Make zfs-share service resilient to stale exports

[mirror_zfs.git] / lib / libzfs / libzfs_mount.c

/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or https://opensource.org/licenses/CDDL-1.0.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */

/*
 * Copyright 2015 Nexenta Systems, Inc.  All rights reserved.
 * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
 * Copyright (c) 2014, 2022 by Delphix. All rights reserved.
 * Copyright 2016 Igor Kozhukhov <ikozhukhov@gmail.com>
 * Copyright 2017 RackTop Systems.
 * Copyright (c) 2018 Datto Inc.
 * Copyright 2018 OmniOS Community Edition (OmniOSce) Association.
 */

/*
 * 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_mount_at()
 *      zfs_unmount()
 *      zfs_unmountall()
 *
 * This file also contains the functions used to manage sharing filesystems:
 *
 *      zfs_is_shared()
 *      zfs_share()
 *      zfs_unshare()
 *      zfs_unshareall()
 *      zfs_commit_shares()
 *
 * 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 <fcntl.h>
#include <libgen.h>
#include <libintl.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <zone.h>
#include <sys/mntent.h>
#include <sys/mount.h>
#include <sys/stat.h>
#include <sys/vfs.h>
#include <sys/dsl_crypt.h>

#include <libzfs.h>

#include "libzfs_impl.h"
#include <thread_pool.h>

#include <libshare.h>
#include <sys/systeminfo.h>
#define MAXISALEN       257     /* based on sysinfo(2) man page */

static int mount_tp_nthr = 512; /* tpool threads for multi-threaded mounting */

static void zfs_mount_task(void *);

static const proto_table_t proto_table[SA_PROTOCOL_COUNT] = {
        [SA_PROTOCOL_NFS] =
            {ZFS_PROP_SHARENFS, EZFS_SHARENFSFAILED, EZFS_UNSHARENFSFAILED},
        [SA_PROTOCOL_SMB] =
            {ZFS_PROP_SHARESMB, EZFS_SHARESMBFAILED, EZFS_UNSHARESMBFAILED},
};

static const enum sa_protocol share_all_proto[SA_PROTOCOL_COUNT + 1] = {
        SA_PROTOCOL_NFS,
        SA_PROTOCOL_SMB,
        SA_NO_PROTOCOL
};



static boolean_t
dir_is_empty_stat(const char *dirname)
{
        struct stat st;

        /*
         * We only want to return false if the given path is a non empty
         * directory, all other errors are handled elsewhere.
         */
        if (stat(dirname, &st) < 0 || !S_ISDIR(st.st_mode)) {
                return (B_TRUE);
        }

        /*
         * An empty directory will still have two entries in it, one
         * entry for each of "." and "..".
         */
        if (st.st_size > 2) {
                return (B_FALSE);
        }

        return (B_TRUE);
}

static boolean_t
dir_is_empty_readdir(const char *dirname)
{
        DIR *dirp;
        struct dirent64 *dp;
        int dirfd;

        if ((dirfd = openat(AT_FDCWD, dirname,
            O_RDONLY | O_NDELAY | O_LARGEFILE | O_CLOEXEC, 0)) < 0) {
                return (B_TRUE);
        }

        if ((dirp = fdopendir(dirfd)) == NULL) {
                (void) close(dirfd);
                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);
}

/*
 * 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)
{
        struct statfs64 st;

        /*
         * If the statvfs call fails or the filesystem is not a ZFS
         * filesystem, fall back to the slow path which uses readdir.
         */
        if ((statfs64(dirname, &st) != 0) ||
            (st.f_type != ZFS_SUPER_MAGIC)) {
                return (dir_is_empty_readdir(dirname));
        }

        /*
         * At this point, we know the provided path is on a ZFS
         * filesystem, so we can use stat instead of readdir to
         * determine if the directory is empty or not. We try to avoid
         * using readdir because that requires opening "dirname"; this
         * open file descriptor can potentially end up in a child
         * process if there's a concurrent fork, thus preventing the
         * zfs_mount() from otherwise succeeding (the open file
         * descriptor inherited by the child process will cause the
         * parent's mount to fail with EBUSY). The performance
         * implications of replacing the open, read, and close with a
         * single stat is nice; but is not the main motivation for the
         * added complexity.
         */
        return (dir_is_empty_stat(dirname));
}

