cl_list, cn)) { + if (zfs_unshare_proto(cn->cn_handle, NULL, proto) != 0) + ret = -1; + } + + return (ret); +} + +/* + * Check if there is any child exported to a local zone in a given changelist. + * This information has already been recorded while gathering the changelist + * via changelist_gather(). + */ +int +changelist_haszonedchild(prop_changelist_t *clp) +{ + return (clp->cl_haszonedchild); +} + +/* + * Remove a node from a gathered list. + */ +void +changelist_remove(prop_changelist_t *clp, const char *name) +{ + prop_changenode_t *cn; + + for (cn = uu_list_first(clp->cl_list); cn != NULL; + cn = uu_list_next(clp->cl_list, cn)) { + + if (strcmp(cn->cn_handle->zfs_name, name) == 0) { + uu_list_remove(clp->cl_list, cn); + zfs_close(cn->cn_handle); + free(cn); + return; + } + } +} + +/* + * Release any memory associated with a changelist. + */ +void +changelist_free(prop_changelist_t *clp) +{ + prop_changenode_t *cn; + void *cookie; + + if (clp->cl_list) { + cookie = NULL; + while ((cn = uu_list_teardown(clp->cl_list, &cookie)) != NULL) { + zfs_close(cn->cn_handle); + free(cn); + } + + uu_list_destroy(clp->cl_list); + } + if (clp->cl_pool) + uu_list_pool_destroy(clp->cl_pool); + + free(clp); +} + +static int +change_one(zfs_handle_t *zhp, void *data) +{ + prop_changelist_t *clp = data; + char property[ZFS_MAXPROPLEN]; + char where[64]; + prop_changenode_t *cn; + zprop_source_t sourcetype; + zprop_source_t share_sourcetype; + + /* + * We only want to unmount/unshare those filesystems that may inherit + * from the target filesystem. If we find any filesystem with a + * locally set mountpoint, we ignore any children since changing the + * property will not affect them. If this is a rename, we iterate + * over all children regardless, since we need them unmounted in + * order to do the rename. Also, if this is a volume and we're doing + * a rename, then always add it to the changelist. + */ + + if (!(ZFS_IS_VOLUME(zhp) && clp->cl_realprop == ZFS_PROP_NAME) && + zfs_prop_get(zhp, clp->cl_prop, property, + sizeof (property), &sourcetype, where, sizeof (where), + B_FALSE) != 0) { + zfs_close(zhp); + return (0); + } + + /* + * If we are "watching" sharenfs or sharesmb + * then check out the companion property which is tracked + * in cl_shareprop + */ + if (clp->cl_shareprop != ZPROP_INVAL && + zfs_prop_get(zhp, clp->cl_shareprop, property, + sizeof (property), &share_sourcetype, where, sizeof (where), + B_FALSE) != 0) { + zfs_close(zhp); + return (0); + } + + if (clp->cl_alldependents || clp->cl_allchildren || + sourcetype == ZPROP_SRC_DEFAULT || + sourcetype == ZPROP_SRC_INHERITED || + (clp->cl_shareprop != ZPROP_INVAL && + (share_sourcetype == ZPROP_SRC_DEFAULT || + share_sourcetype == ZPROP_SRC_INHERITED))) { + if ((cn = zfs_alloc(zfs_get_handle(zhp), + sizeof (prop_changenode_t))) == NULL) { + zfs_close(zhp); + return (-1); + } + + cn->cn_handle = zhp; + cn->cn_mounted = (clp->cl_gflags & CL_GATHER_MOUNT_ALWAYS) || + zfs_is_mounted(zhp, NULL); + cn->cn_shared = zfs_is_shared(zhp); + cn->cn_zoned = zfs_prop_get_int(zhp, ZFS_PROP_ZONED); + cn->cn_needpost = B_TRUE; + + /* Indicate if any child is exported to a local zone. */ + if (getzoneid() == GLOBAL_ZONEID && cn->cn_zoned) + clp->cl_haszonedchild = B_TRUE; + + uu_list_node_init(cn, &cn->cn_listnode, clp->cl_pool); + + if (clp->cl_sorted) { + uu_list_index_t idx; + + (void) uu_list_find(clp->cl_list, cn, NULL, + &idx); + uu_list_insert(clp->cl_list, cn, idx); + } else { + /* + * Add this child to beginning of the list. Children + * below this one in the hierarchy will get added above + * this one in the list. This produces a list in + * reverse dataset name order. + * This is necessary when the original mountpoint + * is legacy or none. + */ + verify(uu_list_insert_before(clp->cl_list, + uu_list_first(clp->cl_list), cn) == 0); + } + + if (!clp->cl_alldependents) + return (zfs_iter_children(zhp, change_one, data)); + } else { + zfs_close(zhp); + } + + return (0); +} + +/*ARGSUSED*/ +static int +compare_mountpoints(const void *a, const void *b, void *unused) +{ + const prop_changenode_t *ca = a; + const prop_changenode_t *cb = b; + + char mounta[MAXPATHLEN]; + char mountb[MAXPATHLEN]; + + boolean_t hasmounta, hasmountb; + + /* + * When unsharing or unmounting filesystems, we need to do it in + * mountpoint order. This allows the user to have a mountpoint + * hierarchy that is different from the dataset hierarchy, and still + * allow it to be changed. However, if either dataset doesn't have a + * mountpoint (because it is a volume or a snapshot), we place it at the + * end of the list, because it doesn't affect our change at all. + */ + hasmounta = (zfs_prop_get(ca->cn_handle, ZFS_PROP_MOUNTPOINT, mounta, + sizeof (mounta), NULL, NULL, 0, B_FALSE) == 0); + hasmountb = (zfs_prop_get(cb->cn_handle, ZFS_PROP_MOUNTPOINT, mountb, + sizeof (mountb), NULL, NULL, 0, B_FALSE) == 0); + + if (!hasmounta && hasmountb) + return (-1); + else if (hasmounta && !hasmountb) + return (1); + else if (!hasmounta && !hasmountb) + return (0); + else + return (strcmp(mountb, mounta)); +} + +/* + * Given a ZFS handle and a property, construct a complete list of datasets + * that need to be modified as part of this process. For anything but the + * 'mountpoint' and 'sharenfs' properties, this just returns an empty list. + * Otherwise, we iterate over all children and look for any datasets that + * inherit the property. For each such dataset, we add it to the list and + * mark whether it was shared beforehand. + */ +prop_changelist_t * +changelist_gather(zfs_handle_t *zhp, zfs_prop_t prop, int gather_flags, + int mnt_flags) +{ + prop_changelist_t *clp; + prop_changenode_t *cn; + zfs_handle_t *temp; + char property[ZFS_MAXPROPLEN]; + uu_compare_fn_t *compare = NULL; + boolean_t legacy = B_FALSE; + + if ((clp = zfs_alloc(zhp->zfs_hdl, sizeof (prop_changelist_t))) == NULL) + return (NULL); + + /* + * For mountpoint-related tasks, we want to sort everything by + * mountpoint, so that we mount and unmount them in the appropriate + * order, regardless of their position in the hierarchy. + */ + if (prop == ZFS_PROP_NAME || prop == ZFS_PROP_ZONED || + prop == ZFS_PROP_MOUNTPOINT || prop == ZFS_PROP_SHARENFS || + prop == ZFS_PROP_SHARESMB) { + + if (zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, + property, sizeof (property), + NULL, NULL, 0, B_FALSE) == 0 && + (strcmp(property, "legacy") == 0 || + strcmp(property, "none") == 0)) { + + legacy = B_TRUE; + } + if (!legacy) { + compare = compare_mountpoints; + clp->cl_sorted = B_TRUE; + } + } + + clp->cl_pool = uu_list_pool_create("changelist_pool", + sizeof (prop_changenode_t), + offsetof(prop_changenode_t, cn_listnode), + compare, 0); + if (clp->cl_pool == NULL) { + assert(uu_error() == UU_ERROR_NO_MEMORY); + (void) zfs_error(zhp->zfs_hdl, EZFS_NOMEM, "internal error"); + changelist_free(clp); + return (NULL); + } + + clp->cl_list = uu_list_create(clp->cl_pool, NULL, + clp->cl_sorted ? UU_LIST_SORTED : 0); + clp->cl_gflags = gather_flags; + clp->cl_mflags = mnt_flags; + + if (clp->cl_list == NULL) { + assert(uu_error() == UU_ERROR_NO_MEMORY); + (void) zfs_error(zhp->zfs_hdl, EZFS_NOMEM, "internal error"); + changelist_free(clp); + return (NULL); + } + + /* + * If this is a rename or the 'zoned' property, we pretend we're + * changing the mountpoint and flag it so we can catch all children in + * change_one(). + * + * Flag cl_alldependents to catch all children plus the dependents + * (clones) that are not in the hierarchy. + */ + if (prop == ZFS_PROP_NAME) { + clp->cl_prop = ZFS_PROP_MOUNTPOINT; + clp->cl_alldependents = B_TRUE; + } else if (prop == ZFS_PROP_ZONED) { + clp->cl_prop = ZFS_PROP_MOUNTPOINT; + clp->cl_allchildren = B_TRUE; + } else if (prop == ZFS_PROP_CANMOUNT) { + clp->cl_prop = ZFS_PROP_MOUNTPOINT; + } else if (prop == ZFS_PROP_VOLSIZE) { + clp->cl_prop = ZFS_PROP_MOUNTPOINT; + } else { + clp->cl_prop = prop; + } + clp->cl_realprop = prop; + + if (clp->cl_prop != ZFS_PROP_MOUNTPOINT && + clp->cl_prop != ZFS_PROP_SHARENFS && + clp->cl_prop != ZFS_PROP_SHARESMB) + return (clp); + + /* + * If watching SHARENFS or SHARESMB then + * also watch its companion property. + */ + if (clp->cl_prop == ZFS_PROP_SHARENFS) + clp->cl_shareprop = ZFS_PROP_SHARESMB; + else if (clp->cl_prop == ZFS_PROP_SHARESMB) + clp->cl_shareprop = ZFS_PROP_SHARENFS; + + if (clp->cl_alldependents) { + if (zfs_iter_dependents(zhp, B_TRUE, change_one, clp) != 0) { + changelist_free(clp); + return (NULL); + } + } else if (zfs_iter_children(zhp, change_one, clp) != 0) { + changelist_free(clp); + return (NULL); + } + + /* + * We have to re-open ourselves because we auto-close all the handles + * and can't tell the difference. + */ + if ((temp = zfs_open(zhp->zfs_hdl, zfs_get_name(zhp), + ZFS_TYPE_DATASET)) == NULL) { + changelist_free(clp); + return (NULL); + } + + /* + * Always add ourself to the list. We add ourselves to the end so that + * we're the last to be unmounted. + */ + if ((cn = zfs_alloc(zhp->zfs_hdl, + sizeof (prop_changenode_t))) == NULL) { + zfs_close(temp); + changelist_free(clp); + return (NULL); + } + + cn->cn_handle = temp; + cn->cn_mounted = (clp->cl_gflags & CL_GATHER_MOUNT_ALWAYS) || + zfs_is_mounted(temp, NULL); + cn->cn_shared = zfs_is_shared(temp); + cn->cn_zoned = zfs_prop_get_int(zhp, ZFS_PROP_ZONED); + cn->cn_needpost = B_TRUE; + + uu_list_node_init(cn, &cn->cn_listnode, clp->cl_pool); + if (clp->cl_sorted) { + uu_list_index_t idx; + (void) uu_list_find(clp->cl_list, cn, NULL, &idx); + uu_list_insert(clp->cl_list, cn, idx); + } else { + /* + * Add the target dataset to the end of the list. + * The list is not really unsorted. The list will be + * in reverse dataset name order. This is necessary + * when the original mountpoint is legacy or none. + */ + verify(uu_list_insert_after(clp->cl_list, + uu_list_last(clp->cl_list), cn) == 0); + } + + /* + * If the mountpoint property was previously 'legacy', or 'none', + * record it as the behavior of changelist_postfix() will be different. + */ + if ((clp->cl_prop == ZFS_PROP_MOUNTPOINT) && legacy) { + /* + * do not automatically mount ex-legacy datasets if + * we specifically set canmount to noauto + */ + if (zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT) != + ZFS_CANMOUNT_NOAUTO) + clp->cl_waslegacy = B_TRUE; + } + + return (clp); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_compat.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_compat.c new file mode 100644 index 0000000..3c8119d --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_compat.c @@ -0,0 +1,105 @@ +/* + * CDDL HEADER SART + * + * 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) 2013 Martin Matuska <mm@FreeBSD.org>. All rights reserved. + */ + +#include "libzfs_compat.h" + +int zfs_ioctl_version = ZFS_IOCVER_UNDEF; +static int zfs_spa_version = -1; + +/* + * Get zfs_ioctl_version + */ +int +get_zfs_ioctl_version(void) +{ + size_t ver_size; + int ver = ZFS_IOCVER_NONE; + + ver_size = sizeof(ver); + sysctlbyname("vfs.zfs.version.ioctl", &ver, &ver_size, NULL, 0); + + return (ver); +} + +/* + * Get the SPA version + */ +static int +get_zfs_spa_version(void) +{ + size_t ver_size; + int ver = 0; + + ver_size = sizeof(ver); + sysctlbyname("vfs.zfs.version.spa", &ver, &ver_size, NULL, 0); + + return (ver); +} + +/* + * This is FreeBSD version of ioctl, because Solaris' ioctl() updates + * zc_nvlist_dst_size even if an error is returned, on FreeBSD if an + * error is returned zc_nvlist_dst_size won't be updated. + */ +int +zcmd_ioctl(int fd, int request, zfs_cmd_t *zc) +{ + size_t oldsize; + int ret, cflag = ZFS_CMD_COMPAT_NONE; + + if (zfs_ioctl_version == ZFS_IOCVER_UNDEF) + zfs_ioctl_version = get_zfs_ioctl_version(); + + if (zfs_ioctl_version == ZFS_IOCVER_LZC) + cflag = ZFS_CMD_COMPAT_LZC; + else if (zfs_ioctl_version == ZFS_IOCVER_DEADMAN) + cflag = ZFS_CMD_COMPAT_DEADMAN; + + /* + * If vfs.zfs.version.ioctl is not defined, assume we have v28 + * compatible binaries and use vfs.zfs.version.spa to test for v15 + */ + if (zfs_ioctl_version < ZFS_IOCVER_DEADMAN) { + cflag = ZFS_CMD_COMPAT_V28; + + if (zfs_spa_version < 0) + zfs_spa_version = get_zfs_spa_version(); + + if (zfs_spa_version == SPA_VERSION_15 || + zfs_spa_version == SPA_VERSION_14 || + zfs_spa_version == SPA_VERSION_13) + cflag = ZFS_CMD_COMPAT_V15; + } + + oldsize = zc->zc_nvlist_dst_size; + ret = zcmd_ioctl_compat(fd, request, zc, cflag); + + if (ret == 0 && oldsize < zc->zc_nvlist_dst_size) { + ret = -1; + errno = ENOMEM; + } + + return (ret); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_compat.h b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_compat.h new file mode 100644 index 0000000..3761668 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_compat.h @@ -0,0 +1,44 @@ +/* + * CDDL HEADER SART + * + * 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) 2013 Martin Matuska <mm@FreeBSD.org>. All rights reserved. + */ + +#ifndef _LIBZFS_COMPAT_H +#define _LIBZFS_COMPAT_H + +#include <zfs_ioctl_compat.h> + +#ifdef __cplusplus +extern "C" { +#endif + +int get_zfs_ioctl_version(void); +int zcmd_ioctl(int fd, int request, zfs_cmd_t *zc); + +#define ioctl(fd, ioc, zc) zcmd_ioctl((fd), (ioc), (zc)) + +#ifdef __cplusplus +} +#endif + +#endif /* _LIBZFS_COMPAT_H */ diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_config.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_config.c new file mode 100644 index 0000000..d5ba20f --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_config.c @@ -0,0 +1,453 @@ +/* + * 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 2009 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +/* + * Copyright (c) 2012 by Delphix. All rights reserved. + */ + +/* + * The pool configuration repository is stored in /etc/zfs/zpool.cache as a + * single packed nvlist. While it would be nice to just read in this + * file from userland, this wouldn't work from a local zone. So we have to have + * a zpool ioctl to return the complete configuration for all pools. In the + * global zone, this will be identical to reading the file and unpacking it in + * userland. + */ + +#include <errno.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <stddef.h> +#include <string.h> +#include <unistd.h> +#include <libintl.h> +#include <libuutil.h> + +#include "libzfs_impl.h" + +typedef struct config_node { + char *cn_name; + nvlist_t *cn_config; + uu_avl_node_t cn_avl; +} config_node_t; + +/* ARGSUSED */ +static int +config_node_compare(const void *a, const void *b, void *unused) +{ + int ret; + + const config_node_t *ca = (config_node_t *)a; + const config_node_t *cb = (config_node_t *)b; + + ret = strcmp(ca->cn_name, cb->cn_name); + + if (ret < 0) + return (-1); + else if (ret > 0) + return (1); + else + return (0); +} + +void +namespace_clear(libzfs_handle_t *hdl) +{ + if (hdl->libzfs_ns_avl) { + config_node_t *cn; + void *cookie = NULL; + + while ((cn = uu_avl_teardown(hdl->libzfs_ns_avl, + &cookie)) != NULL) { + nvlist_free(cn->cn_config); + free(cn->cn_name); + free(cn); + } + + uu_avl_destroy(hdl->libzfs_ns_avl); + hdl->libzfs_ns_avl = NULL; + } + + if (hdl->libzfs_ns_avlpool) { + uu_avl_pool_destroy(hdl->libzfs_ns_avlpool); + hdl->libzfs_ns_avlpool = NULL; + } +} + +/* + * Loads the pool namespace, or re-loads it if the cache has changed. + */ +static int +namespace_reload(libzfs_handle_t *hdl) +{ + nvlist_t *config; + config_node_t *cn; + nvpair_t *elem; + zfs_cmd_t zc = { 0 }; + void *cookie; + + if (hdl->libzfs_ns_gen == 0) { + /* + * This is the first time we've accessed the configuration + * cache. Initialize the AVL tree and then fall through to the + * common code. + */ + if ((hdl->libzfs_ns_avlpool = uu_avl_pool_create("config_pool", + sizeof (config_node_t), + offsetof(config_node_t, cn_avl), + config_node_compare, UU_DEFAULT)) == NULL) + return (no_memory(hdl)); + + if ((hdl->libzfs_ns_avl = uu_avl_create(hdl->libzfs_ns_avlpool, + NULL, UU_DEFAULT)) == NULL) + return (no_memory(hdl)); + } + + if (zcmd_alloc_dst_nvlist(hdl, &zc, 0) != 0) + return (-1); + + for (;;) { + zc.zc_cookie = hdl->libzfs_ns_gen; + if (ioctl(hdl->libzfs_fd, ZFS_IOC_POOL_CONFIGS, &zc) != 0) { + switch (errno) { + case EEXIST: + /* + * The namespace hasn't changed. + */ + zcmd_free_nvlists(&zc); + return (0); + + case ENOMEM: + if (zcmd_expand_dst_nvlist(hdl, &zc) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + break; + + default: + zcmd_free_nvlists(&zc); + return (zfs_standard_error(hdl, errno, + dgettext(TEXT_DOMAIN, "failed to read " + "pool configuration"))); + } + } else { + hdl->libzfs_ns_gen = zc.zc_cookie; + break; + } + } + + if (zcmd_read_dst_nvlist(hdl, &zc, &config) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + + zcmd_free_nvlists(&zc); + + /* + * Clear out any existing configuration information. + */ + cookie = NULL; + while ((cn = uu_avl_teardown(hdl->libzfs_ns_avl, &cookie)) != NULL) { + nvlist_free(cn->cn_config); + free(cn->cn_name); + free(cn); + } + + elem = NULL; + while ((elem = nvlist_next_nvpair(config, elem)) != NULL) { + nvlist_t *child; + uu_avl_index_t where; + + if ((cn = zfs_alloc(hdl, sizeof (config_node_t))) == NULL) { + nvlist_free(config); + return (-1); + } + + if ((cn->cn_name = zfs_strdup(hdl, + nvpair_name(elem))) == NULL) { + free(cn); + nvlist_free(config); + return (-1); + } + + verify(nvpair_value_nvlist(elem, &child) == 0); + if (nvlist_dup(child, &cn->cn_config, 0) != 0) { + free(cn->cn_name); + free(cn); + nvlist_free(config); + return (no_memory(hdl)); + } + verify(uu_avl_find(hdl->libzfs_ns_avl, cn, NULL, &where) + == NULL); + + uu_avl_insert(hdl->libzfs_ns_avl, cn, where); + } + + nvlist_free(config); + return (0); +} + +/* + * Retrieve the configuration for the given pool. The configuration is a nvlist + * describing the vdevs, as well as the statistics associated with each one. + */ +nvlist_t * +zpool_get_config(zpool_handle_t *zhp, nvlist_t **oldconfig) +{ + if (oldconfig) + *oldconfig = zhp->zpool_old_config; + return (zhp->zpool_config); +} + +/* + * Retrieves a list of enabled features and their refcounts and caches it in + * the pool handle. + */ +nvlist_t * +zpool_get_features(zpool_handle_t *zhp) +{ + nvlist_t *config, *features; + + config = zpool_get_config(zhp, NULL); + + if (config == NULL || !nvlist_exists(config, + ZPOOL_CONFIG_FEATURE_STATS)) { + int error; + boolean_t missing = B_FALSE; + + error = zpool_refresh_stats(zhp, &missing); + + if (error != 0 || missing) + return (NULL); + + config = zpool_get_config(zhp, NULL); + } + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_FEATURE_STATS, + &features) == 0); + + return (features); +} + +/* + * Refresh the vdev statistics associated with the given pool. This is used in + * iostat to show configuration changes and determine the delta from the last + * time the function was called. This function can fail, in case the pool has + * been destroyed. + */ +int +zpool_refresh_stats(zpool_handle_t *zhp, boolean_t *missing) +{ + zfs_cmd_t zc = { 0 }; + int error; + nvlist_t *config; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + *missing = B_FALSE; + (void) strcpy(zc.zc_name, zhp->zpool_name); + + if (zhp->zpool_config_size == 0) + zhp->zpool_config_size = 1 << 16; + + if (zcmd_alloc_dst_nvlist(hdl, &zc, zhp->zpool_config_size) != 0) + return (-1); + + for (;;) { + if (ioctl(zhp->zpool_hdl->libzfs_fd, ZFS_IOC_POOL_STATS, + &zc) == 0) { + /* + * The real error is returned in the zc_cookie field. + */ + error = zc.zc_cookie; + break; + } + + if (errno == ENOMEM) { + if (zcmd_expand_dst_nvlist(hdl, &zc) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + } else { + zcmd_free_nvlists(&zc); + if (errno == ENOENT || errno == EINVAL) + *missing = B_TRUE; + zhp->zpool_state = POOL_STATE_UNAVAIL; + return (0); + } + } + + if (zcmd_read_dst_nvlist(hdl, &zc, &config) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + + zcmd_free_nvlists(&zc); + + zhp->zpool_config_size = zc.zc_nvlist_dst_size; + + if (zhp->zpool_config != NULL) { + uint64_t oldtxg, newtxg; + + verify(nvlist_lookup_uint64(zhp->zpool_config, + ZPOOL_CONFIG_POOL_TXG, &oldtxg) == 0); + verify(nvlist_lookup_uint64(config, + ZPOOL_CONFIG_POOL_TXG, &newtxg) == 0); + + if (zhp->zpool_old_config != NULL) + nvlist_free(zhp->zpool_old_config); + + if (oldtxg != newtxg) { + nvlist_free(zhp->zpool_config); + zhp->zpool_old_config = NULL; + } else { + zhp->zpool_old_config = zhp->zpool_config; + } + } + + zhp->zpool_config = config; + if (error) + zhp->zpool_state = POOL_STATE_UNAVAIL; + else + zhp->zpool_state = POOL_STATE_ACTIVE; + + return (0); +} + +/* + * If the __ZFS_POOL_RESTRICT environment variable is set we only iterate over + * pools it lists. + * + * This is an undocumented feature for use during testing only. + * + * This function returns B_TRUE if the pool should be skipped + * during iteration. + */ +static boolean_t +check_restricted(const char *poolname) +{ + static boolean_t initialized = B_FALSE; + static char *restricted = NULL; + + const char *cur, *end; + int len, namelen; + + if (!initialized) { + initialized = B_TRUE; + restricted = getenv("__ZFS_POOL_RESTRICT"); + } + + if (NULL == restricted) + return (B_FALSE); + + cur = restricted; + namelen = strlen(poolname); + do { + end = strchr(cur, ' '); + len = (NULL == end) ? strlen(cur) : (end - cur); + + if (len == namelen && 0 == strncmp(cur, poolname, len)) { + return (B_FALSE); + } + + cur += (len + 1); + } while (NULL != end); + + return (B_TRUE); +} + +/* + * Iterate over all pools in the system. + */ +int +zpool_iter(libzfs_handle_t *hdl, zpool_iter_f func, void *data) +{ + config_node_t *cn; + zpool_handle_t *zhp; + int ret; + + /* + * If someone makes a recursive call to zpool_iter(), we want to avoid + * refreshing the namespace because that will invalidate the parent + * context. We allow recursive calls, but simply re-use the same + * namespace AVL tree. + */ + if (!hdl->libzfs_pool_iter && namespace_reload(hdl) != 0) + return (-1); + + hdl->libzfs_pool_iter++; + for (cn = uu_avl_first(hdl->libzfs_ns_avl); cn != NULL; + cn = uu_avl_next(hdl->libzfs_ns_avl, cn)) { + + if (check_restricted(cn->cn_name)) + continue; + + if (zpool_open_silent(hdl, cn->cn_name, &zhp) != 0) { + hdl->libzfs_pool_iter--; + return (-1); + } + + if (zhp == NULL) + continue; + + if ((ret = func(zhp, data)) != 0) { + hdl->libzfs_pool_iter--; + return (ret); + } + } + hdl->libzfs_pool_iter--; + + return (0); +} + +/* + * Iterate over root datasets, calling the given function for each. The zfs + * handle passed each time must be explicitly closed by the callback. + */ +int +zfs_iter_root(libzfs_handle_t *hdl, zfs_iter_f func, void *data) +{ + config_node_t *cn; + zfs_handle_t *zhp; + int ret; + + if (namespace_reload(hdl) != 0) + return (-1); + + for (cn = uu_avl_first(hdl->libzfs_ns_avl); cn != NULL; + cn = uu_avl_next(hdl->libzfs_ns_avl, cn)) { + + if (check_restricted(cn->cn_name)) + continue; + + if ((zhp = make_dataset_handle(hdl, cn->cn_name)) == NULL) + continue; + + if ((ret = func(zhp, data)) != 0) + return (ret); + } + + return (0); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c new file mode 100644 index 0000000..11378a2 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c @@ -0,0 +1,4743 @@ +/* + * 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) 2013, Joyent, Inc. All rights reserved. + * Copyright (c) 2013 by Delphix. All rights reserved. + * Copyright (c) 2012 DEY Storage Systems, Inc. All rights reserved. + * Copyright (c) 2011-2012 Pawel Jakub Dawidek <pawel@dawidek.net>. + * All rights reserved. + * Copyright (c) 2012 Martin Matuska <mm@FreeBSD.org>. All rights reserved. + * Copyright (c) 2013 Steven Hartland. All rights reserved. + * Copyright 2013 Nexenta Systems, Inc. All rights reserved. + */ + +#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> +#include <idmap.h> + +#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")); + } + + return (NULL); +} + +/* + * Given a path and mask of ZFS types, return a string describing this dataset. + * This is used when we fail to open a dataset and we cannot get an exact type. + * We guess what the type would have been based on the path and the mask of + * acceptable types. + */ +static const char * +path_to_str(const char *path, int types) +{ + /* + * When given a single type, always report the exact type. + */ + if (types == ZFS_TYPE_SNAPSHOT) + return (dgettext(TEXT_DOMAIN, "snapshot")); + if (types == ZFS_TYPE_FILESYSTEM) + return (dgettext(TEXT_DOMAIN, "filesystem")); + if (types == ZFS_TYPE_VOLUME) + return (dgettext(TEXT_DOMAIN, "volume")); + + /* + * The user is requesting more than one type of dataset. If this is the + * case, consult the path itself. If we're looking for a snapshot, and + * a '@' is found, then report it as "snapshot". Otherwise, remove the + * snapshot attribute and try again. + */ + if (types & ZFS_TYPE_SNAPSHOT) { + if (strchr(path, '@') != NULL) + return (dgettext(TEXT_DOMAIN, "snapshot")); + return (path_to_str(path, types & ~ZFS_TYPE_SNAPSHOT)); + } + + /* + * The user has requested either filesystems or volumes. + * We have no way of knowing a priori what type this would be, so always + * report it as "filesystem" or "volume", our two primitive types. + */ + if (types & ZFS_TYPE_FILESYSTEM) + return (dgettext(TEXT_DOMAIN, "filesystem")); + + assert(types & ZFS_TYPE_VOLUME); + return (dgettext(TEXT_DOMAIN, "volume")); +} + +/* + * 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; + + (void) zfs_prop_get_table(); + if (dataset_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_AT: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "multiple '@' 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; + } + } + + return (0); + } + + if (!(type & ZFS_TYPE_SNAPSHOT) && strchr(path, '@') != NULL) { + if (hdl != NULL) + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "snapshot delimiter '@' in filesystem name")); + 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 (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 + 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_MAXNAMELEN]; + 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); +} + +/* + * Opens the given snapshot, 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]; + + (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, ZFS_TYPE_DATASET, B_FALSE)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "invalid dataset name")); + (void) zfs_error(hdl, EZFS_INVALIDNAME, errbuf); + return (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); + } + + 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 = arg1; + const mnttab_node_t *mtn2 = arg2; + int rv; + + rv = strcmp(mtn1->mtn_mt.mnt_special, mtn2->mtn_mt.mnt_special); + + if (rv == 0) + return (0); + return (rv > 0 ? 1 : -1); +} + +void +libzfs_mnttab_init(libzfs_handle_t *hdl) +{ + 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)) { + 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 +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; + + 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); + } + + 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; + return (0); + } + return (ENOENT); +} + +void +libzfs_mnttab_add(libzfs_handle_t *hdl, const char *special, + const char *mountp, const char *mntopts) +{ + mnttab_node_t *mtn; + + if (avl_numnodes(&hdl->libzfs_mnttab_cache) == 0) + return; + 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); +} + +void +libzfs_mnttab_remove(libzfs_handle_t *hdl, const char *fsname) +{ + mnttab_node_t find; + mnttab_node_t *ret; + + find.mtn_mt.mnt_special = (char *)fsname; + if (ret = avl_find(&hdl->libzfs_mnttab_cache, (void *)&find, 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); + } +} + +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, 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_RECORDSIZE: + case ZFS_PROP_VOLBLOCKSIZE: + /* must be power of two within SPA_{MIN,MAX}BLOCKSIZE */ + if (intval < SPA_MINBLOCKSIZE || + intval > SPA_MAXBLOCKSIZE || !ISP2(intval)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "'%s' must be power of 2 from %u " + "to %uk"), propname, + (uint_t)SPA_MINBLOCKSIZE, + (uint_t)SPA_MAXBLOCKSIZE >> 10); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); + goto error; + } + break; + + case ZFS_PROP_MLSLABEL: + { +#ifdef sun + /* + * 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 /* !sun */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "mlslabel is not supported on FreeBSD")); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); +#endif /* !sun */ + 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; + } + (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; + } + + /* + * 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: + case ZFS_PROP_REFRESERVATION: + 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_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; + } + } + } + + /* + * 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; + + /* + * 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); + if ((zvol_volsize_to_reservation(old_volsize, zhp->zfs_props) != + old_reservation) || nvlist_lookup_uint64(nvl, + zfs_prop_to_name(resv_prop), &new_reservation) != ENOENT) { + return (0); + } + if (nvlist_lookup_uint64(nvl, zfs_prop_to_name(ZFS_PROP_VOLSIZE), + &new_volsize) != 0) + return (-1); + new_reservation = zvol_volsize_to_reservation(new_volsize, + zhp->zfs_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); +} + +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 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: + if (prop == ZFS_PROP_COMPRESSION) { + (void) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property setting is not allowed on " + "bootable datasets")); + (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) +{ + zfs_cmd_t zc = { 0 }; + int ret = -1; + prop_changelist_t *cl = NULL; + char errbuf[1024]; + libzfs_handle_t *hdl = zhp->zfs_hdl; + nvlist_t *nvl = NULL, *realprops; + zfs_prop_t prop; + boolean_t do_prefix = B_TRUE; + int added_resv; + + (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; + } + + if ((realprops = zfs_valid_proplist(hdl, zhp->zfs_type, nvl, + zfs_prop_get_int(zhp, ZFS_PROP_ZONED), zhp, errbuf)) == NULL) + goto error; + + nvlist_free(nvl); + nvl = realprops; + + prop = zfs_name_to_prop(propname); + + /* 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", propname); + ret = zfs_error(hdl, EZFS_PERM, errbuf); + goto error; + } + + if (prop == ZFS_PROP_VOLSIZE) { + if ((added_resv = zfs_add_synthetic_resv(zhp, nvl)) == -1) + goto error; + } + + if ((cl = changelist_gather(zhp, prop, 0, 0)) == NULL) + goto error; + + 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; + } + + /* + * 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) { + uint64_t idx; + int err = zprop_string_to_index(prop, propval, &idx, + ZFS_TYPE_DATASET); + if (err == 0 && idx != ZFS_CANMOUNT_OFF) + do_prefix = B_FALSE; + } + + if (do_prefix && (ret = changelist_prefix(cl)) != 0) + goto error; + + /* + * Execute the corresponding ioctl() to set this property. + */ + (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); + + if (zcmd_write_src_nvlist(hdl, &zc, nvl) != 0) + goto error; + + ret = zfs_ioctl(hdl, ZFS_IOC_SET_PROP, &zc); + + if (ret != 0) { + zfs_setprop_error(hdl, prop, errno, errbuf); + 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); + 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 { + if (do_prefix) + ret = changelist_postfix(cl); + + /* + * 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 (cl) + changelist_free(cl); + 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 char * +getprop_string(zfs_handle_t *zhp, zfs_prop_t prop, char **source) +{ + nvlist_t *nv; + char *value; + + *source = NULL; + if (nvlist_lookup_nvlist(zhp->zfs_props, + zfs_prop_to_name(prop), &nv) == 0) { + verify(nvlist_lookup_string(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); + if ((value = (char *)zfs_prop_default_string(prop)) == NULL) + value = ""; + *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; + } + + /* + * 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); + } + if (zplprops) + 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. + */ + if (zfs_prop_readonly(prop) && + *source != NULL && (*source)[0] == '\0') { + *source = NULL; + } + 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_MAXNAMELEN]; +}; + +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_MAXNAMELEN]; + 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); +} + +/* + * 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; + 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); + } + 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); + } + + break; + + case ZFS_PROP_ORIGIN: + (void) strlcpy(propbuf, getprop_string(zhp, prop, &source), + proplen); + /* + * If there is no parent at all, return failure to indicate that + * it doesn't apply to this dataset. + */ + if (propbuf[0] == '\0') + return (-1); + 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); + } + 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); + } + 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)); + 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); + 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); + break; + + case ZFS_PROP_MLSLABEL: + { +#ifdef sun + 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 /* !sun */ + propbuf[0] = '\0'; +#endif /* !sun */ + } + break; + + case ZFS_PROP_GUID: + /* + * 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); + 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); + break; + + case PROP_TYPE_STRING: + (void) strlcpy(propbuf, + getprop_string(zhp, prop, &source), proplen); + 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); + 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 sun + 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 /* !sun */ + assert(!"invalid code path"); + return (EINVAL); // silence compiler warning +#endif /* !sun */ +} + +/* + * 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'; + + /* 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 sun + /* + * It's a SID name (eg "user@domain") that needs to be + * turned into S-1-domainID-RID. + */ + directory_error_t e; + if (zoned && getzoneid() == GLOBAL_ZONEID) + return (ENOENT); + if (isuser) { + e = directory_sid_from_user_name(NULL, + cp, &numericsid); + } else { + e = directory_sid_from_group_name(NULL, + cp, &numericsid); + } + if (e != NULL) { + directory_error_free(e); + return (ENOENT); + } + if (numericsid == NULL) + return (ENOENT); + cp = numericsid; + /* will be further decoded below */ +#else /* !sun */ + return (ENOENT); +#endif /* !sun */ + } + + if (strncmp(cp, "S-1-", 4) == 0) { + /* It's a numeric SID (eg "S-1-234-567-89") */ + (void) strlcpy(domain, cp, domainlen); + cp = strrchr(domain, '-'); + *cp = '\0'; + cp++; + + errno = 0; + *ridp = strtoull(cp, &end, 10); + 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 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_MAXNAMELEN]; + 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, '/'); *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; + int rc; + + 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; + dmu_objset_type_t ost; + + (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)); + + /* 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 = DMU_OST_ZVOL; + else + ost = DMU_OST_ZFS; + + if (props && (props = zfs_valid_proplist(hdl, type, props, + zoned, NULL, errbuf)) == 0) + return (-1); + + 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_MAXNAMELEN]; + (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 EDOM: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "volume block size must be power of 2 from " + "%u to %uk"), + (uint_t)SPA_MINBLOCKSIZE, + (uint_t)SPA_MAXBLOCKSIZE >> 10); + + return (zfs_error(hdl, EZFS_BADPROP, 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)); +#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) +{ + zfs_cmd_t zc = { 0 }; + + if (zhp->zfs_type == ZFS_TYPE_BOOKMARK) { + nvlist_t *nv = fnvlist_alloc(); + fnvlist_add_boolean(nv, zhp->zfs_name); + int error = lzc_destroy_bookmarks(nv, NULL); + fnvlist_free(nv); + if (error != 0) { + return (zfs_standard_error_fmt(zhp->zfs_hdl, errno, + dgettext(TEXT_DOMAIN, "cannot destroy '%s'"), + zhp->zfs_name)); + } + return (0); + } + + (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); + + if (ZFS_IS_VOLUME(zhp)) { + zc.zc_objset_type = DMU_OST_ZVOL; + } else { + zc.zc_objset_type = DMU_OST_ZFS; + } + + zc.zc_defer_destroy = defer; + if (zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_DESTROY, &zc) != 0 && + errno != 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_MAXNAMELEN]; + 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; + + ret = lzc_destroy_snaps(snaps, defer, &errlist); + + if (ret == 0) + 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; + } + } + + 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_MAXNAMELEN]; + 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, errbuf)) == NULL) + 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; + zfs_cmd_t zc = { 0 }; + char parent[MAXPATHLEN]; + 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)); + } + + (void) strlcpy(parent, zhp->zfs_dmustats.dds_origin, sizeof (parent)); + if (parent[0] == '\0') { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "not a cloned filesystem")); + return (zfs_error(hdl, EZFS_BADTYPE, errbuf)); + } + + (void) strlcpy(zc.zc_value, zhp->zfs_dmustats.dds_origin, + sizeof (zc.zc_value)); + (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); + ret = zfs_ioctl(hdl, ZFS_IOC_PROMOTE, &zc); + + if (ret != 0) { + int save_errno = errno; + + switch (save_errno) { + case EEXIST: + /* There is a conflicting snapshot name. */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "conflicting snapshot '%s' from parent '%s'"), + zc.zc_string, parent); + return (zfs_error(hdl, EZFS_EXISTS, errbuf)); + + default: + return (zfs_standard_error(hdl, save_errno, 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_MAXNAMELEN]; + 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); +} + +/* + * 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)); + } + } + + if (props != NULL && + (props = zfs_valid_proplist(hdl, ZFS_TYPE_SNAPSHOT, + props, B_FALSE, NULL, errbuf)) == NULL) { + return (-1); + } + + 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_MAXNAMELEN]; + 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 old_volsize, 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); + (void) zfs_iter_snapshots(zhp, B_FALSE, rollback_destroy, &cb); + (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)); + } + + /* + * We rely on zfs_iter_children() to verify that there are no + * newer snapshots for the given dataset. Therefore, we can + * simply pass the name on to the ioctl() call. There is still + * an unlikely race condition where the user has taken a + * snapshot since we verified that this was the most recent. + */ + err = lzc_rollback(zhp->zfs_name, NULL, 0); + if (err != 0) { + (void) zfs_standard_error_fmt(zhp->zfs_hdl, errno, + dgettext(TEXT_DOMAIN, "cannot rollback '%s'"), + zhp->zfs_name); + 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; + zfs_cmd_t zc = { 0 }; + char *delim; + prop_changelist_t *cl = NULL; + zfs_handle_t *zhrp = NULL; + char *parentname = NULL; + char parent[ZFS_MAXNAMELEN]; + 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 the target name is valid + */ + if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) { + if ((strchr(target, '@') == NULL) || + *target == '@') { + /* + * Snapshot target name is abbreviated, + * reconstruct full dataset name + */ + (void) strlcpy(parent, zhp->zfs_name, + sizeof (parent)); + delim = strchr(parent, '@'); + if (strchr(target, '@') == 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, '@'); + if (strncmp(zhp->zfs_name, target, delim - target) + != 0 || zhp->zfs_name[delim - target] != '@') { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "snapshots must be part of same " + "dataset")); + 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 ((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); + 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 { + (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 (!flags.recurse) + (void) changelist_postfix(cl); + } else { + if (!flags.recurse) { + changelist_rename(cl, zfs_get_name(zhp), target); + ret = changelist_postfix(cl); + } + } + +error: + if (parentname) { + free(parentname); + } + if (zhrp) { + zfs_close(zhrp); + } + if (cl) { + 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 sun +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 (NULL); + } + } + + 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); + if (nvlist) + 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 /* sun */ + +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_MAXNAMELEN]; + 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_MAXNAMELEN]; + 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, ZFS_MAXNAMELEN); + + 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); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_diff.