HAMMER VFS - Disallow rebalancing on small-memory machines

[dragonfly.git] / sys / vfs / hammer / hammer_ioctl.c

/*
 * Copyright (c) 2008 The DragonFly Project.  All rights reserved.
 * 
 * This code is derived from software contributed to The DragonFly Project
 * by Matthew Dillon <dillon@backplane.com>
 * 
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in
 *    the documentation and/or other materials provided with the
 *    distribution.
 * 3. Neither the name of The DragonFly Project nor the names of its
 *    contributors may be used to endorse or promote products derived
 *    from this software without specific, prior written permission.
 * 
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE
 * COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
 * INCIDENTAL, SPECIAL, EXEMPLARY OR CONSEQUENTIAL DAMAGES (INCLUDING,
 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
 * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
 * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
 * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 * 
 * $DragonFly: src/sys/vfs/hammer/hammer_ioctl.c,v 1.32 2008/11/13 02:23:29 dillon Exp $
 */

#include "hammer.h"

static int hammer_ioc_gethistory(hammer_transaction_t trans, hammer_inode_t ip,
                                struct hammer_ioc_history *hist);
static int hammer_ioc_synctid(hammer_transaction_t trans, hammer_inode_t ip,
                                struct hammer_ioc_synctid *std);
static int hammer_ioc_get_version(hammer_transaction_t trans,
                                hammer_inode_t ip,
                                struct hammer_ioc_version *ver);
static int hammer_ioc_set_version(hammer_transaction_t trans,
                                hammer_inode_t ip,
                                struct hammer_ioc_version *ver);
static int hammer_ioc_get_info(hammer_transaction_t trans,
                                struct hammer_ioc_info *info);
static int hammer_ioc_add_snapshot(hammer_transaction_t trans, hammer_inode_t ip,
                                struct hammer_ioc_snapshot *snap);
static int hammer_ioc_del_snapshot(hammer_transaction_t trans, hammer_inode_t ip,
                                struct hammer_ioc_snapshot *snap);
static int hammer_ioc_get_snapshot(hammer_transaction_t trans, hammer_inode_t ip,
                                struct hammer_ioc_snapshot *snap);
static int hammer_ioc_get_config(hammer_transaction_t trans, hammer_inode_t ip,
                                struct hammer_ioc_config *snap);
static int hammer_ioc_set_config(hammer_transaction_t trans, hammer_inode_t ip,
                                struct hammer_ioc_config *snap);

int
hammer_ioctl(hammer_inode_t ip, u_long com, caddr_t data, int fflag,
             struct ucred *cred)
{
        struct hammer_transaction trans;
        int error;

        error = priv_check_cred(cred, PRIV_HAMMER_IOCTL, 0);

        hammer_start_transaction(&trans, ip->hmp);