/*
 * 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));
}

/*
 * Checks any higher order concerns about whether the given dataset is
 * mountable, false otherwise.  zfs_is_mountable_internal specifically assumes
 * that the caller has verified the sanity of mounting the dataset at
 * its mountpoint to the extent the caller wants.
 */
static boolean_t
zfs_is_mountable_internal(zfs_handle_t *zhp)
{
        if (zfs_prop_get_int(zhp, ZFS_PROP_ZONED) &&
            getzoneid() == GLOBAL_ZONEID)
                return (B_FALSE);

        return (B_TRUE);
}

/*
 * 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, int flags)
{
        char sourceloc[MAXNAMELEN];
        zprop_source_t sourcetype;

        if (!zfs_prop_valid_for_type(ZFS_PROP_MOUNTPOINT, zhp->zfs_type,
            B_FALSE))
                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_is_mountable_internal(zhp))
                return (B_FALSE);

        if (zfs_prop_get_int(zhp, ZFS_PROP_REDACTED) && !(flags & MS_FORCE))
                return (B_FALSE);

        if (source)
                *source = sourcetype;

        return (B_TRUE);
}

/*
 * The filesystem is mounted by invoking the system mount utility rather
 * than by the system call mount(2).  This ensures that the /etc/mtab
 * file is correctly locked for the update.  Performing our own locking
 * and /etc/mtab update requires making an unsafe assumption about how
 * the mount utility performs its locking.  Unfortunately, this also means
 * in the case of a mount failure we do not have the exact errno.  We must
 * make due with return value from the mount process.
 *
 * In the long term a shared library called libmount is under development
 * which provides a common API to address the locking and errno issues.
 * Once the standard mount utility has been updated to use this library
 * we can add an autoconf check to conditionally use it.
 *
 * http://www.kernel.org/pub/linux/utils/util-linux/libmount-docs/index.html
 */

static int
zfs_add_option(zfs_handle_t *zhp, char *options, int len,
    zfs_prop_t prop, const char *on, const char *off)
{
        char *source;
        uint64_t value;

        /* Skip adding duplicate default options */
        if ((strstr(options, on) != NULL) || (strstr(options, off) != NULL))
                return (0);

        /*
         * zfs_prop_get_int() is not used to ensure our mount options
         * are not influenced by the current /proc/self/mounts contents.
         */
        value = getprop_uint64(zhp, prop, &source);

        (void) strlcat(options, ",", len);
        (void) strlcat(options, value ? on : off, len);

        return (0);
}

static int
zfs_add_options(zfs_handle_t *zhp, char *options, int len)
{
        int error = 0;

        error = zfs_add_option(zhp, options, len,
            ZFS_PROP_ATIME, MNTOPT_ATIME, MNTOPT_NOATIME);
        /*
         * don't add relatime/strictatime when atime=off, otherwise strictatime
         * will force atime=on
         */
        if (strstr(options, MNTOPT_NOATIME) == NULL) {
                error = zfs_add_option(zhp, options, len,
                    ZFS_PROP_RELATIME, MNTOPT_RELATIME, MNTOPT_STRICTATIME);
        }
        error = error ? error : zfs_add_option(zhp, options, len,
            ZFS_PROP_DEVICES, MNTOPT_DEVICES, MNTOPT_NODEVICES);
        error = error ? error : zfs_add_option(zhp, options, len,
            ZFS_PROP_EXEC, MNTOPT_EXEC, MNTOPT_NOEXEC);
        error = error ? error : zfs_add_option(zhp, options, len,
            ZFS_PROP_READONLY, MNTOPT_RO, MNTOPT_RW);
        error = error ? error : zfs_add_option(zhp, options, len,
            ZFS_PROP_SETUID, MNTOPT_SETUID, MNTOPT_NOSETUID);
        error = error ? error : zfs_add_option(zhp, options, len,
            ZFS_PROP_NBMAND, MNTOPT_NBMAND, MNTOPT_NONBMAND);

        return (error);
}

int
zfs_mount(zfs_handle_t *zhp, const char *options, int flags)
{
        char mountpoint[ZFS_MAXPROPLEN];

        if (!zfs_is_mountable(zhp, mountpoint, sizeof (mountpoint), NULL,
            flags))
                return (0);

        return (zfs_mount_at(zhp, options, flags, mountpoint));
}

/*
 * Mount the given filesystem.
 */
int
zfs_mount_at(zfs_handle_t *zhp, const char *options, int flags,
    const char *mountpoint)
{
        struct stat buf;
        char mntopts[MNT_LINE_MAX];
        char overlay[ZFS_MAXPROPLEN];
        char prop_encroot[MAXNAMELEN];
        boolean_t is_encroot;
        zfs_handle_t *encroot_hp = zhp;
        libzfs_handle_t *hdl = zhp->zfs_hdl;
        uint64_t keystatus;
        int remount = 0, rc;

        if (options == NULL) {
                (void) strlcpy(mntopts, MNTOPT_DEFAULTS, sizeof (mntopts));
        } else {
                (void) strlcpy(mntopts, options, sizeof (mntopts));
        }

        if (strstr(mntopts, MNTOPT_REMOUNT) != NULL)
                remount = 1;

        /* Potentially duplicates some checks if invoked by zfs_mount(). */
        if (!zfs_is_mountable_internal(zhp))
                return (0);

        /*
         * 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))
                (void) strlcat(mntopts, "," MNTOPT_RO, sizeof (mntopts));

        /*
         * Append default mount options which apply to the mount point.
         * This is done because under Linux (unlike Solaris) multiple mount
         * points may reference a single super block.  This means that just
         * given a super block there is no back reference to update the per
         * mount point options.
         */
        rc = zfs_add_options(zhp, mntopts, sizeof (mntopts));
        if (rc) {
                zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
                    "default options unavailable"));
                return (zfs_error_fmt(hdl, EZFS_MOUNTFAILED,
                    dgettext(TEXT_DOMAIN, "cannot mount '%s'"),
                    mountpoint));
        }