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_diff.c new file mode 100644 index 0000000..ab2007d --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_diff.c @@ -0,0 +1,834 @@ +/* + * 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) 2010, Oracle and/or its affiliates. All rights reserved. + */ + +/* + * zfs diff support + */ +#include <ctype.h> +#include <errno.h> +#include <libintl.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <stddef.h> +#include <unistd.h> +#include <stdio.h> +#include <stdlib.h> +#include <pthread.h> +#include <sys/zfs_ioctl.h> +#include <libzfs.h> +#include "libzfs_impl.h" + +#define ZDIFF_SNAPDIR "/.zfs/snapshot/" +#define ZDIFF_SHARESDIR "/.zfs/shares/" +#define ZDIFF_PREFIX "zfs-diff-%d" + +#define ZDIFF_ADDED '+' +#define ZDIFF_MODIFIED 'M' +#define ZDIFF_REMOVED '-' +#define ZDIFF_RENAMED 'R' + +static boolean_t +do_name_cmp(const char *fpath, const char *tpath) +{ + char *fname, *tname; + fname = strrchr(fpath, '/') + 1; + tname = strrchr(tpath, '/') + 1; + return (strcmp(fname, tname) == 0); +} + +typedef struct differ_info { + zfs_handle_t *zhp; + char *fromsnap; + char *frommnt; + char *tosnap; + char *tomnt; + char *ds; + char *dsmnt; + char *tmpsnap; + char errbuf[1024]; + boolean_t isclone; + boolean_t scripted; + boolean_t classify; + boolean_t timestamped; + uint64_t shares; + int zerr; + int cleanupfd; + int outputfd; + int datafd; +} differ_info_t; + +/* + * Given a {dsname, object id}, get the object path + */ +static int +get_stats_for_obj(differ_info_t *di, const char *dsname, uint64_t obj, + char *pn, int maxlen, zfs_stat_t *sb) +{ + zfs_cmd_t zc = { 0 }; + int error; + + (void) strlcpy(zc.zc_name, dsname, sizeof (zc.zc_name)); + zc.zc_obj = obj; + + errno = 0; + error = ioctl(di->zhp->zfs_hdl->libzfs_fd, ZFS_IOC_OBJ_TO_STATS, &zc); + di->zerr = errno; + + /* we can get stats even if we failed to get a path */ + (void) memcpy(sb, &zc.zc_stat, sizeof (zfs_stat_t)); + if (error == 0) { + ASSERT(di->zerr == 0); + (void) strlcpy(pn, zc.zc_value, maxlen); + return (0); + } + + if (di->zerr == EPERM) { + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "The sys_config privilege or diff delegated permission " + "is needed\nto discover path names")); + return (-1); + } else { + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "Unable to determine path or stats for " + "object %lld in %s"), obj, dsname); + return (-1); + } +} + +/* + * stream_bytes + * + * Prints a file name out a character at a time. If the character is + * not in the range of what we consider "printable" ASCII, display it + * as an escaped 3-digit octal value. ASCII values less than a space + * are all control characters and we declare the upper end as the + * DELete character. This also is the last 7-bit ASCII character. + * We choose to treat all 8-bit ASCII as not printable for this + * application. + */ +static void +stream_bytes(FILE *fp, const char *string) +{ + while (*string) { + if (*string > ' ' && *string != '\\' && *string < '\177') + (void) fprintf(fp, "%c", *string++); + else { + (void) fprintf(fp, "\\%03hho", + (unsigned char)*string++); + } + } +} + +static void +print_what(FILE *fp, mode_t what) +{ + char symbol; + + switch (what & S_IFMT) { + case S_IFBLK: + symbol = 'B'; + break; + case S_IFCHR: + symbol = 'C'; + break; + case S_IFDIR: + symbol = '/'; + break; +#ifdef S_IFDOOR + case S_IFDOOR: + symbol = '>'; + break; +#endif + case S_IFIFO: + symbol = '|'; + break; + case S_IFLNK: + symbol = '@'; + break; +#ifdef S_IFPORT + case S_IFPORT: + symbol = 'P'; + break; +#endif + case S_IFSOCK: + symbol = '='; + break; + case S_IFREG: + symbol = 'F'; + break; + default: + symbol = '?'; + break; + } + (void) fprintf(fp, "%c", symbol); +} + +static void +print_cmn(FILE *fp, differ_info_t *di, const char *file) +{ + stream_bytes(fp, di->dsmnt); + stream_bytes(fp, file); +} + +static void +print_rename(FILE *fp, differ_info_t *di, const char *old, const char *new, + zfs_stat_t *isb) +{ + if (di->timestamped) + (void) fprintf(fp, "%10lld.%09lld\t", + (longlong_t)isb->zs_ctime[0], + (longlong_t)isb->zs_ctime[1]); + (void) fprintf(fp, "%c\t", ZDIFF_RENAMED); + if (di->classify) { + print_what(fp, isb->zs_mode); + (void) fprintf(fp, "\t"); + } + print_cmn(fp, di, old); + if (di->scripted) + (void) fprintf(fp, "\t"); + else + (void) fprintf(fp, " -> "); + print_cmn(fp, di, new); + (void) fprintf(fp, "\n"); +} + +static void +print_link_change(FILE *fp, differ_info_t *di, int delta, const char *file, + zfs_stat_t *isb) +{ + if (di->timestamped) + (void) fprintf(fp, "%10lld.%09lld\t", + (longlong_t)isb->zs_ctime[0], + (longlong_t)isb->zs_ctime[1]); + (void) fprintf(fp, "%c\t", ZDIFF_MODIFIED); + if (di->classify) { + print_what(fp, isb->zs_mode); + (void) fprintf(fp, "\t"); + } + print_cmn(fp, di, file); + (void) fprintf(fp, "\t(%+d)", delta); + (void) fprintf(fp, "\n"); +} + +static void +print_file(FILE *fp, differ_info_t *di, char type, const char *file, + zfs_stat_t *isb) +{ + if (di->timestamped) + (void) fprintf(fp, "%10lld.%09lld\t", + (longlong_t)isb->zs_ctime[0], + (longlong_t)isb->zs_ctime[1]); + (void) fprintf(fp, "%c\t", type); + if (di->classify) { + print_what(fp, isb->zs_mode); + (void) fprintf(fp, "\t"); + } + print_cmn(fp, di, file); + (void) fprintf(fp, "\n"); +} + +static int +write_inuse_diffs_one(FILE *fp, differ_info_t *di, uint64_t dobj) +{ + struct zfs_stat fsb, tsb; + boolean_t same_name; + mode_t fmode, tmode; + char fobjname[MAXPATHLEN], tobjname[MAXPATHLEN]; + int fobjerr, tobjerr; + int change; + + if (dobj == di->shares) + return (0); + + /* + * Check the from and to snapshots for info on the object. If + * we get ENOENT, then the object just didn't exist in that + * snapshot. If we get ENOTSUP, then we tried to get + * info on a non-ZPL object, which we don't care about anyway. + */ + fobjerr = get_stats_for_obj(di, di->fromsnap, dobj, fobjname, + MAXPATHLEN, &fsb); + if (fobjerr && di->zerr != ENOENT && di->zerr != ENOTSUP) + return (-1); + + tobjerr = get_stats_for_obj(di, di->tosnap, dobj, tobjname, + MAXPATHLEN, &tsb); + if (tobjerr && di->zerr != ENOENT && di->zerr != ENOTSUP) + return (-1); + + /* + * Unallocated object sharing the same meta dnode block + */ + if (fobjerr && tobjerr) { + ASSERT(di->zerr == ENOENT || di->zerr == ENOTSUP); + di->zerr = 0; + return (0); + } + + di->zerr = 0; /* negate get_stats_for_obj() from side that failed */ + fmode = fsb.zs_mode & S_IFMT; + tmode = tsb.zs_mode & S_IFMT; + if (fmode == S_IFDIR || tmode == S_IFDIR || fsb.zs_links == 0 || + tsb.zs_links == 0) + change = 0; + else + change = tsb.zs_links - fsb.zs_links; + + if (fobjerr) { + if (change) { + print_link_change(fp, di, change, tobjname, &tsb); + return (0); + } + print_file(fp, di, ZDIFF_ADDED, tobjname, &tsb); + return (0); + } else if (tobjerr) { + if (change) { + print_link_change(fp, di, change, fobjname, &fsb); + return (0); + } + print_file(fp, di, ZDIFF_REMOVED, fobjname, &fsb); + return (0); + } + + if (fmode != tmode && fsb.zs_gen == tsb.zs_gen) + tsb.zs_gen++; /* Force a generational difference */ + same_name = do_name_cmp(fobjname, tobjname); + + /* Simple modification or no change */ + if (fsb.zs_gen == tsb.zs_gen) { + /* No apparent changes. Could we assert !this? */ + if (fsb.zs_ctime[0] == tsb.zs_ctime[0] && + fsb.zs_ctime[1] == tsb.zs_ctime[1]) + return (0); + if (change) { + print_link_change(fp, di, change, + change > 0 ? fobjname : tobjname, &tsb); + } else if (same_name) { + print_file(fp, di, ZDIFF_MODIFIED, fobjname, &tsb); + } else { + print_rename(fp, di, fobjname, tobjname, &tsb); + } + return (0); + } else { + /* file re-created or object re-used */ + print_file(fp, di, ZDIFF_REMOVED, fobjname, &fsb); + print_file(fp, di, ZDIFF_ADDED, tobjname, &tsb); + return (0); + } +} + +static int +write_inuse_diffs(FILE *fp, differ_info_t *di, dmu_diff_record_t *dr) +{ + uint64_t o; + int err; + + for (o = dr->ddr_first; o <= dr->ddr_last; o++) { + if (err = write_inuse_diffs_one(fp, di, o)) + return (err); + } + return (0); +} + +static int +describe_free(FILE *fp, differ_info_t *di, uint64_t object, char *namebuf, + int maxlen) +{ + struct zfs_stat sb; + + if (get_stats_for_obj(di, di->fromsnap, object, namebuf, + maxlen, &sb) != 0) { + /* Let it slide, if in the delete queue on from side */ + if (di->zerr == ENOENT && sb.zs_links == 0) { + di->zerr = 0; + return (0); + } + return (-1); + } + + print_file(fp, di, ZDIFF_REMOVED, namebuf, &sb); + return (0); +} + +static int +write_free_diffs(FILE *fp, differ_info_t *di, dmu_diff_record_t *dr) +{ + zfs_cmd_t zc = { 0 }; + libzfs_handle_t *lhdl = di->zhp->zfs_hdl; + char fobjname[MAXPATHLEN]; + + (void) strlcpy(zc.zc_name, di->fromsnap, sizeof (zc.zc_name)); + zc.zc_obj = dr->ddr_first - 1; + + ASSERT(di->zerr == 0); + + while (zc.zc_obj < dr->ddr_last) { + int err; + + err = ioctl(lhdl->libzfs_fd, ZFS_IOC_NEXT_OBJ, &zc); + if (err == 0) { + if (zc.zc_obj == di->shares) { + zc.zc_obj++; + continue; + } + if (zc.zc_obj > dr->ddr_last) { + break; + } + err = describe_free(fp, di, zc.zc_obj, fobjname, + MAXPATHLEN); + if (err) + break; + } else if (errno == ESRCH) { + break; + } else { + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "next allocated object (> %lld) find failure"), + zc.zc_obj); + di->zerr = errno; + break; + } + } + if (di->zerr) + return (-1); + return (0); +} + +static void * +differ(void *arg) +{ + differ_info_t *di = arg; + dmu_diff_record_t dr; + FILE *ofp; + int err = 0; + + if ((ofp = fdopen(di->outputfd, "w")) == NULL) { + di->zerr = errno; + (void) strerror_r(errno, di->errbuf, sizeof (di->errbuf)); + (void) close(di->datafd); + return ((void *)-1); + } + + for (;;) { + char *cp = (char *)&dr; + int len = sizeof (dr); + int rv; + + do { + rv = read(di->datafd, cp, len); + cp += rv; + len -= rv; + } while (len > 0 && rv > 0); + + if (rv < 0 || (rv == 0 && len != sizeof (dr))) { + di->zerr = EPIPE; + break; + } else if (rv == 0) { + /* end of file at a natural breaking point */ + break; + } + + switch (dr.ddr_type) { + case DDR_FREE: + err = write_free_diffs(ofp, di, &dr); + break; + case DDR_INUSE: + err = write_inuse_diffs(ofp, di, &dr); + break; + default: + di->zerr = EPIPE; + break; + } + + if (err || di->zerr) + break; + } + + (void) fclose(ofp); + (void) close(di->datafd); + if (err) + return ((void *)-1); + if (di->zerr) { + ASSERT(di->zerr == EINVAL); + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "Internal error: bad data from diff IOCTL")); + return ((void *)-1); + } + return ((void *)0); +} + +static int +find_shares_object(differ_info_t *di) +{ + char fullpath[MAXPATHLEN]; + struct stat64 sb = { 0 }; + + (void) strlcpy(fullpath, di->dsmnt, MAXPATHLEN); + (void) strlcat(fullpath, ZDIFF_SHARESDIR, MAXPATHLEN); + + if (stat64(fullpath, &sb) != 0) { +#ifdef sun + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, "Cannot stat %s"), fullpath); + return (zfs_error(di->zhp->zfs_hdl, EZFS_DIFF, di->errbuf)); +#else + return (0); +#endif + } + + di->shares = (uint64_t)sb.st_ino; + return (0); +} + +static int +make_temp_snapshot(differ_info_t *di) +{ + libzfs_handle_t *hdl = di->zhp->zfs_hdl; + zfs_cmd_t zc = { 0 }; + + (void) snprintf(zc.zc_value, sizeof (zc.zc_value), + ZDIFF_PREFIX, getpid()); + (void) strlcpy(zc.zc_name, di->ds, sizeof (zc.zc_name)); + zc.zc_cleanup_fd = di->cleanupfd; + + if (ioctl(hdl->libzfs_fd, ZFS_IOC_TMP_SNAPSHOT, &zc) != 0) { + int err = errno; + if (err == EPERM) { + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, "The diff delegated " + "permission is needed in order\nto create a " + "just-in-time snapshot for diffing\n")); + return (zfs_error(hdl, EZFS_DIFF, di->errbuf)); + } else { + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, "Cannot create just-in-time " + "snapshot of '%s'"), zc.zc_name); + return (zfs_standard_error(hdl, err, di->errbuf)); + } + } + + di->tmpsnap = zfs_strdup(hdl, zc.zc_value); + di->tosnap = zfs_asprintf(hdl, "%s@%s", di->ds, di->tmpsnap); + return (0); +} + +static void +teardown_differ_info(differ_info_t *di) +{ + free(di->ds); + free(di->dsmnt); + free(di->fromsnap); + free(di->frommnt); + free(di->tosnap); + free(di->tmpsnap); + free(di->tomnt); + (void) close(di->cleanupfd); +} + +static int +get_snapshot_names(differ_info_t *di, const char *fromsnap, + const char *tosnap) +{ + libzfs_handle_t *hdl = di->zhp->zfs_hdl; + char *atptrf = NULL; + char *atptrt = NULL; + int fdslen, fsnlen; + int tdslen, tsnlen; + + /* + * Can accept + * dataset@snap1 + * dataset@snap1 dataset@snap2 + * dataset@snap1 @snap2 + * dataset@snap1 dataset + * @snap1 dataset@snap2 + */ + if (tosnap == NULL) { + /* only a from snapshot given, must be valid */ + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "Badly formed snapshot name %s"), fromsnap); + + if (!zfs_validate_name(hdl, fromsnap, ZFS_TYPE_SNAPSHOT, + B_FALSE)) { + return (zfs_error(hdl, EZFS_INVALIDNAME, + di->errbuf)); + } + + atptrf = strchr(fromsnap, '@'); + ASSERT(atptrf != NULL); + fdslen = atptrf - fromsnap; + + di->fromsnap = zfs_strdup(hdl, fromsnap); + di->ds = zfs_strdup(hdl, fromsnap); + di->ds[fdslen] = '\0'; + + /* the to snap will be a just-in-time snap of the head */ + return (make_temp_snapshot(di)); + } + + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "Unable to determine which snapshots to compare")); + + atptrf = strchr(fromsnap, '@'); + atptrt = strchr(tosnap, '@'); + fdslen = atptrf ? atptrf - fromsnap : strlen(fromsnap); + tdslen = atptrt ? atptrt - tosnap : strlen(tosnap); + fsnlen = strlen(fromsnap) - fdslen; /* includes @ sign */ + tsnlen = strlen(tosnap) - tdslen; /* includes @ sign */ + + if (fsnlen <= 1 || tsnlen == 1 || (fdslen == 0 && tdslen == 0) || + (fsnlen == 0 && tsnlen == 0)) { + return (zfs_error(hdl, EZFS_INVALIDNAME, di->errbuf)); + } else if ((fdslen > 0 && tdslen > 0) && + ((tdslen != fdslen || strncmp(fromsnap, tosnap, fdslen) != 0))) { + /* + * not the same dataset name, might be okay if + * tosnap is a clone of a fromsnap descendant. + */ + char origin[ZFS_MAXNAMELEN]; + zprop_source_t src; + zfs_handle_t *zhp; + + di->ds = zfs_alloc(di->zhp->zfs_hdl, tdslen + 1); + (void) strncpy(di->ds, tosnap, tdslen); + di->ds[tdslen] = '\0'; + + zhp = zfs_open(hdl, di->ds, ZFS_TYPE_FILESYSTEM); + while (zhp != NULL) { + (void) zfs_prop_get(zhp, ZFS_PROP_ORIGIN, + origin, sizeof (origin), &src, NULL, 0, B_FALSE); + + if (strncmp(origin, fromsnap, fsnlen) == 0) + break; + + (void) zfs_close(zhp); + zhp = zfs_open(hdl, origin, ZFS_TYPE_FILESYSTEM); + } + + if (zhp == NULL) { + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "Not an earlier snapshot from the same fs")); + return (zfs_error(hdl, EZFS_INVALIDNAME, di->errbuf)); + } else { + (void) zfs_close(zhp); + } + + di->isclone = B_TRUE; + di->fromsnap = zfs_strdup(hdl, fromsnap); + if (tsnlen) { + di->tosnap = zfs_strdup(hdl, tosnap); + } else { + return (make_temp_snapshot(di)); + } + } else { + int dslen = fdslen ? fdslen : tdslen; + + di->ds = zfs_alloc(hdl, dslen + 1); + (void) strncpy(di->ds, fdslen ? fromsnap : tosnap, dslen); + di->ds[dslen] = '\0'; + + di->fromsnap = zfs_asprintf(hdl, "%s%s", di->ds, atptrf); + if (tsnlen) { + di->tosnap = zfs_asprintf(hdl, "%s%s", di->ds, atptrt); + } else { + return (make_temp_snapshot(di)); + } + } + return (0); +} + +static int +get_mountpoint(differ_info_t *di, char *dsnm, char **mntpt) +{ + boolean_t mounted; + + mounted = is_mounted(di->zhp->zfs_hdl, dsnm, mntpt); + if (mounted == B_FALSE) { + (void) snprintf(di->errbuf, sizeof (di->errbuf), + dgettext(TEXT_DOMAIN, + "Cannot diff an unmounted snapshot")); + return (zfs_error(di->zhp->zfs_hdl, EZFS_BADTYPE, di->errbuf)); + } + + /* Avoid a double slash at the beginning of root-mounted datasets */ + if (**mntpt == '/' && *(*mntpt + 1) == '\0') + **mntpt = '\0'; + return (0); +} + +static int +get_mountpoints(differ_info_t *di) +{ + char *strptr; + char *frommntpt; + + /* + * first get the mountpoint for the parent dataset + */ + if (get_mountpoint(di, di->ds, &di->dsmnt) != 0) + return (-1); + + strptr = strchr(di->tosnap, '@'); + ASSERT3P(strptr, !=, NULL); + di->tomnt = zfs_asprintf(di->zhp->zfs_hdl, "%s%s%s", di->dsmnt, + ZDIFF_SNAPDIR, ++strptr); + + strptr = strchr(di->fromsnap, '@'); + ASSERT3P(strptr, !=, NULL); + + frommntpt = di->dsmnt; + if (di->isclone) { + char *mntpt; + int err; + + *strptr = '\0'; + err = get_mountpoint(di, di->fromsnap, &mntpt); + *strptr = '@'; + if (err != 0) + return (-1); + frommntpt = mntpt; + } + + di->frommnt = zfs_asprintf(di->zhp->zfs_hdl, "%s%s%s", frommntpt, + ZDIFF_SNAPDIR, ++strptr); + + if (di->isclone) + free(frommntpt); + + return (0); +} + +static int +setup_differ_info(zfs_handle_t *zhp, const char *fromsnap, + const char *tosnap, differ_info_t *di) +{ + di->zhp = zhp; + + di->cleanupfd = open(ZFS_DEV, O_RDWR|O_EXCL); + VERIFY(di->cleanupfd >= 0); + + if (get_snapshot_names(di, fromsnap, tosnap) != 0) + return (-1); + + if (get_mountpoints(di) != 0) + return (-1); + + if (find_shares_object(di) != 0) + return (-1); + + return (0); +} + +int +zfs_show_diffs(zfs_handle_t *zhp, int outfd, const char *fromsnap, + const char *tosnap, int flags) +{ + zfs_cmd_t zc = { 0 }; + char errbuf[1024]; + differ_info_t di = { 0 }; + pthread_t tid; + int pipefd[2]; + int iocerr; + + (void) snprintf(errbuf, sizeof (errbuf), + dgettext(TEXT_DOMAIN, "zfs diff failed")); + + if (setup_differ_info(zhp, fromsnap, tosnap, &di)) { + teardown_differ_info(&di); + return (-1); + } + + if (pipe(pipefd)) { + zfs_error_aux(zhp->zfs_hdl, strerror(errno)); + teardown_differ_info(&di); + return (zfs_error(zhp->zfs_hdl, EZFS_PIPEFAILED, errbuf)); + } + + di.scripted = (flags & ZFS_DIFF_PARSEABLE); + di.classify = (flags & ZFS_DIFF_CLASSIFY); + di.timestamped = (flags & ZFS_DIFF_TIMESTAMP); + + di.outputfd = outfd; + di.datafd = pipefd[0]; + + if (pthread_create(&tid, NULL, differ, &di)) { + zfs_error_aux(zhp->zfs_hdl, strerror(errno)); + (void) close(pipefd[0]); + (void) close(pipefd[1]); + teardown_differ_info(&di); + return (zfs_error(zhp->zfs_hdl, + EZFS_THREADCREATEFAILED, errbuf)); + } + + /* do the ioctl() */ + (void) strlcpy(zc.zc_value, di.fromsnap, strlen(di.fromsnap) + 1); + (void) strlcpy(zc.zc_name, di.tosnap, strlen(di.tosnap) + 1); + zc.zc_cookie = pipefd[1]; + + iocerr = ioctl(zhp->zfs_hdl->libzfs_fd, ZFS_IOC_DIFF, &zc); + if (iocerr != 0) { + (void) snprintf(errbuf, sizeof (errbuf), + dgettext(TEXT_DOMAIN, "Unable to obtain diffs")); + if (errno == EPERM) { + zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN, + "\n The sys_mount privilege or diff delegated " + "permission is needed\n to execute the " + "diff ioctl")); + } else if (errno == EXDEV) { + zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN, + "\n Not an earlier snapshot from the same fs")); + } else if (errno != EPIPE || di.zerr == 0) { + zfs_error_aux(zhp->zfs_hdl, strerror(errno)); + } + (void) close(pipefd[1]); + (void) pthread_cancel(tid); + (void) pthread_join(tid, NULL); + teardown_differ_info(&di); + if (di.zerr != 0 && di.zerr != EPIPE) { + zfs_error_aux(zhp->zfs_hdl, strerror(di.zerr)); + return (zfs_error(zhp->zfs_hdl, EZFS_DIFF, di.errbuf)); + } else { + return (zfs_error(zhp->zfs_hdl, EZFS_DIFFDATA, errbuf)); + } + } + + (void) close(pipefd[1]); + (void) pthread_join(tid, NULL); + + if (di.zerr != 0) { + zfs_error_aux(zhp->zfs_hdl, strerror(di.zerr)); + return (zfs_error(zhp->zfs_hdl, EZFS_DIFF, di.errbuf)); + } + teardown_differ_info(&di); + return (0); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_fru.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_fru.c new file mode 100644 index 0000000..788fa2c --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_fru.c @@ -0,0 +1,452 @@ +/* + * 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 2009 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +#include <dlfcn.h> +#include <errno.h> +#include <libintl.h> +#include <link.h> +#include <pthread.h> +#include <strings.h> +#include <unistd.h> + +#include <libzfs.h> + +#include <fm/libtopo.h> +#include <sys/fm/protocol.h> +#include <sys/systeminfo.h> + +#include "libzfs_impl.h" + +/* + * This file is responsible for determining the relationship between I/O + * devices paths and physical locations. In the world of MPxIO and external + * enclosures, the device path is not synonymous with the physical location. + * If you remove a drive and insert it into a different slot, it will end up + * with the same path under MPxIO. If you recable storage enclosures, the + * device paths may change. All of this makes it difficult to implement the + * 'autoreplace' property, which is supposed to automatically manage disk + * replacement based on physical slot. + * + * In order to work around these limitations, we have a per-vdev FRU property + * that is the libtopo path (minus disk-specific authority information) to the + * physical location of the device on the system. This is an optional + * property, and is only needed when using the 'autoreplace' property or when + * generating FMA faults against vdevs. + */ + +/* + * Because the FMA packages depend on ZFS, we have to dlopen() libtopo in case + * it is not present. We only need this once per library instance, so it is + * not part of the libzfs handle. + */ +static void *_topo_dlhandle; +static topo_hdl_t *(*_topo_open)(int, const char *, int *); +static void (*_topo_close)(topo_hdl_t *); +static char *(*_topo_snap_hold)(topo_hdl_t *, const char *, int *); +static void (*_topo_snap_release)(topo_hdl_t *); +static topo_walk_t *(*_topo_walk_init)(topo_hdl_t *, const char *, + topo_walk_cb_t, void *, int *); +static int (*_topo_walk_step)(topo_walk_t *, int); +static void (*_topo_walk_fini)(topo_walk_t *); +static void (*_topo_hdl_strfree)(topo_hdl_t *, char *); +static char *(*_topo_node_name)(tnode_t *); +static int (*_topo_prop_get_string)(tnode_t *, const char *, const char *, + char **, int *); +static int (*_topo_node_fru)(tnode_t *, nvlist_t **, nvlist_t *, int *); +static int (*_topo_fmri_nvl2str)(topo_hdl_t *, nvlist_t *, char **, int *); +static int (*_topo_fmri_strcmp_noauth)(topo_hdl_t *, const char *, + const char *); + +#define ZFS_FRU_HASH_SIZE 257 + +static size_t +fru_strhash(const char *key) +{ + ulong_t g, h = 0; + const char *p; + + for (p = key; *p != '\0'; p++) { + h = (h << 4) + *p; + + if ((g = (h & 0xf0000000)) != 0) { + h ^= (g >> 24); + h ^= g; + } + } + + return (h % ZFS_FRU_HASH_SIZE); +} + +static int +libzfs_fru_gather(topo_hdl_t *thp, tnode_t *tn, void *arg) +{ + libzfs_handle_t *hdl = arg; + nvlist_t *fru; + char *devpath, *frustr; + int err; + libzfs_fru_t *frup; + size_t idx; + + /* + * If this is the chassis node, and we don't yet have the system + * chassis ID, then fill in this value now. + */ + if (hdl->libzfs_chassis_id[0] == '\0' && + strcmp(_topo_node_name(tn), "chassis") == 0) { + if (_topo_prop_get_string(tn, FM_FMRI_AUTHORITY, + FM_FMRI_AUTH_CHASSIS, &devpath, &err) == 0) + (void) strlcpy(hdl->libzfs_chassis_id, devpath, + sizeof (hdl->libzfs_chassis_id)); + } + + /* + * Skip non-disk nodes. + */ + if (strcmp(_topo_node_name(tn), "disk") != 0) + return (TOPO_WALK_NEXT); + + /* + * Get the devfs path and FRU. + */ + if (_topo_prop_get_string(tn, "io", "devfs-path", &devpath, &err) != 0) + return (TOPO_WALK_NEXT); + + if (libzfs_fru_lookup(hdl, devpath) != NULL) { + _topo_hdl_strfree(thp, devpath); + return (TOPO_WALK_NEXT); + } + + if (_topo_node_fru(tn, &fru, NULL, &err) != 0) { + _topo_hdl_strfree(thp, devpath); + return (TOPO_WALK_NEXT); + } + + /* + * Convert the FRU into a string. + */ + if (_topo_fmri_nvl2str(thp, fru, &frustr, &err) != 0) { + nvlist_free(fru); + _topo_hdl_strfree(thp, devpath); + return (TOPO_WALK_NEXT); + } + + nvlist_free(fru); + + /* + * Finally, we have a FRU string and device path. Add it to the hash. + */ + if ((frup = calloc(sizeof (libzfs_fru_t), 1)) == NULL) { + _topo_hdl_strfree(thp, devpath); + _topo_hdl_strfree(thp, frustr); + return (TOPO_WALK_NEXT); + } + + if ((frup->zf_device = strdup(devpath)) == NULL || + (frup->zf_fru = strdup(frustr)) == NULL) { + free(frup->zf_device); + free(frup); + _topo_hdl_strfree(thp, devpath); + _topo_hdl_strfree(thp, frustr); + return (TOPO_WALK_NEXT); + } + + _topo_hdl_strfree(thp, devpath); + _topo_hdl_strfree(thp, frustr); + + idx = fru_strhash(frup->zf_device); + frup->zf_chain = hdl->libzfs_fru_hash[idx]; + hdl->libzfs_fru_hash[idx] = frup; + frup->zf_next = hdl->libzfs_fru_list; + hdl->libzfs_fru_list = frup; + + return (TOPO_WALK_NEXT); +} + +/* + * Called during initialization to setup the dynamic libtopo connection. + */ +#pragma init(libzfs_init_fru) +static void +libzfs_init_fru(void) +{ + char path[MAXPATHLEN]; + char isa[257]; + +#if defined(_LP64) + if (sysinfo(SI_ARCHITECTURE_64, isa, sizeof (isa)) < 0) + isa[0] = '\0'; +#else + isa[0] = '\0'; +#endif + (void) snprintf(path, sizeof (path), + "/usr/lib/fm/%s/libtopo.so", isa); + + if ((_topo_dlhandle = dlopen(path, RTLD_LAZY)) == NULL) + return; + + _topo_open = (topo_hdl_t *(*)()) + dlsym(_topo_dlhandle, "topo_open"); + _topo_close = (void (*)()) + dlsym(_topo_dlhandle, "topo_close"); + _topo_snap_hold = (char *(*)()) + dlsym(_topo_dlhandle, "topo_snap_hold"); + _topo_snap_release = (void (*)()) + dlsym(_topo_dlhandle, "topo_snap_release"); + _topo_walk_init = (topo_walk_t *(*)()) + dlsym(_topo_dlhandle, "topo_walk_init"); + _topo_walk_step = (int (*)()) + dlsym(_topo_dlhandle, "topo_walk_step"); + _topo_walk_fini = (void (*)()) + dlsym(_topo_dlhandle, "topo_walk_fini"); + _topo_hdl_strfree = (void (*)()) + dlsym(_topo_dlhandle, "topo_hdl_strfree"); + _topo_node_name = (char *(*)()) + dlsym(_topo_dlhandle, "topo_node_name"); + _topo_prop_get_string = (int (*)()) + dlsym(_topo_dlhandle, "topo_prop_get_string"); + _topo_node_fru = (int (*)()) + dlsym(_topo_dlhandle, "topo_node_fru"); + _topo_fmri_nvl2str = (int (*)()) + dlsym(_topo_dlhandle, "topo_fmri_nvl2str"); + _topo_fmri_strcmp_noauth = (int (*)()) + dlsym(_topo_dlhandle, "topo_fmri_strcmp_noauth"); + + if (_topo_open == NULL || _topo_close == NULL || + _topo_snap_hold == NULL || _topo_snap_release == NULL || + _topo_walk_init == NULL || _topo_walk_step == NULL || + _topo_walk_fini == NULL || _topo_hdl_strfree == NULL || + _topo_node_name == NULL || _topo_prop_get_string == NULL || + _topo_node_fru == NULL || _topo_fmri_nvl2str == NULL || + _topo_fmri_strcmp_noauth == NULL) { + (void) dlclose(_topo_dlhandle); + _topo_dlhandle = NULL; + } +} + +/* + * Refresh the mappings from device path -> FMRI. We do this by walking the + * hc topology looking for disk nodes, and recording the io/devfs-path and FRU. + * Note that we strip out the disk-specific authority information (serial, + * part, revision, etc) so that we are left with only the identifying + * characteristics of the slot (hc path and chassis-id). + */ +void +libzfs_fru_refresh(libzfs_handle_t *hdl) +{ + int err; + char *uuid; + topo_hdl_t *thp; + topo_walk_t *twp; + + if (_topo_dlhandle == NULL) + return; + + /* + * Clear the FRU hash and initialize our basic structures. + */ + libzfs_fru_clear(hdl, B_FALSE); + + if ((hdl->libzfs_topo_hdl = _topo_open(TOPO_VERSION, + NULL, &err)) == NULL) + return; + + thp = hdl->libzfs_topo_hdl; + + if ((uuid = _topo_snap_hold(thp, NULL, &err)) == NULL) + return; + + _topo_hdl_strfree(thp, uuid); + + if (hdl->libzfs_fru_hash == NULL && + (hdl->libzfs_fru_hash = + calloc(ZFS_FRU_HASH_SIZE * sizeof (void *), 1)) == NULL) + return; + + /* + * We now have a topo snapshot, so iterate over the hc topology looking + * for disks to add to the hash. + */ + twp = _topo_walk_init(thp, FM_FMRI_SCHEME_HC, + libzfs_fru_gather, hdl, &err); + if (twp != NULL) { + (void) _topo_walk_step(twp, TOPO_WALK_CHILD); + _topo_walk_fini(twp); + } +} + +/* + * Given a devfs path, return the FRU for the device, if known. This will + * automatically call libzfs_fru_refresh() if it hasn't already been called by + * the consumer. The string returned is valid until the next call to + * libzfs_fru_refresh(). + */ +const char * +libzfs_fru_lookup(libzfs_handle_t *hdl, const char *devpath) +{ + size_t idx = fru_strhash(devpath); + libzfs_fru_t *frup; + + if (hdl->libzfs_fru_hash == NULL) + libzfs_fru_refresh(hdl); + + if (hdl->libzfs_fru_hash == NULL) + return (NULL); + + for (frup = hdl->libzfs_fru_hash[idx]; frup != NULL; + frup = frup->zf_chain) { + if (strcmp(devpath, frup->zf_device) == 0) + return (frup->zf_fru); + } + + return (NULL); +} + +/* + * Given a fru path, return the device path. This will automatically call + * libzfs_fru_refresh() if it hasn't already been called by the consumer. The + * string returned is valid until the next call to libzfs_fru_refresh(). + */ +const char * +libzfs_fru_devpath(libzfs_handle_t *hdl, const char *fru) +{ + libzfs_fru_t *frup; + size_t idx; + + if (hdl->libzfs_fru_hash == NULL) + libzfs_fru_refresh(hdl); + + if (hdl->libzfs_fru_hash == NULL) + return (NULL); + + for (idx = 0; idx < ZFS_FRU_HASH_SIZE; idx++) { + for (frup = hdl->libzfs_fru_hash[idx]; frup != NULL; + frup = frup->zf_next) { + if (_topo_fmri_strcmp_noauth(hdl->libzfs_topo_hdl, + fru, frup->zf_fru)) + return (frup->zf_device); + } + } + + return (NULL); +} + +/* + * Change the stored FRU for the given vdev. + */ +int +zpool_fru_set(zpool_handle_t *zhp, uint64_t vdev_guid, const char *fru) +{ + zfs_cmd_t zc = { 0 }; + + (void) strncpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + (void) strncpy(zc.zc_value, fru, sizeof (zc.zc_value)); + zc.zc_guid = vdev_guid; + + if (zfs_ioctl(zhp->zpool_hdl, ZFS_IOC_VDEV_SETFRU, &zc) != 0) + return (zpool_standard_error_fmt(zhp->zpool_hdl, errno, + dgettext(TEXT_DOMAIN, "cannot set FRU"))); + + return (0); +} + +/* + * Compare to two FRUs, ignoring any authority information. + */ +boolean_t +libzfs_fru_compare(libzfs_handle_t *hdl, const char *a, const char *b) +{ + if (hdl->libzfs_fru_hash == NULL) + libzfs_fru_refresh(hdl); + + if (hdl->libzfs_fru_hash == NULL) + return (strcmp(a, b) == 0); + + return (_topo_fmri_strcmp_noauth(hdl->libzfs_topo_hdl, a, b)); +} + +/* + * This special function checks to see whether the FRU indicates it's supposed + * to be in the system chassis, but the chassis-id doesn't match. This can + * happen in a clustered case, where both head nodes have the same logical + * disk, but opening the device on the other head node is meaningless. + */ +boolean_t +libzfs_fru_notself(libzfs_handle_t *hdl, const char *fru) +{ + const char *chassisid; + size_t len; + + if (hdl->libzfs_fru_hash == NULL) + libzfs_fru_refresh(hdl); + + if (hdl->libzfs_chassis_id[0] == '\0') + return (B_FALSE); + + if (strstr(fru, "/chassis=0/") == NULL) + return (B_FALSE); + + if ((chassisid = strstr(fru, ":chassis-id=")) == NULL) + return (B_FALSE); + + chassisid += 12; + len = strlen(hdl->libzfs_chassis_id); + if (strncmp(chassisid, hdl->libzfs_chassis_id, len) == 0 && + (chassisid[len] == '/' || chassisid[len] == ':')) + return (B_FALSE); + + return (B_TRUE); +} + +/* + * Clear memory associated with the FRU hash. + */ +void +libzfs_fru_clear(libzfs_handle_t *hdl, boolean_t final) +{ + libzfs_fru_t *frup; + + while ((frup = hdl->libzfs_fru_list) != NULL) { + hdl->libzfs_fru_list = frup->zf_next; + free(frup->zf_device); + free(frup->zf_fru); + free(frup); + } + + hdl->libzfs_fru_list = NULL; + + if (hdl->libzfs_topo_hdl != NULL) { + _topo_snap_release(hdl->libzfs_topo_hdl); + _topo_close(hdl->libzfs_topo_hdl); + hdl->libzfs_topo_hdl = NULL; + } + + if (final) { + free(hdl->libzfs_fru_hash); + } else if (hdl->libzfs_fru_hash != NULL) { + bzero(hdl->libzfs_fru_hash, + ZFS_FRU_HASH_SIZE * sizeof (void *)); + } +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_impl.h b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_impl.h new file mode 100644 index 0000000..481ae52 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_impl.h @@ -0,0 +1,223 @@ +/* + * CDDL HEADER SART + * + * 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) 2013 by Delphix. All rights reserved. + * Copyright (c) 2013 Martin Matuska <mm@FreeBSD.org>. All rights reserved. + */ + +#ifndef _LIBZFS_IMPL_H +#define _LIBZFS_IMPL_H + +#include <sys/dmu.h> +#include <sys/fs/zfs.h> +#include <sys/zfs_ioctl.h> +#include <sys/spa.h> +#include <sys/nvpair.h> + +#include <libshare.h> +#include <libuutil.h> +#include <libzfs.h> +#include <libzfs_core.h> +#include <libzfs_compat.h> + +#ifdef __cplusplus +extern "C" { +#endif + +#ifdef VERIFY +#undef VERIFY +#endif +#define VERIFY verify + +typedef struct libzfs_fru { + char *zf_device; + char *zf_fru; + struct libzfs_fru *zf_chain; + struct libzfs_fru *zf_next; +} libzfs_fru_t; + +struct libzfs_handle { + int libzfs_error; + int libzfs_fd; + FILE *libzfs_mnttab; + FILE *libzfs_sharetab; + zpool_handle_t *libzfs_pool_handles; + uu_avl_pool_t *libzfs_ns_avlpool; + uu_avl_t *libzfs_ns_avl; + uint64_t libzfs_ns_gen; + int libzfs_desc_active; + char libzfs_action[1024]; + char libzfs_desc[1024]; + int libzfs_printerr; + int libzfs_storeerr; /* stuff error messages into buffer */ + void *libzfs_sharehdl; /* libshare handle */ + uint_t libzfs_shareflags; + boolean_t libzfs_mnttab_enable; + avl_tree_t libzfs_mnttab_cache; + int libzfs_pool_iter; + libzfs_fru_t **libzfs_fru_hash; + libzfs_fru_t *libzfs_fru_list; + char libzfs_chassis_id[256]; +}; + +#define ZFSSHARE_MISS 0x01 /* Didn't find entry in cache */ + +struct zfs_handle { + libzfs_handle_t *zfs_hdl; + zpool_handle_t *zpool_hdl; + char zfs_name[ZFS_MAXNAMELEN]; + zfs_type_t zfs_type; /* type including snapshot */ + zfs_type_t zfs_head_type; /* type excluding snapshot */ + dmu_objset_stats_t zfs_dmustats; + nvlist_t *zfs_props; + nvlist_t *zfs_user_props; + nvlist_t *zfs_recvd_props; + boolean_t zfs_mntcheck; + char *zfs_mntopts; + uint8_t *zfs_props_table; +}; + +/* + * This is different from checking zfs_type, because it will also catch + * snapshots of volumes. + */ +#define ZFS_IS_VOLUME(zhp) ((zhp)->zfs_head_type == ZFS_TYPE_VOLUME) + +struct zpool_handle { + libzfs_handle_t *zpool_hdl; + zpool_handle_t *zpool_next; + char zpool_name[ZPOOL_MAXNAMELEN]; + int zpool_state; + size_t zpool_config_size; + nvlist_t *zpool_config; + nvlist_t *zpool_old_config; + nvlist_t *zpool_props; + diskaddr_t zpool_start_block; +}; + +typedef enum { + PROTO_NFS = 0, + PROTO_SMB = 1, + PROTO_END = 2 +} zfs_share_proto_t; + +/* + * The following can be used as a bitmask and any new values + * added must preserve that capability. + */ +typedef enum { + SHARED_NOT_SHARED = 0x0, + SHARED_NFS = 0x2, + SHARED_SMB = 0x4 +} zfs_share_type_t; + +int zfs_error(libzfs_handle_t *, int, const char *); +int zfs_error_fmt(libzfs_handle_t *, int, const char *, ...); +void zfs_error_aux(libzfs_handle_t *, const char *, ...); +void *zfs_alloc(libzfs_handle_t *, size_t); +void *zfs_realloc(libzfs_handle_t *, void *, size_t, size_t); +char *zfs_asprintf(libzfs_handle_t *, const char *, ...); +char *zfs_strdup(libzfs_handle_t *, const char *); +int no_memory(libzfs_handle_t *); + +int zfs_standard_error(libzfs_handle_t *, int, const char *); +int zfs_standard_error_fmt(libzfs_handle_t *, int, const char *, ...); +int zpool_standard_error(libzfs_handle_t *, int, const char *); +int zpool_standard_error_fmt(libzfs_handle_t *, int, const char *, ...); + +int get_dependents(libzfs_handle_t *, boolean_t, const char *, char ***, + size_t *); +zfs_handle_t *make_dataset_handle_zc(libzfs_handle_t *, zfs_cmd_t *); +zfs_handle_t *make_dataset_simple_handle_zc(zfs_handle_t *, zfs_cmd_t *); + +int zprop_parse_value(libzfs_handle_t *, nvpair_t *, int, zfs_type_t, + nvlist_t *, char **, uint64_t *, const char *); +int zprop_expand_list(libzfs_handle_t *hdl, zprop_list_t **plp, + zfs_type_t type); + +/* + * Use this changelist_gather() flag to force attempting mounts + * on each change node regardless of whether or not it is currently + * mounted. + */ +#define CL_GATHER_MOUNT_ALWAYS 0x01 +/* + * Use this changelist_gather() flag to prevent unmounting of file systems. + */ +#define CL_GATHER_DONT_UNMOUNT 0x02 + +typedef struct prop_changelist prop_changelist_t; + +int zcmd_alloc_dst_nvlist(libzfs_handle_t *, zfs_cmd_t *, size_t); +int zcmd_write_src_nvlist(libzfs_handle_t *, zfs_cmd_t *, nvlist_t *); +int zcmd_write_conf_nvlist(libzfs_handle_t *, zfs_cmd_t *, nvlist_t *); +int zcmd_expand_dst_nvlist(libzfs_handle_t *, zfs_cmd_t *); +int zcmd_read_dst_nvlist(libzfs_handle_t *, zfs_cmd_t *, nvlist_t **); +void zcmd_free_nvlists(zfs_cmd_t *); + +int changelist_prefix(prop_changelist_t *); +int changelist_postfix(prop_changelist_t *); +void changelist_rename(prop_changelist_t *, const char *, const char *); +void changelist_remove(prop_changelist_t *, const char *); +void changelist_free(prop_changelist_t *); +prop_changelist_t *changelist_gather(zfs_handle_t *, zfs_prop_t, int, int); +int changelist_unshare(prop_changelist_t *, zfs_share_proto_t *); +int changelist_haszonedchild(prop_changelist_t *); + +void remove_mountpoint(zfs_handle_t *); +int create_parents(libzfs_handle_t *, char *, int); +boolean_t isa_child_of(const char *dataset, const char *parent); + +zfs_handle_t *make_dataset_handle(libzfs_handle_t *, const char *); +zfs_handle_t *make_bookmark_handle(zfs_handle_t *, const char *, + nvlist_t *props); + +int zpool_open_silent(libzfs_handle_t *, const char *, zpool_handle_t **); + +boolean_t zpool_name_valid(libzfs_handle_t *, boolean_t, const char *); + +int zfs_validate_name(libzfs_handle_t *hdl, const char *path, int type, + boolean_t modifying); + +void namespace_clear(libzfs_handle_t *); + +/* + * libshare (sharemgr) interfaces used internally. + */ + +extern int zfs_init_libshare(libzfs_handle_t *, int); +extern void zfs_uninit_libshare(libzfs_handle_t *); +extern int zfs_parse_options(char *, zfs_share_proto_t); + +extern int zfs_unshare_proto(zfs_handle_t *, + const char *, zfs_share_proto_t *); + +extern void libzfs_fru_clear(libzfs_handle_t *, boolean_t); + +#ifdef __cplusplus +} +#endif + +#endif /* _LIBZFS_IMPL_H */ diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_import.