        switch(com) {
        case HAMMERIOC_PRUNE:
                if (error == 0) {
                        error = hammer_ioc_prune(&trans, ip,
                                        (struct hammer_ioc_prune *)data);
                }
                break;
        case HAMMERIOC_GETHISTORY:
                error = hammer_ioc_gethistory(&trans, ip,
                                        (struct hammer_ioc_history *)data);
                break;
        case HAMMERIOC_REBLOCK:
                if (error == 0) {
                        error = hammer_ioc_reblock(&trans, ip,
                                        (struct hammer_ioc_reblock *)data);
                }
                break;
        case HAMMERIOC_REBALANCE:
                /*
                 * Rebalancing needs to lock a lot of B-Tree nodes.  The
                 * children and children's children.  Systems with very
                 * little memory will not be able to do it.
                 */
                if (error == 0 && nbuf < HAMMER_REBALANCE_MIN_BUFS) {
                        kprintf("hammer: System has insufficient buffers "
                                "to rebalance the tree.  nbuf < %d\n",
                                HAMMER_REBALANCE_MIN_BUFS);
                        error = ENOSPC;
                }
                if (error == 0) {
                        error = hammer_ioc_rebalance(&trans, ip,
                                        (struct hammer_ioc_rebalance *)data);
                }
                break;
        case HAMMERIOC_SYNCTID:
                error = hammer_ioc_synctid(&trans, ip,
                                        (struct hammer_ioc_synctid *)data);
                break;
        case HAMMERIOC_GET_PSEUDOFS:
                error = hammer_ioc_get_pseudofs(&trans, ip,
                                    (struct hammer_ioc_pseudofs_rw *)data);
                break;
        case HAMMERIOC_SET_PSEUDOFS:
                if (error == 0) {
                        error = hammer_ioc_set_pseudofs(&trans, ip, cred,
                                    (struct hammer_ioc_pseudofs_rw *)data);
                }
                break;
        case HAMMERIOC_UPG_PSEUDOFS:
                if (error == 0) {
                        error = hammer_ioc_upgrade_pseudofs(&trans, ip, 
                                    (struct hammer_ioc_pseudofs_rw *)data);
                }
                break;
        case HAMMERIOC_DGD_PSEUDOFS:
                if (error == 0) {
                        error = hammer_ioc_downgrade_pseudofs(&trans, ip,
                                    (struct hammer_ioc_pseudofs_rw *)data);
                }
                break;
        case HAMMERIOC_RMR_PSEUDOFS:
                if (error == 0) {
                        error = hammer_ioc_destroy_pseudofs(&trans, ip,
                                    (struct hammer_ioc_pseudofs_rw *)data);
                }
                break;
        case HAMMERIOC_WAI_PSEUDOFS:
                if (error == 0) {
                        error = hammer_ioc_wait_pseudofs(&trans, ip,
                                    (struct hammer_ioc_pseudofs_rw *)data);
                }
                break;
        case HAMMERIOC_MIRROR_READ:
                if (error == 0) {
                        error = hammer_ioc_mirror_read(&trans, ip,
                                    (struct hammer_ioc_mirror_rw *)data);
                }
                break;
        case HAMMERIOC_MIRROR_WRITE:
                if (error == 0) {
                        error = hammer_ioc_mirror_write(&trans, ip,
                                    (struct hammer_ioc_mirror_rw *)data);
                }
                break;
        case HAMMERIOC_GET_VERSION:
                error = hammer_ioc_get_version(&trans, ip, 
                                    (struct hammer_ioc_version *)data);
                break;
        case HAMMERIOC_GET_INFO:
                error = hammer_ioc_get_info(&trans,
                                    (struct hammer_ioc_info *)data);
                break;
        case HAMMERIOC_SET_VERSION:
                if (error == 0) {
                        error = hammer_ioc_set_version(&trans, ip, 
                                            (struct hammer_ioc_version *)data);
                }
                break;
        case HAMMERIOC_ADD_VOLUME:
                if (error == 0) {
                        error = priv_check_cred(cred, PRIV_HAMMER_VOLUME, 0);
                        if (error == 0)
                                error = hammer_ioc_volume_add(&trans, ip,
                                            (struct hammer_ioc_volume *)data);
                }
                break;
        case HAMMERIOC_DEL_VOLUME:
                if (error == 0) {
                        error = priv_check_cred(cred, PRIV_HAMMER_VOLUME, 0);
                        if (error == 0)
                                error = hammer_ioc_volume_del(&trans, ip,
                                            (struct hammer_ioc_volume *)data);
                }
                break;
        case HAMMERIOC_ADD_SNAPSHOT:
                if (error == 0) {
                        error = hammer_ioc_add_snapshot(
                                        &trans, ip, (struct hammer_ioc_snapshot *)data);
                }
                break;
        case HAMMERIOC_DEL_SNAPSHOT:
                if (error == 0) {
                        error = hammer_ioc_del_snapshot(
                                        &trans, ip, (struct hammer_ioc_snapshot *)data);
                }
                break;
        case HAMMERIOC_GET_SNAPSHOT:
                error = hammer_ioc_get_snapshot(
                                        &trans, ip, (struct hammer_ioc_snapshot *)data);
                break;
        case HAMMERIOC_GET_CONFIG:
                error = hammer_ioc_get_config(
                                        &trans, ip, (struct hammer_ioc_config *)data);
                break;
        case HAMMERIOC_SET_CONFIG:
                if (error == 0) {
                        error = hammer_ioc_set_config(
                                        &trans, ip, (struct hammer_ioc_config *)data);
                }
                break;
        default:
                error = EOPNOTSUPP;
                break;
        }
        hammer_done_transaction(&trans);
        return (error);
}

/*
 * Iterate through an object's inode or an object's records and record
 * modification TIDs.
 */
static void add_history(hammer_inode_t ip, struct hammer_ioc_history *hist,
                        hammer_btree_elm_t elm);

static
int
hammer_ioc_gethistory(hammer_transaction_t trans, hammer_inode_t ip,
                      struct hammer_ioc_history *hist)
{
        struct hammer_cursor cursor;
        hammer_btree_elm_t elm;
        int error;