        /*
         * If the filesystem is encrypted the key must be loaded  in order to
         * mount. If the key isn't loaded, the MS_CRYPT flag decides whether
         * or not we attempt to load the keys. Note: we must call
         * zfs_refresh_properties() here since some callers of this function
         * (most notably zpool_enable_datasets()) may implicitly load our key
         * by loading the parent's key first.
         */
        if (zfs_prop_get_int(zhp, ZFS_PROP_ENCRYPTION) != ZIO_CRYPT_OFF) {
                zfs_refresh_properties(zhp);
                keystatus = zfs_prop_get_int(zhp, ZFS_PROP_KEYSTATUS);

                /*
                 * If the key is unavailable and MS_CRYPT is set give the
                 * user a chance to enter the key. Otherwise just fail
                 * immediately.
                 */
                if (keystatus == ZFS_KEYSTATUS_UNAVAILABLE) {
                        if (flags & MS_CRYPT) {
                                rc = zfs_crypto_get_encryption_root(zhp,
                                    &is_encroot, prop_encroot);
                                if (rc) {
                                        zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
                                            "Failed to get encryption root for "
                                            "'%s'."), zfs_get_name(zhp));
                                        return (rc);
                                }

                                if (!is_encroot) {
                                        encroot_hp = zfs_open(hdl, prop_encroot,
                                            ZFS_TYPE_DATASET);
                                        if (encroot_hp == NULL)
                                                return (hdl->libzfs_error);
                                }

                                rc = zfs_crypto_load_key(encroot_hp,
                                    B_FALSE, NULL);

                                if (!is_encroot)
                                        zfs_close(encroot_hp);
                                if (rc)
                                        return (rc);
                        } else {
                                zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
                                    "encryption key not loaded"));
                                return (zfs_error_fmt(hdl, EZFS_MOUNTFAILED,
                                    dgettext(TEXT_DOMAIN, "cannot mount '%s'"),
                                    mountpoint));
                        }
                }

        }

        /*
         * Append zfsutil option so the mount helper allow the mount
         */
        strlcat(mntopts, "," MNTOPT_ZFSUTIL, sizeof (mntopts));

        /* 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: %s"),
                            strerror(errno));
                        return (zfs_error_fmt(hdl, EZFS_MOUNTFAILED,
                            dgettext(TEXT_DOMAIN, "cannot mount '%s'"),
                            mountpoint));
                }
        }

        /*
         * Overlay mounts are enabled by default but may be disabled
         * via the 'overlay' property. The -O flag remains for compatibility.
         */
        if (!(flags & MS_OVERLAY)) {
                if (zfs_prop_get(zhp, ZFS_PROP_OVERLAY, overlay,
                    sizeof (overlay), NULL, NULL, 0, B_FALSE) == 0) {
                        if (strcmp(overlay, "on") == 0) {
                                flags |= MS_OVERLAY;
                        }
                }
        }

        /*
         * Determine if the mountpoint is empty.  If so, refuse to perform the
         * mount.  We don't perform this check if 'remount' is
         * specified or if overlay option (-O) is given
         */
        if ((flags & MS_OVERLAY) == 0 && !remount &&
            !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));
        }

        /* perform the mount */
        rc = do_mount(zhp, mountpoint, mntopts, flags);
        if (rc) {
                /*
                 * 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 (rc == EBUSY) {
                        zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
                            "mountpoint or dataset is busy"));
                } else if (rc == EPERM) {
                        zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
                            "Insufficient privileges"));
                } else if (rc == ENOTSUP) {
                        int spa_version;

                        VERIFY(zfs_spa_version(zhp, &spa_version) == 0);
                        zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
                            "Can't mount a version %llu "
                            "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);
                } else {
                        zfs_error_aux(hdl, "%s", strerror(rc));
                }
                return (zfs_error_fmt(hdl, EZFS_MOUNTFAILED,
                    dgettext(TEXT_DOMAIN, "cannot mount '%s'"),
                    zhp->zfs_name));
        }

        /* remove the mounted entry before re-adding on remount */
        if (remount)
                libzfs_mnttab_remove(hdl, 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(zfs_handle_t *zhp, const char *mountpoint, int flags)
{
        int error;

        error = do_unmount(zhp, mountpoint, flags);
        if (error != 0) {
                int libzfs_err;

                switch (error) {
                case EBUSY:
                        libzfs_err = EZFS_BUSY;
                        break;
                case EIO:
                        libzfs_err = EZFS_IO;
                        break;
                case ENOENT:
                        libzfs_err = EZFS_NOENT;
                        break;
                case ENOMEM:
                        libzfs_err = EZFS_NOMEM;
                        break;
                case EPERM:
                        libzfs_err = EZFS_PERM;
                        break;
                default:
                        libzfs_err = EZFS_UMOUNTFAILED;
                }
                if (zhp) {
                        return (zfs_error_fmt(zhp->zfs_hdl, libzfs_err,
                            dgettext(TEXT_DOMAIN, "cannot unmount '%s'"),
                            mountpoint));
                } else {
                        return (-1);
                }
        }

