+static int find_next_devid(struct btrfs_root *root, struct btrfs_path *path,
+ u64 *objectid)
+{
+ int ret;
+ struct btrfs_key key;
+ struct btrfs_key found_key;
+
+ key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
+ key.type = BTRFS_DEV_ITEM_KEY;
+ key.offset = (u64)-1;
+
+ ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+ if (ret < 0)
+ goto error;
+
+ BUG_ON(ret == 0);
+
+ ret = btrfs_previous_item(root, path, BTRFS_DEV_ITEMS_OBJECTID,
+ BTRFS_DEV_ITEM_KEY);
+ if (ret) {
+ *objectid = 1;
+ } else {
+ btrfs_item_key_to_cpu(path->nodes[0], &found_key,
+ path->slots[0]);
+ *objectid = found_key.offset + 1;
+ }
+ ret = 0;
+error:
+ btrfs_release_path(root, path);
+ return ret;
+}
+
+/*
+ * the device information is stored in the chunk root
+ * the btrfs_device struct should be fully filled in
+ */
+int btrfs_add_device(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root,
+ struct btrfs_device *device)
+{
+ int ret;
+ struct btrfs_path *path;
+ struct btrfs_dev_item *dev_item;
+ struct extent_buffer *leaf;
+ struct btrfs_key key;
+ unsigned long ptr;
+ u64 free_devid = 0;
+
+ root = root->fs_info->chunk_root;
+
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ ret = find_next_devid(root, path, &free_devid);
+ if (ret)
+ goto out;
+
+ key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
+ key.type = BTRFS_DEV_ITEM_KEY;
+ key.offset = free_devid;
+
+ ret = btrfs_insert_empty_item(trans, root, path, &key,
+ sizeof(*dev_item));
+ if (ret)
+ goto out;
+
+ leaf = path->nodes[0];
+ dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);
+
+ device->devid = free_devid;
+ btrfs_set_device_id(leaf, dev_item, device->devid);
+ btrfs_set_device_type(leaf, dev_item, device->type);
+ btrfs_set_device_io_align(leaf, dev_item, device->io_align);
+ btrfs_set_device_io_width(leaf, dev_item, device->io_width);
+ btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
+ btrfs_set_device_total_bytes(leaf, dev_item, device->total_bytes);
+ btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used);
+ btrfs_set_device_group(leaf, dev_item, 0);
+ btrfs_set_device_seek_speed(leaf, dev_item, 0);
+ btrfs_set_device_bandwidth(leaf, dev_item, 0);
+
+ ptr = (unsigned long)btrfs_device_uuid(dev_item);
+ write_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
+ btrfs_mark_buffer_dirty(leaf);
+ ret = 0;
+
+out:
+ btrfs_free_path(path);
+ return ret;
+}
+
+static int btrfs_rm_dev_item(struct btrfs_root *root,
+ struct btrfs_device *device)
+{
+ int ret;
+ struct btrfs_path *path;
+ struct block_device *bdev = device->bdev;
+ struct btrfs_device *next_dev;
+ struct btrfs_key key;
+ u64 total_bytes;
+ struct btrfs_fs_devices *fs_devices;
+ struct btrfs_trans_handle *trans;
+
+ root = root->fs_info->chunk_root;
+
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ trans = btrfs_start_transaction(root, 1);
+ key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
+ key.type = BTRFS_DEV_ITEM_KEY;
+ key.offset = device->devid;
+
+ ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
+ if (ret < 0)
+ goto out;
+
+ if (ret > 0) {
+ ret = -ENOENT;
+ goto out;
+ }
+
+ ret = btrfs_del_item(trans, root, path);
+ if (ret)
+ goto out;
+
+ /*
+ * at this point, the device is zero sized. We want to
+ * remove it from the devices list and zero out the old super
+ */
+ list_del_init(&device->dev_list);
+ list_del_init(&device->dev_alloc_list);
+ fs_devices = root->fs_info->fs_devices;
+
+ next_dev = list_entry(fs_devices->devices.next, struct btrfs_device,
+ dev_list);
+ if (bdev == root->fs_info->sb->s_bdev)
+ root->fs_info->sb->s_bdev = next_dev->bdev;