        /*
         * Validate the structure and initialize for return.
         */
        if (hist->beg_tid > hist->end_tid)
                return(EINVAL);
        if (hist->head.flags & HAMMER_IOC_HISTORY_ATKEY) {
                if (hist->key > hist->nxt_key)
                        return(EINVAL);
        }

        hist->obj_id = ip->obj_id;
        hist->count = 0;
        hist->nxt_tid = hist->end_tid;
        hist->head.flags &= ~HAMMER_IOC_HISTORY_NEXT_TID;
        hist->head.flags &= ~HAMMER_IOC_HISTORY_NEXT_KEY;
        hist->head.flags &= ~HAMMER_IOC_HISTORY_EOF;
        hist->head.flags &= ~HAMMER_IOC_HISTORY_UNSYNCED;
        if ((ip->flags & HAMMER_INODE_MODMASK) & 
            ~(HAMMER_INODE_ATIME | HAMMER_INODE_MTIME)) {
                hist->head.flags |= HAMMER_IOC_HISTORY_UNSYNCED;
        }

        /*
         * Setup the cursor.  We can't handle undeletable records
         * (create_tid of 0) at the moment.  A create_tid of 0 has
         * a special meaning and cannot be specified in the cursor.
         */
        error = hammer_init_cursor(trans, &cursor, &ip->cache[0], NULL);
        if (error) {
                hammer_done_cursor(&cursor);
                return(error);
        }

        cursor.key_beg.obj_id = hist->obj_id;
        cursor.key_beg.create_tid = hist->beg_tid;
        cursor.key_beg.delete_tid = 0;
        cursor.key_beg.obj_type = 0;
        if (cursor.key_beg.create_tid == HAMMER_MIN_TID)
                cursor.key_beg.create_tid = 1;

        cursor.key_end.obj_id = hist->obj_id;
        cursor.key_end.create_tid = hist->end_tid;
        cursor.key_end.delete_tid = 0;
        cursor.key_end.obj_type = 0;

        cursor.flags |= HAMMER_CURSOR_END_EXCLUSIVE;

        if (hist->head.flags & HAMMER_IOC_HISTORY_ATKEY) {
                /*
                 * key-range within the file.  For a regular file the
                 * on-disk key represents BASE+LEN, not BASE, so the
                 * first possible record containing the offset 'key'
                 * has an on-disk key of (key + 1).
                 */
                cursor.key_beg.key = hist->key;
                cursor.key_end.key = HAMMER_MAX_KEY;
                cursor.key_beg.localization = ip->obj_localization + 
                                              HAMMER_LOCALIZE_MISC;
                cursor.key_end.localization = ip->obj_localization + 
                                              HAMMER_LOCALIZE_MISC;

                switch(ip->ino_data.obj_type) {
                case HAMMER_OBJTYPE_REGFILE:
                        ++cursor.key_beg.key;
                        cursor.key_beg.rec_type = HAMMER_RECTYPE_DATA;
                        break;
                case HAMMER_OBJTYPE_DIRECTORY:
                        cursor.key_beg.rec_type = HAMMER_RECTYPE_DIRENTRY;
                        cursor.key_beg.localization = ip->obj_localization +
                                                hammer_dir_localization(ip);
                        cursor.key_end.localization = ip->obj_localization +
                                                hammer_dir_localization(ip);
                        break;
                case HAMMER_OBJTYPE_DBFILE:
                        cursor.key_beg.rec_type = HAMMER_RECTYPE_DB;
                        break;
                default:
                        error = EINVAL;
                        break;
                }
                cursor.key_end.rec_type = cursor.key_beg.rec_type;
        } else {
                /*
                 * The inode itself.
                 */
                cursor.key_beg.key = 0;
                cursor.key_end.key = 0;
                cursor.key_beg.rec_type = HAMMER_RECTYPE_INODE;
                cursor.key_end.rec_type = HAMMER_RECTYPE_INODE;
                cursor.key_beg.localization = ip->obj_localization +
                                              HAMMER_LOCALIZE_INODE;
                cursor.key_end.localization = ip->obj_localization +
                                              HAMMER_LOCALIZE_INODE;
        }

        error = hammer_btree_first(&cursor);
        while (error == 0) {
                elm = &cursor.node->ondisk->elms[cursor.index];

                add_history(ip, hist, elm);
                if (hist->head.flags & (HAMMER_IOC_HISTORY_NEXT_TID |
                                        HAMMER_IOC_HISTORY_NEXT_KEY |
                                        HAMMER_IOC_HISTORY_EOF)) {
                        break;
                }
                error = hammer_btree_iterate(&cursor);
        }
        if (error == ENOENT) {
                hist->head.flags |= HAMMER_IOC_HISTORY_EOF;
                error = 0;
        }
        hammer_done_cursor(&cursor);
        return(error);
}