        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;
        boolean_t encroot, unmounted = B_FALSE;

        /* 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(zhp, mntpt, share_all_proto) != 0) {
                        free(mntpt);
                        return (-1);
                }
                zfs_commit_shares(NULL);

                if (unmount_one(zhp, mntpt, flags) != 0) {
                        free(mntpt);
                        (void) zfs_share(zhp, NULL);
                        zfs_commit_shares(NULL);
                        return (-1);
                }

                libzfs_mnttab_remove(hdl, zhp->zfs_name);
                free(mntpt);
                unmounted = B_TRUE;
        }

        /*
         * If the MS_CRYPT flag is provided we must ensure we attempt to
         * unload the dataset's key regardless of whether we did any work
         * to unmount it. We only do this for encryption roots.
         */
        if ((flags & MS_CRYPT) != 0 &&
            zfs_prop_get_int(zhp, ZFS_PROP_ENCRYPTION) != ZIO_CRYPT_OFF) {
                zfs_refresh_properties(zhp);

                if (zfs_crypto_get_encryption_root(zhp, &encroot, NULL) != 0 &&
                    unmounted) {
                        (void) zfs_mount(zhp, NULL, 0);
                        return (-1);
                }

                if (encroot && zfs_prop_get_int(zhp, ZFS_PROP_KEYSTATUS) ==
                    ZFS_KEYSTATUS_AVAILABLE &&
                    zfs_crypto_unload_key(zhp) != 0) {
                        (void) zfs_mount(zhp, NULL, 0);
                        return (-1);
                }
        }

        zpool_disable_volume_os(zhp->zfs_name);

        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,
            CL_GATHER_ITER_MOUNTED, flags);
        if (clp == NULL)
                return (-1);

        ret = changelist_prefix(clp);
        changelist_free(clp);

        return (ret);
}

/*
 * Unshare a filesystem by mountpoint.
 */
static int
unshare_one(libzfs_handle_t *hdl, const char *name, const char *mountpoint,
    enum sa_protocol proto)
{
        int err = sa_disable_share(mountpoint, proto);
        if (err != SA_OK)
                return (zfs_error_fmt(hdl, proto_table[proto].p_unshare_err,
                    dgettext(TEXT_DOMAIN, "cannot unshare '%s': %s"),
                    name, sa_errorstr(err)));

        return (0);
}

/*
 * Share the given filesystem according to the options in the specified
 * protocol specific properties (sharenfs, sharesmb).  We rely
 * on "libshare" to do the dirty work for us.
 */
int
zfs_share(zfs_handle_t *zhp, const enum sa_protocol *proto)
{
        char mountpoint[ZFS_MAXPROPLEN];
        char shareopts[ZFS_MAXPROPLEN];
        char sourcestr[ZFS_MAXPROPLEN];
        const enum sa_protocol *curr_proto;
        zprop_source_t sourcetype;
        int err = 0;

        if (proto == NULL)
                proto = share_all_proto;

        if (!zfs_is_mountable(zhp, mountpoint, sizeof (mountpoint), NULL, 0))
                return (0);

        for (curr_proto = proto; *curr_proto != SA_NO_PROTOCOL; 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;

                err = sa_enable_share(zfs_get_name(zhp), mountpoint, shareopts,
                    *curr_proto);
                if (err != SA_OK) {
                        return (zfs_error_fmt(zhp->zfs_hdl,
                            proto_table[*curr_proto].p_share_err,
                            dgettext(TEXT_DOMAIN, "cannot share '%s: %s'"),
                            zfs_get_name(zhp), sa_errorstr(err)));
                }

        }
        return (0);
}

/*
 * Check to see if the filesystem is currently shared.
 */
boolean_t
zfs_is_shared(zfs_handle_t *zhp, char **where,
    const enum sa_protocol *proto)
{
        char *mountpoint;
        if (proto == NULL)
                proto = share_all_proto;

        if (ZFS_IS_VOLUME(zhp))
                return (B_FALSE);

        if (!zfs_is_mounted(zhp, &mountpoint))
                return (B_FALSE);

        for (const enum sa_protocol *p = proto; *p != SA_NO_PROTOCOL; ++p)
                if (sa_is_shared(mountpoint, *p)) {
                        if (where != NULL)
                                *where = mountpoint;
                        else
                                free(mountpoint);
                        return (B_TRUE);
                }

        free(mountpoint);
        return (B_FALSE);
}

void
zfs_commit_shares(const enum sa_protocol *proto)
{
        if (proto == NULL)
                proto = share_all_proto;

        for (const enum sa_protocol *p = proto; *p != SA_NO_PROTOCOL; ++p)
                sa_commit_shares(*p);
}

void
zfs_truncate_shares(const enum sa_protocol *proto)
{
        if (proto == NULL)
                proto = share_all_proto;

        for (const enum sa_protocol *p = proto; *p != SA_NO_PROTOCOL; ++p)
                sa_truncate_shares(*p);
}