+ if (bdev == fs_devices->latest_bdev)
+ fs_devices->latest_bdev = next_dev->bdev;
+
+ total_bytes = btrfs_super_num_devices(&root->fs_info->super_copy);
+ btrfs_set_super_num_devices(&root->fs_info->super_copy,
+ total_bytes - 1);
+out:
+ btrfs_free_path(path);
+ btrfs_commit_transaction(trans, root);
+ return ret;
+}
+
+int btrfs_rm_device(struct btrfs_root *root, char *device_path)
+{
+ struct btrfs_device *device;
+ struct block_device *bdev;
+ struct buffer_head *bh = NULL;
+ struct btrfs_super_block *disk_super;
+ u64 all_avail;
+ u64 devid;
+ int ret = 0;
+
+ mutex_lock(&root->fs_info->alloc_mutex);
+ mutex_lock(&root->fs_info->chunk_mutex);
+ mutex_lock(&uuid_mutex);
+
+ all_avail = root->fs_info->avail_data_alloc_bits |
+ root->fs_info->avail_system_alloc_bits |
+ root->fs_info->avail_metadata_alloc_bits;
+
+ if ((all_avail & BTRFS_BLOCK_GROUP_RAID10) &&
+ btrfs_super_num_devices(&root->fs_info->super_copy) <= 4) {
+ printk("btrfs: unable to go below four devices on raid10\n");
+ ret = -EINVAL;
+ goto out;
+ }
+
+ if ((all_avail & BTRFS_BLOCK_GROUP_RAID1) &&
+ btrfs_super_num_devices(&root->fs_info->super_copy) <= 2) {
+ printk("btrfs: unable to go below two devices on raid1\n");
+ ret = -EINVAL;
+ goto out;
+ }
+
+ if (strcmp(device_path, "missing") == 0) {
+ struct list_head *cur;
+ struct list_head *devices;
+ struct btrfs_device *tmp;
+
+ device = NULL;
+ devices = &root->fs_info->fs_devices->devices;
+ list_for_each(cur, devices) {
+ tmp = list_entry(cur, struct btrfs_device, dev_list);
+ if (tmp->in_fs_metadata && !tmp->bdev) {
+ device = tmp;
+ break;
+ }
+ }
+ bdev = NULL;
+ bh = NULL;
+ disk_super = NULL;
+ if (!device) {
+ printk("btrfs: no missing devices found to remove\n");
+ goto out;
+ }
+
+ } else {
+ bdev = open_bdev_excl(device_path, 0,
+ root->fs_info->bdev_holder);
+ if (IS_ERR(bdev)) {
+ ret = PTR_ERR(bdev);
+ goto out;
+ }
+
+ bh = __bread(bdev, BTRFS_SUPER_INFO_OFFSET / 4096, 4096);
+ if (!bh) {
+ ret = -EIO;
+ goto error_close;
+ }
+ disk_super = (struct btrfs_super_block *)bh->b_data;
+ if (strncmp((char *)(&disk_super->magic), BTRFS_MAGIC,
+ sizeof(disk_super->magic))) {
+ ret = -ENOENT;
+ goto error_brelse;
+ }
+ if (memcmp(disk_super->fsid, root->fs_info->fsid,
+ BTRFS_FSID_SIZE)) {
+ ret = -ENOENT;
+ goto error_brelse;
+ }
+ devid = le64_to_cpu(disk_super->dev_item.devid);
+ device = btrfs_find_device(root, devid, NULL);
+ if (!device) {
+ ret = -ENOENT;
+ goto error_brelse;
+ }
+
+ }
+ root->fs_info->fs_devices->num_devices--;
+ root->fs_info->fs_devices->open_devices--;
+
+ ret = btrfs_shrink_device(device, 0);
+ if (ret)
+ goto error_brelse;
+
+
+ ret = btrfs_rm_dev_item(root->fs_info->chunk_root, device);
+ if (ret)
+ goto error_brelse;
+
+ if (bh) {
+ /* make sure this device isn't detected as part of
+ * the FS anymore
+ */
+ memset(&disk_super->magic, 0, sizeof(disk_super->magic));
+ set_buffer_dirty(bh);
+ sync_dirty_buffer(bh);
+
+ brelse(bh);
+ }
+
+ if (device->bdev) {
+ /* one close for the device struct or super_block */
+ close_bdev_excl(device->bdev);
+ }
+ if (bdev) {
+ /* one close for us */
+ close_bdev_excl(bdev);
+ }
+ kfree(device->name);
+ kfree(device);
+ ret = 0;
+ goto out;
+
+error_brelse:
+ brelse(bh);
+error_close:
+ if (bdev)
+ close_bdev_excl(bdev);
+out:
+ mutex_unlock(&uuid_mutex);
+ mutex_unlock(&root->fs_info->chunk_mutex);
+ mutex_unlock(&root->fs_info->alloc_mutex);
+ return ret;
+}
+
+int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
+{
+ struct btrfs_trans_handle *trans;
+ struct btrfs_device *device;
+ struct block_device *bdev;
+ struct list_head *cur;
+ struct list_head *devices;