/*
 * Add the scanned element to the ioctl return structure.  Some special
 * casing is required for regular files to accomodate how data ranges are
 * stored on-disk.
 */
static void
add_history(hammer_inode_t ip, struct hammer_ioc_history *hist,
            hammer_btree_elm_t elm)
{
        int i;

        if (elm->base.btype != HAMMER_BTREE_TYPE_RECORD)
                return;
        if ((hist->head.flags & HAMMER_IOC_HISTORY_ATKEY) &&
            ip->ino_data.obj_type == HAMMER_OBJTYPE_REGFILE) {
                /*
                 * Adjust nxt_key
                 */
                if (hist->nxt_key > elm->leaf.base.key - elm->leaf.data_len &&
                    hist->key < elm->leaf.base.key - elm->leaf.data_len) {
                        hist->nxt_key = elm->leaf.base.key - elm->leaf.data_len;
                }
                if (hist->nxt_key > elm->leaf.base.key)
                        hist->nxt_key = elm->leaf.base.key;

                /*
                 * Record is beyond MAXPHYS, there won't be any more records
                 * in the iteration covering the requested offset (key).
                 */
                if (elm->leaf.base.key >= MAXPHYS &&
                    elm->leaf.base.key - MAXPHYS > hist->key) {
                        hist->head.flags |= HAMMER_IOC_HISTORY_NEXT_KEY;
                }

                /*
                 * Data-range of record does not cover the key.
                 */
                if (elm->leaf.base.key - elm->leaf.data_len > hist->key)
                        return;

        } else if (hist->head.flags & HAMMER_IOC_HISTORY_ATKEY) {
                /*
                 * Adjust nxt_key
                 */
                if (hist->nxt_key > elm->leaf.base.key &&
                    hist->key < elm->leaf.base.key) {
                        hist->nxt_key = elm->leaf.base.key;
                }

                /*
                 * Record is beyond the requested key.
                 */
                if (elm->leaf.base.key > hist->key)
                        hist->head.flags |= HAMMER_IOC_HISTORY_NEXT_KEY;
        }

        /*
         * Add create_tid if it is in-bounds.
         */
        i = hist->count;
        if ((i == 0 ||
             elm->leaf.base.create_tid != hist->hist_ary[i - 1].tid) &&
            elm->leaf.base.create_tid >= hist->beg_tid &&
            elm->leaf.base.create_tid < hist->end_tid) {
                if (hist->count == HAMMER_MAX_HISTORY_ELMS) {
                        hist->nxt_tid = elm->leaf.base.create_tid;
                        hist->head.flags |= HAMMER_IOC_HISTORY_NEXT_TID;
                        return;
                }
                hist->hist_ary[i].tid = elm->leaf.base.create_tid;
                hist->hist_ary[i].time32 = elm->leaf.create_ts;
                ++hist->count;
        }

        /*
         * Add delete_tid if it is in-bounds.  Note that different portions
         * of the history may have overlapping data ranges with different
         * delete_tid's.  If this case occurs the delete_tid may match the
         * create_tid of a following record.  XXX
         *
         *      [        ]
         *            [     ]
         */
        i = hist->count;
        if (elm->leaf.base.delete_tid &&
            elm->leaf.base.delete_tid >= hist->beg_tid &&
            elm->leaf.base.delete_tid < hist->end_tid) {
                if (i == HAMMER_MAX_HISTORY_ELMS) {
                        hist->nxt_tid = elm->leaf.base.delete_tid;
                        hist->head.flags |= HAMMER_IOC_HISTORY_NEXT_TID;
                        return;
                }
                hist->hist_ary[i].tid = elm->leaf.base.delete_tid;
                hist->hist_ary[i].time32 = elm->leaf.delete_ts;
                ++hist->count;
        }
}

/*
 * Acquire synchronization TID
 */
static
int
hammer_ioc_synctid(hammer_transaction_t trans, hammer_inode_t ip,
                   struct hammer_ioc_synctid *std)
{
        hammer_mount_t hmp = ip->hmp;
        int error = 0;

        switch(std->op) {
        case HAMMER_SYNCTID_NONE:
                std->tid = hmp->flusher.tid;    /* inaccurate */
                break;
        case HAMMER_SYNCTID_ASYNC:
                hammer_queue_inodes_flusher(hmp, MNT_NOWAIT);
                hammer_flusher_async(hmp, NULL);
                std->tid = hmp->flusher.tid;    /* inaccurate */
                break;
        case HAMMER_SYNCTID_SYNC1:
                hammer_queue_inodes_flusher(hmp, MNT_WAIT);
                hammer_flusher_sync(hmp);
                std->tid = hmp->flusher.tid;
                break;
        case HAMMER_SYNCTID_SYNC2:
                hammer_queue_inodes_flusher(hmp, MNT_WAIT);
                hammer_flusher_sync(hmp);
                std->tid = hmp->flusher.tid;
                hammer_flusher_sync(hmp);
                break;
        default:
                error = EOPNOTSUPP;
                break;
        }
        return(error);
}