/*
 * Unshare the given filesystem.
 */
int
zfs_unshare(zfs_handle_t *zhp, const char *mountpoint,
    const enum sa_protocol *proto)
{
        libzfs_handle_t *hdl = zhp->zfs_hdl;
        struct mnttab entry;

        if (proto == NULL)
                proto = share_all_proto;

        if (mountpoint != NULL || ((zfs_get_type(zhp) == ZFS_TYPE_FILESYSTEM) &&
            libzfs_mnttab_find(hdl, zfs_get_name(zhp), &entry) == 0)) {

                /* check to see if need to unmount the filesystem */
                const char *mntpt = mountpoint ?: entry.mnt_mountp;

                for (const enum sa_protocol *curr_proto = proto;
                    *curr_proto != SA_NO_PROTOCOL; curr_proto++)
                        if (sa_is_shared(mntpt, *curr_proto) &&
                            unshare_one(hdl, zhp->zfs_name,
                            mntpt, *curr_proto) != 0)
                                        return (-1);
        }

        return (0);
}

/*
 * Same as zfs_unmountall(), but for NFS and SMB unshares.
 */
int
zfs_unshareall(zfs_handle_t *zhp, const enum sa_protocol *proto)
{
        prop_changelist_t *clp;
        int ret;

        if (proto == NULL)
                proto = share_all_proto;

        clp = changelist_gather(zhp, ZFS_PROP_SHARENFS, 0, 0);
        if (clp == NULL)
                return (-1);

        ret = changelist_unshare(clp, proto);
        changelist_free(clp);

        return (ret);
}

/*
 * 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, 0))
                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);
        }
}

/*
 * Add the given zfs handle to the cb_handles array, dynamically reallocating
 * the array if it is out of space.
 */
void
libzfs_add_handle(get_all_cb_t *cbp, zfs_handle_t *zhp)
{
        if (cbp->cb_alloc == cbp->cb_used) {
                size_t newsz;
                zfs_handle_t **newhandles;

                newsz = cbp->cb_alloc != 0 ? cbp->cb_alloc * 2 : 64;
                newhandles = zfs_realloc(zhp->zfs_hdl,
                    cbp->cb_handles, cbp->cb_alloc * sizeof (zfs_handle_t *),
                    newsz * sizeof (zfs_handle_t *));
                cbp->cb_handles = newhandles;
                cbp->cb_alloc = newsz;
        }
        cbp->cb_handles[cbp->cb_used++] = zhp;
}

/*
 * Recursive helper function used during file system enumeration
 */
static int
zfs_iter_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);
        }

        if (zfs_prop_get_int(zhp, ZFS_PROP_KEYSTATUS) ==
            ZFS_KEYSTATUS_UNAVAILABLE) {
                zfs_close(zhp);
                return (0);
        }

        /*
         * If this filesystem is inconsistent and has a receive resume
         * token, we can not mount it.
         */
        if (zfs_prop_get_int(zhp, ZFS_PROP_INCONSISTENT) &&
            zfs_prop_get(zhp, ZFS_PROP_RECEIVE_RESUME_TOKEN,
            NULL, 0, NULL, NULL, 0, B_TRUE) == 0) {
                zfs_close(zhp);
                return (0);
        }

        libzfs_add_handle(cbp, zhp);
        if (zfs_iter_filesystems(zhp, zfs_iter_cb, cbp) != 0) {
                zfs_close(zhp);
                return (-1);
        }
        return (0);
}

/*
 * Sort comparator that compares two mountpoint paths. We sort these paths so
 * that subdirectories immediately follow their parents. This means that we
 * effectively treat the '/' character as the lowest value non-nul char.
 * Since filesystems from non-global zones can have the same mountpoint
 * as other filesystems, the comparator sorts global zone filesystems to
 * the top of the list. This means that the global zone will traverse the
 * filesystem list in the correct order and can stop when it sees the
 * first zoned filesystem. In a non-global zone, only the delegated
 * filesystems are seen.
 *
 * An example sorted list using this comparator would look like:
 *
 * /foo
 * /foo/bar
 * /foo/bar/baz
 * /foo/baz
 * /foo.bar
 * /foo (NGZ1)
 * /foo (NGZ2)
 *
 * The mounting code depends on this ordering to deterministically iterate
 * over filesystems in order to spawn parallel mount tasks.
 */
static int
mountpoint_cmp(const void *arga, const void *argb)
{
        zfs_handle_t *const *zap = arga;
        zfs_handle_t *za = *zap;
        zfs_handle_t *const *zbp = argb;
        zfs_handle_t *zb = *zbp;
        char mounta[MAXPATHLEN];
        char mountb[MAXPATHLEN];
        const char *a = mounta;
        const char *b = mountb;
        boolean_t gota, gotb;
        uint64_t zoneda, zonedb;