+ u64 total_bytes;
+ int ret = 0;
+
+
+ bdev = open_bdev_excl(device_path, 0, root->fs_info->bdev_holder);
+ if (!bdev) {
+ return -EIO;
+ }
+
+ mutex_lock(&root->fs_info->alloc_mutex);
+ mutex_lock(&root->fs_info->chunk_mutex);
+
+ trans = btrfs_start_transaction(root, 1);
+ devices = &root->fs_info->fs_devices->devices;
+ list_for_each(cur, devices) {
+ device = list_entry(cur, struct btrfs_device, dev_list);
+ if (device->bdev == bdev) {
+ ret = -EEXIST;
+ goto out;
+ }
+ }
+
+ device = kzalloc(sizeof(*device), GFP_NOFS);
+ if (!device) {
+ /* we can safely leave the fs_devices entry around */
+ ret = -ENOMEM;
+ goto out_close_bdev;
+ }
+
+ device->barriers = 1;
+ device->work.func = pending_bios_fn;
+ generate_random_uuid(device->uuid);
+ spin_lock_init(&device->io_lock);
+ device->name = kstrdup(device_path, GFP_NOFS);
+ if (!device->name) {
+ kfree(device);
+ goto out_close_bdev;
+ }
+ device->io_width = root->sectorsize;
+ device->io_align = root->sectorsize;
+ device->sector_size = root->sectorsize;
+ device->total_bytes = i_size_read(bdev->bd_inode);
+ device->dev_root = root->fs_info->dev_root;
+ device->bdev = bdev;
+ device->in_fs_metadata = 1;
+
+ ret = btrfs_add_device(trans, root, device);
+ if (ret)
+ goto out_close_bdev;
+
+ total_bytes = btrfs_super_total_bytes(&root->fs_info->super_copy);
+ btrfs_set_super_total_bytes(&root->fs_info->super_copy,
+ total_bytes + device->total_bytes);
+
+ total_bytes = btrfs_super_num_devices(&root->fs_info->super_copy);
+ btrfs_set_super_num_devices(&root->fs_info->super_copy,
+ total_bytes + 1);
+
+ list_add(&device->dev_list, &root->fs_info->fs_devices->devices);
+ list_add(&device->dev_alloc_list,
+ &root->fs_info->fs_devices->alloc_list);
+ root->fs_info->fs_devices->num_devices++;
+ root->fs_info->fs_devices->open_devices++;
+out:
+ btrfs_end_transaction(trans, root);
+ mutex_unlock(&root->fs_info->chunk_mutex);
+ mutex_unlock(&root->fs_info->alloc_mutex);
+
+ return ret;
+
+out_close_bdev:
+ close_bdev_excl(bdev);
+ goto out;
+}
+
+int btrfs_update_device(struct btrfs_trans_handle *trans,
+ struct btrfs_device *device)
+{
+ int ret;
+ struct btrfs_path *path;
+ struct btrfs_root *root;
+ struct btrfs_dev_item *dev_item;
+ struct extent_buffer *leaf;
+ struct btrfs_key key;
+
+ root = device->dev_root->fs_info->chunk_root;
+
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
+ key.type = BTRFS_DEV_ITEM_KEY;
+ key.offset = device->devid;
+
+ ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
+ if (ret < 0)
+ goto out;
+
+ if (ret > 0) {
+ ret = -ENOENT;
+ goto out;
+ }
+
+ leaf = path->nodes[0];
+ dev_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_item);
+
+ btrfs_set_device_id(leaf, dev_item, device->devid);
+ btrfs_set_device_type(leaf, dev_item, device->type);
+ btrfs_set_device_io_align(leaf, dev_item, device->io_align);
+ btrfs_set_device_io_width(leaf, dev_item, device->io_width);
+ btrfs_set_device_sector_size(leaf, dev_item, device->sector_size);
+ btrfs_set_device_total_bytes(leaf, dev_item, device->total_bytes);
+ btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used);
+ btrfs_mark_buffer_dirty(leaf);
+
+out:
+ btrfs_free_path(path);
+ return ret;
+}
+
+int btrfs_grow_device(struct btrfs_trans_handle *trans,
+ struct btrfs_device *device, u64 new_size)
+{
+ struct btrfs_super_block *super_copy =
+ &device->dev_root->fs_info->super_copy;
+ u64 old_total = btrfs_super_total_bytes(super_copy);
+ u64 diff = new_size - device->total_bytes;
+
+ btrfs_set_super_total_bytes(super_copy, old_total + diff);
+ return btrfs_update_device(trans, device);
+}
+