/*
 * Retrieve version info.
 *
 * Load min_version, wip_version, and max_versino.  If cur_version is passed
 * as 0 then load the current version into cur_version.  Load the description
 * for cur_version into the description array.
 *
 * Returns 0 on success, EINVAL if cur_version is non-zero and set to an
 * unsupported value.
 */
static
int
hammer_ioc_get_version(hammer_transaction_t trans, hammer_inode_t ip,
                   struct hammer_ioc_version *ver)
{
        int error = 0;

        ver->min_version = HAMMER_VOL_VERSION_MIN;
        ver->wip_version = HAMMER_VOL_VERSION_WIP;
        ver->max_version = HAMMER_VOL_VERSION_MAX;
        if (ver->cur_version == 0)
                ver->cur_version = trans->hmp->version;
        switch(ver->cur_version) {
        case 1:
                ksnprintf(ver->description, sizeof(ver->description),
                         "First HAMMER release (DragonFly 2.0+)");
                break;
        case 2:
                ksnprintf(ver->description, sizeof(ver->description),
                         "New directory entry layout (DragonFly 2.3+)");
                break;
        case 3:
                ksnprintf(ver->description, sizeof(ver->description),
                         "New snapshot management (DragonFly 2.5+)");
                break;
        case 4:
                ksnprintf(ver->description, sizeof(ver->description),
                         "New undo/flush, faster flush/sync (DragonFly 2.5+)");
                break;
        default:
                ksnprintf(ver->description, sizeof(ver->description),
                         "Unknown");
                error = EINVAL;
                break;
        }
        return(error);
};

/*
 * Set version info
 */
static
int
hammer_ioc_set_version(hammer_transaction_t trans, hammer_inode_t ip,
                   struct hammer_ioc_version *ver)
{
        hammer_mount_t hmp = trans->hmp;
        struct hammer_cursor cursor;
        hammer_volume_t volume;
        int error;
        int over = hmp->version;

        /*
         * Generally do not allow downgrades.  However, version 4 can
         * be downgraded to version 3.
         */
        if (ver->cur_version < hmp->version) {
                if (!(ver->cur_version == 3 && hmp->version == 4))
                        return(EINVAL);
        }
        if (ver->cur_version == hmp->version)
                return(0);
        if (ver->cur_version > HAMMER_VOL_VERSION_MAX)
                return(EINVAL);
        if (hmp->ronly)
                return(EROFS);

        /*
         * Update the root volume header and the version cached in
         * the hammer_mount structure.
         */
        error = hammer_init_cursor(trans, &cursor, NULL, NULL);
        if (error)
                goto failed;
        hammer_lock_ex(&hmp->flusher.finalize_lock);
        hammer_sync_lock_ex(trans);
        hmp->version = ver->cur_version;

        /*
         * If upgrading from version < 4 to version >= 4 the UNDO FIFO
         * must be reinitialized.
         */
        if (over < HAMMER_VOL_VERSION_FOUR &&
            ver->cur_version >= HAMMER_VOL_VERSION_FOUR) {
                kprintf("upgrade undo to version 4\n");
                error = hammer_upgrade_undo_4(trans);
                if (error)
                        goto failed;
        }

        /*
         * Adjust the version in the volume header
         */
        volume = hammer_get_root_volume(hmp, &error);
        KKASSERT(error == 0);
        hammer_modify_volume_field(cursor.trans, volume, vol_version);
        volume->ondisk->vol_version = ver->cur_version;
        hammer_modify_volume_done(volume);
        hammer_rel_volume(volume, 0);

        hammer_sync_unlock(trans);
        hammer_unlock(&hmp->flusher.finalize_lock);
failed:
        ver->head.error = error;
        hammer_done_cursor(&cursor);
        return(0);
}

/*
 * Get information
 */
static
int
hammer_ioc_get_info(hammer_transaction_t trans, struct hammer_ioc_info *info) {

        struct hammer_volume_ondisk     *od = trans->hmp->rootvol->ondisk;
        struct hammer_mount             *hm = trans->hmp;