        zoneda = zfs_prop_get_int(za, ZFS_PROP_ZONED);
        zonedb = zfs_prop_get_int(zb, ZFS_PROP_ZONED);
        if (zoneda && !zonedb)
                return (1);
        if (!zoneda && zonedb)
                return (-1);

        gota = (zfs_get_type(za) == ZFS_TYPE_FILESYSTEM);
        if (gota) {
                verify(zfs_prop_get(za, ZFS_PROP_MOUNTPOINT, mounta,
                    sizeof (mounta), NULL, NULL, 0, B_FALSE) == 0);
        }
        gotb = (zfs_get_type(zb) == ZFS_TYPE_FILESYSTEM);
        if (gotb) {
                verify(zfs_prop_get(zb, ZFS_PROP_MOUNTPOINT, mountb,
                    sizeof (mountb), NULL, NULL, 0, B_FALSE) == 0);
        }

        if (gota && gotb) {
                while (*a != '\0' && (*a == *b)) {
                        a++;
                        b++;
                }
                if (*a == *b)
                        return (0);
                if (*a == '\0')
                        return (-1);
                if (*b == '\0')
                        return (1);
                if (*a == '/')
                        return (-1);
                if (*b == '/')
                        return (1);
                return (*a < *b ? -1 : *a > *b);
        }

        if (gota)
                return (-1);
        if (gotb)
                return (1);

        /*
         * If neither filesystem has a mountpoint, revert to sorting by
         * dataset name.
         */
        return (strcmp(zfs_get_name(za), zfs_get_name(zb)));
}

/*
 * Return true if path2 is a child of path1 or path2 equals path1 or
 * path1 is "/" (path2 is always a child of "/").
 */
static boolean_t
libzfs_path_contains(const char *path1, const char *path2)
{
        return (strcmp(path1, path2) == 0 || strcmp(path1, "/") == 0 ||
            (strstr(path2, path1) == path2 && path2[strlen(path1)] == '/'));
}

/*
 * Given a mountpoint specified by idx in the handles array, find the first
 * non-descendent of that mountpoint and return its index. Descendant paths
 * start with the parent's path. This function relies on the ordering
 * enforced by mountpoint_cmp().
 */
static int
non_descendant_idx(zfs_handle_t **handles, size_t num_handles, int idx)
{
        char parent[ZFS_MAXPROPLEN];
        char child[ZFS_MAXPROPLEN];
        int i;

        verify(zfs_prop_get(handles[idx], ZFS_PROP_MOUNTPOINT, parent,
            sizeof (parent), NULL, NULL, 0, B_FALSE) == 0);

        for (i = idx + 1; i < num_handles; i++) {
                verify(zfs_prop_get(handles[i], ZFS_PROP_MOUNTPOINT, child,
                    sizeof (child), NULL, NULL, 0, B_FALSE) == 0);
                if (!libzfs_path_contains(parent, child))
                        break;
        }
        return (i);
}

typedef struct mnt_param {
        libzfs_handle_t *mnt_hdl;
        tpool_t         *mnt_tp;
        zfs_handle_t    **mnt_zhps; /* filesystems to mount */
        size_t          mnt_num_handles;
        int             mnt_idx;        /* Index of selected entry to mount */
        zfs_iter_f      mnt_func;
        void            *mnt_data;
} mnt_param_t;

/*
 * Allocate and populate the parameter struct for mount function, and
 * schedule mounting of the entry selected by idx.
 */
static void
zfs_dispatch_mount(libzfs_handle_t *hdl, zfs_handle_t **handles,
    size_t num_handles, int idx, zfs_iter_f func, void *data, tpool_t *tp)
{
        mnt_param_t *mnt_param = zfs_alloc(hdl, sizeof (mnt_param_t));

        mnt_param->mnt_hdl = hdl;
        mnt_param->mnt_tp = tp;
        mnt_param->mnt_zhps = handles;
        mnt_param->mnt_num_handles = num_handles;
        mnt_param->mnt_idx = idx;
        mnt_param->mnt_func = func;
        mnt_param->mnt_data = data;

        (void) tpool_dispatch(tp, zfs_mount_task, (void*)mnt_param);
}

/*
 * This is the structure used to keep state of mounting or sharing operations
 * during a call to zpool_enable_datasets().
 */
typedef struct mount_state {
        /*
         * ms_mntstatus is set to -1 if any mount fails. While multiple threads
         * could update this variable concurrently, no synchronization is
         * needed as it's only ever set to -1.
         */
        int             ms_mntstatus;
        int             ms_mntflags;
        const char      *ms_mntopts;
} mount_state_t;

static int
zfs_mount_one(zfs_handle_t *zhp, void *arg)
{
        mount_state_t *ms = arg;
        int ret = 0;

        /*
         * don't attempt to mount encrypted datasets with
         * unloaded keys
         */
        if (zfs_prop_get_int(zhp, ZFS_PROP_KEYSTATUS) ==
            ZFS_KEYSTATUS_UNAVAILABLE)
                return (0);

        if (zfs_mount(zhp, ms->ms_mntopts, ms->ms_mntflags) != 0)
                ret = ms->ms_mntstatus = -1;
        return (ret);
}

static int
zfs_share_one(zfs_handle_t *zhp, void *arg)
{
        mount_state_t *ms = arg;
        int ret = 0;

        if (zfs_share(zhp, NULL) != 0)
                ret = ms->ms_mntstatus = -1;
        return (ret);
}