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_import.c new file mode 100644 index 0000000..868961d --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_import.c @@ -0,0 +1,1747 @@ +/* + * 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 by Delphix. All rights reserved. + * Copyright 2014 Nexenta Systems, Inc. All rights reserved. + */ + +/* + * Pool import support functions. + * + * To import a pool, we rely on reading the configuration information from the + * ZFS label of each device. If we successfully read the label, then we + * organize the configuration information in the following hierarchy: + * + * pool guid -> toplevel vdev guid -> label txg + * + * Duplicate entries matching this same tuple will be discarded. Once we have + * examined every device, we pick the best label txg config for each toplevel + * vdev. We then arrange these toplevel vdevs into a complete pool config, and + * update any paths that have changed. Finally, we attempt to import the pool + * using our derived config, and record the results. + */ + +#include <ctype.h> +#include <devid.h> +#include <dirent.h> +#include <errno.h> +#include <libintl.h> +#include <stddef.h> +#include <stdlib.h> +#include <string.h> +#include <sys/stat.h> +#include <unistd.h> +#include <fcntl.h> +#include <thread_pool.h> +#include <libgeom.h> + +#include <sys/vdev_impl.h> + +#include "libzfs.h" +#include "libzfs_impl.h" + +/* + * Intermediate structures used to gather configuration information. + */ +typedef struct config_entry { + uint64_t ce_txg; + nvlist_t *ce_config; + struct config_entry *ce_next; +} config_entry_t; + +typedef struct vdev_entry { + uint64_t ve_guid; + config_entry_t *ve_configs; + struct vdev_entry *ve_next; +} vdev_entry_t; + +typedef struct pool_entry { + uint64_t pe_guid; + vdev_entry_t *pe_vdevs; + struct pool_entry *pe_next; +} pool_entry_t; + +typedef struct name_entry { + char *ne_name; + uint64_t ne_guid; + struct name_entry *ne_next; +} name_entry_t; + +typedef struct pool_list { + pool_entry_t *pools; + name_entry_t *names; +} pool_list_t; + +static char * +get_devid(const char *path) +{ +#ifdef have_devid + int fd; + ddi_devid_t devid; + char *minor, *ret; + + if ((fd = open(path, O_RDONLY)) < 0) + return (NULL); + + minor = NULL; + ret = NULL; + if (devid_get(fd, &devid) == 0) { + if (devid_get_minor_name(fd, &minor) == 0) + ret = devid_str_encode(devid, minor); + if (minor != NULL) + devid_str_free(minor); + devid_free(devid); + } + (void) close(fd); + + return (ret); +#else + return (NULL); +#endif +} + + +/* + * Go through and fix up any path and/or devid information for the given vdev + * configuration. + */ +static int +fix_paths(nvlist_t *nv, name_entry_t *names) +{ + nvlist_t **child; + uint_t c, children; + uint64_t guid; + name_entry_t *ne, *best; + char *path, *devid; + int matched; + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) == 0) { + for (c = 0; c < children; c++) + if (fix_paths(child[c], names) != 0) + return (-1); + return (0); + } + + /* + * This is a leaf (file or disk) vdev. In either case, go through + * the name list and see if we find a matching guid. If so, replace + * the path and see if we can calculate a new devid. + * + * There may be multiple names associated with a particular guid, in + * which case we have overlapping slices or multiple paths to the same + * disk. If this is the case, then we want to pick the path that is + * the most similar to the original, where "most similar" is the number + * of matching characters starting from the end of the path. This will + * preserve slice numbers even if the disks have been reorganized, and + * will also catch preferred disk names if multiple paths exist. + */ + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, &guid) == 0); + if (nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) != 0) + path = NULL; + + matched = 0; + best = NULL; + for (ne = names; ne != NULL; ne = ne->ne_next) { + if (ne->ne_guid == guid) { + const char *src, *dst; + int count; + + if (path == NULL) { + best = ne; + break; + } + + src = ne->ne_name + strlen(ne->ne_name) - 1; + dst = path + strlen(path) - 1; + for (count = 0; src >= ne->ne_name && dst >= path; + src--, dst--, count++) + if (*src != *dst) + break; + + /* + * At this point, 'count' is the number of characters + * matched from the end. + */ + if (count > matched || best == NULL) { + best = ne; + matched = count; + } + } + } + + if (best == NULL) + return (0); + + if (nvlist_add_string(nv, ZPOOL_CONFIG_PATH, best->ne_name) != 0) + return (-1); + + if ((devid = get_devid(best->ne_name)) == NULL) { + (void) nvlist_remove_all(nv, ZPOOL_CONFIG_DEVID); + } else { + if (nvlist_add_string(nv, ZPOOL_CONFIG_DEVID, devid) != 0) + return (-1); + devid_str_free(devid); + } + + return (0); +} + +/* + * Add the given configuration to the list of known devices. + */ +static int +add_config(libzfs_handle_t *hdl, pool_list_t *pl, const char *path, + nvlist_t *config) +{ + uint64_t pool_guid, vdev_guid, top_guid, txg, state; + pool_entry_t *pe; + vdev_entry_t *ve; + config_entry_t *ce; + name_entry_t *ne; + + /* + * If this is a hot spare not currently in use or level 2 cache + * device, add it to the list of names to translate, but don't do + * anything else. + */ + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE, + &state) == 0 && + (state == POOL_STATE_SPARE || state == POOL_STATE_L2CACHE) && + nvlist_lookup_uint64(config, ZPOOL_CONFIG_GUID, &vdev_guid) == 0) { + if ((ne = zfs_alloc(hdl, sizeof (name_entry_t))) == NULL) + return (-1); + + if ((ne->ne_name = zfs_strdup(hdl, path)) == NULL) { + free(ne); + return (-1); + } + ne->ne_guid = vdev_guid; + ne->ne_next = pl->names; + pl->names = ne; + return (0); + } + + /* + * If we have a valid config but cannot read any of these fields, then + * it means we have a half-initialized label. In vdev_label_init() + * we write a label with txg == 0 so that we can identify the device + * in case the user refers to the same disk later on. If we fail to + * create the pool, we'll be left with a label in this state + * which should not be considered part of a valid pool. + */ + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, + &pool_guid) != 0 || + nvlist_lookup_uint64(config, ZPOOL_CONFIG_GUID, + &vdev_guid) != 0 || + nvlist_lookup_uint64(config, ZPOOL_CONFIG_TOP_GUID, + &top_guid) != 0 || + nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_TXG, + &txg) != 0 || txg == 0) { + nvlist_free(config); + return (0); + } + + /* + * First, see if we know about this pool. If not, then add it to the + * list of known pools. + */ + for (pe = pl->pools; pe != NULL; pe = pe->pe_next) { + if (pe->pe_guid == pool_guid) + break; + } + + if (pe == NULL) { + if ((pe = zfs_alloc(hdl, sizeof (pool_entry_t))) == NULL) { + nvlist_free(config); + return (-1); + } + pe->pe_guid = pool_guid; + pe->pe_next = pl->pools; + pl->pools = pe; + } + + /* + * Second, see if we know about this toplevel vdev. Add it if its + * missing. + */ + for (ve = pe->pe_vdevs; ve != NULL; ve = ve->ve_next) { + if (ve->ve_guid == top_guid) + break; + } + + if (ve == NULL) { + if ((ve = zfs_alloc(hdl, sizeof (vdev_entry_t))) == NULL) { + nvlist_free(config); + return (-1); + } + ve->ve_guid = top_guid; + ve->ve_next = pe->pe_vdevs; + pe->pe_vdevs = ve; + } + + /* + * Third, see if we have a config with a matching transaction group. If + * so, then we do nothing. Otherwise, add it to the list of known + * configs. + */ + for (ce = ve->ve_configs; ce != NULL; ce = ce->ce_next) { + if (ce->ce_txg == txg) + break; + } + + if (ce == NULL) { + if ((ce = zfs_alloc(hdl, sizeof (config_entry_t))) == NULL) { + nvlist_free(config); + return (-1); + } + ce->ce_txg = txg; + ce->ce_config = config; + ce->ce_next = ve->ve_configs; + ve->ve_configs = ce; + } else { + nvlist_free(config); + } + + /* + * At this point we've successfully added our config to the list of + * known configs. The last thing to do is add the vdev guid -> path + * mappings so that we can fix up the configuration as necessary before + * doing the import. + */ + if ((ne = zfs_alloc(hdl, sizeof (name_entry_t))) == NULL) + return (-1); + + if ((ne->ne_name = zfs_strdup(hdl, path)) == NULL) { + free(ne); + return (-1); + } + + ne->ne_guid = vdev_guid; + ne->ne_next = pl->names; + pl->names = ne; + + return (0); +} + +/* + * Returns true if the named pool matches the given GUID. + */ +static int +pool_active(libzfs_handle_t *hdl, const char *name, uint64_t guid, + boolean_t *isactive) +{ + zpool_handle_t *zhp; + uint64_t theguid; + + if (zpool_open_silent(hdl, name, &zhp) != 0) + return (-1); + + if (zhp == NULL) { + *isactive = B_FALSE; + return (0); + } + + verify(nvlist_lookup_uint64(zhp->zpool_config, ZPOOL_CONFIG_POOL_GUID, + &theguid) == 0); + + zpool_close(zhp); + + *isactive = (theguid == guid); + return (0); +} + +static nvlist_t * +refresh_config(libzfs_handle_t *hdl, nvlist_t *config) +{ + nvlist_t *nvl; + zfs_cmd_t zc = { 0 }; + int err; + + if (zcmd_write_conf_nvlist(hdl, &zc, config) != 0) + return (NULL); + + if (zcmd_alloc_dst_nvlist(hdl, &zc, + zc.zc_nvlist_conf_size * 2) != 0) { + zcmd_free_nvlists(&zc); + return (NULL); + } + + while ((err = ioctl(hdl->libzfs_fd, ZFS_IOC_POOL_TRYIMPORT, + &zc)) != 0 && errno == ENOMEM) { + if (zcmd_expand_dst_nvlist(hdl, &zc) != 0) { + zcmd_free_nvlists(&zc); + return (NULL); + } + } + + if (err) { + zcmd_free_nvlists(&zc); + return (NULL); + } + + if (zcmd_read_dst_nvlist(hdl, &zc, &nvl) != 0) { + zcmd_free_nvlists(&zc); + return (NULL); + } + + zcmd_free_nvlists(&zc); + return (nvl); +} + +/* + * Determine if the vdev id is a hole in the namespace. + */ +boolean_t +vdev_is_hole(uint64_t *hole_array, uint_t holes, uint_t id) +{ + for (int c = 0; c < holes; c++) { + + /* Top-level is a hole */ + if (hole_array[c] == id) + return (B_TRUE); + } + return (B_FALSE); +} + +/* + * Convert our list of pools into the definitive set of configurations. We + * start by picking the best config for each toplevel vdev. Once that's done, + * we assemble the toplevel vdevs into a full config for the pool. We make a + * pass to fix up any incorrect paths, and then add it to the main list to + * return to the user. + */ +static nvlist_t * +get_configs(libzfs_handle_t *hdl, pool_list_t *pl, boolean_t active_ok) +{ + pool_entry_t *pe; + vdev_entry_t *ve; + config_entry_t *ce; + nvlist_t *ret = NULL, *config = NULL, *tmp, *nvtop, *nvroot; + nvlist_t **spares, **l2cache; + uint_t i, nspares, nl2cache; + boolean_t config_seen; + uint64_t best_txg; + char *name, *hostname; + uint64_t guid; + uint_t children = 0; + nvlist_t **child = NULL; + uint_t holes; + uint64_t *hole_array, max_id; + uint_t c; + boolean_t isactive; + uint64_t hostid; + nvlist_t *nvl; + boolean_t found_one = B_FALSE; + boolean_t valid_top_config = B_FALSE; + + if (nvlist_alloc(&ret, 0, 0) != 0) + goto nomem; + + for (pe = pl->pools; pe != NULL; pe = pe->pe_next) { + uint64_t id, max_txg = 0; + + if (nvlist_alloc(&config, NV_UNIQUE_NAME, 0) != 0) + goto nomem; + config_seen = B_FALSE; + + /* + * Iterate over all toplevel vdevs. Grab the pool configuration + * from the first one we find, and then go through the rest and + * add them as necessary to the 'vdevs' member of the config. + */ + for (ve = pe->pe_vdevs; ve != NULL; ve = ve->ve_next) { + + /* + * Determine the best configuration for this vdev by + * selecting the config with the latest transaction + * group. + */ + best_txg = 0; + for (ce = ve->ve_configs; ce != NULL; + ce = ce->ce_next) { + + if (ce->ce_txg > best_txg) { + tmp = ce->ce_config; + best_txg = ce->ce_txg; + } + } + + /* + * We rely on the fact that the max txg for the + * pool will contain the most up-to-date information + * about the valid top-levels in the vdev namespace. + */ + if (best_txg > max_txg) { + (void) nvlist_remove(config, + ZPOOL_CONFIG_VDEV_CHILDREN, + DATA_TYPE_UINT64); + (void) nvlist_remove(config, + ZPOOL_CONFIG_HOLE_ARRAY, + DATA_TYPE_UINT64_ARRAY); + + max_txg = best_txg; + hole_array = NULL; + holes = 0; + max_id = 0; + valid_top_config = B_FALSE; + + if (nvlist_lookup_uint64(tmp, + ZPOOL_CONFIG_VDEV_CHILDREN, &max_id) == 0) { + verify(nvlist_add_uint64(config, + ZPOOL_CONFIG_VDEV_CHILDREN, + max_id) == 0); + valid_top_config = B_TRUE; + } + + if (nvlist_lookup_uint64_array(tmp, + ZPOOL_CONFIG_HOLE_ARRAY, &hole_array, + &holes) == 0) { + verify(nvlist_add_uint64_array(config, + ZPOOL_CONFIG_HOLE_ARRAY, + hole_array, holes) == 0); + } + } + + if (!config_seen) { + /* + * Copy the relevant pieces of data to the pool + * configuration: + * + * version + * pool guid + * name + * comment (if available) + * pool state + * hostid (if available) + * hostname (if available) + */ + uint64_t state, version; + char *comment = NULL; + + version = fnvlist_lookup_uint64(tmp, + ZPOOL_CONFIG_VERSION); + fnvlist_add_uint64(config, + ZPOOL_CONFIG_VERSION, version); + guid = fnvlist_lookup_uint64(tmp, + ZPOOL_CONFIG_POOL_GUID); + fnvlist_add_uint64(config, + ZPOOL_CONFIG_POOL_GUID, guid); + name = fnvlist_lookup_string(tmp, + ZPOOL_CONFIG_POOL_NAME); + fnvlist_add_string(config, + ZPOOL_CONFIG_POOL_NAME, name); + + if (nvlist_lookup_string(tmp, + ZPOOL_CONFIG_COMMENT, &comment) == 0) + fnvlist_add_string(config, + ZPOOL_CONFIG_COMMENT, comment); + + state = fnvlist_lookup_uint64(tmp, + ZPOOL_CONFIG_POOL_STATE); + fnvlist_add_uint64(config, + ZPOOL_CONFIG_POOL_STATE, state); + + hostid = 0; + if (nvlist_lookup_uint64(tmp, + ZPOOL_CONFIG_HOSTID, &hostid) == 0) { + fnvlist_add_uint64(config, + ZPOOL_CONFIG_HOSTID, hostid); + hostname = fnvlist_lookup_string(tmp, + ZPOOL_CONFIG_HOSTNAME); + fnvlist_add_string(config, + ZPOOL_CONFIG_HOSTNAME, hostname); + } + + config_seen = B_TRUE; + } + + /* + * Add this top-level vdev to the child array. + */ + verify(nvlist_lookup_nvlist(tmp, + ZPOOL_CONFIG_VDEV_TREE, &nvtop) == 0); + verify(nvlist_lookup_uint64(nvtop, ZPOOL_CONFIG_ID, + &id) == 0); + + if (id >= children) { + nvlist_t **newchild; + + newchild = zfs_alloc(hdl, (id + 1) * + sizeof (nvlist_t *)); + if (newchild == NULL) + goto nomem; + + for (c = 0; c < children; c++) + newchild[c] = child[c]; + + free(child); + child = newchild; + children = id + 1; + } + if (nvlist_dup(nvtop, &child[id], 0) != 0) + goto nomem; + + } + + /* + * If we have information about all the top-levels then + * clean up the nvlist which we've constructed. This + * means removing any extraneous devices that are + * beyond the valid range or adding devices to the end + * of our array which appear to be missing. + */ + if (valid_top_config) { + if (max_id < children) { + for (c = max_id; c < children; c++) + nvlist_free(child[c]); + children = max_id; + } else if (max_id > children) { + nvlist_t **newchild; + + newchild = zfs_alloc(hdl, (max_id) * + sizeof (nvlist_t *)); + if (newchild == NULL) + goto nomem; + + for (c = 0; c < children; c++) + newchild[c] = child[c]; + + free(child); + child = newchild; + children = max_id; + } + } + + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, + &guid) == 0); + + /* + * The vdev namespace may contain holes as a result of + * device removal. We must add them back into the vdev + * tree before we process any missing devices. + */ + if (holes > 0) { + ASSERT(valid_top_config); + + for (c = 0; c < children; c++) { + nvlist_t *holey; + + if (child[c] != NULL || + !vdev_is_hole(hole_array, holes, c)) + continue; + + if (nvlist_alloc(&holey, NV_UNIQUE_NAME, + 0) != 0) + goto nomem; + + /* + * Holes in the namespace are treated as + * "hole" top-level vdevs and have a + * special flag set on them. + */ + if (nvlist_add_string(holey, + ZPOOL_CONFIG_TYPE, + VDEV_TYPE_HOLE) != 0 || + nvlist_add_uint64(holey, + ZPOOL_CONFIG_ID, c) != 0 || + nvlist_add_uint64(holey, + ZPOOL_CONFIG_GUID, 0ULL) != 0) + goto nomem; + child[c] = holey; + } + } + + /* + * Look for any missing top-level vdevs. If this is the case, + * create a faked up 'missing' vdev as a placeholder. We cannot + * simply compress the child array, because the kernel performs + * certain checks to make sure the vdev IDs match their location + * in the configuration. + */ + for (c = 0; c < children; c++) { + if (child[c] == NULL) { + nvlist_t *missing; + if (nvlist_alloc(&missing, NV_UNIQUE_NAME, + 0) != 0) + goto nomem; + if (nvlist_add_string(missing, + ZPOOL_CONFIG_TYPE, + VDEV_TYPE_MISSING) != 0 || + nvlist_add_uint64(missing, + ZPOOL_CONFIG_ID, c) != 0 || + nvlist_add_uint64(missing, + ZPOOL_CONFIG_GUID, 0ULL) != 0) { + nvlist_free(missing); + goto nomem; + } + child[c] = missing; + } + } + + /* + * Put all of this pool's top-level vdevs into a root vdev. + */ + if (nvlist_alloc(&nvroot, NV_UNIQUE_NAME, 0) != 0) + goto nomem; + if (nvlist_add_string(nvroot, ZPOOL_CONFIG_TYPE, + VDEV_TYPE_ROOT) != 0 || + nvlist_add_uint64(nvroot, ZPOOL_CONFIG_ID, 0ULL) != 0 || + nvlist_add_uint64(nvroot, ZPOOL_CONFIG_GUID, guid) != 0 || + nvlist_add_nvlist_array(nvroot, ZPOOL_CONFIG_CHILDREN, + child, children) != 0) { + nvlist_free(nvroot); + goto nomem; + } + + for (c = 0; c < children; c++) + nvlist_free(child[c]); + free(child); + children = 0; + child = NULL; + + /* + * Go through and fix up any paths and/or devids based on our + * known list of vdev GUID -> path mappings. + */ + if (fix_paths(nvroot, pl->names) != 0) { + nvlist_free(nvroot); + goto nomem; + } + + /* + * Add the root vdev to this pool's configuration. + */ + if (nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + nvroot) != 0) { + nvlist_free(nvroot); + goto nomem; + } + nvlist_free(nvroot); + + /* + * zdb uses this path to report on active pools that were + * imported or created using -R. + */ + if (active_ok) + goto add_pool; + + /* + * Determine if this pool is currently active, in which case we + * can't actually import it. + */ + verify(nvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME, + &name) == 0); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, + &guid) == 0); + + if (pool_active(hdl, name, guid, &isactive) != 0) + goto error; + + if (isactive) { + nvlist_free(config); + config = NULL; + continue; + } + + if ((nvl = refresh_config(hdl, config)) == NULL) { + nvlist_free(config); + config = NULL; + continue; + } + + nvlist_free(config); + config = nvl; + + /* + * Go through and update the paths for spares, now that we have + * them. + */ + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + if (nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_SPARES, + &spares, &nspares) == 0) { + for (i = 0; i < nspares; i++) { + if (fix_paths(spares[i], pl->names) != 0) + goto nomem; + } + } + + /* + * Update the paths for l2cache devices. + */ + if (nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_L2CACHE, + &l2cache, &nl2cache) == 0) { + for (i = 0; i < nl2cache; i++) { + if (fix_paths(l2cache[i], pl->names) != 0) + goto nomem; + } + } + + /* + * Restore the original information read from the actual label. + */ + (void) nvlist_remove(config, ZPOOL_CONFIG_HOSTID, + DATA_TYPE_UINT64); + (void) nvlist_remove(config, ZPOOL_CONFIG_HOSTNAME, + DATA_TYPE_STRING); + if (hostid != 0) { + verify(nvlist_add_uint64(config, ZPOOL_CONFIG_HOSTID, + hostid) == 0); + verify(nvlist_add_string(config, ZPOOL_CONFIG_HOSTNAME, + hostname) == 0); + } + +add_pool: + /* + * Add this pool to the list of configs. + */ + verify(nvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME, + &name) == 0); + if (nvlist_add_nvlist(ret, name, config) != 0) + goto nomem; + + found_one = B_TRUE; + nvlist_free(config); + config = NULL; + } + + if (!found_one) { + nvlist_free(ret); + ret = NULL; + } + + return (ret); + +nomem: + (void) no_memory(hdl); +error: + nvlist_free(config); + nvlist_free(ret); + for (c = 0; c < children; c++) + nvlist_free(child[c]); + free(child); + + return (NULL); +} + +/* + * Return the offset of the given label. + */ +static uint64_t +label_offset(uint64_t size, int l) +{ + ASSERT(P2PHASE_TYPED(size, sizeof (vdev_label_t), uint64_t) == 0); + return (l * sizeof (vdev_label_t) + (l < VDEV_LABELS / 2 ? + 0 : size - VDEV_LABELS * sizeof (vdev_label_t))); +} + +/* + * Given a file descriptor, read the label information and return an nvlist + * describing the configuration, if there is one. + */ +int +zpool_read_label(int fd, nvlist_t **config) +{ + struct stat64 statbuf; + int l; + vdev_label_t *label; + uint64_t state, txg, size; + + *config = NULL; + + if (fstat64(fd, &statbuf) == -1) + return (0); + size = P2ALIGN_TYPED(statbuf.st_size, sizeof (vdev_label_t), uint64_t); + + if ((label = malloc(sizeof (vdev_label_t))) == NULL) + return (-1); + + for (l = 0; l < VDEV_LABELS; l++) { + if (pread64(fd, label, sizeof (vdev_label_t), + label_offset(size, l)) != sizeof (vdev_label_t)) + continue; + + if (nvlist_unpack(label->vl_vdev_phys.vp_nvlist, + sizeof (label->vl_vdev_phys.vp_nvlist), config, 0) != 0) + continue; + + if (nvlist_lookup_uint64(*config, ZPOOL_CONFIG_POOL_STATE, + &state) != 0 || state > POOL_STATE_L2CACHE) { + nvlist_free(*config); + continue; + } + + if (state != POOL_STATE_SPARE && state != POOL_STATE_L2CACHE && + (nvlist_lookup_uint64(*config, ZPOOL_CONFIG_POOL_TXG, + &txg) != 0 || txg == 0)) { + nvlist_free(*config); + continue; + } + + free(label); + return (0); + } + + free(label); + *config = NULL; + return (0); +} + +typedef struct rdsk_node { + char *rn_name; + int rn_dfd; + libzfs_handle_t *rn_hdl; + nvlist_t *rn_config; + avl_tree_t *rn_avl; + avl_node_t rn_node; + boolean_t rn_nozpool; +} rdsk_node_t; + +static int +slice_cache_compare(const void *arg1, const void *arg2) +{ + const char *nm1 = ((rdsk_node_t *)arg1)->rn_name; + const char *nm2 = ((rdsk_node_t *)arg2)->rn_name; + char *nm1slice, *nm2slice; + int rv; + + /* + * slices zero and two are the most likely to provide results, + * so put those first + */ + nm1slice = strstr(nm1, "s0"); + nm2slice = strstr(nm2, "s0"); + if (nm1slice && !nm2slice) { + return (-1); + } + if (!nm1slice && nm2slice) { + return (1); + } + nm1slice = strstr(nm1, "s2"); + nm2slice = strstr(nm2, "s2"); + if (nm1slice && !nm2slice) { + return (-1); + } + if (!nm1slice && nm2slice) { + return (1); + } + + rv = strcmp(nm1, nm2); + if (rv == 0) + return (0); + return (rv > 0 ? 1 : -1); +} + +#ifdef sun +static void +check_one_slice(avl_tree_t *r, char *diskname, uint_t partno, + diskaddr_t size, uint_t blksz) +{ + rdsk_node_t tmpnode; + rdsk_node_t *node; + char sname[MAXNAMELEN]; + + tmpnode.rn_name = &sname[0]; + (void) snprintf(tmpnode.rn_name, MAXNAMELEN, "%s%u", + diskname, partno); + /* + * protect against division by zero for disk labels that + * contain a bogus sector size + */ + if (blksz == 0) + blksz = DEV_BSIZE; + /* too small to contain a zpool? */ + if ((size < (SPA_MINDEVSIZE / blksz)) && + (node = avl_find(r, &tmpnode, NULL))) + node->rn_nozpool = B_TRUE; +} +#endif /* sun */ + +static void +nozpool_all_slices(avl_tree_t *r, const char *sname) +{ +#ifdef sun + char diskname[MAXNAMELEN]; + char *ptr; + int i; + + (void) strncpy(diskname, sname, MAXNAMELEN); + if (((ptr = strrchr(diskname, 's')) == NULL) && + ((ptr = strrchr(diskname, 'p')) == NULL)) + return; + ptr[0] = 's'; + ptr[1] = '\0'; + for (i = 0; i < NDKMAP; i++) + check_one_slice(r, diskname, i, 0, 1); + ptr[0] = 'p'; + for (i = 0; i <= FD_NUMPART; i++) + check_one_slice(r, diskname, i, 0, 1); +#endif /* sun */ +} + +#ifdef sun +static void +check_slices(avl_tree_t *r, int fd, const char *sname) +{ + struct extvtoc vtoc; + struct dk_gpt *gpt; + char diskname[MAXNAMELEN]; + char *ptr; + int i; + + (void) strncpy(diskname, sname, MAXNAMELEN); + if ((ptr = strrchr(diskname, 's')) == NULL || !isdigit(ptr[1])) + return; + ptr[1] = '\0'; + + if (read_extvtoc(fd, &vtoc) >= 0) { + for (i = 0; i < NDKMAP; i++) + check_one_slice(r, diskname, i, + vtoc.v_part[i].p_size, vtoc.v_sectorsz); + } else if (efi_alloc_and_read(fd, &gpt) >= 0) { + /* + * on x86 we'll still have leftover links that point + * to slices s[9-15], so use NDKMAP instead + */ + for (i = 0; i < NDKMAP; i++) + check_one_slice(r, diskname, i, + gpt->efi_parts[i].p_size, gpt->efi_lbasize); + /* nodes p[1-4] are never used with EFI labels */ + ptr[0] = 'p'; + for (i = 1; i <= FD_NUMPART; i++) + check_one_slice(r, diskname, i, 0, 1); + efi_free(gpt); + } +} +#endif /* sun */ + +static void +zpool_open_func(void *arg) +{ + rdsk_node_t *rn = arg; + struct stat64 statbuf; + nvlist_t *config; + int fd; + + if (rn->rn_nozpool) + return; + if ((fd = openat64(rn->rn_dfd, rn->rn_name, O_RDONLY)) < 0) { + /* symlink to a device that's no longer there */ + if (errno == ENOENT) + nozpool_all_slices(rn->rn_avl, rn->rn_name); + return; + } + /* + * Ignore failed stats. We only want regular + * files, character devs and block devs. + */ + if (fstat64(fd, &statbuf) != 0 || + (!S_ISREG(statbuf.st_mode) && + !S_ISCHR(statbuf.st_mode) && + !S_ISBLK(statbuf.st_mode))) { + (void) close(fd); + return; + } + /* this file is too small to hold a zpool */ +#ifdef sun + if (S_ISREG(statbuf.st_mode) && + statbuf.st_size < SPA_MINDEVSIZE) { + (void) close(fd); + return; + } else if (!S_ISREG(statbuf.st_mode)) { + /* + * Try to read the disk label first so we don't have to + * open a bunch of minor nodes that can't have a zpool. + */ + check_slices(rn->rn_avl, fd, rn->rn_name); + } +#else /* !sun */ + if (statbuf.st_size < SPA_MINDEVSIZE) { + (void) close(fd); + return; + } +#endif /* sun */ + + if ((zpool_read_label(fd, &config)) != 0) { + (void) close(fd); + (void) no_memory(rn->rn_hdl); + return; + } + (void) close(fd); + + + rn->rn_config = config; + if (config != NULL) { + assert(rn->rn_nozpool == B_FALSE); + } +} + +/* + * Given a file descriptor, clear (zero) the label information. This function + * is used in the appliance stack as part of the ZFS sysevent module and + * to implement the "zpool labelclear" command. + */ +int +zpool_clear_label(int fd) +{ + struct stat64 statbuf; + int l; + vdev_label_t *label; + uint64_t size; + + if (fstat64(fd, &statbuf) == -1) + return (0); + size = P2ALIGN_TYPED(statbuf.st_size, sizeof (vdev_label_t), uint64_t); + + if ((label = calloc(sizeof (vdev_label_t), 1)) == NULL) + return (-1); + + for (l = 0; l < VDEV_LABELS; l++) { + if (pwrite64(fd, label, sizeof (vdev_label_t), + label_offset(size, l)) != sizeof (vdev_label_t)) + return (-1); + } + + free(label); + return (0); +} + +/* + * Given a list of directories to search, find all pools stored on disk. This + * includes partial pools which are not available to import. If no args are + * given (argc is 0), then the default directory (/dev/dsk) is searched. + * poolname or guid (but not both) are provided by the caller when trying + * to import a specific pool. + */ +static nvlist_t * +zpool_find_import_impl(libzfs_handle_t *hdl, importargs_t *iarg) +{ + int i, dirs = iarg->paths; + DIR *dirp = NULL; + struct dirent64 *dp; + char path[MAXPATHLEN]; + char *end, **dir = iarg->path; + size_t pathleft; + nvlist_t *ret = NULL; + static char *default_dir = "/dev"; + pool_list_t pools = { 0 }; + pool_entry_t *pe, *penext; + vdev_entry_t *ve, *venext; + config_entry_t *ce, *cenext; + name_entry_t *ne, *nenext; + avl_tree_t slice_cache; + rdsk_node_t *slice; + void *cookie; + + if (dirs == 0) { + dirs = 1; + dir = &default_dir; + } + + /* + * Go through and read the label configuration information from every + * possible device, organizing the information according to pool GUID + * and toplevel GUID. + */ + for (i = 0; i < dirs; i++) { + tpool_t *t; + char *rdsk; + int dfd; + + /* use realpath to normalize the path */ + if (realpath(dir[i], path) == 0) { + (void) zfs_error_fmt(hdl, EZFS_BADPATH, + dgettext(TEXT_DOMAIN, "cannot open '%s'"), dir[i]); + goto error; + } + end = &path[strlen(path)]; + *end++ = '/'; + *end = 0; + pathleft = &path[sizeof (path)] - end; + + /* + * Using raw devices instead of block devices when we're + * reading the labels skips a bunch of slow operations during + * close(2) processing, so we replace /dev/dsk with /dev/rdsk. + */ + if (strcmp(path, "/dev/dsk/") == 0) + rdsk = "/dev/"; + else + rdsk = path; + + if ((dfd = open64(rdsk, O_RDONLY)) < 0 || + (dirp = fdopendir(dfd)) == NULL) { + zfs_error_aux(hdl, strerror(errno)); + (void) zfs_error_fmt(hdl, EZFS_BADPATH, + dgettext(TEXT_DOMAIN, "cannot open '%s'"), + rdsk); + goto error; + } + + avl_create(&slice_cache, slice_cache_compare, + sizeof (rdsk_node_t), offsetof(rdsk_node_t, rn_node)); + + if (strcmp(rdsk, "/dev/") == 0) { + struct gmesh mesh; + struct gclass *mp; + struct ggeom *gp; + struct gprovider *pp; + + errno = geom_gettree(&mesh); + if (errno != 0) { + zfs_error_aux(hdl, strerror(errno)); + (void) zfs_error_fmt(hdl, EZFS_BADPATH, + dgettext(TEXT_DOMAIN, "cannot get GEOM tree")); + goto error; + } + + LIST_FOREACH(mp, &mesh.lg_class, lg_class) { + LIST_FOREACH(gp, &mp->lg_geom, lg_geom) { + LIST_FOREACH(pp, &gp->lg_provider, lg_provider) { + slice = zfs_alloc(hdl, sizeof (rdsk_node_t)); + slice->rn_name = zfs_strdup(hdl, pp->lg_name); + slice->rn_avl = &slice_cache; + slice->rn_dfd = dfd; + slice->rn_hdl = hdl; + slice->rn_nozpool = B_FALSE; + avl_add(&slice_cache, slice); + } + } + } + + geom_deletetree(&mesh); + goto skipdir; + } + + /* + * This is not MT-safe, but we have no MT consumers of libzfs + */ + while ((dp = readdir64(dirp)) != NULL) { + const char *name = dp->d_name; + if (name[0] == '.' && + (name[1] == 0 || (name[1] == '.' && name[2] == 0))) + continue; + + slice = zfs_alloc(hdl, sizeof (rdsk_node_t)); + slice->rn_name = zfs_strdup(hdl, name); + slice->rn_avl = &slice_cache; + slice->rn_dfd = dfd; + slice->rn_hdl = hdl; + slice->rn_nozpool = B_FALSE; + avl_add(&slice_cache, slice); + } +skipdir: + /* + * create a thread pool to do all of this in parallel; + * rn_nozpool is not protected, so this is racy in that + * multiple tasks could decide that the same slice can + * not hold a zpool, which is benign. Also choose + * double the number of processors; we hold a lot of + * locks in the kernel, so going beyond this doesn't + * buy us much. + */ + t = tpool_create(1, 2 * sysconf(_SC_NPROCESSORS_ONLN), + 0, NULL); + for (slice = avl_first(&slice_cache); slice; + (slice = avl_walk(&slice_cache, slice, + AVL_AFTER))) + (void) tpool_dispatch(t, zpool_open_func, slice); + tpool_wait(t); + tpool_destroy(t); + + cookie = NULL; + while ((slice = avl_destroy_nodes(&slice_cache, + &cookie)) != NULL) { + if (slice->rn_config != NULL) { + nvlist_t *config = slice->rn_config; + boolean_t matched = B_TRUE; + + if (iarg->poolname != NULL) { + char *pname; + + matched = nvlist_lookup_string(config, + ZPOOL_CONFIG_POOL_NAME, + &pname) == 0 && + strcmp(iarg->poolname, pname) == 0; + } else if (iarg->guid != 0) { + uint64_t this_guid; + + matched = nvlist_lookup_uint64(config, + ZPOOL_CONFIG_POOL_GUID, + &this_guid) == 0 && + iarg->guid == this_guid; + } + if (!matched) { + nvlist_free(config); + config = NULL; + continue; + } + /* use the non-raw path for the config */ + (void) strlcpy(end, slice->rn_name, pathleft); + if (add_config(hdl, &pools, path, config) != 0) + goto error; + } + free(slice->rn_name); + free(slice); + } + avl_destroy(&slice_cache); + + (void) closedir(dirp); + dirp = NULL; + } + + ret = get_configs(hdl, &pools, iarg->can_be_active); + +error: + for (pe = pools.pools; pe != NULL; pe = penext) { + penext = pe->pe_next; + for (ve = pe->pe_vdevs; ve != NULL; ve = venext) { + venext = ve->ve_next; + for (ce = ve->ve_configs; ce != NULL; ce = cenext) { + cenext = ce->ce_next; + if (ce->ce_config) + nvlist_free(ce->ce_config); + free(ce); + } + free(ve); + } + free(pe); + } + + for (ne = pools.names; ne != NULL; ne = nenext) { + nenext = ne->ne_next; + if (ne->ne_name) + free(ne->ne_name); + free(ne); + } + + if (dirp) + (void) closedir(dirp); + + return (ret); +} + +nvlist_t * +zpool_find_import(libzfs_handle_t *hdl, int argc, char **argv) +{ + importargs_t iarg = { 0 }; + + iarg.paths = argc; + iarg.path = argv; + + return (zpool_find_import_impl(hdl, &iarg)); +} + +/* + * Given a cache file, return the contents as a list of importable pools. + * poolname or guid (but not both) are provided by the caller when trying + * to import a specific pool. + */ +nvlist_t * +zpool_find_import_cached(libzfs_handle_t *hdl, const char *cachefile, + char *poolname, uint64_t guid) +{ + char *buf; + int fd; + struct stat64 statbuf; + nvlist_t *raw, *src, *dst; + nvlist_t *pools; + nvpair_t *elem; + char *name; + uint64_t this_guid; + boolean_t active; + + verify(poolname == NULL || guid == 0); + + if ((fd = open(cachefile, O_RDONLY)) < 0) { + zfs_error_aux(hdl, "%s", strerror(errno)); + (void) zfs_error(hdl, EZFS_BADCACHE, + dgettext(TEXT_DOMAIN, "failed to open cache file")); + return (NULL); + } + + if (fstat64(fd, &statbuf) != 0) { + zfs_error_aux(hdl, "%s", strerror(errno)); + (void) close(fd); + (void) zfs_error(hdl, EZFS_BADCACHE, + dgettext(TEXT_DOMAIN, "failed to get size of cache file")); + return (NULL); + } + + if ((buf = zfs_alloc(hdl, statbuf.st_size)) == NULL) { + (void) close(fd); + return (NULL); + } + + if (read(fd, buf, statbuf.st_size) != statbuf.st_size) { + (void) close(fd); + free(buf); + (void) zfs_error(hdl, EZFS_BADCACHE, + dgettext(TEXT_DOMAIN, + "failed to read cache file contents")); + return (NULL); + } + + (void) close(fd); + + if (nvlist_unpack(buf, statbuf.st_size, &raw, 0) != 0) { + free(buf); + (void) zfs_error(hdl, EZFS_BADCACHE, + dgettext(TEXT_DOMAIN, + "invalid or corrupt cache file contents")); + return (NULL); + } + + free(buf); + + /* + * Go through and get the current state of the pools and refresh their + * state. + */ + if (nvlist_alloc(&pools, 0, 0) != 0) { + (void) no_memory(hdl); + nvlist_free(raw); + return (NULL); + } + + elem = NULL; + while ((elem = nvlist_next_nvpair(raw, elem)) != NULL) { + verify(nvpair_value_nvlist(elem, &src) == 0); + + verify(nvlist_lookup_string(src, ZPOOL_CONFIG_POOL_NAME, + &name) == 