        /* Fill the structure with the necessary information */
        _hammer_checkspace(hm, HAMMER_CHKSPC_WRITE, &info->rsvbigblocks);
        info->rsvbigblocks = info->rsvbigblocks >> HAMMER_LARGEBLOCK_BITS;
        strlcpy(info->vol_name, od->vol_name, sizeof(od->vol_name));

        info->vol_fsid = hm->fsid;
        info->vol_fstype = od->vol_fstype;
        info->version = hm->version;

        info->inodes = od->vol0_stat_inodes;
        info->bigblocks = od->vol0_stat_bigblocks;
        info->freebigblocks = od->vol0_stat_freebigblocks;
        info->nvolumes = hm->nvolumes;

        return 0;
}

/*
 * Add a snapshot transction id(s) to the list of snapshots.
 *
 * NOTE: Records are created with an allocated TID.  If a flush cycle
 *       is in progress the record may be synced in the current flush
 *       cycle and the volume header will reflect the allocation of the
 *       TID, but the synchronization point may not catch up to the
 *       TID until the next flush cycle.
 */
static
int
hammer_ioc_add_snapshot(hammer_transaction_t trans, hammer_inode_t ip,
                        struct hammer_ioc_snapshot *snap)
{
        hammer_mount_t hmp = ip->hmp;
        struct hammer_btree_leaf_elm leaf;
        struct hammer_cursor cursor;
        int error;

        /*
         * Validate structure
         */
        if (snap->count > HAMMER_SNAPS_PER_IOCTL)
                return (EINVAL);
        if (snap->index > snap->count)
                return (EINVAL);

        hammer_lock_ex(&hmp->snapshot_lock);
again:
        /*
         * Look for keys starting after the previous iteration, or at
         * the beginning if snap->count is 0.
         */
        error = hammer_init_cursor(trans, &cursor, &ip->cache[0], NULL);
        if (error) {
                hammer_done_cursor(&cursor);
                return(error);
        }

        cursor.asof = HAMMER_MAX_TID;
        cursor.flags |= HAMMER_CURSOR_BACKEND | HAMMER_CURSOR_ASOF;

        bzero(&leaf, sizeof(leaf));
        leaf.base.obj_id = HAMMER_OBJID_ROOT;
        leaf.base.rec_type = HAMMER_RECTYPE_SNAPSHOT;
        leaf.base.create_tid = hammer_alloc_tid(hmp, 1);
        leaf.base.btype = HAMMER_BTREE_TYPE_RECORD;
        leaf.base.localization = ip->obj_localization + HAMMER_LOCALIZE_INODE;
        leaf.data_len = sizeof(struct hammer_snapshot_data);

        while (snap->index < snap->count) {
                leaf.base.key = (int64_t)snap->snaps[snap->index].tid;
                cursor.key_beg = leaf.base;
                error = hammer_btree_lookup(&cursor);
                if (error == 0) {
                        error = EEXIST;
                        break;
                }

                cursor.flags &= ~HAMMER_CURSOR_ASOF;
                error = hammer_create_at_cursor(&cursor, &leaf,
                                                &snap->snaps[snap->index],
                                                HAMMER_CREATE_MODE_SYS);
                if (error == EDEADLK) {
                        hammer_done_cursor(&cursor);
                        goto again;
                }
                cursor.flags |= HAMMER_CURSOR_ASOF;
                if (error)
                        break;
                ++snap->index;
        }
        snap->head.error = error;
        hammer_done_cursor(&cursor);
        hammer_unlock(&hmp->snapshot_lock);
        return(0);
}

/*
 * Delete snapshot transaction id(s) from the list of snapshots.
 */
static
int
hammer_ioc_del_snapshot(hammer_transaction_t trans, hammer_inode_t ip,
                        struct hammer_ioc_snapshot *snap)
{
        hammer_mount_t hmp = ip->hmp;
        struct hammer_cursor cursor;
        int error;