/*
 * Thread pool function to mount one file system. On completion, it finds and
 * schedules its children to be mounted. This depends on the sorting done in
 * zfs_foreach_mountpoint(). Note that the degenerate case (chain of entries
 * each descending from the previous) will have no parallelism since we always
 * have to wait for the parent to finish mounting before we can schedule
 * its children.
 */
static void
zfs_mount_task(void *arg)
{
        mnt_param_t *mp = arg;
        int idx = mp->mnt_idx;
        zfs_handle_t **handles = mp->mnt_zhps;
        size_t num_handles = mp->mnt_num_handles;
        char mountpoint[ZFS_MAXPROPLEN];

        verify(zfs_prop_get(handles[idx], ZFS_PROP_MOUNTPOINT, mountpoint,
            sizeof (mountpoint), NULL, NULL, 0, B_FALSE) == 0);

        if (mp->mnt_func(handles[idx], mp->mnt_data) != 0)
                goto out;

        /*
         * We dispatch tasks to mount filesystems with mountpoints underneath
         * this one. We do this by dispatching the next filesystem with a
         * descendant mountpoint of the one we just mounted, then skip all of
         * its descendants, dispatch the next descendant mountpoint, and so on.
         * The non_descendant_idx() function skips over filesystems that are
         * descendants of the filesystem we just dispatched.
         */
        for (int i = idx + 1; i < num_handles;
            i = non_descendant_idx(handles, num_handles, i)) {
                char child[ZFS_MAXPROPLEN];
                verify(zfs_prop_get(handles[i], ZFS_PROP_MOUNTPOINT,
                    child, sizeof (child), NULL, NULL, 0, B_FALSE) == 0);

                if (!libzfs_path_contains(mountpoint, child))
                        break; /* not a descendant, return */
                zfs_dispatch_mount(mp->mnt_hdl, handles, num_handles, i,
                    mp->mnt_func, mp->mnt_data, mp->mnt_tp);
        }

out:
        free(mp);
}

/*
 * Issue the func callback for each ZFS handle contained in the handles
 * array. This function is used to mount all datasets, and so this function
 * guarantees that filesystems for parent mountpoints are called before their
 * children. As such, before issuing any callbacks, we first sort the array
 * of handles by mountpoint.
 *
 * Callbacks are issued in one of two ways:
 *
 * 1. Sequentially: If the parallel argument is B_FALSE or the ZFS_SERIAL_MOUNT
 *    environment variable is set, then we issue callbacks sequentially.
 *
 * 2. In parallel: If the parallel argument is B_TRUE and the ZFS_SERIAL_MOUNT
 *    environment variable is not set, then we use a tpool to dispatch threads
 *    to mount filesystems in parallel. This function dispatches tasks to mount
 *    the filesystems at the top-level mountpoints, and these tasks in turn
 *    are responsible for recursively mounting filesystems in their children
 *    mountpoints.
 */
void
zfs_foreach_mountpoint(libzfs_handle_t *hdl, zfs_handle_t **handles,
    size_t num_handles, zfs_iter_f func, void *data, boolean_t parallel)
{
        zoneid_t zoneid = getzoneid();

        /*
         * The ZFS_SERIAL_MOUNT environment variable is an undocumented
         * variable that can be used as a convenience to do a/b comparison
         * of serial vs. parallel mounting.
         */
        boolean_t serial_mount = !parallel ||
            (getenv("ZFS_SERIAL_MOUNT") != NULL);

        /*
         * Sort the datasets by mountpoint. See mountpoint_cmp for details
         * of how these are sorted.
         */
        qsort(handles, num_handles, sizeof (zfs_handle_t *), mountpoint_cmp);

        if (serial_mount) {
                for (int i = 0; i < num_handles; i++) {
                        func(handles[i], data);
                }
                return;
        }

        /*
         * Issue the callback function for each dataset using a parallel
         * algorithm that uses a thread pool to manage threads.
         */
        tpool_t *tp = tpool_create(1, mount_tp_nthr, 0, NULL);

        /*
         * There may be multiple "top level" mountpoints outside of the pool's
         * root mountpoint, e.g.: /foo /bar. Dispatch a mount task for each of
         * these.
         */
        for (int i = 0; i < num_handles;
            i = non_descendant_idx(handles, num_handles, i)) {
                /*
                 * Since the mountpoints have been sorted so that the zoned
                 * filesystems are at the end, a zoned filesystem seen from
                 * the global zone means that we're done.
                 */
                if (zoneid == GLOBAL_ZONEID &&
                    zfs_prop_get_int(handles[i], ZFS_PROP_ZONED))
                        break;
                zfs_dispatch_mount(hdl, handles, num_handles, i, func, data,
                    tp);
        }

        tpool_wait(tp); /* wait for all scheduled mounts to complete */
        tpool_destroy(tp);
}