0); + if (poolname != NULL && strcmp(poolname, name) != 0) + continue; + + verify(nvlist_lookup_uint64(src, ZPOOL_CONFIG_POOL_GUID, + &this_guid) == 0); + if (guid != 0) { + verify(nvlist_lookup_uint64(src, ZPOOL_CONFIG_POOL_GUID, + &this_guid) == 0); + if (guid != this_guid) + continue; + } + + if (pool_active(hdl, name, this_guid, &active) != 0) { + nvlist_free(raw); + nvlist_free(pools); + return (NULL); + } + + if (active) + continue; + + if ((dst = refresh_config(hdl, src)) == NULL) { + nvlist_free(raw); + nvlist_free(pools); + return (NULL); + } + + if (nvlist_add_nvlist(pools, nvpair_name(elem), dst) != 0) { + (void) no_memory(hdl); + nvlist_free(dst); + nvlist_free(raw); + nvlist_free(pools); + return (NULL); + } + nvlist_free(dst); + } + + nvlist_free(raw); + return (pools); +} + +static int +name_or_guid_exists(zpool_handle_t *zhp, void *data) +{ + importargs_t *import = data; + int found = 0; + + if (import->poolname != NULL) { + char *pool_name; + + verify(nvlist_lookup_string(zhp->zpool_config, + ZPOOL_CONFIG_POOL_NAME, &pool_name) == 0); + if (strcmp(pool_name, import->poolname) == 0) + found = 1; + } else { + uint64_t pool_guid; + + verify(nvlist_lookup_uint64(zhp->zpool_config, + ZPOOL_CONFIG_POOL_GUID, &pool_guid) == 0); + if (pool_guid == import->guid) + found = 1; + } + + zpool_close(zhp); + return (found); +} + +nvlist_t * +zpool_search_import(libzfs_handle_t *hdl, importargs_t *import) +{ + verify(import->poolname == NULL || import->guid == 0); + + if (import->unique) + import->exists = zpool_iter(hdl, name_or_guid_exists, import); + + if (import->cachefile != NULL) + return (zpool_find_import_cached(hdl, import->cachefile, + import->poolname, import->guid)); + + return (zpool_find_import_impl(hdl, import)); +} + +boolean_t +find_guid(nvlist_t *nv, uint64_t guid) +{ + uint64_t tmp; + nvlist_t **child; + uint_t c, children; + + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, &tmp) == 0); + if (tmp == guid) + return (B_TRUE); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) == 0) { + for (c = 0; c < children; c++) + if (find_guid(child[c], guid)) + return (B_TRUE); + } + + return (B_FALSE); +} + +typedef struct aux_cbdata { + const char *cb_type; + uint64_t cb_guid; + zpool_handle_t *cb_zhp; +} aux_cbdata_t; + +static int +find_aux(zpool_handle_t *zhp, void *data) +{ + aux_cbdata_t *cbp = data; + nvlist_t **list; + uint_t i, count; + uint64_t guid; + nvlist_t *nvroot; + + verify(nvlist_lookup_nvlist(zhp->zpool_config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + + if (nvlist_lookup_nvlist_array(nvroot, cbp->cb_type, + &list, &count) == 0) { + for (i = 0; i < count; i++) { + verify(nvlist_lookup_uint64(list[i], + ZPOOL_CONFIG_GUID, &guid) == 0); + if (guid == cbp->cb_guid) { + cbp->cb_zhp = zhp; + return (1); + } + } + } + + zpool_close(zhp); + return (0); +} + +/* + * Determines if the pool is in use. If so, it returns true and the state of + * the pool as well as the name of the pool. Both strings are allocated and + * must be freed by the caller. + */ +int +zpool_in_use(libzfs_handle_t *hdl, int fd, pool_state_t *state, char **namestr, + boolean_t *inuse) +{ + nvlist_t *config; + char *name; + boolean_t ret; + uint64_t guid, vdev_guid; + zpool_handle_t *zhp; + nvlist_t *pool_config; + uint64_t stateval, isspare; + aux_cbdata_t cb = { 0 }; + boolean_t isactive; + + *inuse = B_FALSE; + + if (zpool_read_label(fd, &config) != 0) { + (void) no_memory(hdl); + return (-1); + } + + if (config == NULL) + return (0); + + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE, + &stateval) == 0); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_GUID, + &vdev_guid) == 0); + + if (stateval != POOL_STATE_SPARE && stateval != POOL_STATE_L2CACHE) { + verify(nvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME, + &name) == 0); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, + &guid) == 0); + } + + switch (stateval) { + case POOL_STATE_EXPORTED: + /* + * A pool with an exported state may in fact be imported + * read-only, so check the in-core state to see if it's + * active and imported read-only. If it is, set + * its state to active. + */ + if (pool_active(hdl, name, guid, &isactive) == 0 && isactive && + (zhp = zpool_open_canfail(hdl, name)) != NULL) { + if (zpool_get_prop_int(zhp, ZPOOL_PROP_READONLY, NULL)) + stateval = POOL_STATE_ACTIVE; + + /* + * All we needed the zpool handle for is the + * readonly prop check. + */ + zpool_close(zhp); + } + + ret = B_TRUE; + break; + + case POOL_STATE_ACTIVE: + /* + * For an active pool, we have to determine if it's really part + * of a currently active pool (in which case the pool will exist + * and the guid will be the same), or whether it's part of an + * active pool that was disconnected without being explicitly + * exported. + */ + if (pool_active(hdl, name, guid, &isactive) != 0) { + nvlist_free(config); + return (-1); + } + + if (isactive) { + /* + * Because the device may have been removed while + * offlined, we only report it as active if the vdev is + * still present in the config. Otherwise, pretend like + * it's not in use. + */ + if ((zhp = zpool_open_canfail(hdl, name)) != NULL && + (pool_config = zpool_get_config(zhp, NULL)) + != NULL) { + nvlist_t *nvroot; + + verify(nvlist_lookup_nvlist(pool_config, + ZPOOL_CONFIG_VDEV_TREE, &nvroot) == 0); + ret = find_guid(nvroot, vdev_guid); + } else { + ret = B_FALSE; + } + + /* + * If this is an active spare within another pool, we + * treat it like an unused hot spare. This allows the + * user to create a pool with a hot spare that currently + * in use within another pool. Since we return B_TRUE, + * libdiskmgt will continue to prevent generic consumers + * from using the device. + */ + if (ret && nvlist_lookup_uint64(config, + ZPOOL_CONFIG_IS_SPARE, &isspare) == 0 && isspare) + stateval = POOL_STATE_SPARE; + + if (zhp != NULL) + zpool_close(zhp); + } else { + stateval = POOL_STATE_POTENTIALLY_ACTIVE; + ret = B_TRUE; + } + break; + + case POOL_STATE_SPARE: + /* + * For a hot spare, it can be either definitively in use, or + * potentially active. To determine if it's in use, we iterate + * over all pools in the system and search for one with a spare + * with a matching guid. + * + * Due to the shared nature of spares, we don't actually report + * the potentially active case as in use. This means the user + * can freely create pools on the hot spares of exported pools, + * but to do otherwise makes the resulting code complicated, and + * we end up having to deal with this case anyway. + */ + cb.cb_zhp = NULL; + cb.cb_guid = vdev_guid; + cb.cb_type = ZPOOL_CONFIG_SPARES; + if (zpool_iter(hdl, find_aux, &cb) == 1) { + name = (char *)zpool_get_name(cb.cb_zhp); + ret = TRUE; + } else { + ret = FALSE; + } + break; + + case POOL_STATE_L2CACHE: + + /* + * Check if any pool is currently using this l2cache device. + */ + cb.cb_zhp = NULL; + cb.cb_guid = vdev_guid; + cb.cb_type = ZPOOL_CONFIG_L2CACHE; + if (zpool_iter(hdl, find_aux, &cb) == 1) { + name = (char *)zpool_get_name(cb.cb_zhp); + ret = TRUE; + } else { + ret = FALSE; + } + break; + + default: + ret = B_FALSE; + } + + + if (ret) { + if ((*namestr = zfs_strdup(hdl, name)) == NULL) { + if (cb.cb_zhp) + zpool_close(cb.cb_zhp); + nvlist_free(config); + return (-1); + } + *state = (pool_state_t)stateval; + } + + if (cb.cb_zhp) + zpool_close(cb.cb_zhp); + + nvlist_free(config); + *inuse = ret; + return (0); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_iter.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_iter.c new file mode 100644 index 0000000..9698a72 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_iter.c @@ -0,0 +1,529 @@ +/* + * 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) 2013 by Delphix. All rights reserved. + * Copyright (c) 2012 Pawel Jakub Dawidek <pawel@dawidek.net>. + * All rights reserved. + * Copyright 2013 Nexenta Systems, Inc. All rights reserved. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> +#include <unistd.h> +#include <stddef.h> +#include <libintl.h> +#include <libzfs.h> + +#include "libzfs_impl.h" + +int +zfs_iter_clones(zfs_handle_t *zhp, zfs_iter_f func, void *data) +{ + nvlist_t *nvl = zfs_get_clones_nvl(zhp); + nvpair_t *pair; + + if (nvl == NULL) + return (0); + + for (pair = nvlist_next_nvpair(nvl, NULL); pair != NULL; + pair = nvlist_next_nvpair(nvl, pair)) { + zfs_handle_t *clone = zfs_open(zhp->zfs_hdl, nvpair_name(pair), + ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME); + if (clone != NULL) { + int err = func(clone, data); + if (err != 0) + return (err); + } + } + return (0); +} + +static int +zfs_do_list_ioctl(zfs_handle_t *zhp, unsigned long arg, zfs_cmd_t *zc) +{ + int rc; + uint64_t orig_cookie; + + orig_cookie = zc->zc_cookie; +top: + (void) strlcpy(zc->zc_name, zhp->zfs_name, sizeof (zc->zc_name)); + rc = ioctl(zhp->zfs_hdl->libzfs_fd, arg, zc); + + if (rc == -1) { + switch (errno) { + case ENOMEM: + /* expand nvlist memory and try again */ + if (zcmd_expand_dst_nvlist(zhp->zfs_hdl, zc) != 0) { + zcmd_free_nvlists(zc); + return (-1); + } + zc->zc_cookie = orig_cookie; + goto top; + /* + * An errno value of ESRCH indicates normal completion. + * If ENOENT is returned, then the underlying dataset + * has been removed since we obtained the handle. + */ + case ESRCH: + case ENOENT: + rc = 1; + break; + default: + rc = zfs_standard_error(zhp->zfs_hdl, errno, + dgettext(TEXT_DOMAIN, + "cannot iterate filesystems")); + break; + } + } + return (rc); +} + +/* + * Iterate over all child filesystems + */ +int +zfs_iter_filesystems(zfs_handle_t *zhp, zfs_iter_f func, void *data) +{ + zfs_cmd_t zc = { 0 }; + zfs_handle_t *nzhp; + int ret; + + if (zhp->zfs_type != ZFS_TYPE_FILESYSTEM) + return (0); + + if (zcmd_alloc_dst_nvlist(zhp->zfs_hdl, &zc, 0) != 0) + return (-1); + + while ((ret = zfs_do_list_ioctl(zhp, ZFS_IOC_DATASET_LIST_NEXT, + &zc)) == 0) { + /* + * Silently ignore errors, as the only plausible explanation is + * that the pool has since been removed. + */ + if ((nzhp = make_dataset_handle_zc(zhp->zfs_hdl, + &zc)) == NULL) { + continue; + } + + if ((ret = func(nzhp, data)) != 0) { + zcmd_free_nvlists(&zc); + return (ret); + } + } + zcmd_free_nvlists(&zc); + return ((ret < 0) ? ret : 0); +} + +/* + * Iterate over all snapshots + */ +int +zfs_iter_snapshots(zfs_handle_t *zhp, boolean_t simple, zfs_iter_f func, + void *data) +{ + zfs_cmd_t zc = { 0 }; + zfs_handle_t *nzhp; + int ret; + + if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT || + zhp->zfs_type == ZFS_TYPE_BOOKMARK) + return (0); + + zc.zc_simple = simple; + + if (zcmd_alloc_dst_nvlist(zhp->zfs_hdl, &zc, 0) != 0) + return (-1); + while ((ret = zfs_do_list_ioctl(zhp, ZFS_IOC_SNAPSHOT_LIST_NEXT, + &zc)) == 0) { + + if (simple) + nzhp = make_dataset_simple_handle_zc(zhp, &zc); + else + nzhp = make_dataset_handle_zc(zhp->zfs_hdl, &zc); + if (nzhp == NULL) + continue; + + if ((ret = func(nzhp, data)) != 0) { + zcmd_free_nvlists(&zc); + return (ret); + } + } + zcmd_free_nvlists(&zc); + return ((ret < 0) ? ret : 0); +} + +/* + * Iterate over all bookmarks + */ +int +zfs_iter_bookmarks(zfs_handle_t *zhp, zfs_iter_f func, void *data) +{ + zfs_handle_t *nzhp; + nvlist_t *props = NULL; + nvlist_t *bmarks = NULL; + int err; + + if ((zfs_get_type(zhp) & (ZFS_TYPE_SNAPSHOT | ZFS_TYPE_BOOKMARK)) != 0) + return (0); + + /* Setup the requested properties nvlist. */ + props = fnvlist_alloc(); + fnvlist_add_boolean(props, zfs_prop_to_name(ZFS_PROP_GUID)); + fnvlist_add_boolean(props, zfs_prop_to_name(ZFS_PROP_CREATETXG)); + fnvlist_add_boolean(props, zfs_prop_to_name(ZFS_PROP_CREATION)); + + /* Allocate an nvlist to hold the bookmarks. */ + bmarks = fnvlist_alloc(); + + if ((err = lzc_get_bookmarks(zhp->zfs_name, props, &bmarks)) != 0) + goto out; + + for (nvpair_t *pair = nvlist_next_nvpair(bmarks, NULL); + pair != NULL; pair = nvlist_next_nvpair(bmarks, pair)) { + char name[ZFS_MAXNAMELEN]; + char *bmark_name; + nvlist_t *bmark_props; + + bmark_name = nvpair_name(pair); + bmark_props = fnvpair_value_nvlist(pair); + + (void) snprintf(name, sizeof (name), "%s#%s", zhp->zfs_name, + bmark_name); + + nzhp = make_bookmark_handle(zhp, name, bmark_props); + if (nzhp == NULL) + continue; + + if ((err = func(nzhp, data)) != 0) + goto out; + } + +out: + fnvlist_free(props); + fnvlist_free(bmarks); + + return (err); +} + +/* + * Routines for dealing with the sorted snapshot functionality + */ +typedef struct zfs_node { + zfs_handle_t *zn_handle; + avl_node_t zn_avlnode; +} zfs_node_t; + +static int +zfs_sort_snaps(zfs_handle_t *zhp, void *data) +{ + avl_tree_t *avl = data; + zfs_node_t *node; + zfs_node_t search; + + search.zn_handle = zhp; + node = avl_find(avl, &search, NULL); + if (node) { + /* + * If this snapshot was renamed while we were creating the + * AVL tree, it's possible that we already inserted it under + * its old name. Remove the old handle before adding the new + * one. + */ + zfs_close(node->zn_handle); + avl_remove(avl, node); + free(node); + } + + node = zfs_alloc(zhp->zfs_hdl, sizeof (zfs_node_t)); + node->zn_handle = zhp; + avl_add(avl, node); + + return (0); +} + +static int +zfs_snapshot_compare(const void *larg, const void *rarg) +{ + zfs_handle_t *l = ((zfs_node_t *)larg)->zn_handle; + zfs_handle_t *r = ((zfs_node_t *)rarg)->zn_handle; + uint64_t lcreate, rcreate; + + /* + * Sort them according to creation time. We use the hidden + * CREATETXG property to get an absolute ordering of snapshots. + */ + lcreate = zfs_prop_get_int(l, ZFS_PROP_CREATETXG); + rcreate = zfs_prop_get_int(r, ZFS_PROP_CREATETXG); + + if (lcreate < rcreate) + return (-1); + else if (lcreate > rcreate) + return (+1); + else + return (0); +} + +int +zfs_iter_snapshots_sorted(zfs_handle_t *zhp, zfs_iter_f callback, void *data) +{ + int ret = 0; + zfs_node_t *node; + avl_tree_t avl; + void *cookie = NULL; + + avl_create(&avl, zfs_snapshot_compare, + sizeof (zfs_node_t), offsetof(zfs_node_t, zn_avlnode)); + + ret = zfs_iter_snapshots(zhp, B_FALSE, zfs_sort_snaps, &avl); + + for (node = avl_first(&avl); node != NULL; node = AVL_NEXT(&avl, node)) + ret |= callback(node->zn_handle, data); + + while ((node = avl_destroy_nodes(&avl, &cookie)) != NULL) + free(node); + + avl_destroy(&avl); + + return (ret); +} + +typedef struct { + char *ssa_first; + char *ssa_last; + boolean_t ssa_seenfirst; + boolean_t ssa_seenlast; + zfs_iter_f ssa_func; + void *ssa_arg; +} snapspec_arg_t; + +static int +snapspec_cb(zfs_handle_t *zhp, void *arg) { + snapspec_arg_t *ssa = arg; + char *shortsnapname; + int err = 0; + + if (ssa->ssa_seenlast) + return (0); + shortsnapname = zfs_strdup(zhp->zfs_hdl, + strchr(zfs_get_name(zhp), '@') + 1); + + if (!ssa->ssa_seenfirst && strcmp(shortsnapname, ssa->ssa_first) == 0) + ssa->ssa_seenfirst = B_TRUE; + + if (ssa->ssa_seenfirst) { + err = ssa->ssa_func(zhp, ssa->ssa_arg); + } else { + zfs_close(zhp); + } + + if (strcmp(shortsnapname, ssa->ssa_last) == 0) + ssa->ssa_seenlast = B_TRUE; + free(shortsnapname); + + return (err); +} + +/* + * spec is a string like "A,B%C,D" + * + * <snaps>, where <snaps> can be: + * <snap> (single snapshot) + * <snap>%<snap> (range of snapshots, inclusive) + * %<snap> (range of snapshots, starting with earliest) + * <snap>% (range of snapshots, ending with last) + * % (all snapshots) + * <snaps>[,...] (comma separated list of the above) + * + * If a snapshot can not be opened, continue trying to open the others, but + * return ENOENT at the end. + */ +int +zfs_iter_snapspec(zfs_handle_t *fs_zhp, const char *spec_orig, + zfs_iter_f func, void *arg) +{ + char *buf, *comma_separated, *cp; + int err = 0; + int ret = 0; + + buf = zfs_strdup(fs_zhp->zfs_hdl, spec_orig); + cp = buf; + + while ((comma_separated = strsep(&cp, ",")) != NULL) { + char *pct = strchr(comma_separated, '%'); + if (pct != NULL) { + snapspec_arg_t ssa = { 0 }; + ssa.ssa_func = func; + ssa.ssa_arg = arg; + + if (pct == comma_separated) + ssa.ssa_seenfirst = B_TRUE; + else + ssa.ssa_first = comma_separated; + *pct = '\0'; + ssa.ssa_last = pct + 1; + + /* + * If there is a lastname specified, make sure it + * exists. + */ + if (ssa.ssa_last[0] != '\0') { + char snapname[ZFS_MAXNAMELEN]; + (void) snprintf(snapname, sizeof (snapname), + "%s@%s", zfs_get_name(fs_zhp), + ssa.ssa_last); + if (!zfs_dataset_exists(fs_zhp->zfs_hdl, + snapname, ZFS_TYPE_SNAPSHOT)) { + ret = ENOENT; + continue; + } + } + + err = zfs_iter_snapshots_sorted(fs_zhp, + snapspec_cb, &ssa); + if (ret == 0) + ret = err; + if (ret == 0 && (!ssa.ssa_seenfirst || + (ssa.ssa_last[0] != '\0' && !ssa.ssa_seenlast))) { + ret = ENOENT; + } + } else { + char snapname[ZFS_MAXNAMELEN]; + zfs_handle_t *snap_zhp; + (void) snprintf(snapname, sizeof (snapname), "%s@%s", + zfs_get_name(fs_zhp), comma_separated); + snap_zhp = make_dataset_handle(fs_zhp->zfs_hdl, + snapname); + if (snap_zhp == NULL) { + ret = ENOENT; + continue; + } + err = func(snap_zhp, arg); + if (ret == 0) + ret = err; + } + } + + free(buf); + return (ret); +} + +/* + * Iterate over all children, snapshots and filesystems + */ +int +zfs_iter_children(zfs_handle_t *zhp, zfs_iter_f func, void *data) +{ + int ret; + + if ((ret = zfs_iter_filesystems(zhp, func, data)) != 0) + return (ret); + + return (zfs_iter_snapshots(zhp, B_FALSE, func, data)); +} + + +typedef struct iter_stack_frame { + struct iter_stack_frame *next; + zfs_handle_t *zhp; +} iter_stack_frame_t; + +typedef struct iter_dependents_arg { + boolean_t first; + boolean_t allowrecursion; + iter_stack_frame_t *stack; + zfs_iter_f func; + void *data; +} iter_dependents_arg_t; + +static int +iter_dependents_cb(zfs_handle_t *zhp, void *arg) +{ + iter_dependents_arg_t *ida = arg; + int err = 0; + boolean_t first = ida->first; + ida->first = B_FALSE; + + if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) { + err = zfs_iter_clones(zhp, iter_dependents_cb, ida); + } else if (zhp->zfs_type != ZFS_TYPE_BOOKMARK) { + iter_stack_frame_t isf; + iter_stack_frame_t *f; + + /* + * check if there is a cycle by seeing if this fs is already + * on the stack. + */ + for (f = ida->stack; f != NULL; f = f->next) { + if (f->zhp->zfs_dmustats.dds_guid == + zhp->zfs_dmustats.dds_guid) { + if (ida->allowrecursion) { + zfs_close(zhp); + return (0); + } else { + zfs_error_aux(zhp->zfs_hdl, + dgettext(TEXT_DOMAIN, + "recursive dependency at '%s'"), + zfs_get_name(zhp)); + err = zfs_error(zhp->zfs_hdl, + EZFS_RECURSIVE, + dgettext(TEXT_DOMAIN, + "cannot determine dependent " + "datasets")); + zfs_close(zhp); + return (err); + } + } + } + + isf.zhp = zhp; + isf.next = ida->stack; + ida->stack = &isf; + err = zfs_iter_filesystems(zhp, iter_dependents_cb, ida); + if (err == 0) { + err = zfs_iter_snapshots(zhp, B_FALSE, + iter_dependents_cb, ida); + } + ida->stack = isf.next; + } + + if (!first && err == 0) + err = ida->func(zhp, ida->data); + else + zfs_close(zhp); + + return (err); +} + +int +zfs_iter_dependents(zfs_handle_t *zhp, boolean_t allowrecursion, + zfs_iter_f func, void *data) +{ + iter_dependents_arg_t ida; + ida.allowrecursion = allowrecursion; + ida.stack = NULL; + ida.func = func; + ida.data = data; + ida.first = B_TRUE; + return (iter_dependents_cb(zfs_handle_dup(zhp), &ida)); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_mount.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_mount.c new file mode 100644 index 0000000..b2959dd --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_mount.c @@ -0,0 +1,1323 @@ +/* + * 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. + */ + +/* + * Routines to manage ZFS mounts. We separate all the nasty routines that have + * to deal with the OS. The following functions are the main entry points -- + * they are used by mount and unmount and when changing a filesystem's + * mountpoint. + * + * zfs_is_mounted() + * zfs_mount() + * zfs_unmount() + * zfs_unmountall() + * + * This file also contains the functions used to manage sharing filesystems via + * NFS and iSCSI: + * + * zfs_is_shared() + * zfs_share() + * zfs_unshare() + * + * zfs_is_shared_nfs() + * zfs_is_shared_smb() + * zfs_share_proto() + * zfs_shareall(); + * zfs_unshare_nfs() + * zfs_unshare_smb() + * zfs_unshareall_nfs() + * zfs_unshareall_smb() + * zfs_unshareall() + * zfs_unshareall_bypath() + * + * The following functions are available for pool consumers, and will + * mount/unmount and share/unshare all datasets within pool: + * + * zpool_enable_datasets() + * zpool_disable_datasets() + */ + +#include <dirent.h> +#include <dlfcn.h> +#include <errno.h> +#include <libgen.h> +#include <libintl.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> +#include <unistd.h> +#include <zone.h> +#include <sys/mntent.h> +#include <sys/mount.h> +#include <sys/stat.h> + +#include <libzfs.h> + +#include "libzfs_impl.h" + +#include <libshare.h> +#define MAXISALEN 257 /* based on sysinfo(2) man page */ + +static int zfs_share_proto(zfs_handle_t *, zfs_share_proto_t *); +zfs_share_type_t zfs_is_shared_proto(zfs_handle_t *, char **, + zfs_share_proto_t); + +/* + * The share protocols table must be in the same order as the zfs_share_prot_t + * enum in libzfs_impl.h + */ +typedef struct { + zfs_prop_t p_prop; + char *p_name; + int p_share_err; + int p_unshare_err; +} proto_table_t; + +proto_table_t proto_table[PROTO_END] = { + {ZFS_PROP_SHARENFS, "nfs", EZFS_SHARENFSFAILED, EZFS_UNSHARENFSFAILED}, + {ZFS_PROP_SHARESMB, "smb", EZFS_SHARESMBFAILED, EZFS_UNSHARESMBFAILED}, +}; + +zfs_share_proto_t nfs_only[] = { + PROTO_NFS, + PROTO_END +}; + +zfs_share_proto_t smb_only[] = { + PROTO_SMB, + PROTO_END +}; +zfs_share_proto_t share_all_proto[] = { + PROTO_NFS, + PROTO_SMB, + PROTO_END +}; + +/* + * Search the sharetab for the given mountpoint and protocol, returning + * a zfs_share_type_t value. + */ +static zfs_share_type_t +is_shared(libzfs_handle_t *hdl, const char *mountpoint, zfs_share_proto_t proto) +{ + char buf[MAXPATHLEN], *tab; + char *ptr; + + if (hdl->libzfs_sharetab == NULL) + return (SHARED_NOT_SHARED); + + (void) fseek(hdl->libzfs_sharetab, 0, SEEK_SET); + + while (fgets(buf, sizeof (buf), hdl->libzfs_sharetab) != NULL) { + + /* the mountpoint is the first entry on each line */ + if ((tab = strchr(buf, '\t')) == NULL) + continue; + + *tab = '\0'; + if (strcmp(buf, mountpoint) == 0) { +#ifdef sun + /* + * the protocol field is the third field + * skip over second field + */ + ptr = ++tab; + if ((tab = strchr(ptr, '\t')) == NULL) + continue; + ptr = ++tab; + if ((tab = strchr(ptr, '\t')) == NULL) + continue; + *tab = '\0'; + if (strcmp(ptr, + proto_table[proto].p_name) == 0) { + switch (proto) { + case PROTO_NFS: + return (SHARED_NFS); + case PROTO_SMB: + return (SHARED_SMB); + default: + return (0); + } + } +#else + if (proto == PROTO_NFS) + return (SHARED_NFS); +#endif + } + } + + return (SHARED_NOT_SHARED); +} + +#ifdef sun +/* + * Returns true if the specified directory is empty. If we can't open the + * directory at all, return true so that the mount can fail with a more + * informative error message. + */ +static boolean_t +dir_is_empty(const char *dirname) +{ + DIR *dirp; + struct dirent64 *dp; + + if ((dirp = opendir(dirname)) == NULL) + return (B_TRUE); + + while ((dp = readdir64(dirp)) != NULL) { + + if (strcmp(dp->d_name, ".") == 0 || + strcmp(dp->d_name, "..") == 0) + continue; + + (void) closedir(dirp); + return (B_FALSE); + } + + (void) closedir(dirp); + return (B_TRUE); +} +#endif + +/* + * Checks to see if the mount is active. If the filesystem is mounted, we fill + * in 'where' with the current mountpoint, and return 1. Otherwise, we return + * 0. + */ +boolean_t +is_mounted(libzfs_handle_t *zfs_hdl, const char *special, char **where) +{ + struct mnttab entry; + + if (libzfs_mnttab_find(zfs_hdl, special, &entry) != 0) + return (B_FALSE); + + if (where != NULL) + *where = zfs_strdup(zfs_hdl, entry.mnt_mountp); + + return (B_TRUE); +} + +boolean_t +zfs_is_mounted(zfs_handle_t *zhp, char **where) +{ + return (is_mounted(zhp->zfs_hdl, zfs_get_name(zhp), where)); +} + +/* + * Returns true if the given dataset is mountable, false otherwise. Returns the + * mountpoint in 'buf'. + */ +static boolean_t +zfs_is_mountable(zfs_handle_t *zhp, char *buf, size_t buflen, + zprop_source_t *source) +{ + char sourceloc[ZFS_MAXNAMELEN]; + zprop_source_t sourcetype; + + if (!zfs_prop_valid_for_type(ZFS_PROP_MOUNTPOINT, zhp->zfs_type)) + return (B_FALSE); + + verify(zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, buf, buflen, + &sourcetype, sourceloc, sizeof (sourceloc), B_FALSE) == 0); + + if (strcmp(buf, ZFS_MOUNTPOINT_NONE) == 0 || + strcmp(buf, ZFS_MOUNTPOINT_LEGACY) == 0) + return (B_FALSE); + + if (zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT) == ZFS_CANMOUNT_OFF) + return (B_FALSE); + + if (zfs_prop_get_int(zhp, ZFS_PROP_ZONED) && + getzoneid() == GLOBAL_ZONEID) + return (B_FALSE); + + if (source) + *source = sourcetype; + + return (B_TRUE); +} + +/* + * Mount the given filesystem. + */ +int +zfs_mount(zfs_handle_t *zhp, const char *options, int flags) +{ + struct stat buf; + char mountpoint[ZFS_MAXPROPLEN]; + char mntopts[MNT_LINE_MAX]; + libzfs_handle_t *hdl = zhp->zfs_hdl; + + if (options == NULL) + mntopts[0] = '\0'; + else + (void) strlcpy(mntopts, options, sizeof (mntopts)); + + /* + * If the pool is imported read-only then all mounts must be read-only + */ + if (zpool_get_prop_int(zhp->zpool_hdl, ZPOOL_PROP_READONLY, NULL)) + flags |= MS_RDONLY; + + if (!zfs_is_mountable(zhp, mountpoint, sizeof (mountpoint), NULL)) + return (0); + + /* Create the directory if it doesn't already exist */ + if (lstat(mountpoint, &buf) != 0) { + if (mkdirp(mountpoint, 0755) != 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "failed to create mountpoint")); + return (zfs_error_fmt(hdl, EZFS_MOUNTFAILED, + dgettext(TEXT_DOMAIN, "cannot mount '%s'"), + mountpoint)); + } + } + +#ifdef sun /* FreeBSD: overlay mounts are not checked. */ + /* + * Determine if the mountpoint is empty. If so, refuse to perform the + * mount. We don't perform this check if MS_OVERLAY is specified, which + * would defeat the point. We also avoid this check if 'remount' is + * specified. + */ + if ((flags & MS_OVERLAY) == 0 && + strstr(mntopts, MNTOPT_REMOUNT) == NULL && + !dir_is_empty(mountpoint)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "directory is not empty")); + return (zfs_error_fmt(hdl, EZFS_MOUNTFAILED, + dgettext(TEXT_DOMAIN, "cannot mount '%s'"), mountpoint)); + } +#endif + + /* perform the mount */ + if (zmount(zfs_get_name(zhp), mountpoint, flags, + MNTTYPE_ZFS, NULL, 0, mntopts, sizeof (mntopts)) != 0) { + /* + * Generic errors are nasty, but there are just way too many + * from mount(), and they're well-understood. We pick a few + * common ones to improve upon. + */ + if (errno == EBUSY) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "mountpoint or dataset is busy")); + } else if (errno == EPERM) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "Insufficient privileges")); + } else if (errno == ENOTSUP) { + char buf[256]; + int spa_version; + + VERIFY(zfs_spa_version(zhp, &spa_version) == 0); + (void) snprintf(buf, sizeof (buf), + dgettext(TEXT_DOMAIN, "Can't mount a version %lld " + "file system on a version %d pool. Pool must be" + " upgraded to mount this file system."), + (u_longlong_t)zfs_prop_get_int(zhp, + ZFS_PROP_VERSION), spa_version); + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, buf)); + } else { + zfs_error_aux(hdl, strerror(errno)); + } + return (zfs_error_fmt(hdl, EZFS_MOUNTFAILED, + dgettext(TEXT_DOMAIN, "cannot mount '%s'"), + zhp->zfs_name)); + } + + /* add the mounted entry into our cache */ + libzfs_mnttab_add(hdl, zfs_get_name(zhp), mountpoint, + mntopts); + return (0); +} + +/* + * Unmount a single filesystem. + */ +static int +unmount_one(libzfs_handle_t *hdl, const char *mountpoint, int flags) +{ + if (umount2(mountpoint, flags) != 0) { + zfs_error_aux(hdl, strerror(errno)); + return (zfs_error_fmt(hdl, EZFS_UMOUNTFAILED, + dgettext(TEXT_DOMAIN, "cannot unmount '%s'"), + mountpoint)); + } + + return (0); +} + +/* + * Unmount the given filesystem. + */ +int +zfs_unmount(zfs_handle_t *zhp, const char *mountpoint, int flags) +{ + libzfs_handle_t *hdl = zhp->zfs_hdl; + struct mnttab entry; + char *mntpt = NULL; + + /* check to see if we need to unmount the filesystem */ + if (mountpoint != NULL || ((zfs_get_type(zhp) == ZFS_TYPE_FILESYSTEM) && + libzfs_mnttab_find(hdl, zhp->zfs_name, &entry) == 0)) { + /* + * mountpoint may have come from a call to + * getmnt/getmntany if it isn't NULL. If it is NULL, + * we know it comes from libzfs_mnttab_find which can + * then get freed later. We strdup it to play it safe. + */ + if (mountpoint == NULL) + mntpt = zfs_strdup(hdl, entry.mnt_mountp); + else + mntpt = zfs_strdup(hdl, mountpoint); + + /* + * Unshare and unmount the filesystem + */ + if (zfs_unshare_proto(zhp, mntpt, share_all_proto) != 0) + return (-1); + + if (unmount_one(hdl, mntpt, flags) != 0) { + free(mntpt); + (void) zfs_shareall(zhp); + return (-1); + } + libzfs_mnttab_remove(hdl, zhp->zfs_name); + free(mntpt); + } + + return (0); +} + +/* + * Unmount this filesystem and any children inheriting the mountpoint property. + * To do this, just act like we're changing the mountpoint property, but don't + * remount the filesystems afterwards. + */ +int +zfs_unmountall(zfs_handle_t *zhp, int flags) +{ + prop_changelist_t *clp; + int ret; + + clp = changelist_gather(zhp, ZFS_PROP_MOUNTPOINT, 0, flags); + if (clp == NULL) + return (-1); + + ret = changelist_prefix(clp); + changelist_free(clp); + + return (ret); +} + +boolean_t +zfs_is_shared(zfs_handle_t *zhp) +{ + zfs_share_type_t rc = 0; + zfs_share_proto_t *curr_proto; + + if (ZFS_IS_VOLUME(zhp)) + return (B_FALSE); + + for (curr_proto = share_all_proto; *curr_proto != PROTO_END; + curr_proto++) + rc |= zfs_is_shared_proto(zhp, NULL, *curr_proto); + + return (rc ? B_TRUE : B_FALSE); +} + +int +zfs_share(zfs_handle_t *zhp) +{ + assert(!ZFS_IS_VOLUME(zhp)); + return (zfs_share_proto(zhp, share_all_proto)); +} + +int +zfs_unshare(zfs_handle_t *zhp) +{ + assert(!ZFS_IS_VOLUME(zhp)); + return (zfs_unshareall(zhp)); +} + +/* + * Check to see if the filesystem is currently shared. + */ +zfs_share_type_t +zfs_is_shared_proto(zfs_handle_t *zhp, char **where, zfs_share_proto_t proto) +{ + char *mountpoint; + zfs_share_type_t rc; + + if (!zfs_is_mounted(zhp, &mountpoint)) + return (SHARED_NOT_SHARED); + + if (rc = is_shared(zhp->zfs_hdl, mountpoint, proto)) { + if (where != NULL) + *where = mountpoint; + else + free(mountpoint); + return (rc); + } else { + free(mountpoint); + return (SHARED_NOT_SHARED); + } +} + +boolean_t +zfs_is_shared_nfs(zfs_handle_t *zhp, char **where) +{ + return (zfs_is_shared_proto(zhp, where, + PROTO_NFS) != SHARED_NOT_SHARED); +} + +boolean_t +zfs_is_shared_smb(zfs_handle_t *zhp, char **where) +{ + return (zfs_is_shared_proto(zhp, where, + PROTO_SMB) != SHARED_NOT_SHARED); +} + +/* + * Make sure things will work if libshare isn't installed by using + * wrapper functions that check to see that the pointers to functions + * initialized in _zfs_init_libshare() are actually present. + */ + +#ifdef sun +static sa_handle_t (*_sa_init)(int); +static void (*_sa_fini)(sa_handle_t); +static sa_share_t (*_sa_find_share)(sa_handle_t, char *); +static int (*_sa_enable_share)(sa_share_t, char *); +static int (*_sa_disable_share)(sa_share_t, char *); +static char *(*_sa_errorstr)(int); +static int (*_sa_parse_legacy_options)(sa_group_t, char *, char *); +static boolean_t (*_sa_needs_refresh)(sa_handle_t *); +static libzfs_handle_t *(*_sa_get_zfs_handle)(sa_handle_t); +static int (*_sa_zfs_process_share)(sa_handle_t, sa_group_t, sa_share_t, + char *, char *, zprop_source_t, char *, char *, char *); +static void (*_sa_update_sharetab_ts)(sa_handle_t); +#endif + +/* + * _zfs_init_libshare() + * + * Find the libshare.so.1 entry points that we use here and save the + * values to be used later. This is triggered by the runtime loader. + * Make sure the correct ISA version is loaded. + */ + +#pragma init(_zfs_init_libshare) +static void +_zfs_init_libshare(void) +{ +#ifdef sun + void *libshare; + char path[MAXPATHLEN]; + char isa[MAXISALEN]; + +#if defined(_LP64) + if (sysinfo(SI_ARCHITECTURE_64, isa, MAXISALEN) == -1) + isa[0] = '\0'; +#else + isa[0] = '\0'; +#endif + (void) snprintf(path, MAXPATHLEN, + "/usr/lib/%s/libshare.so.1", isa); + + if ((libshare = dlopen(path, RTLD_LAZY | RTLD_GLOBAL)) != NULL) { + _sa_init = (sa_handle_t (*)(int))dlsym(libshare, "sa_init"); + _sa_fini = (void (*)(sa_handle_t))dlsym(libshare, "sa_fini"); + _sa_find_share = (sa_share_t (*)(sa_handle_t, char *)) + dlsym(libshare, "sa_find_share"); + _sa_enable_share = (int (*)(sa_share_t, char *))dlsym(libshare, + "sa_enable_share"); + _sa_disable_share = (int (*)(sa_share_t, char *))dlsym(libshare, + "sa_disable_share"); + _sa_errorstr = (char *(*)(int))dlsym(libshare, "sa_errorstr"); + _sa_parse_legacy_options = (int (*)(sa_group_t, char *, char *)) + dlsym(libshare, "sa_parse_legacy_options"); + _sa_needs_refresh = (boolean_t (*)(sa_handle_t *)) + dlsym(libshare, "sa_needs_refresh"); + _sa_get_zfs_handle = (libzfs_handle_t *(*)(sa_handle_t)) + dlsym(libshare, "sa_get_zfs_handle"); + _sa_zfs_process_share = (int (*)(sa_handle_t, sa_group_t, + sa_share_t, char *, char *, zprop_source_t, char *, + char *, char *))dlsym(libshare, "sa_zfs_process_share"); + _sa_update_sharetab_ts = (void (*)(sa_handle_t)) + dlsym(libshare, "sa_update_sharetab_ts"); + if (_sa_init == NULL || _sa_fini == NULL || + _sa_find_share == NULL || _sa_enable_share == NULL || + _sa_disable_share == NULL || _sa_errorstr == NULL || + _sa_parse_legacy_options == NULL || + _sa_needs_refresh == NULL || _sa_get_zfs_handle == NULL || + _sa_zfs_process_share == NULL || + _sa_update_sharetab_ts == NULL) { + _sa_init = NULL; + _sa_fini = NULL; + _sa_disable_share = NULL; + _sa_enable_share = NULL; + _sa_errorstr = NULL; + _sa_parse_legacy_options = NULL; + (void) dlclose(libshare); + _sa_needs_refresh = NULL; + _sa_get_zfs_handle = NULL; + _sa_zfs_process_share = NULL; + _sa_update_sharetab_ts = NULL; + } + } +#endif +} + +/* + * zfs_init_libshare(zhandle, service) + * + * Initialize the libshare API if it hasn't already been initialized. + * In all cases it returns 0 if it succeeded and an error if not. The + * service value is which part(s) of the API to initialize and is a + * direct map to the libshare sa_init(service) interface. + */ +int +zfs_init_libshare(libzfs_handle_t *zhandle, int service) +{ + int ret = SA_OK; + +#ifdef sun + if (_sa_init == NULL) + ret = SA_CONFIG_ERR; + + if (ret == SA_OK && zhandle->libzfs_shareflags & ZFSSHARE_MISS) { + /* + * We had a cache miss. Most likely it is a new ZFS + * dataset that was just created. We want to make sure + * so check timestamps to see if a different process + * has updated any of the configuration. If there was + * some non-ZFS change, we need to re-initialize the + * internal cache. + */ + zhandle->libzfs_shareflags &= ~ZFSSHARE_MISS; + if (_sa_needs_refresh != NULL && + _sa_needs_refresh(zhandle->libzfs_sharehdl)) { + zfs_uninit_libshare(zhandle); + zhandle->libzfs_sharehdl = _sa_init(service); + } + } + + if (ret == SA_OK && zhandle && zhandle->libzfs_sharehdl == NULL) + zhandle->libzfs_sharehdl = _sa_init(service); + + if (ret == SA_OK && zhandle->libzfs_sharehdl == NULL) + ret = SA_NO_MEMORY; +#endif + + return (ret); +} + +/* + * zfs_uninit_libshare(zhandle) + * + * Uninitialize the libshare API if it hasn't already been + * uninitialized. It is OK to call multiple times. + */ +void +zfs_uninit_libshare(libzfs_handle_t *zhandle) +{ + if (zhandle != NULL && zhandle->libzfs_sharehdl != NULL) { +#ifdef sun + if (_sa_fini != NULL) + _sa_fini(zhandle->libzfs_sharehdl); +#endif + zhandle->libzfs_sharehdl = NULL; + } +} + +/* + * zfs_parse_options(options, proto) + * + * Call the legacy parse interface to get the protocol specific + * options using the NULL arg to indicate that this is a "parse" only. + */ +int +zfs_parse_options(char *options, zfs_share_proto_t proto) +{ +#ifdef sun + if (_sa_parse_legacy_options != NULL) { + return (_sa_parse_legacy_options(NULL, options, + proto_table[proto].p_name)); + } + return (SA_CONFIG_ERR); +#else + return (SA_OK); +#endif +} + +#ifdef sun +/* + * zfs_sa_find_share(handle, path) + * + * wrapper around sa_find_share to find a share path in the + * configuration. + */ +static sa_share_t +zfs_sa_find_share(sa_handle_t handle, char *path) +{ + if (_sa_find_share != NULL) + return (_sa_find_share(handle, path)); + return (NULL); +} + +/* + * zfs_sa_enable_share(share, proto) + * + * Wrapper for sa_enable_share which enables a share for a specified + * protocol. + */ +static int +zfs_sa_enable_share(sa_share_t share, char *proto) +{ + if (_sa_enable_share != NULL) + return (_sa_enable_share(share, proto)); + return (SA_CONFIG_ERR); +} + +/* + * zfs_sa_disable_share(share, proto) + * + * Wrapper for sa_enable_share which disables a share for a specified + * protocol. + */ +static int +zfs_sa_disable_share(sa_share_t share, char *proto) +{ + if (_sa_disable_share != NULL) + return (_sa_disable_share(share, proto)); + return (SA_CONFIG_ERR); +} +#endif /* sun */ + +/* + * Share the given filesystem according to the options in the specified + * protocol specific properties (sharenfs, sharesmb). We rely + * on "libshare" to the dirty work for us. + */ +static int +zfs_share_proto(zfs_handle_t *zhp, zfs_share_proto_t *proto) +{ + char mountpoint[ZFS_MAXPROPLEN]; + char shareopts[ZFS_MAXPROPLEN]; + char sourcestr[ZFS_MAXPROPLEN]; + libzfs_handle_t *hdl = zhp->zfs_hdl; + zfs_share_proto_t *curr_proto; + zprop_source_t sourcetype; + int error, ret; + + if (!zfs_is_mountable(zhp, mountpoint, sizeof (mountpoint), NULL)) + return (0); + +#ifdef sun + if ((ret = zfs_init_libshare(hdl, SA_INIT_SHARE_API)) != SA_OK) { + (void) zfs_error_fmt(hdl, EZFS_SHARENFSFAILED, + dgettext(TEXT_DOMAIN, "cannot share '%s': %s"), + zfs_get_name(zhp), _sa_errorstr != NULL ? + _sa_errorstr(ret) : ""); + return (-1); + } +#endif + + for (curr_proto = proto; *curr_proto != PROTO_END; curr_proto++) { + /* + * Return success if there are no share options. + */ + if (zfs_prop_get(zhp, proto_table[*curr_proto].p_prop, + shareopts, sizeof (shareopts), &sourcetype, sourcestr, + ZFS_MAXPROPLEN, B_FALSE) != 0 || + strcmp(shareopts, "off") == 0) + continue; + + /* + * If the 'zoned' property is set, then zfs_is_mountable() + * will have already bailed out if we are in the global zone. + * But local zones cannot be NFS servers, so we ignore it for + * local zones as well. + */ + if (zfs_prop_get_int(zhp, ZFS_PROP_ZONED)) + continue; + +#ifdef sun + share = zfs_sa_find_share(hdl->libzfs_sharehdl, mountpoint); + if (share == NULL) { + /* + * This may be a new file system that was just + * created so isn't in the internal cache + * (second time through). Rather than + * reloading the entire configuration, we can + * assume ZFS has done the checking and it is + * safe to add this to the internal + * configuration. + */ + if (_sa_zfs_process_share(hdl->libzfs_sharehdl, + NULL, NULL, mountpoint, + proto_table[*curr_proto].p_name, sourcetype, + shareopts, sourcestr, zhp->zfs_name) != SA_OK) { + (void) zfs_error_fmt(hdl, + proto_table[*curr_proto].p_share_err, + dgettext(TEXT_DOMAIN, "cannot share '%s'"), + zfs_get_name(zhp)); + return (-1); + } + hdl->libzfs_shareflags |= ZFSSHARE_MISS; + share = zfs_sa_find_share(hdl->libzfs_sharehdl, + mountpoint); + } + if (share != NULL) { + int err; + err = zfs_sa_enable_share(share, + proto_table[*curr_proto].p_name); + if (err != SA_OK) { + (void) zfs_error_fmt(hdl, + proto_table[*curr_proto].p_share_err, + dgettext(TEXT_DOMAIN, "cannot share '%s'"), + zfs_get_name(zhp)); + return (-1); + } + } else +#else + if (*curr_proto != PROTO_NFS) { + fprintf(stderr, "Unsupported share protocol: %d.