        /*
         * Validate structure
         */
        if (snap->count > HAMMER_SNAPS_PER_IOCTL)
                return (EINVAL);
        if (snap->index > snap->count)
                return (EINVAL);

        hammer_lock_ex(&hmp->snapshot_lock);
again:
        /*
         * Look for keys starting after the previous iteration, or at
         * the beginning if snap->count is 0.
         */
        error = hammer_init_cursor(trans, &cursor, &ip->cache[0], NULL);
        if (error) {
                hammer_done_cursor(&cursor);
                return(error);
        }

        cursor.key_beg.obj_id = HAMMER_OBJID_ROOT;
        cursor.key_beg.create_tid = 0;
        cursor.key_beg.delete_tid = 0;
        cursor.key_beg.obj_type = 0;
        cursor.key_beg.rec_type = HAMMER_RECTYPE_SNAPSHOT;
        cursor.key_beg.localization = ip->obj_localization + HAMMER_LOCALIZE_INODE;
        cursor.asof = HAMMER_MAX_TID;
        cursor.flags |= HAMMER_CURSOR_ASOF;

        while (snap->index < snap->count) {
                cursor.key_beg.key = (int64_t)snap->snaps[snap->index].tid;
                error = hammer_btree_lookup(&cursor);
                if (error)
                        break;
                error = hammer_btree_extract(&cursor, HAMMER_CURSOR_GET_LEAF);
                if (error)
                        break;
                error = hammer_delete_at_cursor(&cursor, HAMMER_DELETE_DESTROY,
                                                0, 0, 0, NULL);
                if (error == EDEADLK) {
                        hammer_done_cursor(&cursor);
                        goto again;
                }
                if (error)
                        break;
                ++snap->index;
        }
        snap->head.error = error;
        hammer_done_cursor(&cursor);
        hammer_unlock(&hmp->snapshot_lock);
        return(0);
}

/*
 * Retrieve as many snapshot ids as possible or until the array is
 * full, starting after the last transction id passed in.  If count
 * is 0 we retrieve starting at the beginning.
 *
 * NOTE: Because the b-tree key field is signed but transaction ids
 *       are unsigned the returned list will be signed-sorted instead
 *       of unsigned sorted.  The Caller must still sort the aggregate
 *       results.
 */
static
int
hammer_ioc_get_snapshot(hammer_transaction_t trans, hammer_inode_t ip,
                        struct hammer_ioc_snapshot *snap)
{
        struct hammer_cursor cursor;
        int error;

        /*
         * Validate structure
         */
        if (snap->index != 0)
                return (EINVAL);
        if (snap->count > HAMMER_SNAPS_PER_IOCTL)
                return (EINVAL);

        /*
         * Look for keys starting after the previous iteration, or at
         * the beginning if snap->count is 0.
         */
        error = hammer_init_cursor(trans, &cursor, &ip->cache[0], NULL);
        if (error) {
                hammer_done_cursor(&cursor);
                return(error);
        }

        cursor.key_beg.obj_id = HAMMER_OBJID_ROOT;
        cursor.key_beg.create_tid = 0;
        cursor.key_beg.delete_tid = 0;
        cursor.key_beg.obj_type = 0;
        cursor.key_beg.rec_type = HAMMER_RECTYPE_SNAPSHOT;
        cursor.key_beg.localization = ip->obj_localization + HAMMER_LOCALIZE_INODE;
        if (snap->count == 0)
                cursor.key_beg.key = HAMMER_MIN_KEY;
        else
                cursor.key_beg.key = (int64_t)snap->snaps[snap->count - 1].tid + 1;

        cursor.key_end = cursor.key_beg;
        cursor.key_end.key = HAMMER_MAX_KEY;
        cursor.asof = HAMMER_MAX_TID;
        cursor.flags |= HAMMER_CURSOR_END_EXCLUSIVE | HAMMER_CURSOR_ASOF;

        snap->count = 0;

        error = hammer_btree_first(&cursor);
        while (error == 0 && snap->count < HAMMER_SNAPS_PER_IOCTL) {
                error = hammer_btree_extract(&cursor, HAMMER_CURSOR_GET_LEAF);
                if (error)
                        break;
                if (cursor.leaf->base.rec_type == HAMMER_RECTYPE_SNAPSHOT) {
                        error = hammer_btree_extract(
                                             &cursor, HAMMER_CURSOR_GET_LEAF |
                                                      HAMMER_CURSOR_GET_DATA);
                        snap->snaps[snap->count] = cursor.data->snap;

                        /*
                         * The snap data tid should match the key but might
                         * not due to a bug in the HAMMER v3 conversion code.
                         *
                         * This error will work itself out over time but we
                         * have to force a match or the snapshot will not
                         * be deletable.
                         */
                        if (cursor.data->snap.tid !=
                            (hammer_tid_t)cursor.leaf->base.key) {
                                kprintf("HAMMER: lo=%08x snapshot key "
                                        "0x%016jx data mismatch 0x%016jx\n",
                                        cursor.key_beg.localization,
                                        (uintmax_t)cursor.data->snap.tid,
                                        cursor.leaf->base.key);
                                kprintf("HAMMER: Probably left over from the "
                                        "original v3 conversion, hammer "
                                        "cleanup should get it eventually\n");
                                snap->snaps[snap->count].tid =
                                        cursor.leaf->base.key;
                        }
                        ++snap->count;
                }
                error = hammer_btree_iterate(&cursor);
        }

        if (error == ENOENT) {
                snap->head.flags |= HAMMER_IOC_SNAPSHOT_EOF;
                error = 0;
        }
        snap->head.error = error;
        hammer_done_cursor(&cursor);
        return(0);
}