/*
 * Mount and share all datasets within the given pool.  This assumes that no
 * datasets within the pool are currently mounted.
 */
int
zpool_enable_datasets(zpool_handle_t *zhp, const char *mntopts, int flags)
{
        get_all_cb_t cb = { 0 };
        mount_state_t ms = { 0 };
        zfs_handle_t *zfsp;
        int ret = 0;

        if ((zfsp = zfs_open(zhp->zpool_hdl, zhp->zpool_name,
            ZFS_TYPE_DATASET)) == NULL)
                goto out;

        /*
         * Gather all non-snapshot datasets within the pool. Start by adding
         * the root filesystem for this pool to the list, and then iterate
         * over all child filesystems.
         */
        libzfs_add_handle(&cb, zfsp);
        if (zfs_iter_filesystems(zfsp, zfs_iter_cb, &cb) != 0)
                goto out;

        /*
         * Mount all filesystems
         */
        ms.ms_mntopts = mntopts;
        ms.ms_mntflags = flags;
        zfs_foreach_mountpoint(zhp->zpool_hdl, cb.cb_handles, cb.cb_used,
            zfs_mount_one, &ms, B_TRUE);
        if (ms.ms_mntstatus != 0)
                ret = ms.ms_mntstatus;

        /*
         * Share all filesystems that need to be shared. This needs to be
         * a separate pass because libshare is not mt-safe, and so we need
         * to share serially.
         */
        ms.ms_mntstatus = 0;
        zfs_foreach_mountpoint(zhp->zpool_hdl, cb.cb_handles, cb.cb_used,
            zfs_share_one, &ms, B_FALSE);
        if (ms.ms_mntstatus != 0)
                ret = ms.ms_mntstatus;
        else
                zfs_commit_shares(NULL);

out:
        for (int i = 0; i < cb.cb_used; i++)
                zfs_close(cb.cb_handles[i]);
        free(cb.cb_handles);

        return (ret);
}

struct sets_s {
        char *mountpoint;
        zfs_handle_t *dataset;
};

static int
mountpoint_compare(const void *a, const void *b)
{
        const struct sets_s *mounta = (struct sets_s *)a;
        const struct sets_s *mountb = (struct sets_s *)b;

        return (strcmp(mountb->mountpoint, mounta->mountpoint));
}

/*
 * 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 /proc/self/mounts
 * and gather all the filesystems that are currently mounted.
 */
int
zpool_disable_datasets(zpool_handle_t *zhp, boolean_t force)
{
        int used, alloc;
        FILE *mnttab;
        struct mnttab entry;
        size_t namelen;
        struct sets_s *sets = NULL;
        libzfs_handle_t *hdl = zhp->zpool_hdl;
        int i;
        int ret = -1;
        int flags = (force ? MS_FORCE : 0);

        namelen = strlen(zhp->zpool_name);

        if ((mnttab = fopen(MNTTAB, "re")) == NULL)
                return (ENOENT);

        used = alloc = 0;
        while (getmntent(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) {
                                sets = zfs_alloc(hdl,
                                    8 * sizeof (struct sets_s));
                                alloc = 8;
                        } else {
                                sets = zfs_realloc(hdl, sets,
                                    alloc * sizeof (struct sets_s),
                                    alloc * 2 * sizeof (struct sets_s));

                                alloc *= 2;
                        }
                }

                sets[used].mountpoint = zfs_strdup(hdl, entry.mnt_mountp);

                /*
                 * 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.
                 */
                sets[used].dataset = make_dataset_handle(hdl,
                    entry.mnt_special);

                used++;
        }

        /*
         * At this point, we have the entire list of filesystems, so sort it by
         * mountpoint.
         */
        if (used != 0)
                qsort(sets, used, sizeof (struct sets_s), mountpoint_compare);

        /*
         * Walk through and first unshare everything.
         */
        for (i = 0; i < used; i++) {
                for (enum sa_protocol i = 0; i < SA_PROTOCOL_COUNT; ++i) {
                        if (sa_is_shared(sets[i].mountpoint, i) &&
                            unshare_one(hdl, sets[i].mountpoint,
                            sets[i].mountpoint, i) != 0)
                                goto out;
                }
        }
        zfs_commit_shares(NULL);

        /*
         * Now unmount everything, removing the underlying directories as
         * appropriate.
         */
        for (i = 0; i < used; i++) {
                if (unmount_one(sets[i].dataset, sets[i].mountpoint,
                    flags) != 0)
                        goto out;
        }

        for (i = 0; i < used; i++) {
                if (sets[i].dataset)
                        remove_mountpoint(sets[i].dataset);
        }

        zpool_disable_datasets_os(zhp, force);

        ret = 0;
out:
        (void) fclose(mnttab);
        for (i = 0; i < used; i++) {
                if (sets[i].dataset)
                        zfs_close(sets[i].dataset);
                free(sets[i].mountpoint);
        }
        free(sets);

        return (ret);
}