\n", + *curr_proto); + continue; + } + + if (strcmp(shareopts, "on") == 0) + error = fsshare(ZFS_EXPORTS_PATH, mountpoint, ""); + else + error = fsshare(ZFS_EXPORTS_PATH, mountpoint, shareopts); + if (error != 0) +#endif + { + (void) zfs_error_fmt(hdl, + proto_table[*curr_proto].p_share_err, + dgettext(TEXT_DOMAIN, "cannot share '%s'"), + zfs_get_name(zhp)); + return (-1); + } + + } + return (0); +} + + +int +zfs_share_nfs(zfs_handle_t *zhp) +{ + return (zfs_share_proto(zhp, nfs_only)); +} + +int +zfs_share_smb(zfs_handle_t *zhp) +{ + return (zfs_share_proto(zhp, smb_only)); +} + +int +zfs_shareall(zfs_handle_t *zhp) +{ + return (zfs_share_proto(zhp, share_all_proto)); +} + +/* + * Unshare a filesystem by mountpoint. + */ +static int +unshare_one(libzfs_handle_t *hdl, const char *name, const char *mountpoint, + zfs_share_proto_t proto) +{ +#ifdef sun + sa_share_t share; + int err; + char *mntpt; + /* + * Mountpoint could get trashed if libshare calls getmntany + * which it does during API initialization, so strdup the + * value. + */ + mntpt = zfs_strdup(hdl, mountpoint); + + /* make sure libshare initialized */ + if ((err = zfs_init_libshare(hdl, SA_INIT_SHARE_API)) != SA_OK) { + free(mntpt); /* don't need the copy anymore */ + return (zfs_error_fmt(hdl, EZFS_SHARENFSFAILED, + dgettext(TEXT_DOMAIN, "cannot unshare '%s': %s"), + name, _sa_errorstr(err))); + } + + share = zfs_sa_find_share(hdl->libzfs_sharehdl, mntpt); + free(mntpt); /* don't need the copy anymore */ + + if (share != NULL) { + err = zfs_sa_disable_share(share, proto_table[proto].p_name); + if (err != SA_OK) { + return (zfs_error_fmt(hdl, EZFS_UNSHARENFSFAILED, + dgettext(TEXT_DOMAIN, "cannot unshare '%s': %s"), + name, _sa_errorstr(err))); + } + } else { + return (zfs_error_fmt(hdl, EZFS_UNSHARENFSFAILED, + dgettext(TEXT_DOMAIN, "cannot unshare '%s': not found"), + name)); + } +#else + char buf[MAXPATHLEN]; + FILE *fp; + int err; + + if (proto != PROTO_NFS) { + fprintf(stderr, "No SMB support in FreeBSD yet.\n"); + return (EOPNOTSUPP); + } + + err = fsunshare(ZFS_EXPORTS_PATH, mountpoint); + if (err != 0) { + zfs_error_aux(hdl, "%s", strerror(err)); + return (zfs_error_fmt(hdl, EZFS_UNSHARENFSFAILED, + dgettext(TEXT_DOMAIN, + "cannot unshare '%s'"), name)); + } +#endif + return (0); +} + +/* + * Unshare the given filesystem. + */ +int +zfs_unshare_proto(zfs_handle_t *zhp, const char *mountpoint, + zfs_share_proto_t *proto) +{ + libzfs_handle_t *hdl = zhp->zfs_hdl; + struct mnttab entry; + char *mntpt = NULL; + + /* check to see if need to unmount the filesystem */ + rewind(zhp->zfs_hdl->libzfs_mnttab); + if (mountpoint != NULL) + mountpoint = mntpt = zfs_strdup(hdl, mountpoint); + + if (mountpoint != NULL || ((zfs_get_type(zhp) == ZFS_TYPE_FILESYSTEM) && + libzfs_mnttab_find(hdl, zfs_get_name(zhp), &entry) == 0)) { + zfs_share_proto_t *curr_proto; + + if (mountpoint == NULL) + mntpt = zfs_strdup(zhp->zfs_hdl, entry.mnt_mountp); + + for (curr_proto = proto; *curr_proto != PROTO_END; + curr_proto++) { + + if (is_shared(hdl, mntpt, *curr_proto) && + unshare_one(hdl, zhp->zfs_name, + mntpt, *curr_proto) != 0) { + if (mntpt != NULL) + free(mntpt); + return (-1); + } + } + } + if (mntpt != NULL) + free(mntpt); + + return (0); +} + +int +zfs_unshare_nfs(zfs_handle_t *zhp, const char *mountpoint) +{ + return (zfs_unshare_proto(zhp, mountpoint, nfs_only)); +} + +int +zfs_unshare_smb(zfs_handle_t *zhp, const char *mountpoint) +{ + return (zfs_unshare_proto(zhp, mountpoint, smb_only)); +} + +/* + * Same as zfs_unmountall(), but for NFS and SMB unshares. + */ +int +zfs_unshareall_proto(zfs_handle_t *zhp, zfs_share_proto_t *proto) +{ + prop_changelist_t *clp; + int ret; + + clp = changelist_gather(zhp, ZFS_PROP_SHARENFS, 0, 0); + if (clp == NULL) + return (-1); + + ret = changelist_unshare(clp, proto); + changelist_free(clp); + + return (ret); +} + +int +zfs_unshareall_nfs(zfs_handle_t *zhp) +{ + return (zfs_unshareall_proto(zhp, nfs_only)); +} + +int +zfs_unshareall_smb(zfs_handle_t *zhp) +{ + return (zfs_unshareall_proto(zhp, smb_only)); +} + +int +zfs_unshareall(zfs_handle_t *zhp) +{ + return (zfs_unshareall_proto(zhp, share_all_proto)); +} + +int +zfs_unshareall_bypath(zfs_handle_t *zhp, const char *mountpoint) +{ + return (zfs_unshare_proto(zhp, mountpoint, share_all_proto)); +} + +/* + * Remove the mountpoint associated with the current dataset, if necessary. + * We only remove the underlying directory if: + * + * - The mountpoint is not 'none' or 'legacy' + * - The mountpoint is non-empty + * - The mountpoint is the default or inherited + * - The 'zoned' property is set, or we're in a local zone + * + * Any other directories we leave alone. + */ +void +remove_mountpoint(zfs_handle_t *zhp) +{ + char mountpoint[ZFS_MAXPROPLEN]; + zprop_source_t source; + + if (!zfs_is_mountable(zhp, mountpoint, sizeof (mountpoint), + &source)) + return; + + if (source == ZPROP_SRC_DEFAULT || + source == ZPROP_SRC_INHERITED) { + /* + * Try to remove the directory, silently ignoring any errors. + * The filesystem may have since been removed or moved around, + * and this error isn't really useful to the administrator in + * any way. + */ + (void) rmdir(mountpoint); + } +} + +void +libzfs_add_handle(get_all_cb_t *cbp, zfs_handle_t *zhp) +{ + if (cbp->cb_alloc == cbp->cb_used) { + size_t newsz; + void *ptr; + + newsz = cbp->cb_alloc ? cbp->cb_alloc * 2 : 64; + ptr = zfs_realloc(zhp->zfs_hdl, + cbp->cb_handles, cbp->cb_alloc * sizeof (void *), + newsz * sizeof (void *)); + cbp->cb_handles = ptr; + cbp->cb_alloc = newsz; + } + cbp->cb_handles[cbp->cb_used++] = zhp; +} + +static int +mount_cb(zfs_handle_t *zhp, void *data) +{ + get_all_cb_t *cbp = data; + + if (!(zfs_get_type(zhp) & ZFS_TYPE_FILESYSTEM)) { + zfs_close(zhp); + return (0); + } + + if (zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT) == ZFS_CANMOUNT_NOAUTO) { + zfs_close(zhp); + return (0); + } + + libzfs_add_handle(cbp, zhp); + if (zfs_iter_filesystems(zhp, mount_cb, cbp) != 0) { + zfs_close(zhp); + return (-1); + } + return (0); +} + +int +libzfs_dataset_cmp(const void *a, const void *b) +{ + zfs_handle_t **za = (zfs_handle_t **)a; + zfs_handle_t **zb = (zfs_handle_t **)b; + char mounta[MAXPATHLEN]; + char mountb[MAXPATHLEN]; + boolean_t gota, gotb; + + if ((gota = (zfs_get_type(*za) == ZFS_TYPE_FILESYSTEM)) != 0) + verify(zfs_prop_get(*za, ZFS_PROP_MOUNTPOINT, mounta, + sizeof (mounta), NULL, NULL, 0, B_FALSE) == 0); + if ((gotb = (zfs_get_type(*zb) == ZFS_TYPE_FILESYSTEM)) != 0) + verify(zfs_prop_get(*zb, ZFS_PROP_MOUNTPOINT, mountb, + sizeof (mountb), NULL, NULL, 0, B_FALSE) == 0); + + if (gota && gotb) + return (strcmp(mounta, mountb)); + + if (gota) + return (-1); + if (gotb) + return (1); + + return (strcmp(zfs_get_name(a), zfs_get_name(b))); +} + +/* + * Mount and share all datasets within the given pool. This assumes that no + * datasets within the pool are currently mounted. Because users can create + * complicated nested hierarchies of mountpoints, we first gather all the + * datasets and mountpoints within the pool, and sort them by mountpoint. Once + * we have the list of all filesystems, we iterate over them in order and mount + * and/or share each one. + */ +#pragma weak zpool_mount_datasets = zpool_enable_datasets +int +zpool_enable_datasets(zpool_handle_t *zhp, const char *mntopts, int flags) +{ + get_all_cb_t cb = { 0 }; + libzfs_handle_t *hdl = zhp->zpool_hdl; + zfs_handle_t *zfsp; + int i, ret = -1; + int *good; + + /* + * Gather all non-snap datasets within the pool. + */ + if ((zfsp = zfs_open(hdl, zhp->zpool_name, ZFS_TYPE_DATASET)) == NULL) + goto out; + + libzfs_add_handle(&cb, zfsp); + if (zfs_iter_filesystems(zfsp, mount_cb, &cb) != 0) + goto out; + /* + * Sort the datasets by mountpoint. + */ + qsort(cb.cb_handles, cb.cb_used, sizeof (void *), + libzfs_dataset_cmp); + + /* + * And mount all the datasets, keeping track of which ones + * succeeded or failed. + */ + if ((good = zfs_alloc(zhp->zpool_hdl, + cb.cb_used * sizeof (int))) == NULL) + goto out; + + ret = 0; + for (i = 0; i < cb.cb_used; i++) { + if (zfs_mount(cb.cb_handles[i], mntopts, flags) != 0) + ret = -1; + else + good[i] = 1; + } + + /* + * Then share all the ones that need to be shared. This needs + * to be a separate pass in order to avoid excessive reloading + * of the configuration. Good should never be NULL since + * zfs_alloc is supposed to exit if memory isn't available. + */ + for (i = 0; i < cb.cb_used; i++) { + if (good[i] && zfs_share(cb.cb_handles[i]) != 0) + ret = -1; + } + + free(good); + +out: + for (i = 0; i < cb.cb_used; i++) + zfs_close(cb.cb_handles[i]); + free(cb.cb_handles); + + return (ret); +} + +static int +mountpoint_compare(const void *a, const void *b) +{ + const char *mounta = *((char **)a); + const char *mountb = *((char **)b); + + return (strcmp(mountb, mounta)); +} + +/* alias for 2002/240 */ +#pragma weak zpool_unmount_datasets = zpool_disable_datasets +/* + * Unshare and unmount all datasets within the given pool. We don't want to + * rely on traversing the DSL to discover the filesystems within the pool, + * because this may be expensive (if not all of them are mounted), and can fail + * arbitrarily (on I/O error, for example). Instead, we walk /etc/mnttab and + * gather all the filesystems that are currently mounted. + */ +int +zpool_disable_datasets(zpool_handle_t *zhp, boolean_t force) +{ + int used, alloc; + struct mnttab entry; + size_t namelen; + char **mountpoints = NULL; + zfs_handle_t **datasets = NULL; + libzfs_handle_t *hdl = zhp->zpool_hdl; + int i; + int ret = -1; + int flags = (force ? MS_FORCE : 0); + + namelen = strlen(zhp->zpool_name); + + rewind(hdl->libzfs_mnttab); + used = alloc = 0; + while (getmntent(hdl->libzfs_mnttab, &entry) == 0) { + /* + * Ignore non-ZFS entries. + */ + if (entry.mnt_fstype == NULL || + strcmp(entry.mnt_fstype, MNTTYPE_ZFS) != 0) + continue; + + /* + * Ignore filesystems not within this pool. + */ + if (entry.mnt_mountp == NULL || + strncmp(entry.mnt_special, zhp->zpool_name, namelen) != 0 || + (entry.mnt_special[namelen] != '/' && + entry.mnt_special[namelen] != '\0')) + continue; + + /* + * At this point we've found a filesystem within our pool. Add + * it to our growing list. + */ + if (used == alloc) { + if (alloc == 0) { + if ((mountpoints = zfs_alloc(hdl, + 8 * sizeof (void *))) == NULL) + goto out; + + if ((datasets = zfs_alloc(hdl, + 8 * sizeof (void *))) == NULL) + goto out; + + alloc = 8; + } else { + void *ptr; + + if ((ptr = zfs_realloc(hdl, mountpoints, + alloc * sizeof (void *), + alloc * 2 * sizeof (void *))) == NULL) + goto out; + mountpoints = ptr; + + if ((ptr = zfs_realloc(hdl, datasets, + alloc * sizeof (void *), + alloc * 2 * sizeof (void *))) == NULL) + goto out; + datasets = ptr; + + alloc *= 2; + } + } + + if ((mountpoints[used] = zfs_strdup(hdl, + entry.mnt_mountp)) == NULL) + goto out; + + /* + * This is allowed to fail, in case there is some I/O error. It + * is only used to determine if we need to remove the underlying + * mountpoint, so failure is not fatal. + */ + datasets[used] = make_dataset_handle(hdl, entry.mnt_special); + + used++; + } + + /* + * At this point, we have the entire list of filesystems, so sort it by + * mountpoint. + */ + qsort(mountpoints, used, sizeof (char *), mountpoint_compare); + + /* + * Walk through and first unshare everything. + */ + for (i = 0; i < used; i++) { + zfs_share_proto_t *curr_proto; + for (curr_proto = share_all_proto; *curr_proto != PROTO_END; + curr_proto++) { + if (is_shared(hdl, mountpoints[i], *curr_proto) && + unshare_one(hdl, mountpoints[i], + mountpoints[i], *curr_proto) != 0) + goto out; + } + } + + /* + * Now unmount everything, removing the underlying directories as + * appropriate. + */ + for (i = 0; i < used; i++) { + if (unmount_one(hdl, mountpoints[i], flags) != 0) + goto out; + } + + for (i = 0; i < used; i++) { + if (datasets[i]) + remove_mountpoint(datasets[i]); + } + + ret = 0; +out: + for (i = 0; i < used; i++) { + if (datasets[i]) + zfs_close(datasets[i]); + free(mountpoints[i]); + } + free(datasets); + free(mountpoints); + + return (ret); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_pool.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_pool.c new file mode 100644 index 0000000..02f0b96 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_pool.c @@ -0,0 +1,4154 @@ +/* + * 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) 2013 by Delphix. All rights reserved. + * Copyright (c) 2013, Joyent, Inc. All rights reserved. + */ + +#include <sys/types.h> +#include <sys/stat.h> +#include <ctype.h> +#include <errno.h> +#include <devid.h> +#include <fcntl.h> +#include <libintl.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> +#include <unistd.h> +#include <libgen.h> +#include <sys/zfs_ioctl.h> +#include <dlfcn.h> + +#include "zfs_namecheck.h" +#include "zfs_prop.h" +#include "libzfs_impl.h" +#include "zfs_comutil.h" +#include "zfeature_common.h" + +static int read_efi_label(nvlist_t *config, diskaddr_t *sb); + +#define DISK_ROOT "/dev/dsk" +#define RDISK_ROOT "/dev/rdsk" +#define BACKUP_SLICE "s2" + +typedef struct prop_flags { + int create:1; /* Validate property on creation */ + int import:1; /* Validate property on import */ +} prop_flags_t; + +/* + * ==================================================================== + * zpool property functions + * ==================================================================== + */ + +static int +zpool_get_all_props(zpool_handle_t *zhp) +{ + zfs_cmd_t zc = { 0 }; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + + if (zcmd_alloc_dst_nvlist(hdl, &zc, 0) != 0) + return (-1); + + while (ioctl(hdl->libzfs_fd, ZFS_IOC_POOL_GET_PROPS, &zc) != 0) { + if (errno == ENOMEM) { + if (zcmd_expand_dst_nvlist(hdl, &zc) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + } else { + zcmd_free_nvlists(&zc); + return (-1); + } + } + + if (zcmd_read_dst_nvlist(hdl, &zc, &zhp->zpool_props) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + + zcmd_free_nvlists(&zc); + + return (0); +} + +static int +zpool_props_refresh(zpool_handle_t *zhp) +{ + nvlist_t *old_props; + + old_props = zhp->zpool_props; + + if (zpool_get_all_props(zhp) != 0) + return (-1); + + nvlist_free(old_props); + return (0); +} + +static char * +zpool_get_prop_string(zpool_handle_t *zhp, zpool_prop_t prop, + zprop_source_t *src) +{ + nvlist_t *nv, *nvl; + uint64_t ival; + char *value; + zprop_source_t source; + + nvl = zhp->zpool_props; + if (nvlist_lookup_nvlist(nvl, zpool_prop_to_name(prop), &nv) == 0) { + verify(nvlist_lookup_uint64(nv, ZPROP_SOURCE, &ival) == 0); + source = ival; + verify(nvlist_lookup_string(nv, ZPROP_VALUE, &value) == 0); + } else { + source = ZPROP_SRC_DEFAULT; + if ((value = (char *)zpool_prop_default_string(prop)) == NULL) + value = "-"; + } + + if (src) + *src = source; + + return (value); +} + +uint64_t +zpool_get_prop_int(zpool_handle_t *zhp, zpool_prop_t prop, zprop_source_t *src) +{ + nvlist_t *nv, *nvl; + uint64_t value; + zprop_source_t source; + + if (zhp->zpool_props == NULL && zpool_get_all_props(zhp)) { + /* + * zpool_get_all_props() has most likely failed because + * the pool is faulted, but if all we need is the top level + * vdev's guid then get it from the zhp config nvlist. + */ + if ((prop == ZPOOL_PROP_GUID) && + (nvlist_lookup_nvlist(zhp->zpool_config, + ZPOOL_CONFIG_VDEV_TREE, &nv) == 0) && + (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, &value) + == 0)) { + return (value); + } + return (zpool_prop_default_numeric(prop)); + } + + nvl = zhp->zpool_props; + if (nvlist_lookup_nvlist(nvl, zpool_prop_to_name(prop), &nv) == 0) { + verify(nvlist_lookup_uint64(nv, ZPROP_SOURCE, &value) == 0); + source = value; + verify(nvlist_lookup_uint64(nv, ZPROP_VALUE, &value) == 0); + } else { + source = ZPROP_SRC_DEFAULT; + value = zpool_prop_default_numeric(prop); + } + + if (src) + *src = source; + + return (value); +} + +/* + * Map VDEV STATE to printed strings. + */ +const char * +zpool_state_to_name(vdev_state_t state, vdev_aux_t aux) +{ + switch (state) { + case VDEV_STATE_CLOSED: + case VDEV_STATE_OFFLINE: + return (gettext("OFFLINE")); + case VDEV_STATE_REMOVED: + return (gettext("REMOVED")); + case VDEV_STATE_CANT_OPEN: + if (aux == VDEV_AUX_CORRUPT_DATA || aux == VDEV_AUX_BAD_LOG) + return (gettext("FAULTED")); + else if (aux == VDEV_AUX_SPLIT_POOL) + return (gettext("SPLIT")); + else + return (gettext("UNAVAIL")); + case VDEV_STATE_FAULTED: + return (gettext("FAULTED")); + case VDEV_STATE_DEGRADED: + return (gettext("DEGRADED")); + case VDEV_STATE_HEALTHY: + return (gettext("ONLINE")); + } + + return (gettext("UNKNOWN")); +} + +/* + * Map POOL STATE to printed strings. + */ +const char * +zpool_pool_state_to_name(pool_state_t state) +{ + switch (state) { + case POOL_STATE_ACTIVE: + return (gettext("ACTIVE")); + case POOL_STATE_EXPORTED: + return (gettext("EXPORTED")); + case POOL_STATE_DESTROYED: + return (gettext("DESTROYED")); + case POOL_STATE_SPARE: + return (gettext("SPARE")); + case POOL_STATE_L2CACHE: + return (gettext("L2CACHE")); + case POOL_STATE_UNINITIALIZED: + return (gettext("UNINITIALIZED")); + case POOL_STATE_UNAVAIL: + return (gettext("UNAVAIL")); + case POOL_STATE_POTENTIALLY_ACTIVE: + return (gettext("POTENTIALLY_ACTIVE")); + } + + return (gettext("UNKNOWN")); +} + +/* + * Get a zpool property value for 'prop' and return the value in + * a pre-allocated buffer. + */ +int +zpool_get_prop(zpool_handle_t *zhp, zpool_prop_t prop, char *buf, size_t len, + zprop_source_t *srctype, boolean_t literal) +{ + uint64_t intval; + const char *strval; + zprop_source_t src = ZPROP_SRC_NONE; + nvlist_t *nvroot; + vdev_stat_t *vs; + uint_t vsc; + + if (zpool_get_state(zhp) == POOL_STATE_UNAVAIL) { + switch (prop) { + case ZPOOL_PROP_NAME: + (void) strlcpy(buf, zpool_get_name(zhp), len); + break; + + case ZPOOL_PROP_HEALTH: + (void) strlcpy(buf, "FAULTED", len); + break; + + case ZPOOL_PROP_GUID: + intval = zpool_get_prop_int(zhp, prop, &src); + (void) snprintf(buf, len, "%llu", intval); + break; + + case ZPOOL_PROP_ALTROOT: + case ZPOOL_PROP_CACHEFILE: + case ZPOOL_PROP_COMMENT: + if (zhp->zpool_props != NULL || + zpool_get_all_props(zhp) == 0) { + (void) strlcpy(buf, + zpool_get_prop_string(zhp, prop, &src), + len); + break; + } + /* FALLTHROUGH */ + default: + (void) strlcpy(buf, "-", len); + break; + } + + if (srctype != NULL) + *srctype = src; + return (0); + } + + if (zhp->zpool_props == NULL && zpool_get_all_props(zhp) && + prop != ZPOOL_PROP_NAME) + return (-1); + + switch (zpool_prop_get_type(prop)) { + case PROP_TYPE_STRING: + (void) strlcpy(buf, zpool_get_prop_string(zhp, prop, &src), + len); + break; + + case PROP_TYPE_NUMBER: + intval = zpool_get_prop_int(zhp, prop, &src); + + switch (prop) { + case ZPOOL_PROP_SIZE: + case ZPOOL_PROP_ALLOCATED: + case ZPOOL_PROP_FREE: + case ZPOOL_PROP_FREEING: + case ZPOOL_PROP_EXPANDSZ: + if (literal) { + (void) snprintf(buf, len, "%llu", + (u_longlong_t)intval); + } else { + (void) zfs_nicenum(intval, buf, len); + } + break; + + case ZPOOL_PROP_CAPACITY: + if (literal) { + (void) snprintf(buf, len, "%llu", + (u_longlong_t)intval); + } else { + (void) snprintf(buf, len, "%llu%%", + (u_longlong_t)intval); + } + break; + + case ZPOOL_PROP_DEDUPRATIO: + (void) snprintf(buf, len, "%llu.%02llux", + (u_longlong_t)(intval / 100), + (u_longlong_t)(intval % 100)); + break; + + case ZPOOL_PROP_HEALTH: + verify(nvlist_lookup_nvlist(zpool_get_config(zhp, NULL), + ZPOOL_CONFIG_VDEV_TREE, &nvroot) == 0); + verify(nvlist_lookup_uint64_array(nvroot, + ZPOOL_CONFIG_VDEV_STATS, (uint64_t **)&vs, &vsc) + == 0); + + (void) strlcpy(buf, zpool_state_to_name(intval, + vs->vs_aux), len); + break; + case ZPOOL_PROP_VERSION: + if (intval >= SPA_VERSION_FEATURES) { + (void) snprintf(buf, len, "-"); + break; + } + /* FALLTHROUGH */ + default: + (void) snprintf(buf, len, "%llu", intval); + } + break; + + case PROP_TYPE_INDEX: + intval = zpool_get_prop_int(zhp, prop, &src); + if (zpool_prop_index_to_string(prop, intval, &strval) + != 0) + return (-1); + (void) strlcpy(buf, strval, len); + break; + + default: + abort(); + } + + if (srctype) + *srctype = src; + + return (0); +} + +/* + * Check if the bootfs name has the same pool name as it is set to. + * Assuming bootfs is a valid dataset name. + */ +static boolean_t +bootfs_name_valid(const char *pool, char *bootfs) +{ + int len = strlen(pool); + + if (!zfs_name_valid(bootfs, ZFS_TYPE_FILESYSTEM|ZFS_TYPE_SNAPSHOT)) + return (B_FALSE); + + if (strncmp(pool, bootfs, len) == 0 && + (bootfs[len] == '/' || bootfs[len] == '\0')) + return (B_TRUE); + + return (B_FALSE); +} + +/* + * Inspect the configuration to determine if any of the devices contain + * an EFI label. + */ +static boolean_t +pool_uses_efi(nvlist_t *config) +{ +#ifdef sun + nvlist_t **child; + uint_t c, children; + + if (nvlist_lookup_nvlist_array(config, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) + return (read_efi_label(config, NULL) >= 0); + + for (c = 0; c < children; c++) { + if (pool_uses_efi(child[c])) + return (B_TRUE); + } +#endif /* sun */ + return (B_FALSE); +} + +boolean_t +zpool_is_bootable(zpool_handle_t *zhp) +{ + char bootfs[ZPOOL_MAXNAMELEN]; + + return (zpool_get_prop(zhp, ZPOOL_PROP_BOOTFS, bootfs, + sizeof (bootfs), NULL, B_FALSE) == 0 && strncmp(bootfs, "-", + sizeof (bootfs)) != 0); +} + + +/* + * Given an nvlist of zpool properties to be set, validate that they are + * correct, and parse any numeric properties (index, boolean, etc) if they are + * specified as strings. + */ +static nvlist_t * +zpool_valid_proplist(libzfs_handle_t *hdl, const char *poolname, + nvlist_t *props, uint64_t version, prop_flags_t flags, char *errbuf) +{ + nvpair_t *elem; + nvlist_t *retprops; + zpool_prop_t prop; + char *strval; + uint64_t intval; + char *slash, *check; + struct stat64 statbuf; + zpool_handle_t *zhp; + nvlist_t *nvroot; + + if (nvlist_alloc(&retprops, NV_UNIQUE_NAME, 0) != 0) { + (void) no_memory(hdl); + return (NULL); + } + + elem = NULL; + while ((elem = nvlist_next_nvpair(props, elem)) != NULL) { + const char *propname = nvpair_name(elem); + + prop = zpool_name_to_prop(propname); + if (prop == ZPROP_INVAL && zpool_prop_feature(propname)) { + int err; + char *fname = strchr(propname, '@') + 1; + + err = zfeature_lookup_name(fname, NULL); + if (err != 0) { + ASSERT3U(err, ==, ENOENT); + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "invalid feature '%s'"), fname); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); + goto error; + } + + 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; + } + + (void) nvpair_value_string(elem, &strval); + if (strcmp(strval, ZFS_FEATURE_ENABLED) != 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property '%s' can only be set to " + "'enabled'"), propname); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); + goto error; + } + + if (nvlist_add_uint64(retprops, propname, 0) != 0) { + (void) no_memory(hdl); + goto error; + } + continue; + } + + /* + * Make sure this property is valid and applies to this type. + */ + 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 (zpool_prop_readonly(prop)) { + 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, ZFS_TYPE_POOL, retprops, + &strval, &intval, errbuf) != 0) + goto error; + + /* + * Perform additional checking for specific properties. + */ + switch (prop) { + case ZPOOL_PROP_VERSION: + if (intval < version || + !SPA_VERSION_IS_SUPPORTED(intval)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property '%s' number %d is invalid."), + propname, intval); + (void) zfs_error(hdl, EZFS_BADVERSION, errbuf); + goto error; + } + break; + + case ZPOOL_PROP_BOOTFS: + if (flags.create || flags.import) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property '%s' cannot be set at creation " + "or import time"), propname); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); + goto error; + } + + if (version < SPA_VERSION_BOOTFS) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "pool must be upgraded to support " + "'%s' property"), propname); + (void) zfs_error(hdl, EZFS_BADVERSION, errbuf); + goto error; + } + + /* + * bootfs property value has to be a dataset name and + * the dataset has to be in the same pool as it sets to. + */ + if (strval[0] != '\0' && !bootfs_name_valid(poolname, + strval)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' " + "is an invalid name"), strval); + (void) zfs_error(hdl, EZFS_INVALIDNAME, errbuf); + goto error; + } + + if ((zhp = zpool_open_canfail(hdl, poolname)) == NULL) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "could not open pool '%s'"), poolname); + (void) zfs_error(hdl, EZFS_OPENFAILED, errbuf); + goto error; + } + verify(nvlist_lookup_nvlist(zpool_get_config(zhp, NULL), + ZPOOL_CONFIG_VDEV_TREE, &nvroot) == 0); + +#ifdef sun + /* + * bootfs property cannot be set on a disk which has + * been EFI labeled. + */ + if (pool_uses_efi(nvroot)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property '%s' not supported on " + "EFI labeled devices"), propname); + (void) zfs_error(hdl, EZFS_POOL_NOTSUP, errbuf); + zpool_close(zhp); + goto error; + } +#endif /* sun */ + zpool_close(zhp); + break; + + case ZPOOL_PROP_ALTROOT: + if (!flags.create && !flags.import) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property '%s' can only be set during pool " + "creation or import"), propname); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); + goto error; + } + + if (strval[0] != '/') { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "bad alternate root '%s'"), strval); + (void) zfs_error(hdl, EZFS_BADPATH, errbuf); + goto error; + } + break; + + case ZPOOL_PROP_CACHEFILE: + if (strval[0] == '\0') + break; + + if (strcmp(strval, "none") == 0) + break; + + if (strval[0] != '/') { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property '%s' must be empty, an " + "absolute path, or 'none'"), propname); + (void) zfs_error(hdl, EZFS_BADPATH, errbuf); + goto error; + } + + slash = strrchr(strval, '/'); + + if (slash[1] == '\0' || strcmp(slash, "/.") == 0 || + strcmp(slash, "/..") == 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "'%s' is not a valid file"), strval); + (void) zfs_error(hdl, EZFS_BADPATH, errbuf); + goto error; + } + + *slash = '\0'; + + if (strval[0] != '\0' && + (stat64(strval, &statbuf) != 0 || + !S_ISDIR(statbuf.st_mode))) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "'%s' is not a valid directory"), + strval); + (void) zfs_error(hdl, EZFS_BADPATH, errbuf); + goto error; + } + + *slash = '/'; + break; + + case ZPOOL_PROP_COMMENT: + for (check = strval; *check != '\0'; check++) { + if (!isprint(*check)) { + zfs_error_aux(hdl, + dgettext(TEXT_DOMAIN, + "comment may only have printable " + "characters")); + (void) zfs_error(hdl, EZFS_BADPROP, + errbuf); + goto error; + } + } + if (strlen(strval) > ZPROP_MAX_COMMENT) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "comment must not exceed %d characters"), + ZPROP_MAX_COMMENT); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); + goto error; + } + break; + case ZPOOL_PROP_READONLY: + if (!flags.import) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "property '%s' can only be set at " + "import time"), propname); + (void) zfs_error(hdl, EZFS_BADPROP, errbuf); + goto error; + } + break; + } + } + + return (retprops); +error: + nvlist_free(retprops); + return (NULL); +} + +/* + * Set zpool property : propname=propval. + */ +int +zpool_set_prop(zpool_handle_t *zhp, const char *propname, const char *propval) +{ + zfs_cmd_t zc = { 0 }; + int ret = -1; + char errbuf[1024]; + nvlist_t *nvl = NULL; + nvlist_t *realprops; + uint64_t version; + prop_flags_t flags = { 0 }; + + (void) snprintf(errbuf, sizeof (errbuf), + dgettext(TEXT_DOMAIN, "cannot set property for '%s'"), + zhp->zpool_name); + + if (nvlist_alloc(&nvl, NV_UNIQUE_NAME, 0) != 0) + return (no_memory(zhp->zpool_hdl)); + + if (nvlist_add_string(nvl, propname, propval) != 0) { + nvlist_free(nvl); + return (no_memory(zhp->zpool_hdl)); + } + + version = zpool_get_prop_int(zhp, ZPOOL_PROP_VERSION, NULL); + if ((realprops = zpool_valid_proplist(zhp->zpool_hdl, + zhp->zpool_name, nvl, version, flags, errbuf)) == NULL) { + nvlist_free(nvl); + return (-1); + } + + nvlist_free(nvl); + nvl = realprops; + + /* + * Execute the corresponding ioctl() to set this property. + */ + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + + if (zcmd_write_src_nvlist(zhp->zpool_hdl, &zc, nvl) != 0) { + nvlist_free(nvl); + return (-1); + } + + ret = zfs_ioctl(zhp->zpool_hdl, ZFS_IOC_POOL_SET_PROPS, &zc); + + zcmd_free_nvlists(&zc); + nvlist_free(nvl); + + if (ret) + (void) zpool_standard_error(zhp->zpool_hdl, errno, errbuf); + else + (void) zpool_props_refresh(zhp); + + return (ret); +} + +int +zpool_expand_proplist(zpool_handle_t *zhp, zprop_list_t **plp) +{ + libzfs_handle_t *hdl = zhp->zpool_hdl; + zprop_list_t *entry; + char buf[ZFS_MAXPROPLEN]; + nvlist_t *features = NULL; + zprop_list_t **last; + boolean_t firstexpand = (NULL == *plp); + + if (zprop_expand_list(hdl, plp, ZFS_TYPE_POOL) != 0) + return (-1); + + last = plp; + while (*last != NULL) + last = &(*last)->pl_next; + + if ((*plp)->pl_all) + features = zpool_get_features(zhp); + + if ((*plp)->pl_all && firstexpand) { + for (int i = 0; i < SPA_FEATURES; i++) { + zprop_list_t *entry = zfs_alloc(hdl, + sizeof (zprop_list_t)); + entry->pl_prop = ZPROP_INVAL; + entry->pl_user_prop = zfs_asprintf(hdl, "feature@%s", + spa_feature_table[i].fi_uname); + entry->pl_width = strlen(entry->pl_user_prop); + entry->pl_all = B_TRUE; + + *last = entry; + last = &entry->pl_next; + } + } + + /* add any unsupported features */ + for (nvpair_t *nvp = nvlist_next_nvpair(features, NULL); + nvp != NULL; nvp = nvlist_next_nvpair(features, nvp)) { + char *propname; + boolean_t found; + zprop_list_t *entry; + + if (zfeature_is_supported(nvpair_name(nvp))) + continue; + + propname = zfs_asprintf(hdl, "unsupported@%s", + nvpair_name(nvp)); + + /* + * Before adding the property to the list make sure that no + * other pool already added the same property. + */ + found = B_FALSE; + entry = *plp; + while (entry != NULL) { + if (entry->pl_user_prop != NULL && + strcmp(propname, entry->pl_user_prop) == 0) { + found = B_TRUE; + break; + } + entry = entry->pl_next; + } + if (found) { + free(propname); + continue; + } + + entry = zfs_alloc(hdl, sizeof (zprop_list_t)); + entry->pl_prop = ZPROP_INVAL; + entry->pl_user_prop = propname; + entry->pl_width = strlen(entry->pl_user_prop); + entry->pl_all = B_TRUE; + + *last = entry; + last = &entry->pl_next; + } + + for (entry = *plp; entry != NULL; entry = entry->pl_next) { + + if (entry->pl_fixed) + continue; + + if (entry->pl_prop != ZPROP_INVAL && + zpool_get_prop(zhp, entry->pl_prop, buf, sizeof (buf), + NULL, B_FALSE) == 0) { + if (strlen(buf) > entry->pl_width) + entry->pl_width = strlen(buf); + } + } + + return (0); +} + +/* + * Get the state for the given feature on the given ZFS pool. + */ +int +zpool_prop_get_feature(zpool_handle_t *zhp, const char *propname, char *buf, + size_t len) +{ + uint64_t refcount; + boolean_t found = B_FALSE; + nvlist_t *features = zpool_get_features(zhp); + boolean_t supported; + const char *feature = strchr(propname, '@') + 1; + + supported = zpool_prop_feature(propname); + ASSERT(supported || zpool_prop_unsupported(propname)); + + /* + * Convert from feature name to feature guid. This conversion is + * unecessary for unsupported@... properties because they already + * use guids. + */ + if (supported) { + int ret; + spa_feature_t fid; + + ret = zfeature_lookup_name(feature, &fid); + if (ret != 0) { + (void) strlcpy(buf, "-", len); + return (ENOTSUP); + } + feature = spa_feature_table[fid].fi_guid; + } + + if (nvlist_lookup_uint64(features, feature, &refcount) == 0) + found = B_TRUE; + + if (supported) { + if (!found) { + (void) strlcpy(buf, ZFS_FEATURE_DISABLED, len); + } else { + if (refcount == 0) + (void) strlcpy(buf, ZFS_FEATURE_ENABLED, len); + else + (void) strlcpy(buf, ZFS_FEATURE_ACTIVE, len); + } + } else { + if (found) { + if (refcount == 0) { + (void) strcpy(buf, ZFS_UNSUPPORTED_INACTIVE); + } else { + (void) strcpy(buf, ZFS_UNSUPPORTED_READONLY); + } + } else { + (void) strlcpy(buf, "-", len); + return (ENOTSUP); + } + } + + return (0); +} + +/* + * Don't start the slice at the default block of 34; many storage + * devices will use a stripe width of 128k, so start there instead. + */ +#define NEW_START_BLOCK 256 + +/* + * Validate the given pool name, optionally putting an extended error message in + * 'buf'. + */ +boolean_t +zpool_name_valid(libzfs_handle_t *hdl, boolean_t isopen, const char *pool) +{ + namecheck_err_t why; + char what; + int ret; + + ret = pool_namecheck(pool, &why, &what); + + /* + * The rules for reserved pool names were extended at a later point. + * But we need to support users with existing pools that may now be + * invalid. So we only check for this expanded set of names during a + * create (or import), and only in userland. + */ + if (ret == 0 && !isopen && + (strncmp(pool, "mirror", 6) == 0 || + strncmp(pool, "raidz", 5) == 0 || + strncmp(pool, "spare", 5) == 0 || + strcmp(pool, "log") == 0)) { + if (hdl != NULL) + zfs_error_aux(hdl, + dgettext(TEXT_DOMAIN, "name is reserved")); + return (B_FALSE); + } + + + if (ret != 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_INVALCHAR: + zfs_error_aux(hdl, + dgettext(TEXT_DOMAIN, "invalid character " + "'%c' in pool name"), what); + break; + + case NAME_ERR_NOLETTER: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "name must 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, + "pool name is reserved")); + 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_MULTIPLE_AT: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "multiple '@' delimiters in name")); + break; + + } + } + return (B_FALSE); + } + + return (B_TRUE); +} + +/* + * Open a handle to the given pool, even if the pool is currently in the FAULTED + * state. + */ +zpool_handle_t * +zpool_open_canfail(libzfs_handle_t *hdl, const char *pool) +{ + zpool_handle_t *zhp; + boolean_t missing; + + /* + * Make sure the pool name is valid. + */ + if (!zpool_name_valid(hdl, B_TRUE, pool)) { + (void) zfs_error_fmt(hdl, EZFS_INVALIDNAME, + dgettext(TEXT_DOMAIN, "cannot open '%s'"), + pool); + return (NULL); + } + + if ((zhp = zfs_alloc(hdl, sizeof (zpool_handle_t))) == NULL) + return (NULL); + + zhp->zpool_hdl = hdl; + (void) strlcpy(zhp->zpool_name, pool, sizeof (zhp->zpool_name)); + + if (zpool_refresh_stats(zhp, &missing) != 0) { + zpool_close(zhp); + return (NULL); + } + + if (missing) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "no such pool")); + (void) zfs_error_fmt(hdl, EZFS_NOENT, + dgettext(TEXT_DOMAIN, "cannot open '%s'"), pool); + zpool_close(zhp); + return (NULL); + } + + return (zhp); +} + +/* + * Like the above, but silent on error. Used when iterating over pools (because + * the configuration cache may be out of date). + */ +int +zpool_open_silent(libzfs_handle_t *hdl, const char *pool, zpool_handle_t **ret) +{ + zpool_handle_t *zhp; + boolean_t missing; + + if ((zhp = zfs_alloc(hdl, sizeof (zpool_handle_t))) == NULL) + return (-1); + + zhp->zpool_hdl = hdl; + (void) strlcpy(zhp->zpool_name, pool, sizeof (zhp->zpool_name)); + + if (zpool_refresh_stats(zhp, &missing) != 0) { + zpool_close(zhp); + return (-1); + } + + if (missing) { + zpool_close(zhp); + *ret = NULL; + return (0); + } + + *ret = zhp; + return (0); +} + +/* + * Similar