/*
 * Retrieve the PFS hammer cleanup utility config record.  This is
 * different (newer than) the PFS config.
 */
static
int
hammer_ioc_get_config(hammer_transaction_t trans, hammer_inode_t ip,
                        struct hammer_ioc_config *config)
{
        struct hammer_cursor cursor;
        int error;

        error = hammer_init_cursor(trans, &cursor, &ip->cache[0], NULL);
        if (error) {
                hammer_done_cursor(&cursor);
                return(error);
        }

        cursor.key_beg.obj_id = HAMMER_OBJID_ROOT;
        cursor.key_beg.create_tid = 0;
        cursor.key_beg.delete_tid = 0;
        cursor.key_beg.obj_type = 0;
        cursor.key_beg.rec_type = HAMMER_RECTYPE_CONFIG;
        cursor.key_beg.localization = ip->obj_localization + HAMMER_LOCALIZE_INODE;
        cursor.key_beg.key = 0;         /* config space page 0 */

        cursor.asof = HAMMER_MAX_TID;
        cursor.flags |= HAMMER_CURSOR_ASOF;

        error = hammer_btree_lookup(&cursor);
        if (error == 0) {
                error = hammer_btree_extract(&cursor, HAMMER_CURSOR_GET_LEAF |
                                                      HAMMER_CURSOR_GET_DATA);
                if (error == 0)
                        config->config = cursor.data->config;
        }
        /* error can be ENOENT */
        config->head.error = error;
        hammer_done_cursor(&cursor);
        return(0);
}

/*
 * Retrieve the PFS hammer cleanup utility config record.  This is
 * different (newer than) the PFS config.
 *
 * This is kinda a hack.
 */
static
int
hammer_ioc_set_config(hammer_transaction_t trans, hammer_inode_t ip,
                        struct hammer_ioc_config *config)
{
        struct hammer_btree_leaf_elm leaf;
        struct hammer_cursor cursor;
        hammer_mount_t hmp = ip->hmp;
        int error;

again:
        error = hammer_init_cursor(trans, &cursor, &ip->cache[0], NULL);
        if (error) {
                hammer_done_cursor(&cursor);
                return(error);
        }

        bzero(&leaf, sizeof(leaf));
        leaf.base.obj_id = HAMMER_OBJID_ROOT;
        leaf.base.rec_type = HAMMER_RECTYPE_CONFIG;
        leaf.base.create_tid = hammer_alloc_tid(hmp, 1);
        leaf.base.btype = HAMMER_BTREE_TYPE_RECORD;
        leaf.base.localization = ip->obj_localization + HAMMER_LOCALIZE_INODE;
        leaf.base.key = 0;      /* page 0 */
        leaf.data_len = sizeof(struct hammer_config_data);

        cursor.key_beg = leaf.base;

        cursor.asof = HAMMER_MAX_TID;
        cursor.flags |= HAMMER_CURSOR_BACKEND | HAMMER_CURSOR_ASOF;

        error = hammer_btree_lookup(&cursor);
        if (error == 0) {
                error = hammer_btree_extract(&cursor, HAMMER_CURSOR_GET_LEAF |
                                                      HAMMER_CURSOR_GET_DATA);
                error = hammer_delete_at_cursor(&cursor, HAMMER_DELETE_DESTROY,
                                                0, 0, 0, NULL);
                if (error == EDEADLK) {
                        hammer_done_cursor(&cursor);
                        goto again;
                }
        }
        if (error == ENOENT)
                error = 0;
        if (error == 0) {
                cursor.flags &= ~HAMMER_CURSOR_ASOF;
                cursor.key_beg = leaf.base;
                error = hammer_create_at_cursor(&cursor, &leaf,
                                                &config->config,
                                                HAMMER_CREATE_MODE_SYS);
                if (error == EDEADLK) {
                        hammer_done_cursor(&cursor);
                        goto again;
                }
        }
        config->head.error = error;
        hammer_done_cursor(&cursor);
        return(0);
}