+static int btrfs_free_chunk(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root,
+ u64 chunk_tree, u64 chunk_objectid,
+ u64 chunk_offset)
+{
+ int ret;
+ struct btrfs_path *path;
+ struct btrfs_key key;
+
+ root = root->fs_info->chunk_root;
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ key.objectid = chunk_objectid;
+ key.offset = chunk_offset;
+ key.type = BTRFS_CHUNK_ITEM_KEY;
+
+ ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
+ BUG_ON(ret);
+
+ ret = btrfs_del_item(trans, root, path);
+ BUG_ON(ret);
+
+ btrfs_free_path(path);
+ return 0;
+}
+
+int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64
+ chunk_offset)
+{
+ struct btrfs_super_block *super_copy = &root->fs_info->super_copy;
+ struct btrfs_disk_key *disk_key;
+ struct btrfs_chunk *chunk;
+ u8 *ptr;
+ int ret = 0;
+ u32 num_stripes;
+ u32 array_size;
+ u32 len = 0;
+ u32 cur;
+ struct btrfs_key key;
+
+ array_size = btrfs_super_sys_array_size(super_copy);
+
+ ptr = super_copy->sys_chunk_array;
+ cur = 0;
+
+ while (cur < array_size) {
+ disk_key = (struct btrfs_disk_key *)ptr;
+ btrfs_disk_key_to_cpu(&key, disk_key);
+
+ len = sizeof(*disk_key);
+
+ if (key.type == BTRFS_CHUNK_ITEM_KEY) {
+ chunk = (struct btrfs_chunk *)(ptr + len);
+ num_stripes = btrfs_stack_chunk_num_stripes(chunk);
+ len += btrfs_chunk_item_size(num_stripes);
+ } else {
+ ret = -EIO;
+ break;
+ }
+ if (key.objectid == chunk_objectid &&
+ key.offset == chunk_offset) {
+ memmove(ptr, ptr + len, array_size - (cur + len));
+ array_size -= len;
+ btrfs_set_super_sys_array_size(super_copy, array_size);
+ } else {
+ ptr += len;
+ cur += len;
+ }
+ }
+ return ret;
+}
+
+
+int btrfs_relocate_chunk(struct btrfs_root *root,
+ u64 chunk_tree, u64 chunk_objectid,
+ u64 chunk_offset)
+{
+ struct extent_map_tree *em_tree;
+ struct btrfs_root *extent_root;
+ struct btrfs_trans_handle *trans;
+ struct extent_map *em;
+ struct map_lookup *map;
+ int ret;
+ int i;
+
+ printk("btrfs relocating chunk %llu\n",
+ (unsigned long long)chunk_offset);
+ root = root->fs_info->chunk_root;
+ extent_root = root->fs_info->extent_root;
+ em_tree = &root->fs_info->mapping_tree.map_tree;
+
+ /* step one, relocate all the extents inside this chunk */
+ ret = btrfs_shrink_extent_tree(extent_root, chunk_offset);
+ BUG_ON(ret);
+
+ trans = btrfs_start_transaction(root, 1);
+ BUG_ON(!trans);
+
+ /*
+ * step two, delete the device extents and the
+ * chunk tree entries
+ */
+ spin_lock(&em_tree->lock);
+ em = lookup_extent_mapping(em_tree, chunk_offset, 1);
+ spin_unlock(&em_tree->lock);
+
+ BUG_ON(em->start > chunk_offset ||
+ em->start + em->len < chunk_offset);
+ map = (struct map_lookup *)em->bdev;
+
+ for (i = 0; i < map->num_stripes; i++) {
+ ret = btrfs_free_dev_extent(trans, map->stripes[i].dev,
+ map->stripes[i].physical);
+ BUG_ON(ret);
+
+ if (map->stripes[i].dev) {
+ ret = btrfs_update_device(trans, map->stripes[i].dev);
+ BUG_ON(ret);
+ }
+ }
+ ret = btrfs_free_chunk(trans, root, chunk_tree, chunk_objectid,
+ chunk_offset);
+
+ BUG_ON(ret);
+
+ if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
+ ret = btrfs_del_sys_chunk(root, chunk_objectid, chunk_offset);
+ BUG_ON(ret);
+ }
+
+ spin_lock(&em_tree->lock);
+ remove_extent_mapping(em_tree, em);
+ kfree(map);
+ em->bdev = NULL;
+
+ /* once for the tree */
+ free_extent_map(em);
+ spin_unlock(&em_tree->lock);
+
+ /* once for us */
+ free_extent_map(em);
+
+ btrfs_end_transaction(trans, root);
+ return 0;
+}
+
+static u64 div_factor(u64 num, int factor)
+{
+ if (factor == 10)
+ return num;
+ num *= factor;
+ do_div(num, 10);
+ return num;
+}
+
+
+int btrfs_balance(struct btrfs_root *dev_root)