to zpool_open_canfail(), but refuses to open pools in the faulted + * state. + */ +zpool_handle_t * +zpool_open(libzfs_handle_t *hdl, const char *pool) +{ + zpool_handle_t *zhp; + + if ((zhp = zpool_open_canfail(hdl, pool)) == NULL) + return (NULL); + + if (zhp->zpool_state == POOL_STATE_UNAVAIL) { + (void) zfs_error_fmt(hdl, EZFS_POOLUNAVAIL, + dgettext(TEXT_DOMAIN, "cannot open '%s'"), zhp->zpool_name); + zpool_close(zhp); + return (NULL); + } + + return (zhp); +} + +/* + * Close the handle. Simply frees the memory associated with the handle. + */ +void +zpool_close(zpool_handle_t *zhp) +{ + if (zhp->zpool_config) + nvlist_free(zhp->zpool_config); + if (zhp->zpool_old_config) + nvlist_free(zhp->zpool_old_config); + if (zhp->zpool_props) + nvlist_free(zhp->zpool_props); + free(zhp); +} + +/* + * Return the name of the pool. + */ +const char * +zpool_get_name(zpool_handle_t *zhp) +{ + return (zhp->zpool_name); +} + + +/* + * Return the state of the pool (ACTIVE or UNAVAILABLE) + */ +int +zpool_get_state(zpool_handle_t *zhp) +{ + return (zhp->zpool_state); +} + +/* + * Create the named pool, using the provided vdev list. It is assumed + * that the consumer has already validated the contents of the nvlist, so we + * don't have to worry about error semantics. + */ +int +zpool_create(libzfs_handle_t *hdl, const char *pool, nvlist_t *nvroot, + nvlist_t *props, nvlist_t *fsprops) +{ + zfs_cmd_t zc = { 0 }; + nvlist_t *zc_fsprops = NULL; + nvlist_t *zc_props = NULL; + char msg[1024]; + int ret = -1; + + (void) snprintf(msg, sizeof (msg), dgettext(TEXT_DOMAIN, + "cannot create '%s'"), pool); + + if (!zpool_name_valid(hdl, B_FALSE, pool)) + return (zfs_error(hdl, EZFS_INVALIDNAME, msg)); + + if (zcmd_write_conf_nvlist(hdl, &zc, nvroot) != 0) + return (-1); + + if (props) { + prop_flags_t flags = { .create = B_TRUE, .import = B_FALSE }; + + if ((zc_props = zpool_valid_proplist(hdl, pool, props, + SPA_VERSION_1, flags, msg)) == NULL) { + goto create_failed; + } + } + + if (fsprops) { + uint64_t zoned; + char *zonestr; + + zoned = ((nvlist_lookup_string(fsprops, + zfs_prop_to_name(ZFS_PROP_ZONED), &zonestr) == 0) && + strcmp(zonestr, "on") == 0); + + if ((zc_fsprops = zfs_valid_proplist(hdl, + ZFS_TYPE_FILESYSTEM, fsprops, zoned, NULL, msg)) == NULL) { + goto create_failed; + } + if (!zc_props && + (nvlist_alloc(&zc_props, NV_UNIQUE_NAME, 0) != 0)) { + goto create_failed; + } + if (nvlist_add_nvlist(zc_props, + ZPOOL_ROOTFS_PROPS, zc_fsprops) != 0) { + goto create_failed; + } + } + + if (zc_props && zcmd_write_src_nvlist(hdl, &zc, zc_props) != 0) + goto create_failed; + + (void) strlcpy(zc.zc_name, pool, sizeof (zc.zc_name)); + + if ((ret = zfs_ioctl(hdl, ZFS_IOC_POOL_CREATE, &zc)) != 0) { + + zcmd_free_nvlists(&zc); + nvlist_free(zc_props); + nvlist_free(zc_fsprops); + + switch (errno) { + case EBUSY: + /* + * This can happen if the user has specified the same + * device multiple times. We can't reliably detect this + * until we try to add it and see we already have a + * label. + */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "one or more vdevs refer to the same device")); + return (zfs_error(hdl, EZFS_BADDEV, msg)); + + case EOVERFLOW: + /* + * This occurs when one of the devices is below + * SPA_MINDEVSIZE. Unfortunately, we can't detect which + * device was the problem device since there's no + * reliable way to determine device size from userland. + */ + { + char buf[64]; + + zfs_nicenum(SPA_MINDEVSIZE, buf, sizeof (buf)); + + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "one or more devices is less than the " + "minimum size (%s)"), buf); + } + return (zfs_error(hdl, EZFS_BADDEV, msg)); + + case ENOSPC: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "one or more devices is out of space")); + return (zfs_error(hdl, EZFS_BADDEV, msg)); + + case ENOTBLK: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "cache device must be a disk or disk slice")); + return (zfs_error(hdl, EZFS_BADDEV, msg)); + + default: + return (zpool_standard_error(hdl, errno, msg)); + } + } + +create_failed: + zcmd_free_nvlists(&zc); + nvlist_free(zc_props); + nvlist_free(zc_fsprops); + return (ret); +} + +/* + * Destroy the given pool. It is up to the caller to ensure that there are no + * datasets left in the pool. + */ +int +zpool_destroy(zpool_handle_t *zhp, const char *log_str) +{ + zfs_cmd_t zc = { 0 }; + zfs_handle_t *zfp = NULL; + libzfs_handle_t *hdl = zhp->zpool_hdl; + char msg[1024]; + + if (zhp->zpool_state == POOL_STATE_ACTIVE && + (zfp = zfs_open(hdl, zhp->zpool_name, ZFS_TYPE_FILESYSTEM)) == NULL) + return (-1); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + zc.zc_history = (uint64_t)(uintptr_t)log_str; + + if (zfs_ioctl(hdl, ZFS_IOC_POOL_DESTROY, &zc) != 0) { + (void) snprintf(msg, sizeof (msg), dgettext(TEXT_DOMAIN, + "cannot destroy '%s'"), zhp->zpool_name); + + if (errno == EROFS) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "one or more devices is read only")); + (void) zfs_error(hdl, EZFS_BADDEV, msg); + } else { + (void) zpool_standard_error(hdl, errno, msg); + } + + if (zfp) + zfs_close(zfp); + return (-1); + } + + if (zfp) { + remove_mountpoint(zfp); + zfs_close(zfp); + } + + return (0); +} + +/* + * Add the given vdevs to the pool. The caller must have already performed the + * necessary verification to ensure that the vdev specification is well-formed. + */ +int +zpool_add(zpool_handle_t *zhp, nvlist_t *nvroot) +{ + zfs_cmd_t zc = { 0 }; + int ret; + libzfs_handle_t *hdl = zhp->zpool_hdl; + char msg[1024]; + nvlist_t **spares, **l2cache; + uint_t nspares, nl2cache; + + (void) snprintf(msg, sizeof (msg), dgettext(TEXT_DOMAIN, + "cannot add to '%s'"), zhp->zpool_name); + + if (zpool_get_prop_int(zhp, ZPOOL_PROP_VERSION, NULL) < + SPA_VERSION_SPARES && + nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_SPARES, + &spares, &nspares) == 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be " + "upgraded to add hot spares")); + return (zfs_error(hdl, EZFS_BADVERSION, msg)); + } + + if (zpool_is_bootable(zhp) && nvlist_lookup_nvlist_array(nvroot, + ZPOOL_CONFIG_SPARES, &spares, &nspares) == 0) { + uint64_t s; + + for (s = 0; s < nspares; s++) { + char *path; + + if (nvlist_lookup_string(spares[s], ZPOOL_CONFIG_PATH, + &path) == 0 && pool_uses_efi(spares[s])) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "device '%s' contains an EFI label and " + "cannot be used on root pools."), + zpool_vdev_name(hdl, NULL, spares[s], + B_FALSE)); + return (zfs_error(hdl, EZFS_POOL_NOTSUP, msg)); + } + } + } + + if (zpool_get_prop_int(zhp, ZPOOL_PROP_VERSION, NULL) < + SPA_VERSION_L2CACHE && + nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_L2CACHE, + &l2cache, &nl2cache) == 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be " + "upgraded to add cache devices")); + return (zfs_error(hdl, EZFS_BADVERSION, msg)); + } + + if (zcmd_write_conf_nvlist(hdl, &zc, nvroot) != 0) + return (-1); + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + + if (zfs_ioctl(hdl, ZFS_IOC_VDEV_ADD, &zc) != 0) { + switch (errno) { + case EBUSY: + /* + * This can happen if the user has specified the same + * device multiple times. We can't reliably detect this + * until we try to add it and see we already have a + * label. + */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "one or more vdevs refer to the same device")); + (void) zfs_error(hdl, EZFS_BADDEV, msg); + break; + + case EOVERFLOW: + /* + * This occurrs when one of the devices is below + * SPA_MINDEVSIZE. Unfortunately, we can't detect which + * device was the problem device since there's no + * reliable way to determine device size from userland. + */ + { + char buf[64]; + + zfs_nicenum(SPA_MINDEVSIZE, buf, sizeof (buf)); + + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "device is less than the minimum " + "size (%s)"), buf); + } + (void) zfs_error(hdl, EZFS_BADDEV, msg); + break; + + case ENOTSUP: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "pool must be upgraded to add these vdevs")); + (void) zfs_error(hdl, EZFS_BADVERSION, msg); + break; + + case EDOM: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "root pool can not have multiple vdevs" + " or separate logs")); + (void) zfs_error(hdl, EZFS_POOL_NOTSUP, msg); + break; + + case ENOTBLK: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "cache device must be a disk or disk slice")); + (void) zfs_error(hdl, EZFS_BADDEV, msg); + break; + + default: + (void) zpool_standard_error(hdl, errno, msg); + } + + ret = -1; + } else { + ret = 0; + } + + zcmd_free_nvlists(&zc); + + return (ret); +} + +/* + * Exports the pool from the system. The caller must ensure that there are no + * mounted datasets in the pool. + */ +static int +zpool_export_common(zpool_handle_t *zhp, boolean_t force, boolean_t hardforce, + const char *log_str) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + + (void) snprintf(msg, sizeof (msg), dgettext(TEXT_DOMAIN, + "cannot export '%s'"), zhp->zpool_name); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + zc.zc_cookie = force; + zc.zc_guid = hardforce; + zc.zc_history = (uint64_t)(uintptr_t)log_str; + + if (zfs_ioctl(zhp->zpool_hdl, ZFS_IOC_POOL_EXPORT, &zc) != 0) { + switch (errno) { + case EXDEV: + zfs_error_aux(zhp->zpool_hdl, dgettext(TEXT_DOMAIN, + "use '-f' to override the following errors:\n" + "'%s' has an active shared spare which could be" + " used by other pools once '%s' is exported."), + zhp->zpool_name, zhp->zpool_name); + return (zfs_error(zhp->zpool_hdl, EZFS_ACTIVE_SPARE, + msg)); + default: + return (zpool_standard_error_fmt(zhp->zpool_hdl, errno, + msg)); + } + } + + return (0); +} + +int +zpool_export(zpool_handle_t *zhp, boolean_t force, const char *log_str) +{ + return (zpool_export_common(zhp, force, B_FALSE, log_str)); +} + +int +zpool_export_force(zpool_handle_t *zhp, const char *log_str) +{ + return (zpool_export_common(zhp, B_TRUE, B_TRUE, log_str)); +} + +static void +zpool_rewind_exclaim(libzfs_handle_t *hdl, const char *name, boolean_t dryrun, + nvlist_t *config) +{ + nvlist_t *nv = NULL; + uint64_t rewindto; + int64_t loss = -1; + struct tm t; + char timestr[128]; + + if (!hdl->libzfs_printerr || config == NULL) + return; + + if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO, &nv) != 0 || + nvlist_lookup_nvlist(nv, ZPOOL_CONFIG_REWIND_INFO, &nv) != 0) { + return; + } + + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_LOAD_TIME, &rewindto) != 0) + return; + (void) nvlist_lookup_int64(nv, ZPOOL_CONFIG_REWIND_TIME, &loss); + + if (localtime_r((time_t *)&rewindto, &t) != NULL && + strftime(timestr, 128, 0, &t) != 0) { + if (dryrun) { + (void) printf(dgettext(TEXT_DOMAIN, + "Would be able to return %s " + "to its state as of %s.\n"), + name, timestr); + } else { + (void) printf(dgettext(TEXT_DOMAIN, + "Pool %s returned to its state as of %s.\n"), + name, timestr); + } + if (loss > 120) { + (void) printf(dgettext(TEXT_DOMAIN, + "%s approximately %lld "), + dryrun ? "Would discard" : "Discarded", + (loss + 30) / 60); + (void) printf(dgettext(TEXT_DOMAIN, + "minutes of transactions.\n")); + } else if (loss > 0) { + (void) printf(dgettext(TEXT_DOMAIN, + "%s approximately %lld "), + dryrun ? "Would discard" : "Discarded", loss); + (void) printf(dgettext(TEXT_DOMAIN, + "seconds of transactions.\n")); + } + } +} + +void +zpool_explain_recover(libzfs_handle_t *hdl, const char *name, int reason, + nvlist_t *config) +{ + nvlist_t *nv = NULL; + int64_t loss = -1; + uint64_t edata = UINT64_MAX; + uint64_t rewindto; + struct tm t; + char timestr[128]; + + if (!hdl->libzfs_printerr) + return; + + if (reason >= 0) + (void) printf(dgettext(TEXT_DOMAIN, "action: ")); + else + (void) printf(dgettext(TEXT_DOMAIN, "\t")); + + /* All attempted rewinds failed if ZPOOL_CONFIG_LOAD_TIME missing */ + if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO, &nv) != 0 || + nvlist_lookup_nvlist(nv, ZPOOL_CONFIG_REWIND_INFO, &nv) != 0 || + nvlist_lookup_uint64(nv, ZPOOL_CONFIG_LOAD_TIME, &rewindto) != 0) + goto no_info; + + (void) nvlist_lookup_int64(nv, ZPOOL_CONFIG_REWIND_TIME, &loss); + (void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_LOAD_DATA_ERRORS, + &edata); + + (void) printf(dgettext(TEXT_DOMAIN, + "Recovery is possible, but will result in some data loss.\n")); + + if (localtime_r((time_t *)&rewindto, &t) != NULL && + strftime(timestr, 128, 0, &t) != 0) { + (void) printf(dgettext(TEXT_DOMAIN, + "\tReturning the pool to its state as of %s\n" + "\tshould correct the problem. "), + timestr); + } else { + (void) printf(dgettext(TEXT_DOMAIN, + "\tReverting the pool to an earlier state " + "should correct the problem.\n\t")); + } + + if (loss > 120) { + (void) printf(dgettext(TEXT_DOMAIN, + "Approximately %lld minutes of data\n" + "\tmust be discarded, irreversibly. "), (loss + 30) / 60); + } else if (loss > 0) { + (void) printf(dgettext(TEXT_DOMAIN, + "Approximately %lld seconds of data\n" + "\tmust be discarded, irreversibly. "), loss); + } + if (edata != 0 && edata != UINT64_MAX) { + if (edata == 1) { + (void) printf(dgettext(TEXT_DOMAIN, + "After rewind, at least\n" + "\tone persistent user-data error will remain. ")); + } else { + (void) printf(dgettext(TEXT_DOMAIN, + "After rewind, several\n" + "\tpersistent user-data errors will remain. ")); + } + } + (void) printf(dgettext(TEXT_DOMAIN, + "Recovery can be attempted\n\tby executing 'zpool %s -F %s'. "), + reason >= 0 ? "clear" : "import", name); + + (void) printf(dgettext(TEXT_DOMAIN, + "A scrub of the pool\n" + "\tis strongly recommended after recovery.\n")); + return; + +no_info: + (void) printf(dgettext(TEXT_DOMAIN, + "Destroy and re-create the pool from\n\ta backup source.\n")); +} + +/* + * zpool_import() is a contracted interface. Should be kept the same + * if possible. + * + * Applications should use zpool_import_props() to import a pool with + * new properties value to be set. + */ +int +zpool_import(libzfs_handle_t *hdl, nvlist_t *config, const char *newname, + char *altroot) +{ + nvlist_t *props = NULL; + int ret; + + if (altroot != NULL) { + if (nvlist_alloc(&props, NV_UNIQUE_NAME, 0) != 0) { + return (zfs_error_fmt(hdl, EZFS_NOMEM, + dgettext(TEXT_DOMAIN, "cannot import '%s'"), + newname)); + } + + if (nvlist_add_string(props, + zpool_prop_to_name(ZPOOL_PROP_ALTROOT), altroot) != 0 || + nvlist_add_string(props, + zpool_prop_to_name(ZPOOL_PROP_CACHEFILE), "none") != 0) { + nvlist_free(props); + return (zfs_error_fmt(hdl, EZFS_NOMEM, + dgettext(TEXT_DOMAIN, "cannot import '%s'"), + newname)); + } + } + + ret = zpool_import_props(hdl, config, newname, props, + ZFS_IMPORT_NORMAL); + if (props) + nvlist_free(props); + return (ret); +} + +static void +print_vdev_tree(libzfs_handle_t *hdl, const char *name, nvlist_t *nv, + int indent) +{ + nvlist_t **child; + uint_t c, children; + char *vname; + uint64_t is_log = 0; + + (void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_IS_LOG, + &is_log); + + if (name != NULL) + (void) printf("\t%*s%s%s\n", indent, "", name, + is_log ? " [log]" : ""); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) + return; + + for (c = 0; c < children; c++) { + vname = zpool_vdev_name(hdl, NULL, child[c], B_TRUE); + print_vdev_tree(hdl, vname, child[c], indent + 2); + free(vname); + } +} + +void +zpool_print_unsup_feat(nvlist_t *config) +{ + nvlist_t *nvinfo, *unsup_feat; + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO, &nvinfo) == + 0); + verify(nvlist_lookup_nvlist(nvinfo, ZPOOL_CONFIG_UNSUP_FEAT, + &unsup_feat) == 0); + + for (nvpair_t *nvp = nvlist_next_nvpair(unsup_feat, NULL); nvp != NULL; + nvp = nvlist_next_nvpair(unsup_feat, nvp)) { + char *desc; + + verify(nvpair_type(nvp) == DATA_TYPE_STRING); + verify(nvpair_value_string(nvp, &desc) == 0); + + if (strlen(desc) > 0) + (void) printf("\t%s (%s)\n", nvpair_name(nvp), desc); + else + (void) printf("\t%s\n", nvpair_name(nvp)); + } +} + +/* + * Import the given pool using the known configuration and a list of + * properties to be set. The configuration should have come from + * zpool_find_import(). The 'newname' parameters control whether the pool + * is imported with a different name. + */ +int +zpool_import_props(libzfs_handle_t *hdl, nvlist_t *config, const char *newname, + nvlist_t *props, int flags) +{ + zfs_cmd_t zc = { 0 }; + zpool_rewind_policy_t policy; + nvlist_t *nv = NULL; + nvlist_t *nvinfo = NULL; + nvlist_t *missing = NULL; + char *thename; + char *origname; + int ret; + int error = 0; + char errbuf[1024]; + + verify(nvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME, + &origname) == 0); + + (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, + "cannot import pool '%s'"), origname); + + if (newname != NULL) { + if (!zpool_name_valid(hdl, B_FALSE, newname)) + return (zfs_error_fmt(hdl, EZFS_INVALIDNAME, + dgettext(TEXT_DOMAIN, "cannot import '%s'"), + newname)); + thename = (char *)newname; + } else { + thename = origname; + } + + if (props) { + uint64_t version; + prop_flags_t flags = { .create = B_FALSE, .import = B_TRUE }; + + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION, + &version) == 0); + + if ((props = zpool_valid_proplist(hdl, origname, + props, version, flags, errbuf)) == NULL) { + return (-1); + } else if (zcmd_write_src_nvlist(hdl, &zc, props) != 0) { + nvlist_free(props); + return (-1); + } + } + + (void) strlcpy(zc.zc_name, thename, sizeof (zc.zc_name)); + + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, + &zc.zc_guid) == 0); + + if (zcmd_write_conf_nvlist(hdl, &zc, config) != 0) { + nvlist_free(props); + return (-1); + } + if (zcmd_alloc_dst_nvlist(hdl, &zc, zc.zc_nvlist_conf_size * 2) != 0) { + nvlist_free(props); + return (-1); + } + + zc.zc_cookie = flags; + while ((ret = zfs_ioctl(hdl, ZFS_IOC_POOL_IMPORT, &zc)) != 0 && + errno == ENOMEM) { + if (zcmd_expand_dst_nvlist(hdl, &zc) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + } + if (ret != 0) + error = errno; + + (void) zcmd_read_dst_nvlist(hdl, &zc, &nv); + zpool_get_rewind_policy(config, &policy); + + if (error) { + char desc[1024]; + + /* + * Dry-run failed, but we print out what success + * looks like if we found a best txg + */ + if (policy.zrp_request & ZPOOL_TRY_REWIND) { + zpool_rewind_exclaim(hdl, newname ? origname : thename, + B_TRUE, nv); + nvlist_free(nv); + return (-1); + } + + if (newname == NULL) + (void) snprintf(desc, sizeof (desc), + dgettext(TEXT_DOMAIN, "cannot import '%s'"), + thename); + else + (void) snprintf(desc, sizeof (desc), + dgettext(TEXT_DOMAIN, "cannot import '%s' as '%s'"), + origname, thename); + + switch (error) { + case ENOTSUP: + if (nv != NULL && nvlist_lookup_nvlist(nv, + ZPOOL_CONFIG_LOAD_INFO, &nvinfo) == 0 && + nvlist_exists(nvinfo, ZPOOL_CONFIG_UNSUP_FEAT)) { + (void) printf(dgettext(TEXT_DOMAIN, "This " + "pool uses the following feature(s) not " + "supported by this system:\n")); + zpool_print_unsup_feat(nv); + if (nvlist_exists(nvinfo, + ZPOOL_CONFIG_CAN_RDONLY)) { + (void) printf(dgettext(TEXT_DOMAIN, + "All unsupported features are only " + "required for writing to the pool." + "\nThe pool can be imported using " + "'-o readonly=on'.\n")); + } + } + /* + * Unsupported version. + */ + (void) zfs_error(hdl, EZFS_BADVERSION, desc); + break; + + case EINVAL: + (void) zfs_error(hdl, EZFS_INVALCONFIG, desc); + break; + + case EROFS: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "one or more devices is read only")); + (void) zfs_error(hdl, EZFS_BADDEV, desc); + break; + + case ENXIO: + if (nv && nvlist_lookup_nvlist(nv, + ZPOOL_CONFIG_LOAD_INFO, &nvinfo) == 0 && + nvlist_lookup_nvlist(nvinfo, + ZPOOL_CONFIG_MISSING_DEVICES, &missing) == 0) { + (void) printf(dgettext(TEXT_DOMAIN, + "The devices below are missing, use " + "'-m' to import the pool anyway:\n")); + print_vdev_tree(hdl, NULL, missing, 2); + (void) printf("\n"); + } + (void) zpool_standard_error(hdl, error, desc); + break; + + case EEXIST: + (void) zpool_standard_error(hdl, error, desc); + break; + + default: + (void) zpool_standard_error(hdl, error, desc); + zpool_explain_recover(hdl, + newname ? origname : thename, -error, nv); + break; + } + + nvlist_free(nv); + ret = -1; + } else { + zpool_handle_t *zhp; + + /* + * This should never fail, but play it safe anyway. + */ + if (zpool_open_silent(hdl, thename, &zhp) != 0) + ret = -1; + else if (zhp != NULL) + zpool_close(zhp); + if (policy.zrp_request & + (ZPOOL_DO_REWIND | ZPOOL_TRY_REWIND)) { + zpool_rewind_exclaim(hdl, newname ? origname : thename, + ((policy.zrp_request & ZPOOL_TRY_REWIND) != 0), nv); + } + nvlist_free(nv); + return (0); + } + + zcmd_free_nvlists(&zc); + nvlist_free(props); + + return (ret); +} + +/* + * Scan the pool. + */ +int +zpool_scan(zpool_handle_t *zhp, pool_scan_func_t func) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + zc.zc_cookie = func; + + if (zfs_ioctl(hdl, ZFS_IOC_POOL_SCAN, &zc) == 0 || + (errno == ENOENT && func != POOL_SCAN_NONE)) + return (0); + + if (func == POOL_SCAN_SCRUB) { + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot scrub %s"), zc.zc_name); + } else if (func == POOL_SCAN_NONE) { + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot cancel scrubbing %s"), + zc.zc_name); + } else { + assert(!"unexpected result"); + } + + if (errno == EBUSY) { + nvlist_t *nvroot; + pool_scan_stat_t *ps = NULL; + uint_t psc; + + verify(nvlist_lookup_nvlist(zhp->zpool_config, + ZPOOL_CONFIG_VDEV_TREE, &nvroot) == 0); + (void) nvlist_lookup_uint64_array(nvroot, + ZPOOL_CONFIG_SCAN_STATS, (uint64_t **)&ps, &psc); + if (ps && ps->pss_func == POOL_SCAN_SCRUB) + return (zfs_error(hdl, EZFS_SCRUBBING, msg)); + else + return (zfs_error(hdl, EZFS_RESILVERING, msg)); + } else if (errno == ENOENT) { + return (zfs_error(hdl, EZFS_NO_SCRUB, msg)); + } else { + return (zpool_standard_error(hdl, errno, msg)); + } +} + +/* + * This provides a very minimal check whether a given string is likely a + * c#t#d# style string. Users of this are expected to do their own + * verification of the s# part. + */ +#define CTD_CHECK(str) (str && str[0] == 'c' && isdigit(str[1])) + +/* + * More elaborate version for ones which may start with "/dev/dsk/" + * and the like. + */ +static int +ctd_check_path(char *str) { + /* + * If it starts with a slash, check the last component. + */ + if (str && str[0] == '/') { + char *tmp = strrchr(str, '/'); + + /* + * If it ends in "/old", check the second-to-last + * component of the string instead. + */ + if (tmp != str && strcmp(tmp, "/old") == 0) { + for (tmp--; *tmp != '/'; tmp--) + ; + } + str = tmp + 1; + } + return (CTD_CHECK(str)); +} + +/* + * Find a vdev that matches the search criteria specified. We use the + * the nvpair name to determine how we should look for the device. + * 'avail_spare' is set to TRUE if the provided guid refers to an AVAIL + * spare; but FALSE if its an INUSE spare. + */ +static nvlist_t * +vdev_to_nvlist_iter(nvlist_t *nv, nvlist_t *search, boolean_t *avail_spare, + boolean_t *l2cache, boolean_t *log) +{ + uint_t c, children; + nvlist_t **child; + nvlist_t *ret; + uint64_t is_log; + char *srchkey; + nvpair_t *pair = nvlist_next_nvpair(search, NULL); + + /* Nothing to look for */ + if (search == NULL || pair == NULL) + return (NULL); + + /* Obtain the key we will use to search */ + srchkey = nvpair_name(pair); + + switch (nvpair_type(pair)) { + case DATA_TYPE_UINT64: + if (strcmp(srchkey, ZPOOL_CONFIG_GUID) == 0) { + uint64_t srchval, theguid; + + verify(nvpair_value_uint64(pair, &srchval) == 0); + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, + &theguid) == 0); + if (theguid == srchval) + return (nv); + } + break; + + case DATA_TYPE_STRING: { + char *srchval, *val; + + verify(nvpair_value_string(pair, &srchval) == 0); + if (nvlist_lookup_string(nv, srchkey, &val) != 0) + break; + + /* + * Search for the requested value. Special cases: + * + * - ZPOOL_CONFIG_PATH for whole disk entries. These end in + * "s0" or "s0/old". The "s0" part is hidden from the user, + * but included in the string, so this matches around it. + * - looking for a top-level vdev name (i.e. ZPOOL_CONFIG_TYPE). + * + * Otherwise, all other searches are simple string compares. + */ + if (strcmp(srchkey, ZPOOL_CONFIG_PATH) == 0 && + ctd_check_path(val)) { + uint64_t wholedisk = 0; + + (void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_WHOLE_DISK, + &wholedisk); + if (wholedisk) { + int slen = strlen(srchval); + int vlen = strlen(val); + + if (slen != vlen - 2) + break; + + /* + * make_leaf_vdev() should only set + * wholedisk for ZPOOL_CONFIG_PATHs which + * will include "/dev/dsk/", giving plenty of + * room for the indices used next. + */ + ASSERT(vlen >= 6); + + /* + * strings identical except trailing "s0" + */ + if (strcmp(&val[vlen - 2], "s0") == 0 && + strncmp(srchval, val, slen) == 0) + return (nv); + + /* + * strings identical except trailing "s0/old" + */ + if (strcmp(&val[vlen - 6], "s0/old") == 0 && + strcmp(&srchval[slen - 4], "/old") == 0 && + strncmp(srchval, val, slen - 4) == 0) + return (nv); + + break; + } + } else if (strcmp(srchkey, ZPOOL_CONFIG_TYPE) == 0 && val) { + char *type, *idx, *end, *p; + uint64_t id, vdev_id; + + /* + * Determine our vdev type, keeping in mind + * that the srchval is composed of a type and + * vdev id pair (i.e. mirror-4). + */ + if ((type = strdup(srchval)) == NULL) + return (NULL); + + if ((p = strrchr(type, '-')) == NULL) { + free(type); + break; + } + idx = p + 1; + *p = '\0'; + + /* + * If the types don't match then keep looking. + */ + if (strncmp(val, type, strlen(val)) != 0) { + free(type); + break; + } + + verify(strncmp(type, VDEV_TYPE_RAIDZ, + strlen(VDEV_TYPE_RAIDZ)) == 0 || + strncmp(type, VDEV_TYPE_MIRROR, + strlen(VDEV_TYPE_MIRROR)) == 0); + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_ID, + &id) == 0); + + errno = 0; + vdev_id = strtoull(idx, &end, 10); + + free(type); + if (errno != 0) + return (NULL); + + /* + * Now verify that we have the correct vdev id. + */ + if (vdev_id == id) + return (nv); + } + + /* + * Common case + */ + if (strcmp(srchval, val) == 0) + return (nv); + break; + } + + default: + break; + } + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) + return (NULL); + + for (c = 0; c < children; c++) { + if ((ret = vdev_to_nvlist_iter(child[c], search, + avail_spare, l2cache, NULL)) != NULL) { + /* + * The 'is_log' value is only set for the toplevel + * vdev, not the leaf vdevs. So we always lookup the + * log device from the root of the vdev tree (where + * 'log' is non-NULL). + */ + if (log != NULL && + nvlist_lookup_uint64(child[c], + ZPOOL_CONFIG_IS_LOG, &is_log) == 0 && + is_log) { + *log = B_TRUE; + } + return (ret); + } + } + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_SPARES, + &child, &children) == 0) { + for (c = 0; c < children; c++) { + if ((ret = vdev_to_nvlist_iter(child[c], search, + avail_spare, l2cache, NULL)) != NULL) { + *avail_spare = B_TRUE; + return (ret); + } + } + } + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_L2CACHE, + &child, &children) == 0) { + for (c = 0; c < children; c++) { + if ((ret = vdev_to_nvlist_iter(child[c], search, + avail_spare, l2cache, NULL)) != NULL) { + *l2cache = B_TRUE; + return (ret); + } + } + } + + return (NULL); +} + +/* + * Given a physical path (minus the "/devices" prefix), find the + * associated vdev. + */ +nvlist_t * +zpool_find_vdev_by_physpath(zpool_handle_t *zhp, const char *ppath, + boolean_t *avail_spare, boolean_t *l2cache, boolean_t *log) +{ + nvlist_t *search, *nvroot, *ret; + + verify(nvlist_alloc(&search, NV_UNIQUE_NAME, KM_SLEEP) == 0); + verify(nvlist_add_string(search, ZPOOL_CONFIG_PHYS_PATH, ppath) == 0); + + verify(nvlist_lookup_nvlist(zhp->zpool_config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + + *avail_spare = B_FALSE; + *l2cache = B_FALSE; + if (log != NULL) + *log = B_FALSE; + ret = vdev_to_nvlist_iter(nvroot, search, avail_spare, l2cache, log); + nvlist_free(search); + + return (ret); +} + +/* + * Determine if we have an "interior" top-level vdev (i.e mirror/raidz). + */ +boolean_t +zpool_vdev_is_interior(const char *name) +{ + if (strncmp(name, VDEV_TYPE_RAIDZ, strlen(VDEV_TYPE_RAIDZ)) == 0 || + strncmp(name, VDEV_TYPE_MIRROR, strlen(VDEV_TYPE_MIRROR)) == 0) + return (B_TRUE); + return (B_FALSE); +} + +nvlist_t * +zpool_find_vdev(zpool_handle_t *zhp, const char *path, boolean_t *avail_spare, + boolean_t *l2cache, boolean_t *log) +{ + char buf[MAXPATHLEN]; + char *end; + nvlist_t *nvroot, *search, *ret; + uint64_t guid; + + verify(nvlist_alloc(&search, NV_UNIQUE_NAME, KM_SLEEP) == 0); + + guid = strtoull(path, &end, 10); + if (guid != 0 && *end == '\0') { + verify(nvlist_add_uint64(search, ZPOOL_CONFIG_GUID, guid) == 0); + } else if (zpool_vdev_is_interior(path)) { + verify(nvlist_add_string(search, ZPOOL_CONFIG_TYPE, path) == 0); + } else if (path[0] != '/') { + (void) snprintf(buf, sizeof (buf), "%s%s", _PATH_DEV, path); + verify(nvlist_add_string(search, ZPOOL_CONFIG_PATH, buf) == 0); + } else { + verify(nvlist_add_string(search, ZPOOL_CONFIG_PATH, path) == 0); + } + + verify(nvlist_lookup_nvlist(zhp->zpool_config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + + *avail_spare = B_FALSE; + *l2cache = B_FALSE; + if (log != NULL) + *log = B_FALSE; + ret = vdev_to_nvlist_iter(nvroot, search, avail_spare, l2cache, log); + nvlist_free(search); + + return (ret); +} + +static int +vdev_online(nvlist_t *nv) +{ + uint64_t ival; + + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_OFFLINE, &ival) == 0 || + nvlist_lookup_uint64(nv, ZPOOL_CONFIG_FAULTED, &ival) == 0 || + nvlist_lookup_uint64(nv, ZPOOL_CONFIG_REMOVED, &ival) == 0) + return (0); + + return (1); +} + +/* + * Helper function for zpool_get_physpaths(). + */ +static int +vdev_get_one_physpath(nvlist_t *config, char *physpath, size_t physpath_size, + size_t *bytes_written) +{ + size_t bytes_left, pos, rsz; + char *tmppath; + const char *format; + + if (nvlist_lookup_string(config, ZPOOL_CONFIG_PHYS_PATH, + &tmppath) != 0) + return (EZFS_NODEVICE); + + pos = *bytes_written; + bytes_left = physpath_size - pos; + format = (pos == 0) ? "%s" : " %s"; + + rsz = snprintf(physpath + pos, bytes_left, format, tmppath); + *bytes_written += rsz; + + if (rsz >= bytes_left) { + /* if physpath was not copied properly, clear it */ + if (bytes_left != 0) { + physpath[pos] = 0; + } + return (EZFS_NOSPC); + } + return (0); +} + +static int +vdev_get_physpaths(nvlist_t *nv, char *physpath, size_t phypath_size, + size_t *rsz, boolean_t is_spare) +{ + char *type; + int ret; + + if (nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &type) != 0) + return (EZFS_INVALCONFIG); + + if (strcmp(type, VDEV_TYPE_DISK) == 0) { + /* + * An active spare device has ZPOOL_CONFIG_IS_SPARE set. + * For a spare vdev, we only want to boot from the active + * spare device. + */ + if (is_spare) { + uint64_t spare = 0; + (void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_IS_SPARE, + &spare); + if (!spare) + return (EZFS_INVALCONFIG); + } + + if (vdev_online(nv)) { + if ((ret = vdev_get_one_physpath(nv, physpath, + phypath_size, rsz)) != 0) + return (ret); + } + } else if (strcmp(type, VDEV_TYPE_MIRROR) == 0 || + strcmp(type, VDEV_TYPE_REPLACING) == 0 || + (is_spare = (strcmp(type, VDEV_TYPE_SPARE) == 0))) { + nvlist_t **child; + uint_t count; + int i, ret; + + if (nvlist_lookup_nvlist_array(nv, + ZPOOL_CONFIG_CHILDREN, &child, &count) != 0) + return (EZFS_INVALCONFIG); + + for (i = 0; i < count; i++) { + ret = vdev_get_physpaths(child[i], physpath, + phypath_size, rsz, is_spare); + if (ret == EZFS_NOSPC) + return (ret); + } + } + + return (EZFS_POOL_INVALARG); +} + +/* + * Get phys_path for a root pool config. + * Return 0 on success; non-zero on failure. + */ +static int +zpool_get_config_physpath(nvlist_t *config, char *physpath, size_t phypath_size) +{ + size_t rsz; + nvlist_t *vdev_root; + nvlist_t **child; + uint_t count; + char *type; + + rsz = 0; + + if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &vdev_root) != 0) + return (EZFS_INVALCONFIG); + + if (nvlist_lookup_string(vdev_root, ZPOOL_CONFIG_TYPE, &type) != 0 || + nvlist_lookup_nvlist_array(vdev_root, ZPOOL_CONFIG_CHILDREN, + &child, &count) != 0) + return (EZFS_INVALCONFIG); + + /* + * root pool can not have EFI labeled disks and can only have + * a single top-level vdev. + */ + if (strcmp(type, VDEV_TYPE_ROOT) != 0 || count != 1 || + pool_uses_efi(vdev_root)) + return (EZFS_POOL_INVALARG); + + (void) vdev_get_physpaths(child[0], physpath, phypath_size, &rsz, + B_FALSE); + + /* No online devices */ + if (rsz == 0) + return (EZFS_NODEVICE); + + return (0); +} + +/* + * Get phys_path for a root pool + * Return 0 on success; non-zero on failure. + */ +int +zpool_get_physpath(zpool_handle_t *zhp, char *physpath, size_t phypath_size) +{ + return (zpool_get_config_physpath(zhp->zpool_config, physpath, + phypath_size)); +} + +/* + * If the device has being dynamically expanded then we need to relabel + * the disk to use the new unallocated space. + */ +static int +zpool_relabel_disk(libzfs_handle_t *hdl, const char *name) +{ +#ifdef sun + char path[MAXPATHLEN]; + char errbuf[1024]; + int fd, error; + int (*_efi_use_whole_disk)(int); + + if ((_efi_use_whole_disk = (int (*)(int))dlsym(RTLD_DEFAULT, + "efi_use_whole_disk")) == NULL) + return (-1); + + (void) snprintf(path, sizeof (path), "%s/%s", RDISK_ROOT, name); + + if ((fd = open(path, O_RDWR | O_NDELAY)) < 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "cannot " + "relabel '%s': unable to open device"), name); + return (zfs_error(hdl, EZFS_OPENFAILED, errbuf)); + } + + /* + * It's possible that we might encounter an error if the device + * does not have any unallocated space left. If so, we simply + * ignore that error and continue on. + */ + error = _efi_use_whole_disk(fd); + (void) close(fd); + if (error && error != VT_ENOSPC) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "cannot " + "relabel '%s': unable to read disk capacity"), name); + return (zfs_error(hdl, EZFS_NOCAP, errbuf)); + } +#endif /* sun */ + return (0); +} + +/* + * Bring the specified vdev online. The 'flags' parameter is a set of the + * ZFS_ONLINE_* flags. + */ +int +zpool_vdev_online(zpool_handle_t *zhp, const char *path, int flags, + vdev_state_t *newstate) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + nvlist_t *tgt; + boolean_t avail_spare, l2cache, islog; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + if (flags & ZFS_ONLINE_EXPAND) { + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot expand %s"), path); + } else { + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot online %s"), path); + } + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if ((tgt = zpool_find_vdev(zhp, path, &avail_spare, &l2cache, + &islog)) == NULL) + return (zfs_error(hdl, EZFS_NODEVICE, msg)); + + verify(nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_GUID, &zc.zc_guid) == 0); + + if (avail_spare) + return (zfs_error(hdl, EZFS_ISSPARE, msg)); + + if (flags & ZFS_ONLINE_EXPAND || + zpool_get_prop_int(zhp, ZPOOL_PROP_AUTOEXPAND, NULL)) { + char *pathname = NULL; + uint64_t wholedisk = 0; + + (void) nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_WHOLE_DISK, + &wholedisk); + verify(nvlist_lookup_string(tgt, ZPOOL_CONFIG_PATH, + &pathname) == 0); + + /* + * XXX - L2ARC 1.0 devices can't support expansion. + */ + if (l2cache) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "cannot expand cache devices")); + return (zfs_error(hdl, EZFS_VDEVNOTSUP, msg)); + } + + if (wholedisk) { + pathname += strlen(DISK_ROOT) + 1; + (void) zpool_relabel_disk(hdl, pathname); + } + } + + zc.zc_cookie = VDEV_STATE_ONLINE; + zc.zc_obj = flags; + + if (zfs_ioctl(hdl, ZFS_IOC_VDEV_SET_STATE, &zc) != 0) { + if (errno == EINVAL) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "was split " + "from this pool into a new one. Use '%s' " + "instead"), "zpool detach"); + return (zfs_error(hdl, EZFS_POSTSPLIT_ONLINE, msg)); + } + return (zpool_standard_error(hdl, errno, msg)); + } + + *newstate = zc.zc_cookie; + return (0); +} + +/* + * Take the specified vdev offline + */ +int +zpool_vdev_offline(zpool_handle_t *zhp, const char *path, boolean_t istmp) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + nvlist_t *tgt; + boolean_t avail_spare, l2cache; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot offline %s"), path); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if ((tgt = zpool_find_vdev(zhp, path, &avail_spare, &l2cache, + NULL)) == NULL) + return (zfs_error(hdl, EZFS_NODEVICE, msg)); + + verify(nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_GUID, &zc.zc_guid) == 0); + + if (avail_spare) + return (zfs_error(hdl, EZFS_ISSPARE, msg)); + + zc.zc_cookie = VDEV_STATE_OFFLINE; + zc.zc_obj = istmp ? ZFS_OFFLINE_TEMPORARY : 0; + + if (zfs_ioctl(hdl, ZFS_IOC_VDEV_SET_STATE, &zc) == 0) + return (0); + + switch (errno) { + case EBUSY: + + /* + * There are no other replicas of this device. + */ + return (zfs_error(hdl, EZFS_NOREPLICAS, msg)); + + case EEXIST: + /* + * The log device has unplayed logs + */ + return (zfs_error(hdl, EZFS_UNPLAYED_LOGS, msg)); + + default: + return (zpool_standard_error(hdl, errno, msg)); + } +} + +/* + * Mark the given vdev faulted. + */ +int +zpool_vdev_fault(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot fault %llu"), guid); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + zc.zc_guid = guid; + zc.zc_cookie = VDEV_STATE_FAULTED; + zc.zc_obj = aux; + + if (ioctl(hdl->libzfs_fd, ZFS_IOC_VDEV_SET_STATE, &zc) == 0) + return (0); + + switch (errno) { + case EBUSY: + + /* + * There are no other replicas of this device. + */ + return (zfs_error(hdl, EZFS_NOREPLICAS, msg)); + + default: + return (zpool_standard_error(hdl, errno, msg)); + } + +} + +/* + * Mark the given vdev degraded. + */ +int +zpool_vdev_degrade(zpool_handle_t *zhp, uint64_t guid, vdev_aux_t aux) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot degrade %llu"), guid); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + zc.zc_guid = guid; + zc.zc_cookie = VDEV_STATE_DEGRADED; + zc.zc_obj = aux; + + if (ioctl(hdl->libzfs_fd, ZFS_IOC_VDEV_SET_STATE, &zc) == 0) + return (0); + + return (zpool_standard_error(hdl, errno, msg)); +} + +/* + * Returns TRUE if the given nvlist is a vdev that was originally swapped in as + * a hot spare. + */ +static boolean_t +is_replacing_spare(nvlist_t *search, nvlist_t *tgt, int which) +{ + nvlist_t **child; + uint_t c, children; + char *type; + + if (nvlist_lookup_nvlist_array(search, ZPOOL_CONFIG_CHILDREN, &child, + &children) == 0) { + verify(nvlist_lookup_string(search, ZPOOL_CONFIG_TYPE, + &type) == 0); + + if (strcmp(type, VDEV_TYPE_SPARE) == 0 && + children == 2 && child[which] == tgt) + return (B_TRUE); + + for (c = 0; c < children; c++) + if (is_replacing_spare(child[c], tgt, which)) + return (B_TRUE); + } + + return (B_FALSE); +} + +/* + * Attach new_disk (fully described by nvroot) to old_disk. + * If 'replacing' is specified, the new disk will replace the old one. + */ +int +zpool_vdev_attach(zpool_handle_t *zhp, + const char *old_disk, const char *new_disk, nvlist_t *nvroot, int replacing) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + int ret; + nvlist_t *tgt; + boolean_t avail_spare, l2cache, islog; + uint64_t val; + char *newname; + nvlist_t **child; + uint_t children; + nvlist_t *config_root; + libzfs_handle_t *hdl = zhp->zpool_hdl; + boolean_t rootpool = zpool_is_bootable(zhp); + + if (replacing) + (void) snprintf(msg, sizeof (msg), dgettext(TEXT_DOMAIN, + "cannot replace %s with %s"), old_disk, new_disk); + else + (void) snprintf(msg, sizeof (msg), dgettext(TEXT_DOMAIN, + "cannot attach %s to %s"), new_disk, old_disk); + + /* + * If this is a root pool, make sure that we're not attaching an + * EFI labeled device. + */ + if (rootpool && pool_uses_efi(nvroot)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "EFI labeled devices are not supported on root pools.")); + return (zfs_error(hdl, EZFS_POOL_NOTSUP, msg)); + } + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if ((tgt = zpool_find_vdev(zhp, old_disk, &avail_spare, &l2cache, + &islog)) == 0) + return (zfs_error(hdl, EZFS_NODEVICE, msg)); + + if (avail_spare) + return (zfs_error(hdl, EZFS_ISSPARE, msg)); + + if (l2cache) + return (zfs_error(hdl, EZFS_ISL2CACHE, msg)); + + verify(nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_GUID, &zc.zc_guid) == 0); + zc.zc_cookie = replacing; + + if (nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0 || children != 1) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "new device must be a single disk")); + return (zfs_error(hdl, EZFS_INVALCONFIG, msg)); + } + + verify(nvlist_lookup_nvlist(zpool_get_config(zhp, NULL), + ZPOOL_CONFIG_VDEV_TREE, &config_root) == 0); + + if ((newname = zpool_vdev_name(NULL, NULL, child[0], B_FALSE)) == NULL) + return (-1); + + /* + * If the target is a hot spare that has been swapped in, we can only + * replace it with another hot spare. + */ + if (replacing && + nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_IS_SPARE, &val) == 0 && + (zpool_find_vdev(zhp, newname, &avail_spare, &l2cache, + NULL) == NULL || !avail_spare) && + is_replacing_spare(config_root, tgt, 1)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "can only be replaced by another hot spare")); + free(newname); + return (zfs_error(hdl, EZFS_BADTARGET, msg)); + } + + free(newname); + + if (zcmd_write_conf_nvlist(hdl, &zc, nvroot) != 0) + return (-1); + + ret = zfs_ioctl(hdl, ZFS_IOC_VDEV_ATTACH, &zc); + + zcmd_free_nvlists(&zc); + + if (ret == 0) { + if (rootpool) { + /* + * XXX need a better way to prevent user from + * booting up a half-baked vdev. + */ + (void) fprintf(stderr, dgettext(TEXT_DOMAIN, "Make " + "sure to wait until resilver is done " + "before rebooting.\n")); + (void) fprintf(stderr, "\n"); + (void) fprintf(stderr, dgettext(TEXT_DOMAIN, "If " + "you boot from pool '%s', you may need to update\n" + "boot code on newly attached disk '%s'.\n\n" + "Assuming you use GPT partitioning and 'da0' is " + "your new boot disk\n" + "you may use the following command:\n\n" + "\tgpart bootcode -b /boot/pmbr -p " + "/boot/gptzfsboot -i 1 da0\n\n"), + zhp->zpool_name, new_disk); + } + return (0); + } + + switch (errno) { + case ENOTSUP: + /* + * Can't attach to or replace this type of vdev. + */ + if (replacing) { + uint64_t version = zpool_get_prop_int(zhp, + ZPOOL_PROP_VERSION, NULL); + + if (islog) + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "cannot replace a log with a spare")); + else if (version >= SPA_VERSION_MULTI_REPLACE) + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "already in replacing/spare config; wait " + "for completion or use 'zpool detach'")); + else + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "cannot replace a replacing device")); + } else { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "can only attach to mirrors and top-level " + "disks")); + } + (void) zfs_error(hdl, EZFS_BADTARGET, msg); + break; + + case EINVAL: + /* + * The new device must be a single disk. + */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "new device must be a single disk")); + (void) zfs_error(hdl, EZFS_INVALCONFIG, msg); + break; + + case EBUSY: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "%s is busy"), + new_disk); + (void) zfs_error(hdl, EZFS_BADDEV, msg); + break; + + case EOVERFLOW: + /* + * The new device is too small. + */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "device is too small")); + (void) zfs_error(hdl, EZFS_BADDEV, msg); + break; + + case EDOM: + /* + * The new device has a different alignment requirement. + */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "devices have different sector alignment")); + (void) zfs_error(hdl, EZFS_BADDEV, msg); + break; + + case ENAMETOOLONG: + /* + * The resulting top-level vdev spec won't fit in the label. + */ + (void) zfs_error(hdl, EZFS_DEVOVERFLOW, msg); + break; + + default: + (void) zpool_standard_error(hdl, errno, msg); + } + + return (-1); +} + +/* + * Detach the specified device. + */ +int +zpool_vdev_detach(zpool_handle_t *zhp, const char *path) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + nvlist_t *tgt; + boolean_t avail_spare, l2cache; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot detach %s"), path); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if ((tgt = zpool_find_vdev(zhp, path, &avail_spare, &l2cache, + NULL)) == 0) + return (zfs_error(hdl, EZFS_NODEVICE, msg)); + + if (avail_spare) + return (zfs_error(hdl, EZFS_ISSPARE, msg)); + + if (l2cache) + return (zfs_error(hdl, EZFS_ISL2CACHE, msg)); + + verify(nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_GUID, &zc.zc_guid) == 0); + + if (zfs_ioctl(hdl, ZFS_IOC_VDEV_DETACH, &zc) == 0) + return (0); + + switch (errno) { + + case ENOTSUP: + /* + * Can't detach from this type of vdev. + */ + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "only " + "applicable to mirror and replacing vdevs")); + (void) zfs_error(hdl, EZFS_BADTARGET, msg); + break; + + case EBUSY: + /* + * There are no other replicas of this device. + */ + (void) zfs_error(hdl, EZFS_NOREPLICAS, msg); + break; + + default: + (void) zpool_standard_error(hdl, errno, msg); + } + + return (-1); +} + +/* + * Find a mirror vdev in the source nvlist. + * + * The mchild array contains a list of disks in one of the top-level mirrors + * of the source pool. The schild array contains a list of disks that the + * user specified on the command line. We loop over the mchild array to + * see if any entry in the schild array matches. + * + * If a disk in the mchild array is found in the schild array, we return + * the index of that entry. Otherwise we return -1. + */ +static int +find_vdev_entry(zpool_handle_t *zhp, nvlist_t **mchild, uint_t mchildren, + nvlist_t **schild, uint_t schildren) +{ + uint_t mc; + + for (mc = 0; mc < mchildren; mc++) { + uint_t sc; + char *mpath = zpool_vdev_name(zhp->zpool_hdl, zhp, + mchild[mc], B_FALSE); + + for (sc = 0; sc < schildren; sc++) { + char *spath = zpool_vdev_name(zhp->zpool_hdl, zhp, + schild[sc], B_FALSE); + boolean_t result = (strcmp(mpath, spath) == 0); + + free(spath); + if (result) { + free(mpath); + return (mc); + } + } + + free(mpath); + } + + return (-1); +} + +/* + * Split a mirror pool. If newroot points to null, then a new nvlist + * is generated and it is the responsibility of the caller to free it. + */ +int +zpool_vdev_split(zpool_handle_t *zhp, char *newname, nvlist_t **newroot, + nvlist_t *props, splitflags_t flags) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + nvlist_t *tree, *config, **child, **newchild, *newconfig = NULL; + nvlist_t **varray = NULL, *zc_props = NULL; + uint_t c, children, newchildren, lastlog = 0, vcount, found = 0; + libzfs_handle_t *hdl = zhp->zpool_hdl; + uint64_t vers; + boolean_t freelist = B_FALSE, memory_err = B_TRUE; + int retval = 0; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "Unable to split %s"), zhp->zpool_name); + + if (!zpool_name_valid(hdl, B_FALSE, newname)) + return (zfs_error(hdl, EZFS_INVALIDNAME, msg)); + + if ((config = zpool_get_config(zhp, NULL)) == NULL) { + (void) fprintf(stderr, gettext("Internal error: unable to " + "retrieve pool configuration\n")); + return (-1); + } + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, &tree) + == 0); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION, &vers) == 0); + + if (props) { + prop_flags_t flags = { .create = B_FALSE, .import = B_TRUE }; + if ((zc_props = zpool_valid_proplist(hdl, zhp->zpool_name, + props, vers, flags, msg)) == NULL) + return (-1); + } + + if (nvlist_lookup_nvlist_array(tree, ZPOOL_CONFIG_CHILDREN, &child, + &children) != 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "Source pool is missing vdev tree")); + if (zc_props) + nvlist_free(zc_props); + return (-1); + } + + varray = zfs_alloc(hdl, children * sizeof (nvlist_t *)); + vcount = 0; + + if (*newroot == NULL || + nvlist_lookup_nvlist_array(*newroot, ZPOOL_CONFIG_CHILDREN, + &newchild, &newchildren) != 0) + newchildren = 0; + + for (c = 0; c < children; c++) { + uint64_t is_log = B_FALSE, is_hole = B_FALSE; + char *type; + nvlist_t **mchild, *vdev; + uint_t mchildren; + int entry; + + /* + * Unlike cache & spares, slogs are stored in the + * ZPOOL_CONFIG_CHILDREN array. We filter them out here. + */ + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_LOG, + &is_log); + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_HOLE, + &is_hole); + if (is_log || is_hole) { + /* + * Create a hole vdev and put it in the config. + */ + if (nvlist_alloc(&vdev, NV_UNIQUE_NAME, 0) != 0) + goto out; + if (nvlist_add_string(vdev, ZPOOL_CONFIG_TYPE, + VDEV_TYPE_HOLE) != 0) + goto out; + if (nvlist_add_uint64(vdev, ZPOOL_CONFIG_IS_HOLE, + 1) != 0) + goto out; + if (lastlog == 0) + lastlog = vcount; + varray[vcount++] = vdev; + continue; + } + lastlog = 0; + verify(nvlist_lookup_string(child[c], ZPOOL_CONFIG_TYPE, &type) + == 0); + if (strcmp(type, VDEV_TYPE_MIRROR) != 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "Source pool must be composed only of mirrors\n")); + retval = zfs_error(hdl, EZFS_INVALCONFIG, msg); + goto out; + } + + verify(nvlist_lookup_nvlist_array(child[c], + ZPOOL_CONFIG_CHILDREN, &mchild, &mchildren) == 0); + + /* find or add an entry for this top-level vdev */ + if (newchildren > 0 && + (entry = find_vdev_entry(zhp, mchild, mchildren, + newchild, newchildren)) >= 0) { + /* We found a disk that the user specified. */ + vdev = mchild[entry]; + ++found; + } else { + /* User didn't specify a disk for this vdev. */ + vdev = mchild[mchildren - 1]; + } + + if (nvlist_dup(vdev, &varray[vcount++], 0) != 0) + goto out; + } + + /* did we find every disk the user specified? */ + if (found != newchildren) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "Device list must " + "include at most one disk from each mirror")); + retval = zfs_error(hdl, EZFS_INVALCONFIG, msg); + goto out; + } + + /* Prepare the nvlist for populating. */ + if (*newroot == NULL) { + if (nvlist_alloc(newroot, NV_UNIQUE_NAME, 0) != 0) + goto out; + freelist = B_TRUE; + if (nvlist_add_string(*newroot, ZPOOL_CONFIG_TYPE, + VDEV_TYPE_ROOT) != 0) + goto out; + } else { + verify(nvlist_remove_all(*newroot, ZPOOL_CONFIG_CHILDREN) == 0); + } + + /* Add all the children we found */ + if (nvlist_add_nvlist_array(*newroot, ZPOOL_CONFIG_CHILDREN, varray, + lastlog == 0 ? vcount : lastlog) != 0) + goto out; + + /* + * If we're just doing a dry run, exit now with success. + */ + if (flags.dryrun) { + memory_err = B_FALSE; + freelist = B_FALSE; + goto out; + } + + /* now build up the config list & call the ioctl */ + if (nvlist_alloc(&newconfig, NV_UNIQUE_NAME, 0) != 0) + goto out; + + if (nvlist_add_nvlist(newconfig, + ZPOOL_CONFIG_VDEV_TREE, *newroot) != 0 || + nvlist_add_string(newconfig, + ZPOOL_CONFIG_POOL_NAME, newname) != 0 || + nvlist_add_uint64(newconfig, ZPOOL_CONFIG_VERSION, vers) != 0) + goto out; + + /* + * The new pool is automatically part of the namespace unless we + * explicitly export it. + */ + if (!flags.import) + zc.zc_cookie = ZPOOL_EXPORT_AFTER_SPLIT; + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + (void) strlcpy(zc.zc_string, newname, sizeof (zc.zc_string)); + if (zcmd_write_conf_nvlist(hdl, &zc, newconfig) != 0) + goto out; + if (zc_props != NULL && zcmd_write_src_nvlist(hdl, &zc, zc_props) != 0) + goto out; + + if (zfs_ioctl(hdl, ZFS_IOC_VDEV_SPLIT, &zc) != 0) { + retval = zpool_standard_error(hdl, errno, msg); + goto out; + } + + freelist = B_FALSE; + memory_err = B_FALSE; + +out: + if (varray != NULL) { + int v; + + for (v = 0; v < vcount; v++) + nvlist_free(varray[v]); + free(varray); + } + zcmd_free_nvlists(&zc); + if (zc_props) + nvlist_free(zc_props); + if (newconfig) + nvlist_free(newconfig); + if (freelist) { + nvlist_free(*newroot); + *newroot = NULL; + } + + if (retval != 0) + return (retval); + + if (memory_err) + return (no_memory(hdl)); + + return (0); +} + +/* + * Remove the given device. Currently, this is supported only for hot spares + * and level 2 cache devices. + */ +int +zpool_vdev_remove(zpool_handle_t *zhp, const char *path) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + nvlist_t *tgt; + boolean_t avail_spare, l2cache, islog; + libzfs_handle_t *hdl = zhp->zpool_hdl; + uint64_t version; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot remove %s"), path); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if ((tgt = zpool_find_vdev(zhp, path, &avail_spare, &l2cache, + &islog)) == 0) + return (zfs_error(hdl, EZFS_NODEVICE, msg)); + /* + * XXX - this should just go away. + */ + if (!avail_spare && !l2cache && !islog) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "only inactive hot spares, cache, top-level, " + "or log devices can be removed")); + return (zfs_error(hdl, EZFS_NODEVICE, msg)); + } + + version = zpool_get_prop_int(zhp, ZPOOL_PROP_VERSION, NULL); + if (islog && version < SPA_VERSION_HOLES) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "pool must be upgrade to support log removal")); + return (zfs_error(hdl, EZFS_BADVERSION, msg)); + } + + verify(nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_GUID, &zc.zc_guid) == 0); + + if (zfs_ioctl(hdl, ZFS_IOC_VDEV_REMOVE, &zc) == 0) + return (0); + + return (zpool_standard_error(hdl, errno, msg)); +} + +/* + * Clear the errors for the pool, or the particular device if specified. + */ +int +zpool_clear(zpool_handle_t *zhp, const char *path, nvlist_t *rewindnvl) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + nvlist_t *tgt; + zpool_rewind_policy_t policy; + boolean_t avail_spare, l2cache; + libzfs_handle_t *hdl = zhp->zpool_hdl; + nvlist_t *nvi = NULL; + int error; + + if (path) + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot clear errors for %s"), + path); + else + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot clear errors for %s"), + zhp->zpool_name); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if (path) { + if ((tgt = zpool_find_vdev(zhp, path, &avail_spare, + &l2cache, NULL)) == 0) + return (zfs_error(hdl, EZFS_NODEVICE, msg)); + + /* + * Don't allow error clearing for hot spares. Do allow + * error clearing for l2cache devices. + */ + if (avail_spare) + return (zfs_error(hdl, EZFS_ISSPARE, msg)); + + verify(nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_GUID, + &zc.zc_guid) == 0); + } + + zpool_get_rewind_policy(rewindnvl, &policy); + zc.zc_cookie = policy.zrp_request; + + if (zcmd_alloc_dst_nvlist(hdl, &zc, zhp->zpool_config_size * 2) != 0) + return (-1); + + if (zcmd_write_src_nvlist(hdl, &zc, rewindnvl) != 0) + return (-1); + + while ((error = zfs_ioctl(hdl, ZFS_IOC_CLEAR, &zc)) != 0 && + errno == ENOMEM) { + if (zcmd_expand_dst_nvlist(hdl, &zc) != 0) { + zcmd_free_nvlists(&zc); + return (-1); + } + } + + if (!error || ((policy.zrp_request & ZPOOL_TRY_REWIND) && + errno != EPERM && errno != EACCES)) { + if (policy.zrp_request & + (ZPOOL_DO_REWIND | ZPOOL_TRY_REWIND)) { + (void) zcmd_read_dst_nvlist(hdl, &zc, &nvi); + zpool_rewind_exclaim(hdl, zc.zc_name, + ((policy.zrp_request & ZPOOL_TRY_REWIND) != 0), + nvi); + nvlist_free(nvi); + } + zcmd_free_nvlists(&zc); + return (0); + } + + zcmd_free_nvlists(&zc); + return (zpool_standard_error(hdl, errno, msg)); +} + +/* + * Similar to zpool_clear(), but takes a GUID (used by fmd). + */ +int +zpool_vdev_clear(zpool_handle_t *zhp, uint64_t guid) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot clear errors for %llx"), + guid); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + zc.zc_guid = guid; + zc.zc_cookie = ZPOOL_NO_REWIND; + + if (ioctl(hdl->libzfs_fd, ZFS_IOC_CLEAR, &zc) == 0) + return (0); + + return (zpool_standard_error(hdl, errno, msg)); +} + +/* + * Change the GUID for a pool. + */ +int +zpool_reguid(zpool_handle_t *zhp) +{ + char msg[1024]; + libzfs_handle_t *hdl = zhp->zpool_hdl; + zfs_cmd_t zc = { 0 }; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot reguid '%s'"), zhp->zpool_name); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if (zfs_ioctl(hdl, ZFS_IOC_POOL_REGUID, &zc) == 0) + return (0); + + return (zpool_standard_error(hdl, errno, msg)); +} + +/* + * Reopen the pool. + */ +int +zpool_reopen(zpool_handle_t *zhp) +{ + zfs_cmd_t zc = { 0 }; + char msg[1024]; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) snprintf(msg, sizeof (msg), + dgettext(TEXT_DOMAIN, "cannot reopen '%s'"), + zhp->zpool_name); + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + if (zfs_ioctl(hdl, ZFS_IOC_POOL_REOPEN, &zc) == 0) + return (0); + return (zpool_standard_error(hdl, errno, msg)); +} + +/* + * Convert from a devid string to a path. + */ +static char * +devid_to_path(char *devid_str) +{ + ddi_devid_t devid; + char *minor; + char *path; + devid_nmlist_t *list = NULL; + int ret; + + if (devid_str_decode(devid_str, &devid, &minor) != 0) + return (NULL); + + ret = devid_deviceid_to_nmlist("/dev", devid, minor, &list); + + devid_str_free(minor); + devid_free(devid); + + if (ret != 0) + return (NULL); + + if ((path = strdup(list[0].devname)) == NULL) + return (NULL); + + devid_free_nmlist(list); + + return (path); +} + +/* + * Convert from a path to a devid string. + */ +static char * +path_to_devid(const char *path) +{ +#ifdef have_devid + int fd; + ddi_devid_t devid; + char *minor, *ret; + + if ((fd = open(path, O_RDONLY)) < 0) + return (NULL); + + minor = NULL; + ret = NULL; + if (devid_get(fd, &devid) == 0) { + if (devid_get_minor_name(fd, &minor) == 0) + ret = devid_str_encode(devid, minor); + if (minor != NULL) + devid_str_free(minor); + devid_free(devid); + } + (void) close(fd); + + return (ret); +#else + return (NULL); +#endif +} + +/* + * Issue the necessary ioctl() to update the stored path value for the vdev. We + * ignore any failure here, since a common case is for an unprivileged user to + * type 'zpool status', and we'll display the correct information anyway. + */ +static void +set_path(zpool_handle_t *zhp, nvlist_t *nv, const char *path) +{ + zfs_cmd_t zc = { 0 }; + + (void) strncpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + (void) strncpy(zc.zc_value, path, sizeof (zc.zc_value)); + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, + &zc.zc_guid) == 0); + + (void) ioctl(zhp->zpool_hdl->libzfs_fd, ZFS_IOC_VDEV_SETPATH, &zc); +} + +/* + * Given a vdev, return the name to display in iostat. If the vdev has a path, + * we use that, stripping off any leading "/dev/dsk/"; if not, we use the type. + * We also check if this is a whole disk, in which case we strip off the + * trailing 's0' slice name. + * + * This routine is also responsible for identifying when disks have been + * reconfigured in a new location. The kernel will have opened the device by + * devid, but the path will still refer to the old location. To catch this, we + * first do a path -> devid translation (which is fast for the common case). If + * the devid matches, we're done. If not, we do a reverse devid -> path + * translation and issue the appropriate ioctl() to update the path of the vdev. + * If 'zhp' is NULL, then this is an exported pool, and we don't need to do any + * of these checks. + */ +char * +zpool_vdev_name(libzfs_handle_t *hdl, zpool_handle_t *zhp, nvlist_t *nv, + boolean_t verbose) +{ + char *path, *devid; + uint64_t value; + char buf[64]; + vdev_stat_t *vs; + uint_t vsc; + int have_stats; + int have_path; + + have_stats = nvlist_lookup_uint64_array(nv, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &vsc) == 0; + have_path = nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) == 0; + + /* + * If the device is not currently present, assume it will not + * come back at the same device path. Display the device by GUID. + */ + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_NOT_PRESENT, &value) == 0 || + have_path && have_stats && vs->vs_state <= VDEV_STATE_CANT_OPEN) { + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, + &value) == 0); + (void) snprintf(buf, sizeof (buf), "%llu", + (u_longlong_t)value); + path = buf; + } else if (have_path) { + + /* + * If the device is dead (faulted, offline, etc) then don't + * bother opening it. Otherwise we may be forcing the user to + * open a misbehaving device, which can have undesirable + * effects. + */ + if ((have_stats == 0 || + vs->vs_state >= VDEV_STATE_DEGRADED) && + zhp != NULL && + nvlist_lookup_string(nv, ZPOOL_CONFIG_DEVID, &devid) == 0) { + /* + * Determine if the current path is correct. + */ + char *newdevid = path_to_devid(path); + + if (newdevid == NULL || + strcmp(devid, newdevid) != 0) { + char *newpath; + + if ((newpath = devid_to_path(devid)) != NULL) { + /* + * Update the path appropriately. + */ + set_path(zhp, nv, newpath); + if (nvlist_add_string(nv, + ZPOOL_CONFIG_PATH, newpath) == 0) + verify(nvlist_lookup_string(nv, + ZPOOL_CONFIG_PATH, + &path) == 0); + free(newpath); + } + } + + if (newdevid) + devid_str_free(newdevid); + } + +#ifdef sun + if (strncmp(path, "/dev/dsk/", 9) == 0) + path += 9; + + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_WHOLE_DISK, + &value) == 0 && value) { + int pathlen = strlen(path); + char *tmp = zfs_strdup(hdl, path); + + /* + * If it starts with c#, and ends with "s0", chop + * the "s0" off, or if it ends with "s0/old", remove + * the "s0" from the middle. + */ + if (CTD_CHECK(tmp)) { + if (strcmp(&tmp[pathlen - 2], "s0") == 0) { + tmp[pathlen - 2] = '\0'; + } else if (pathlen > 6 && + strcmp(&tmp[pathlen - 6], "s0/old") == 0) { + (void) strcpy(&tmp[pathlen - 6], + "/old"); + } + } + return (tmp); + } +#else /* !sun */ + if (strncmp(path, _PATH_DEV, sizeof(_PATH_DEV) - 1) == 0) + path += sizeof(_PATH_DEV) - 1; +#endif /* !sun */ + } else { + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &path) == 0); + + /* + * If it's a raidz device, we need to stick in the parity level. + */ + if (strcmp(path, VDEV_TYPE_RAIDZ) == 0) { + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_NPARITY, + &value) == 0); + (void) snprintf(buf, sizeof (buf), "%s%llu", path, + (u_longlong_t)value); + path = buf; + } + + /* + * We identify each top-level vdev by using a <type-id> + * naming convention. + */ + if (verbose) { + uint64_t id; + + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_ID, + &id) == 0); + (void) snprintf(buf, sizeof (buf), "%s-%llu", path, + (u_longlong_t)id); + path = buf; + } + } + + return (zfs_strdup(hdl, path)); +} + +static int +zbookmark_compare(const void *a, const void *b) +{ + return (memcmp(a, b, sizeof (zbookmark_t))); +} + +/* + * Retrieve the persistent error log, uniquify the members, and return to the + * caller. + */ +int +zpool_get_errlog(zpool_handle_t *zhp, nvlist_t **nverrlistp) +{ + zfs_cmd_t zc = { 0 }; + uint64_t count; + zbookmark_t *zb = NULL; + int i; + + /* + * Retrieve the raw error list from the kernel. If the number of errors + * has increased, allocate more space and continue until we get the + * entire list. + */ + verify(nvlist_lookup_uint64(zhp->zpool_config, ZPOOL_CONFIG_ERRCOUNT, + &count) == 0); + if (count == 0) + return (0); + if ((zc.zc_nvlist_dst = (uintptr_t)zfs_alloc(zhp->zpool_hdl, + count * sizeof (zbookmark_t))) == (uintptr_t)NULL) + return (-1); + zc.zc_nvlist_dst_size = count; + (void) strcpy(zc.zc_name, zhp->zpool_name); + for (;;) { + if (ioctl(zhp->zpool_hdl->libzfs_fd, ZFS_IOC_ERROR_LOG, + &zc) != 0) { + free((void *)(uintptr_t)zc.zc_nvlist_dst); + if (errno == ENOMEM) { + count = zc.zc_nvlist_dst_size; + if ((zc.zc_nvlist_dst = (uintptr_t) + zfs_alloc(zhp->zpool_hdl, count * + sizeof (zbookmark_t))) == (uintptr_t)NULL) + return (-1); + } else { + return (-1); + } + } else { + break; + } + } + + /* + * Sort the resulting bookmarks. This is a little confusing due to the + * implementation of ZFS_IOC_ERROR_LOG. The bookmarks are copied last + * to first, and 'zc_nvlist_dst_size' indicates the number of boomarks + * _not_ copied as part of the process. So we point the start of our + * array appropriate and decrement the total number of elements. + */ + zb = ((zbookmark_t *)(uintptr_t)zc.zc_nvlist_dst) + + zc.zc_nvlist_dst_size; + count -= zc.zc_nvlist_dst_size; + + qsort(zb, count, sizeof (zbookmark_t), zbookmark_compare); + + verify(nvlist_alloc(nverrlistp, 0, KM_SLEEP) == 0); + + /* + * Fill in the nverrlistp with nvlist's of dataset and object numbers. + */ + for (i = 0; i < count; i++) { + nvlist_t *nv; + + /* ignoring zb_blkid and zb_level for now */ + if (i > 0 && zb[i-1].zb_objset == zb[i].zb_objset && + zb[i-1].zb_object == zb[i].zb_object) + continue; + + if (nvlist_alloc(&nv, NV_UNIQUE_NAME, KM_SLEEP) != 0) + goto nomem; + if (nvlist_add_uint64(nv, ZPOOL_ERR_DATASET, + zb[i].zb_objset) != 0) { + nvlist_free(nv); + goto nomem; + } + if (nvlist_add_uint64(nv, ZPOOL_ERR_OBJECT, + zb[i].zb_object) != 0) { + nvlist_free(nv); + goto nomem; + } + if (nvlist_add_nvlist(*nverrlistp, "ejk", nv) != 0) { + nvlist_free(nv); + goto nomem; + } + nvlist_free(nv); + } + + free((void *)(uintptr_t)zc.zc_nvlist_dst); + return (0); + +nomem: + free((void *)(uintptr_t)zc.zc_nvlist_dst); + return (no_memory(zhp->zpool_hdl)); +} + +/* + * Upgrade a ZFS pool to the latest on-disk version. + */ +int +zpool_upgrade(zpool_handle_t *zhp, uint64_t new_version) +{ + zfs_cmd_t zc = { 0 }; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) strcpy(zc.zc_name, zhp->zpool_name); + zc.zc_cookie = new_version; + + if (zfs_ioctl(hdl, ZFS_IOC_POOL_UPGRADE, &zc) != 0) + return (zpool_standard_error_fmt(hdl, errno, + dgettext(TEXT_DOMAIN, "cannot upgrade '%s'"), + zhp->zpool_name)); + return (0); +} + +void +zfs_save_arguments(int argc, char **argv, char *string, int len) +{ + (void) strlcpy(string, basename(argv[0]), len); + for (int i = 1; i < argc; i++) { + (void) strlcat(string, " ", len); + (void) strlcat(string, argv[i], len); + } +} + +int +zpool_log_history(libzfs_handle_t *hdl, const char *message) +{ + zfs_cmd_t zc = { 0 }; + nvlist_t *args; + int err; + + args = fnvlist_alloc(); + fnvlist_add_string(args, "message", message); + err = zcmd_write_src_nvlist(hdl, &zc, args); + if (err == 0) + err = ioctl(hdl->libzfs_fd, ZFS_IOC_LOG_HISTORY, &zc); + nvlist_free(args); + zcmd_free_nvlists(&zc); + return (err); +} + +/* + * Perform ioctl to get some command history of a pool. + * + * 'buf' is the buffer to fill up to 'len' bytes. 'off' is the + * logical offset of the history buffer to start reading from. + * + * Upon return, 'off' is the next logical offset to read from and + * 'len' is the actual amount of bytes read into 'buf'. + */ +static int +get_history(zpool_handle_t *zhp, char *buf, uint64_t *off, uint64_t *len) +{ + zfs_cmd_t zc = { 0 }; + libzfs_handle_t *hdl = zhp->zpool_hdl; + + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + + zc.zc_history = (uint64_t)(uintptr_t)buf; + zc.zc_history_len = *len; + zc.zc_history_offset = *off; + + if (ioctl(hdl->libzfs_fd, ZFS_IOC_POOL_GET_HISTORY, &zc) != 0) { + switch (errno) { + case EPERM: + return (zfs_error_fmt(hdl, EZFS_PERM, + dgettext(TEXT_DOMAIN, + "cannot show history for pool '%s'"), + zhp->zpool_name)); + case ENOENT: + return (zfs_error_fmt(hdl, EZFS_NOHISTORY, + dgettext(TEXT_DOMAIN, "cannot get history for pool " + "'%s'"), zhp->zpool_name)); + case ENOTSUP: + return (zfs_error_fmt(hdl, EZFS_BADVERSION, + dgettext(TEXT_DOMAIN, "cannot get history for pool " + "'%s', pool must be upgraded"), zhp->zpool_name)); + default: + return (zpool_standard_error_fmt(hdl, errno, + dgettext(TEXT_DOMAIN, + "cannot get history for '%s'"), zhp->zpool_name)); + } + } + + *len = zc.zc_history_len; + *off = zc.zc_history_offset; + + return (0); +} + +/* + * Process the buffer of nvlists, unpacking and storing each nvlist record + * into 'records'. 'leftover' is set to the number of bytes that weren't + * processed as there wasn't a complete record. + */ +int +zpool_history_unpack(char *buf, uint64_t bytes_read, uint64_t *leftover, + nvlist_t ***records, uint_t *numrecords) +{ + uint64_t reclen; + nvlist_t *nv; + int i; + + while (bytes_read > sizeof (reclen)) { + + /* get length of packed record (stored as little endian) */ + for (i = 0, reclen = 0; i < sizeof (reclen); i++) + reclen += (uint64_t)(((uchar_t *)buf)[i]) << (8*i); + + if (bytes_read < sizeof (reclen) + reclen) + break; + + /* unpack record */ + if (nvlist_unpack(buf + sizeof (reclen), reclen, &nv, 0) != 0) + return (ENOMEM); + bytes_read -= sizeof (reclen) + reclen; + buf += sizeof (reclen) + reclen; + + /* add record to nvlist array */ + (*numrecords)++; + if (ISP2(*numrecords + 1)) { + *records = realloc(*records, + *numrecords * 2 * sizeof (nvlist_t *)); + } + (*records)[*numrecords - 1] = nv; + } + + *leftover = bytes_read; + return (0); +} + +/* from spa_history.c: spa_history_create_obj() */ +#define HIS_BUF_LEN_DEF (128 << 10) +#define HIS_BUF_LEN_MAX (1 << 30) + +/* + * Retrieve the command history of a pool. + */ +int +zpool_get_history(zpool_handle_t *zhp, nvlist_t **nvhisp) +{ + char *buf = NULL; + uint64_t bufsize = HIS_BUF_LEN_DEF; + uint64_t off = 0; + nvlist_t **records = NULL; + uint_t numrecords = 0; + int err, i; + + if ((buf = malloc(bufsize)) == NULL) + return (ENOMEM); + do { + uint64_t bytes_read = bufsize; + uint64_t leftover; + + if ((err = get_history(zhp, buf, &off, &bytes_read)) != 0) + break; + + /* if nothing else was read in, we're at EOF, just return */ + if (bytes_read == 0) + break; + + if ((err = zpool_history_unpack(buf, bytes_read, + &leftover, &records, &numrecords)) != 0) + break; + off -= leftover; + + /* + * If the history block is too big, double the buffer + * size and try again. + */ + if (leftover == bytes_read) { + free(buf); + buf = NULL; + + bufsize <<= 1; + if ((bufsize >= HIS_BUF_LEN_MAX) || + ((buf = malloc(bufsize)) == NULL)) { + err = ENOMEM; + break; + } + } + + /* CONSTCOND */ + } while (1); + free(buf); + + if (!err) { + verify(nvlist_alloc(nvhisp, NV_UNIQUE_NAME, 0) == 0); + verify(nvlist_add_nvlist_array(*nvhisp, ZPOOL_HIST_RECORD, + records, numrecords) == 0); + } + for (i = 0; i < numrecords; i++) + nvlist_free(records[i]); + free(records); + + return (err); +} + +void +zpool_obj_to_path(zpool_handle_t *zhp, uint64_t dsobj, uint64_t obj, + char *pathname, size_t len) +{ + zfs_cmd_t zc = { 0 }; + boolean_t mounted = B_FALSE; + char *mntpnt = NULL; + char dsname[MAXNAMELEN]; + + if (dsobj == 0) { + /* special case for the MOS */ + (void) snprintf(pathname, len, "<metadata>:<0x%llx>", obj); + return; + } + + /* get the dataset's name */ + (void) strlcpy(zc.zc_name, zhp->zpool_name, sizeof (zc.zc_name)); + zc.zc_obj = dsobj; + if (ioctl(zhp->zpool_hdl->libzfs_fd, + ZFS_IOC_DSOBJ_TO_DSNAME, &zc) != 0) { + /* just write out a path of two object numbers */ + (void) snprintf(pathname, len, "<0x%llx>:<0x%llx>", + dsobj, obj); + return; + } + (void) strlcpy(dsname, zc.zc_value, sizeof (dsname)); + + /* find out if the dataset is mounted */ + mounted = is_mounted(zhp->zpool_hdl, dsname, &mntpnt); + + /* get the corrupted object's path */ + (void) strlcpy(zc.zc_name, dsname, sizeof (zc.zc_name)); + zc.zc_obj = obj; + if (ioctl(zhp->zpool_hdl->libzfs_fd, ZFS_IOC_OBJ_TO_PATH, + &zc) == 0) { + if (mounted) { + (void) snprintf(pathname, len, "%s%s", mntpnt, + zc.zc_value); + } else { + (void) snprintf(pathname, len, "%s:%s", + dsname, zc.zc_value); + } + } else { + (void) snprintf(pathname, len, "%s:<0x%llx>", dsname, obj); + } + free(mntpnt); +} + +#ifdef sun +/* + * Read the EFI label from the config, if a label does not exist then + * pass back the error to the caller. If the caller has passed a non-NULL + * diskaddr argument then we set it to the starting address of the EFI + * partition. + */ +static int +read_efi_label(nvlist_t *config, diskaddr_t *sb) +{ + char *path; + int fd; + char diskname[MAXPATHLEN]; + int err = -1; + + if (nvlist_lookup_string(config, ZPOOL_CONFIG_PATH, &path) != 0) + return (err); + + (void) snprintf(diskname, sizeof (diskname), "%s%s", RDISK_ROOT, + strrchr(path, '/')); + if ((fd = open(diskname, O_RDONLY|O_NDELAY)) >= 0) { + struct dk_gpt *vtoc; + + if ((err = efi_alloc_and_read(fd, &vtoc)) >= 0) { + if (sb != NULL) + *sb = vtoc->efi_parts[0].p_start; + efi_free(vtoc); + } + (void) close(fd); + } + return (err); +} + +/* + * determine where a partition starts on a disk in the current + * configuration + */ +static diskaddr_t +find_start_block(nvlist_t *config) +{ + nvlist_t **child; + uint_t c, children; + diskaddr_t sb = MAXOFFSET_T; + uint64_t wholedisk; + + if (nvlist_lookup_nvlist_array(config, + ZPOOL_CONFIG_CHILDREN, &child, &children) != 0) { + if (nvlist_lookup_uint64(config, + ZPOOL_CONFIG_WHOLE_DISK, + &wholedisk) != 0 || !wholedisk) { + return (MAXOFFSET_T); + } + if (read_efi_label(config, &sb) < 0) + sb = MAXOFFSET_T; + return (sb); + } + + for (c = 0; c < children; c++) { + sb = find_start_block(child[c]); + if (sb != MAXOFFSET_T) { + return (sb); + } + } + return (MAXOFFSET_T); +} +#endif /* sun */ + +/* + * Label an individual disk. The name provided is the short name, + * stripped of any leading /dev path. + */ +int +zpool_label_disk(libzfs_handle_t *hdl, zpool_handle_t *zhp, const char *name) +{ +#ifdef sun + char path[MAXPATHLEN]; + struct dk_gpt *vtoc; + int fd; + size_t resv = EFI_MIN_RESV_SIZE; + uint64_t slice_size; + diskaddr_t start_block; + char errbuf[1024]; + + /* prepare an error message just in case */ + (void) snprintf(errbuf, sizeof (errbuf), + dgettext(TEXT_DOMAIN, "cannot label '%s'"), name); + + if (zhp) { + nvlist_t *nvroot; + + if (zpool_is_bootable(zhp)) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "EFI labeled devices are not supported on root " + "pools.")); + return (zfs_error(hdl, EZFS_POOL_NOTSUP, errbuf)); + } + + verify(nvlist_lookup_nvlist(zhp->zpool_config, + ZPOOL_CONFIG_VDEV_TREE, &nvroot) == 0); + + if (zhp->zpool_start_block == 0) + start_block = find_start_block(nvroot); + else + start_block = zhp->zpool_start_block; + zhp->zpool_start_block = start_block; + } else { + /* new pool */ + start_block = NEW_START_BLOCK; + } + + (void) snprintf(path, sizeof (path), "%s/%s%s", RDISK_ROOT, name, + BACKUP_SLICE); + + if ((fd = open(path, O_RDWR | O_NDELAY)) < 0) { + /* + * This shouldn't happen. We've long since verified that this + * is a valid device. + */ + zfs_error_aux(hdl, + dgettext(TEXT_DOMAIN, "unable to open device")); + return (zfs_error(hdl, EZFS_OPENFAILED, errbuf)); + } + + if (efi_alloc_and_init(fd, EFI_NUMPAR, &vtoc) != 0) { + /* + * The only way this can fail is if we run out of memory, or we + * were unable to read the disk's capacity + */ + if (errno == ENOMEM) + (void) no_memory(hdl); + + (void) close(fd); + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "unable to read disk capacity"), name); + + return (zfs_error(hdl, EZFS_NOCAP, errbuf)); + } + + slice_size = vtoc->efi_last_u_lba + 1; + slice_size -= EFI_MIN_RESV_SIZE; + if (start_block == MAXOFFSET_T) + start_block = NEW_START_BLOCK; + slice_size -= start_block; + + vtoc->efi_parts[0].p_start = start_block; + vtoc->efi_parts[0].p_size = slice_size; + + /* + * Why we use V_USR: V_BACKUP confuses users, and is considered + * disposable by some EFI utilities (since EFI doesn't have a backup + * slice). V_UNASSIGNED is supposed to be used only for zero size + * partitions, and efi_write() will fail if we use it. V_ROOT, V_BOOT, + * etc. were all pretty specific. V_USR is as close to reality as we + * can get, in the absence of V_OTHER. + */ + vtoc->efi_parts[0].p_tag = V_USR; + (void) strcpy(vtoc->efi_parts[0].p_name, "zfs"); + + vtoc->efi_parts[8].p_start = slice_size + start_block; + vtoc->efi_parts[8].p_size = resv; + vtoc->efi_parts[8].p_tag = V_RESERVED; + + if (efi_write(fd, vtoc) != 0) { + /* + * Some block drivers (like pcata) may not support EFI + * GPT labels. Print out a helpful error message dir- + * ecting the user to manually label the disk and give + * a specific slice. + */ + (void) close(fd); + efi_free(vtoc); + + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "try using fdisk(1M) and then provide a specific slice")); + return (zfs_error(hdl, EZFS_LABELFAILED, errbuf)); + } + + (void) close(fd); + efi_free(vtoc); +#endif /* sun */ + return (0); +} + +static boolean_t +supported_dump_vdev_type(libzfs_handle_t *hdl, nvlist_t *config, char *errbuf) +{ + char *type; + nvlist_t **child; + uint_t children, c; + + verify(nvlist_lookup_string(config, ZPOOL_CONFIG_TYPE, &type) == 0); + if (strcmp(type, VDEV_TYPE_FILE) == 0 || + strcmp(type, VDEV_TYPE_HOLE) == 0 || + strcmp(type, VDEV_TYPE_MISSING) == 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "vdev type '%s' is not supported"), type); + (void) zfs_error(hdl, EZFS_VDEVNOTSUP, errbuf); + return (B_FALSE); + } + if (nvlist_lookup_nvlist_array(config, ZPOOL_CONFIG_CHILDREN, + &child, &children) == 0) { + for (c = 0; c < children; c++) { + if (!supported_dump_vdev_type(hdl, child[c], errbuf)) + return (B_FALSE); + } + } + return (B_TRUE); +} + +/* + * Check if this zvol is allowable for use as a dump device; zero if + * it is, > 0 if it isn't, < 0 if it isn't a zvol. + * + * Allowable storage configurations include mirrors, all raidz variants, and + * pools with log, cache, and spare devices. Pools which are backed by files or + * have missing/hole vdevs are not suitable. + */ +int +zvol_check_dump_config(char *arg) +{ + zpool_handle_t *zhp = NULL; + nvlist_t *config, *nvroot; + char *p, *volname; + nvlist_t **top; + uint_t toplevels; + libzfs_handle_t *hdl; + char errbuf[1024]; + char poolname[ZPOOL_MAXNAMELEN]; + int pathlen = strlen(ZVOL_FULL_DEV_DIR); + int ret = 1; + + if (strncmp(arg, ZVOL_FULL_DEV_DIR, pathlen)) { + return (-1); + } + + (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, + "dump is not supported on device '%s'"), arg); + + if ((hdl = libzfs_init()) == NULL) + return (1); + libzfs_print_on_error(hdl, B_TRUE); + + volname = arg + pathlen; + + /* check the configuration of the pool */ + if ((p = strchr(volname, '/')) == NULL) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "malformed dataset name")); + (void) zfs_error(hdl, EZFS_INVALIDNAME, errbuf); + return (1); + } else if (p - volname >= ZFS_MAXNAMELEN) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "dataset name is too long")); + (void) zfs_error(hdl, EZFS_NAMETOOLONG, errbuf); + return (1); + } else { + (void) strncpy(poolname, volname, p - volname); + poolname[p - volname] = '\0'; + } + + if ((zhp = zpool_open(hdl, poolname)) == NULL) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "could not open pool '%s'"), poolname); + (void) zfs_error(hdl, EZFS_OPENFAILED, errbuf); + goto out; + } + config = zpool_get_config(zhp, NULL); + if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) != 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "could not obtain vdev configuration for '%s'"), poolname); + (void) zfs_error(hdl, EZFS_INVALCONFIG, errbuf); + goto out; + } + + verify(nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_CHILDREN, + &top, &toplevels) == 0); + + if (!supported_dump_vdev_type(hdl, top[0], errbuf)) { + goto out; + } + ret = 0; + +out: + if (zhp) + zpool_close(zhp); + libzfs_fini(hdl); + return (ret); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c new file mode 100644 index 0000000..97f18d7 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c @@ -0,0 +1,3331 @@ +/* + * 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, 2014 by Delphix. All rights reserved. + * Copyright (c) 2012, Joyent, Inc. All rights reserved. + * Copyright (c) 2012 Pawel Jakub Dawidek <pawel@dawidek.net>. + * All rights reserved. + * Copyright (c) 2013 Steven Hartland. All rights reserved. + */ + +#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 <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 *, recvflags_t *, + int, const char *, nvlist_t *, avl_tree_t *, char **, int, uint64_t *); + +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; +} 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 +cksum_and_write(const void *buf, uint64_t len, zio_cksum_t *zc, int outfd) +{ + fletcher_4_incremental_native(buf, len, zc); + return (write(outfd, buf, len)); +} + +/* + * 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 = malloc(1<<20); + dmu_replay_record_t thedrr; + dmu_replay_record_t *drr = &thedrr; + struct drr_begin *drrb = &thedrr.drr_u.drr_begin; + struct drr_end *drre = &thedrr.drr_u.drr_end; + struct drr_object *drro = &thedrr.drr_u.drr_object; + struct drr_write *drrw = &thedrr.drr_u.drr_write; + struct drr_spill *drrs = &thedrr.drr_u.drr_spill; + struct drr_write_embedded *drrwe = &thedrr.drr_u.drr_write_embedded; + FILE *ofp; + int outfd; + dmu_replay_record_t wbr_drr = {0}; + struct drr_write_byref *wbr_drrr = &wbr_drr.drr_u.drr_write_byref; + 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; + + /* Initialize the write-by-reference block. */ + wbr_drr.drr_type = DRR_WRITE_BYREF; + wbr_drr.drr_payloadlen = 0; + + outfd = dda->outputfd; + ofp = fdopen(dda->inputfd, "r"); + while (ssread(drr, sizeof (dmu_replay_record_t), ofp) != 0) { + + switch (drr->drr_type) { + case DRR_BEGIN: + { + int fflags; + ZIO_SET_CHECKSUM(&stream_cksum, 0, 0, 0, 0); + + /* 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 (cksum_and_write(drr, sizeof (dmu_replay_record_t), + &stream_cksum, outfd) == -1) + goto out; + if (DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo) == + DMU_COMPOUNDSTREAM && drr->drr_payloadlen != 0) { + int sz = drr->drr_payloadlen; + + if (sz > 1<<20) { + free(buf); + buf = malloc(sz); + } + (void) ssread(buf, sz, ofp); + if (ferror(stdin)) + perror("fread"); + if (cksum_and_write(buf, sz, &stream_cksum, + outfd) == -1) + goto out; + } + break; + } + + case DRR_END: + { + /* use the recalculated checksum */ + ZIO_SET_CHECKSUM(&drre->drr_checksum, + stream_cksum.zc_word[0], stream_cksum.zc_word[1], + stream_cksum.zc_word[2], stream_cksum.zc_word[3]); + if ((write(outfd, drr, + sizeof (dmu_replay_record_t))) == -1) + goto out; + break; + } + + case DRR_OBJECT: + { + if (cksum_and_write(drr, sizeof (dmu_replay_record_t), + &stream_cksum, outfd) == -1) + goto out; + if (drro->drr_bonuslen > 0) { + (void) ssread(buf, + P2ROUNDUP((uint64_t)drro->drr_bonuslen, 8), + ofp); + if (cksum_and_write(buf, + P2ROUNDUP((uint64_t)drro->drr_bonuslen, 8), + &stream_cksum, outfd) == -1) + goto out; + } + break; + } + + case DRR_SPILL: + { + if (cksum_and_write(drr, sizeof (dmu_replay_record_t), + &stream_cksum, outfd) == -1) + goto out; + (void) ssread(buf, drrs->drr_length, ofp); + if (cksum_and_write(buf, drrs->drr_length, + &stream_cksum, outfd) == -1) + goto out; + break; + } + + case DRR_FREEOBJECTS: + { + if (cksum_and_write(drr, sizeof (dmu_replay_record_t), + &stream_cksum, outfd) == -1) + goto out; + break; + } + + case DRR_WRITE: + { + dataref_t dataref; + + (void) ssread(buf, drrw->drr_length, 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, drrw->drr_length); + 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)) { + /* block already present in stream */ + wbr_drrr->drr_object = drrw->drr_object; + wbr_drrr->drr_offset = drrw->drr_offset; + wbr_drrr->drr_length = drrw->drr_length; + 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 (cksum_and_write(&wbr_drr, + sizeof (dmu_replay_record_t), &stream_cksum, + outfd) == -1) + goto out; + } else { + /* block not previously seen */ + if (cksum_and_write(drr, + sizeof (dmu_replay_record_t), &stream_cksum, + outfd) == -1) + goto out; + if (cksum_and_write(buf, + drrw->drr_length, + &stream_cksum, outfd) == -1) + goto out; + } + break; + } + + case DRR_WRITE_EMBEDDED: + { + if (cksum_and_write(drr, sizeof (dmu_replay_record_t), + &stream_cksum, outfd) == -1) + goto out; + (void) ssread(buf, + P2ROUNDUP((uint64_t)drrwe->drr_psize, 8), ofp); + if (cksum_and_write(buf, + P2ROUNDUP((uint64_t)drrwe->drr_psize, 8), + &stream_cksum, outfd) == -1) + goto out; + break; + } + + case DRR_FREE: + { + if (cksum_and_write(drr, sizeof (dmu_replay_record_t), + &stream_cksum, outfd) == -1) + goto out; + break; + } + + default: + (void) printf("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 = arg1; + const fsavl_node_t *fn2 = arg2; + + if (fn1->fn_guid > fn2->fn_guid) + return (+1); + else if (fn1->fn_guid < fn2->fn_guid) + return (-1); + else + return (0); +} + +/* + * 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 { + uint64_t parent_fromsnap_guid; + nvlist_t *parent_snaps; + nvlist_t *fss; + nvlist_t *snapprops; + const char *fromsnap; + const char *tosnap; + boolean_t recursive; + + /* + * 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; + char *snapname; + nvlist_t *nv; + + snapname = strrchr(zhp->zfs_name, '@')+1; + + 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)); + } + } +} + +/* + * 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 parent_fromsnap_guid_save = sd->parent_fromsnap_guid; + uint64_t guid = zhp->zfs_dmustats.dds_guid; + char guidstring[64]; + + VERIFY(0 == nvlist_alloc(&nvfs, NV_UNIQUE_NAME, 0)); + VERIFY(0 == nvlist_add_string(nvfs, "name", zhp->zfs_name)); + VERIFY(0 == nvlist_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) + return (-1); + 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 */ + 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); + VERIFY(0 == nvlist_add_nvlist(nvfs, "snaps", sd->parent_snaps)); + VERIFY(0 == nvlist_add_nvlist(nvfs, "snapprops", sd->snapprops)); + nvlist_free(sd->parent_snaps); + nvlist_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); + + sd->parent_fromsnap_guid = parent_fromsnap_guid_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, nvlist_t **nvlp, avl_tree_t **avlp) +{ + zfs_handle_t *zhp; + send_data_t sd = { 0 }; + int error; + + 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.fromsnap = fromsnap; + sd.tosnap = tosnap; + sd.recursive = recursive; + + 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_MAXNAMELEN]; + uint64_t prevsnap_obj; + boolean_t seenfrom, seento, replicate, doall, fromorigin; + boolean_t verbose, dryrun, parsable, progress, embed_data; + 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_MAXNAMELEN]; + int cleanup_fd; + uint64_t size; +} send_dump_data_t; + +static int +estimate_ioctl(zfs_handle_t *zhp, uint64_t fromsnap_obj, + boolean_t fromorigin, uint64_t *sizep) +{ + zfs_cmd_t zc = { 0 }; + libzfs_handle_t *hdl = zhp->zfs_hdl; + + 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_obj = fromorigin; + zc.zc_sendobj = zfs_prop_get_int(zhp, ZFS_PROP_OBJSETID); + zc.zc_fromobj = fromsnap_obj; + zc.zc_guid = 1; /* estimate flag */ + + if (zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_SEND, &zc) != 0) { + char errbuf[1024]; + (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, + "warning: cannot estimate space for '%s'"), zhp->zfs_name); + + 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: + 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)); + } + } + + *sizep = zc.zc_objset_type; + + 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 sun + 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; + char buf[16]; + + time_t t; + struct tm *tm; + + assert(zhp->zfs_type == ZFS_TYPE_SNAPSHOT); + (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); + + if (!pa->pa_parsable) + (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_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 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; + int err; + boolean_t isfromsnap, istosnap, fromorigin; + boolean_t exclude = B_FALSE; + + 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->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) { + uint64_t size; + err = estimate_ioctl(zhp, sdd->prevsnap_obj, + fromorigin, &size); + + if (sdd->parsable) { + if (sdd->prevsnap[0] != '\0') { + (void) fprintf(stderr, "incremental\t%s\t%s", + sdd->prevsnap, zhp->zfs_name); + } else { + (void) fprintf(stderr, "full\t%s", + zhp->zfs_name); + } + } else { + (void) fprintf(stderr, dgettext(TEXT_DOMAIN, + "send from @%s to %s"), + sdd->prevsnap, zhp->zfs_name); + } + if (err == 0) { + if (sdd->parsable) { + (void) fprintf(stderr, "\t%llu\n", + (longlong_t)size); + } else { + char buf[16]; + zfs_nicenum(size, buf, sizeof (buf)); + (void) fprintf(stderr, dgettext(TEXT_DOMAIN, + " estimated size is %s\n"), buf); + } + sdd->size += size; + } else { + (void) fprintf(stderr, "\n"); + } + } + + 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; + + if (err = pthread_create(&tid, NULL, + send_progress_thread, &pa)) { + zfs_close(zhp); + return (err); + } + } + + enum lzc_send_flags flags = 0; + if (sdd->embed_data) + flags |= LZC_SEND_FLAG_EMBED_DATA; + + 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; + 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; + + rv = zfs_iter_snapshots_sorted(zhp, dump_snapshot, arg); + 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); +} + +/* + * 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; + + (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)) { + 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)) { + (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, &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 = cksum_and_write(&drr, sizeof (drr), &zc, outfd); + + /* write header nvlist */ + if (err != -1 && packbuf != NULL) { + err = cksum_and_write(packbuf, buflen, &zc, + outfd); + } + free(packbuf); + if (err == -1) { + err = errno; + 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.dryrun = flags->dryrun; + sdd.embed_data = flags->embed_data; + sdd.filter_cb = filter_func; + sdd.filter_cb_arg = cb_arg; + if (debugnvp) + sdd.debugnv = *debugnvp; + + /* + * 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->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(stderr, "size\t%llu\n", + (longlong_t)sdd.size); + } else { + char buf[16]; + zfs_nicenum(sdd.size, buf, sizeof (buf)); + (void) fprintf(stderr, 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, + enum lzc_send_flags flags) +{ + int err; + libzfs_handle_t *hdl = zhp->zfs_hdl; + + char errbuf[1024]; + (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, + "warning: cannot send '%s'"), zhp->zfs_name); + + err = lzc_send(zhp->zfs_name, from, fd, 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; + + 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) + fletcher_4_incremental_byteswap(buf, ilen, zc); + else + 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; + zfs_cmd_t zc = { 0 }; + 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); + + zc.zc_objset_type = DMU_OST_ZFS; + (void) strlcpy(zc.zc_name, name, sizeof (zc.zc_name)); + + if (tryname) { + (void) strcpy(newname, tryname); + + (void) strlcpy(zc.zc_value, tryname, sizeof (zc.zc_value)); + + if (flags->verbose) { + (void) printf("attempting rename %s to %s\n", + zc.zc_name, zc.zc_value); + } + err = ioctl(hdl->libzfs_fd, ZFS_IOC_RENAME, &zc); + 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_MAXNAMELEN, "%.*srecv-%u-%u", + baselen, name, getpid(), seq); + (void) strlcpy(zc.zc_value, newname, sizeof (zc.zc_value)); + + if (flags->verbose) { + (void) printf("failed - trying rename %s to %s\n", + zc.zc_name, zc.zc_value); + } + err = ioctl(hdl->libzfs_fd, ZFS_IOC_RENAME, &zc); + 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) +{ + zfs_cmd_t zc = { 0 }; + 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); + + zc.zc_objset_type = DMU_OST_ZFS; + zc.zc_defer_destroy = defer; + (void) strlcpy(zc.zc_name, name, sizeof (zc.zc_name)); + + if (flags->verbose) + (void) printf("attempting destroy %s\n", zc.zc_name); + err = ioctl(hdl->libzfs_fd, ZFS_IOC_DESTROY, &zc); + if (err == 0) { + if (flags->verbose) + (void) printf("success\n"); + changelist_remove(clp, zc.zc_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; + 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; + int err; + + if (gtnd->skip != NULL && + strcmp(zhp->zfs_name, gtnd->skip) == 0) { + return (0); + } + + if (zhp->zfs_dmustats.dds_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); + 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, + char *name) +{ + /* exhaustive search all local snapshots */ + char pname[ZFS_MAXNAMELEN]; + guid_to_name_data_t gtnd; + int err = 0; + zfs_handle_t *zhp; + char *cp; + + gtnd.guid = guid; + gtnd.name = name; + gtnd.skip = NULL; + + (void) strlcpy(pname, parent, sizeof (pname)); + + /* + * Search progressively larger portions of the hierarchy. This will + * select the "most local" version of the origin snapshot in the case + * that there are multiple matching snapshots in the system. + */ + while ((cp = strrchr(pname, '/')) != NULL) { + + /* Chop off the last component and open the parent */ + *cp = '\0'; + zhp = make_dataset_handle(hdl, pname); + + if (zhp == NULL) + continue; + + err = zfs_iter_children(zhp, guid_to_name_cb, >nd); + zfs_close(zhp); + if (err == EEXIST) + return (0); + + /* + * Remember the dataset that we already searched, so we + * skip it next time through. + */ + gtnd.skip = pname; + } + + 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_MAXNAMELEN]; + 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_MAXNAMELEN]; + 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, &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_MAXNAMELEN]; + + 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, "%lu", 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_MAXNAMELEN]; + char tryname[ZFS_MAXNAMELEN]; + + (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, "%lu", 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, "%lu", 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_MAXNAMELEN]; + + 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 *cp; + char tofs[ZFS_MAXNAMELEN]; + char sendfs[ZFS_MAXNAMELEN]; + 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) { + nvlist_t *renamed = NULL; + nvpair_t *pair = NULL; + + (void) strlcpy(tofs, destname, ZFS_MAXNAMELEN); + 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, "/", + ZFS_MAXNAMELEN); + 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], + ZFS_MAXNAMELEN); + *strchr(tofs, '@') = '\0'; + } + + if (recursive && !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, + ZFS_MAXNAMELEN); + if ((cp = strchr(sendfs, '@')) != NULL) + *cp = '\0'; + + /* 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, flags, fd, + sendfs, stream_nv, stream_avl, top_zfs, cleanup_fd, + action_handlep); + if (error == ENODATA) { + error = 0; + break; + } + anyerr |= error; + } while (error == 0); + + if (drr->drr_payloadlen != 0 && 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); + if (stream_nv) + 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 = malloc(1<<20); + 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: + /* NB: not to be used on v2 stream packages */ + if (drr->drr_payloadlen != 0) { + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "invalid substream header")); + return (zfs_error(hdl, EZFS_BADSTREAM, errbuf)); + } + 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_length = + BSWAP_64(drr->drr_u.drr_write.drr_length); + } + (void) recv_read(hdl, fd, buf, + drr->drr_u.drr_write.drr_length, B_FALSE, NULL); + break; + case DRR_SPILL: + if (byteswap) { + drr->drr_u.drr_write.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); +} + +/* + * 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, + 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) +{ + 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; + + 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) { + char *snapname; + 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 (drrb->drr_flags & DRR_FLAG_CLONE) { + if (guid_to_name(hdl, zc.zc_value, + drrb->drr_fromguid, 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); + } + + stream_wantsnewfs = (drrb->drr_fromguid == 0 || + (drrb->drr_flags & DRR_FLAG_CLONE)); + + 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_MAXNAMELEN]; + (void) strcpy(suffix, strrchr(zc.zc_value, '/')); + if (guid_to_name(hdl, zc.zc_name, parent_snapguid, + 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 + */ + (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_MAXNAMELEN]; + (void) strcpy(snap, strchr(zc.zc_value, '@')); + if (guid_to_name(hdl, zc.zc_name, drrb->drr_fromguid, + 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. Therefore this should either + * be an incremental, or the stream specifies a new fs + * (full stream or clone) and they want us to blow it + * away (and have therefore specified -F and removed any + * snapshots). + */ + if (stream_wantsnewfs) { + 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 ((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); + } + } + zfs_close(zhp); + } else { + /* + * 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)); + } + + newfs = B_TRUE; + } + + zc.zc_begin_record = drr_noswap->drr_u.drr_begin; + zc.zc_cookie = infd; + zc.zc_guid = flags->force; + 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 { + (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, + &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: + zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, + "invalid stream (checksum mismatch)")); + (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) { + 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, 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) +{ + 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)); + } + + /* 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)); + 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_MAXNAMELEN]; + 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, ZFS_MAXNAMELEN); + if ((cp = strchr(nonpackage_sendfs, '@')) != NULL) + *cp = '\0'; + sendfs = nonpackage_sendfs; + } + return (zfs_receive_one(hdl, infd, tosnap, flags, + &drr, &drr_noswap, sendfs, stream_nv, stream_avl, + top_zfs, cleanup_fd, action_handlep)); + } 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). + */ +int +zfs_receive(libzfs_handle_t *hdl, const char *tosnap, recvflags_t *flags, + int infd, avl_tree_t *stream_avl) +{ + char *top_zfs = NULL; + int err; + int cleanup_fd; + uint64_t action_handle = 0; + + cleanup_fd = open(ZFS_DEV, O_RDWR|O_EXCL); + VERIFY(cleanup_fd >= 0); + + err = zfs_receive_impl(hdl, tosnap, flags, infd, NULL, NULL, + stream_avl, &top_zfs, cleanup_fd, &action_handle); + + 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); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c new file mode 100644 index 0000000..906883c --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c @@ -0,0 +1,467 @@ +/* + * 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 by Delphix. All rights reserved. + * Copyright (c) 2013 Steven Hartland. All rights reserved. + */ + +/* + * This file contains the functions which analyze the status of a pool. This + * include both the status of an active pool, as well as the status exported + * pools. Returns one of the ZPOOL_STATUS_* defines describing the status of + * the pool. This status is independent (to a certain degree) from the state of + * the pool. A pool's state describes only whether or not it is capable of + * providing the necessary fault tolerance for data. The status describes the + * overall status of devices. A pool that is online can still have a device + * that is experiencing errors. + * + * Only a subset of the possible faults can be detected using 'zpool status', + * and not all possible errors correspond to a FMA message ID. The explanation + * is left up to the caller, depending on whether it is a live pool or an + * import. + */ + +#include <libzfs.h> +#include <string.h> +#include <unistd.h> +#include "libzfs_impl.h" +#include "zfeature_common.h" + +/* + * Message ID table. This must be kept in sync with the ZPOOL_STATUS_* defines + * in libzfs.h. Note that there are some status results which go past the end + * of this table, and hence have no associated message ID. + */ +static char *zfs_msgid_table[] = { + "ZFS-8000-14", + "ZFS-8000-2Q", + "ZFS-8000-3C", + "ZFS-8000-4J", + "ZFS-8000-5E", + "ZFS-8000-6X", + "ZFS-8000-72", + "ZFS-8000-8A", + "ZFS-8000-9P", + "ZFS-8000-A5", + "ZFS-8000-EY", + "ZFS-8000-HC", + "ZFS-8000-JQ", + "ZFS-8000-K4", +}; + +#define NMSGID (sizeof (zfs_msgid_table) / sizeof (zfs_msgid_table[0])) + +/* ARGSUSED */ +static int +vdev_missing(vdev_stat_t *vs, uint_t vsc) +{ + return (vs->vs_state == VDEV_STATE_CANT_OPEN && + vs->vs_aux == VDEV_AUX_OPEN_FAILED); +} + +/* ARGSUSED */ +static int +vdev_faulted(vdev_stat_t *vs, uint_t vsc) +{ + return (vs->vs_state == VDEV_STATE_FAULTED); +} + +/* ARGSUSED */ +static int +vdev_errors(vdev_stat_t *vs, uint_t vsc) +{ + return (vs->vs_state == VDEV_STATE_DEGRADED || + vs->vs_read_errors != 0 || vs->vs_write_errors != 0 || + vs->vs_checksum_errors != 0); +} + +/* ARGSUSED */ +static int +vdev_broken(vdev_stat_t *vs, uint_t vsc) +{ + return (vs->vs_state == VDEV_STATE_CANT_OPEN); +} + +/* ARGSUSED */ +static int +vdev_offlined(vdev_stat_t *vs, uint_t vsc) +{ + return (vs->vs_state == VDEV_STATE_OFFLINE); +} + +/* ARGSUSED */ +static int +vdev_removed(vdev_stat_t *vs, uint_t vsc) +{ + return (vs->vs_state == VDEV_STATE_REMOVED); +} + +static int +vdev_non_native_ashift(vdev_stat_t *vs, uint_t vsc) +{ + return (VDEV_STAT_VALID(vs_physical_ashift, vsc) && + vs->vs_configured_ashift < vs->vs_physical_ashift); +} + +/* + * Detect if any leaf devices that have seen errors or could not be opened. + */ +static boolean_t +find_vdev_problem(nvlist_t *vdev, int (*func)(vdev_stat_t *, uint_t), + boolean_t ignore_replacing) +{ + nvlist_t **child; + vdev_stat_t *vs; + uint_t c, vsc, children; + + /* + * Ignore problems within a 'replacing' vdev, since we're presumably in + * the process of repairing any such errors, and don't want to call them + * out again. We'll pick up the fact that a resilver is happening + * later. + */ + if (ignore_replacing == B_TRUE) { + char *type; + + verify(nvlist_lookup_string(vdev, ZPOOL_CONFIG_TYPE, + &type) == 0); + if (strcmp(type, VDEV_TYPE_REPLACING) == 0) + return (B_FALSE); + } + + if (nvlist_lookup_nvlist_array(vdev, ZPOOL_CONFIG_CHILDREN, &child, + &children) == 0) { + for (c = 0; c < children; c++) + if (find_vdev_problem(child[c], func, ignore_replacing)) + return (B_TRUE); + } else { + verify(nvlist_lookup_uint64_array(vdev, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &vsc) == 0); + + if (func(vs, vsc) != 0) + return (B_TRUE); + } + + /* + * Check any L2 cache devs + */ + if (nvlist_lookup_nvlist_array(vdev, ZPOOL_CONFIG_L2CACHE, &child, + &children) == 0) { + for (c = 0; c < children; c++) + if (find_vdev_problem(child[c], func, ignore_replacing)) + return (B_TRUE); + } + + return (B_FALSE); +} + +/* + * Active pool health status. + * + * To determine the status for a pool, we make several passes over the config, + * picking the most egregious error we find. In order of importance, we do the + * following: + * + * - Check for a complete and valid configuration + * - Look for any faulted or missing devices in a non-replicated config + * - Check for any data errors + * - Check for any faulted or missing devices in a replicated config + * - Look for any devices showing errors + * - Check for any resilvering devices + * + * There can obviously be multiple errors within a single pool, so this routine + * only picks the most damaging of all the current errors to report. + */ +static zpool_status_t +check_status(nvlist_t *config, boolean_t isimport) +{ + nvlist_t *nvroot; + vdev_stat_t *vs; + pool_scan_stat_t *ps = NULL; + uint_t vsc, psc; + uint64_t nerr; + uint64_t version; + uint64_t stateval; + uint64_t suspended; + uint64_t hostid = 0; + + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION, + &version) == 0); + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + verify(nvlist_lookup_uint64_array(nvroot, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &vsc) == 0); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE, + &stateval) == 0); + + /* + * Currently resilvering a vdev + */ + (void) nvlist_lookup_uint64_array(nvroot, ZPOOL_CONFIG_SCAN_STATS, + (uint64_t **)&ps, &psc); + if (ps && ps->pss_func == POOL_SCAN_RESILVER && + ps->pss_state == DSS_SCANNING) + return (ZPOOL_STATUS_RESILVERING); + + /* + * Pool last accessed by another system. + */ + (void) nvlist_lookup_uint64(config, ZPOOL_CONFIG_HOSTID, &hostid); + if (hostid != 0 && (unsigned long)hostid != gethostid() && + stateval == POOL_STATE_ACTIVE) + return (ZPOOL_STATUS_HOSTID_MISMATCH); + + /* + * Newer on-disk version. + */ + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + vs->vs_aux == VDEV_AUX_VERSION_NEWER) + return (ZPOOL_STATUS_VERSION_NEWER); + + /* + * Unsupported feature(s). + */ + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + vs->vs_aux == VDEV_AUX_UNSUP_FEAT) { + nvlist_t *nvinfo; + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO, + &nvinfo) == 0); + if (nvlist_exists(nvinfo, ZPOOL_CONFIG_CAN_RDONLY)) + return (ZPOOL_STATUS_UNSUP_FEAT_WRITE); + return (ZPOOL_STATUS_UNSUP_FEAT_READ); + } + + /* + * Check that the config is complete. + */ + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + vs->vs_aux == VDEV_AUX_BAD_GUID_SUM) + return (ZPOOL_STATUS_BAD_GUID_SUM); + + /* + * Check whether the pool has suspended due to failed I/O. + */ + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_SUSPENDED, + &suspended) == 0) { + if (suspended == ZIO_FAILURE_MODE_CONTINUE) + return (ZPOOL_STATUS_IO_FAILURE_CONTINUE); + return (ZPOOL_STATUS_IO_FAILURE_WAIT); + } + + /* + * Could not read a log. + */ + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + vs->vs_aux == VDEV_AUX_BAD_LOG) { + return (ZPOOL_STATUS_BAD_LOG); + } + + /* + * Bad devices in non-replicated config. + */ + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + find_vdev_problem(nvroot, vdev_faulted, B_TRUE)) + return (ZPOOL_STATUS_FAULTED_DEV_NR); + + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + find_vdev_problem(nvroot, vdev_missing, B_TRUE)) + return (ZPOOL_STATUS_MISSING_DEV_NR); + + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + find_vdev_problem(nvroot, vdev_broken, B_TRUE)) + return (ZPOOL_STATUS_CORRUPT_LABEL_NR); + + /* + * Corrupted pool metadata + */ + if (vs->vs_state == VDEV_STATE_CANT_OPEN && + vs->vs_aux == VDEV_AUX_CORRUPT_DATA) + return (ZPOOL_STATUS_CORRUPT_POOL); + + /* + * Persistent data errors. + */ + if (!isimport) { + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_ERRCOUNT, + &nerr) == 0 && nerr != 0) + return (ZPOOL_STATUS_CORRUPT_DATA); + } + + /* + * Missing devices in a replicated config. + */ + if (find_vdev_problem(nvroot, vdev_faulted, B_TRUE)) + return (ZPOOL_STATUS_FAULTED_DEV_R); + if (find_vdev_problem(nvroot, vdev_missing, B_TRUE)) + return (ZPOOL_STATUS_MISSING_DEV_R); + if (find_vdev_problem(nvroot, vdev_broken, B_TRUE)) + return (ZPOOL_STATUS_CORRUPT_LABEL_R); + + /* + * Devices with errors + */ + if (!isimport && find_vdev_problem(nvroot, vdev_errors, B_TRUE)) + return (ZPOOL_STATUS_FAILING_DEV); + + /* + * Offlined devices + */ + if (find_vdev_problem(nvroot, vdev_offlined, B_TRUE)) + return (ZPOOL_STATUS_OFFLINE_DEV); + + /* + * Removed device + */ + if (find_vdev_problem(nvroot, vdev_removed, B_TRUE)) + return (ZPOOL_STATUS_REMOVED_DEV); + + /* + * Suboptimal, but usable, ashift configuration. + */ + if (find_vdev_problem(nvroot, vdev_non_native_ashift, B_FALSE)) + return (ZPOOL_STATUS_NON_NATIVE_ASHIFT); + + /* + * Outdated, but usable, version + */ + if (SPA_VERSION_IS_SUPPORTED(version) && version != SPA_VERSION) + return (ZPOOL_STATUS_VERSION_OLDER); + + /* + * Usable pool with disabled features + */ + if (version >= SPA_VERSION_FEATURES) { + int i; + nvlist_t *feat; + + if (isimport) { + feat = fnvlist_lookup_nvlist(config, + ZPOOL_CONFIG_LOAD_INFO); + feat = fnvlist_lookup_nvlist(feat, + ZPOOL_CONFIG_ENABLED_FEAT); + } else { + feat = fnvlist_lookup_nvlist(config, + ZPOOL_CONFIG_FEATURE_STATS); + } + + for (i = 0; i < SPA_FEATURES; i++) { + zfeature_info_t *fi = &spa_feature_table[i]; + if (!nvlist_exists(feat, fi->fi_guid)) + return (ZPOOL_STATUS_FEAT_DISABLED); + } + } + + return (ZPOOL_STATUS_OK); +} + +zpool_status_t +zpool_get_status(zpool_handle_t *zhp, char **msgid) +{ + zpool_status_t ret = check_status(zhp->zpool_config, B_FALSE); + + if (ret >= NMSGID) + *msgid = NULL; + else + *msgid = zfs_msgid_table[ret]; + + return (ret); +} + +zpool_status_t +zpool_import_status(nvlist_t *config, char **msgid) +{ + zpool_status_t ret = check_status(config, B_TRUE); + + if (ret >= NMSGID) + *msgid = NULL; + else + *msgid = zfs_msgid_table[ret]; + + return (ret); +} + +static void +dump_ddt_stat(const ddt_stat_t *dds, int h) +{ + char refcnt[6]; + char blocks[6], lsize[6], psize[6], dsize[6]; + char ref_blocks[6], ref_lsize[6], ref_psize[6], ref_dsize[6]; + + if (dds == NULL || dds->dds_blocks == 0) + return; + + if (h == -1) + (void) strcpy(refcnt, "Total"); + else + zfs_nicenum(1ULL << h, refcnt, sizeof (refcnt)); + + zfs_nicenum(dds->dds_blocks, blocks, sizeof (blocks)); + zfs_nicenum(dds->dds_lsize, lsize, sizeof (lsize)); + zfs_nicenum(dds->dds_psize, psize, sizeof (psize)); + zfs_nicenum(dds->dds_dsize, dsize, sizeof (dsize)); + zfs_nicenum(dds->dds_ref_blocks, ref_blocks, sizeof (ref_blocks)); + zfs_nicenum(dds->dds_ref_lsize, ref_lsize, sizeof (ref_lsize)); + zfs_nicenum(dds->dds_ref_psize, ref_psize, sizeof (ref_psize)); + zfs_nicenum(dds->dds_ref_dsize, ref_dsize, sizeof (ref_dsize)); + + (void) printf("%6s %6s %5s %5s %5s %6s %5s %5s %5s\n", + refcnt, + blocks, lsize, psize, dsize, + ref_blocks, ref_lsize, ref_psize, ref_dsize); +} + +/* + * Print the DDT histogram and the column totals. + */ +void +zpool_dump_ddt(const ddt_stat_t *dds_total, const ddt_histogram_t *ddh) +{ + int h; + + (void) printf("\n"); + + (void) printf("bucket " + " allocated " + " referenced \n"); + (void) printf("______ " + "______________________________ " + "______________________________\n"); + + (void) printf("%6s %6s %5s %5s %5s %6s %5s %5s %5s\n", + "refcnt", + "blocks", "LSIZE", "PSIZE", "DSIZE", + "blocks", "LSIZE", "PSIZE", "DSIZE"); + + (void) printf("%6s %6s %5s %5s %5s %6s %5s %5s %5s\n", + "------", + "------", "-----", "-----", "-----", + "------", "-----", "-----", "-----"); + + for (h = 0; h < 64; h++) + dump_ddt_stat(&ddh->ddh_stat[h], h); + + dump_ddt_stat(dds_total, -1); + + (void) printf("\n"); +} diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_util.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_util.c new file mode 100644 index 0000000..3b59914 --- /dev/null +++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_util.c @@ -0,0 +1,1546 @@ +/* + * 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) 2013, Joyent, Inc. All rights reserved. + * Copyright (c) 2012 by Delphix. All rights reserved. + */ + +/* + * 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 <libzfs.h> +#include <libzfs_core.h> + +#include "libzfs_impl.h" +#include "zfs_prop.h" +#include "zfeature_common.h" + +int aok; + +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_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_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; + 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; + + 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) +{ + uint64_t n = num; + int index = 0; + char u; + + while (n >= 1024) { + n /= 1024; + index++; + } + + u = " KMGTPE"[index]; + + if (index == 0) { + (void) snprintf(buf, buflen, "%llu", n); + } else if ((num & ((1ULL << 10 * index) - 1)) == 0) { + /* + * If this is an even multiple of the base, always display + * without any decimal precision. + */ + (void) snprintf(buf, buflen, "%llu%c", n, u); + } else { + /* + * We want to choose a precision that reflects the best choice + * for fitting in 5 characters. This can get rather tricky when + * we have numbers that are very close to an order of magnitude. + * For example, when displaying 10239 (which is really 9.999K), + * we want only a single place of precision for 10.0K. We could + * develop some complex heuristics for this, but it's much + * easier just to try each combination in turn. + */ + int i; + for (i = 2; i >= 0; i--) { + if (snprintf(buf, buflen, "%.*f%c", i, + (double)num / (1ULL << 10 * index), u) <= 5) + break; + } + } +} + +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); + + 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 sun + 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 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 sun + 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 /* sun */ + 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; + if ((zc->zc_nvlist_dst = (uint64_t)(uintptr_t) + zfs_alloc(hdl, zc->zc_nvlist_dst_size)) == 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); + if ((zc->zc_nvlist_dst = (uint64_t)(uintptr_t) + zfs_alloc(hdl, zc->zc_nvlist_dst_size)) + == 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); +} + +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; + 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; + } + 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; + + 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 (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; + } + 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)); +} |
