/*
* insert a defrag record for this inode if auto defrag is
* enabled
*/
int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
struct inode *inode)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct inode_defrag *defrag;
u64 transid;
if (!btrfs_test_opt(root, AUTO_DEFRAG))
return 0;
if (btrfs_fs_closing(root->fs_info))
return 0;
if (test_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags))
return 0;
if (trans)
transid = trans->transid;
else
transid = BTRFS_I(inode)->root->last_trans;
defrag = kzalloc(sizeof(*defrag), GFP_NOFS);
if (!defrag)
return -ENOMEM;
defrag->ino = btrfs_ino(inode);
defrag->transid = transid;
defrag->root = root->root_key.objectid;
spin_lock(&root->fs_info->defrag_inodes_lock);
if (!test_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags))
__btrfs_add_inode_defrag(inode, defrag);
else
kfree(defrag);
spin_unlock(&root->fs_info->defrag_inodes_lock);
return 0;
}
int btrfs_init_dev_replace(struct btrfs_fs_info *fs_info)
{
struct btrfs_key key;
struct btrfs_root *dev_root = fs_info->dev_root;
struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace;
struct extent_buffer *eb;
int slot;
int ret = 0;
struct btrfs_path *path = NULL;
int item_size;
struct btrfs_dev_replace_item *ptr;
u64 src_devid;
path = btrfs_alloc_path();
if (!path) {
ret = -ENOMEM;
goto out;
}
key.objectid = 0;
key.type = BTRFS_DEV_REPLACE_KEY;
key.offset = 0;
ret = btrfs_search_slot(NULL, dev_root, &key, path, 0, 0);
if (ret) {
no_valid_dev_replace_entry_found:
ret = 0;
dev_replace->replace_state =
BTRFS_DEV_REPLACE_ITEM_STATE_NEVER_STARTED;
dev_replace->cont_reading_from_srcdev_mode =
BTRFS_DEV_REPLACE_ITEM_CONT_READING_FROM_SRCDEV_MODE_ALWAYS;
dev_replace->replace_state = 0;
dev_replace->time_started = 0;
dev_replace->time_stopped = 0;
atomic64_set(&dev_replace->num_write_errors, 0);
atomic64_set(&dev_replace->num_uncorrectable_read_errors, 0);
dev_replace->cursor_left = 0;
dev_replace->committed_cursor_left = 0;
dev_replace->cursor_left_last_write_of_item = 0;
dev_replace->cursor_right = 0;
dev_replace->srcdev = NULL;
dev_replace->tgtdev = NULL;
dev_replace->is_valid = 0;
dev_replace->item_needs_writeback = 0;
goto out;
}
slot = path->slots[0];
eb = path->nodes[0];
item_size = btrfs_item_size_nr(eb, slot);
ptr = btrfs_item_ptr(eb, slot, struct btrfs_dev_replace_item);
if (item_size != sizeof(struct btrfs_dev_replace_item)) {
pr_warn("btrfs: dev_replace entry found has unexpected size, ignore entry\n");
goto no_valid_dev_replace_entry_found;
}
src_devid = btrfs_dev_replace_src_devid(eb, ptr);
dev_replace->cont_reading_from_srcdev_mode =
btrfs_dev_replace_cont_reading_from_srcdev_mode(eb, ptr);
dev_replace->replace_state = btrfs_dev_replace_replace_state(eb, ptr);
dev_replace->time_started = btrfs_dev_replace_time_started(eb, ptr);
dev_replace->time_stopped =
btrfs_dev_replace_time_stopped(eb, ptr);
atomic64_set(&dev_replace->num_write_errors,
btrfs_dev_replace_num_write_errors(eb, ptr));
atomic64_set(&dev_replace->num_uncorrectable_read_errors,
btrfs_dev_replace_num_uncorrectable_read_errors(eb, ptr));
dev_replace->cursor_left = btrfs_dev_replace_cursor_left(eb, ptr);
dev_replace->committed_cursor_left = dev_replace->cursor_left;
dev_replace->cursor_left_last_write_of_item = dev_replace->cursor_left;
dev_replace->cursor_right = btrfs_dev_replace_cursor_right(eb, ptr);
dev_replace->is_valid = 1;
dev_replace->item_needs_writeback = 0;
switch (dev_replace->replace_state) {
case BTRFS_IOCTL_DEV_REPLACE_STATE_NEVER_STARTED:
case BTRFS_IOCTL_DEV_REPLACE_STATE_FINISHED:
case BTRFS_IOCTL_DEV_REPLACE_STATE_CANCELED:
dev_replace->srcdev = NULL;
dev_replace->tgtdev = NULL;
break;
case BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED:
case BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED:
dev_replace->srcdev = btrfs_find_device(fs_info, src_devid,
NULL, NULL);
dev_replace->tgtdev = btrfs_find_device(fs_info,
BTRFS_DEV_REPLACE_DEVID,
NULL, NULL);
/*
* allow 'btrfs dev replace_cancel' if src/tgt device is
* missing
*/
if (!dev_replace->srcdev &&
!btrfs_test_opt(dev_root, DEGRADED)) {
ret = -EIO;
pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "srcdev (devid %llu) is missing, need to run 'btrfs dev scan'?\n",
(unsigned long long)src_devid);
}
if (!dev_replace->tgtdev &&
!btrfs_test_opt(dev_root, DEGRADED)) {
ret = -EIO;
pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "tgtdev (devid %llu) is missing, need to run btrfs dev scan?\n",
(unsigned long long)BTRFS_DEV_REPLACE_DEVID);
}
if (dev_replace->tgtdev) {
if (dev_replace->srcdev) {
dev_replace->tgtdev->total_bytes =
dev_replace->srcdev->total_bytes;
dev_replace->tgtdev->disk_total_bytes =
dev_replace->srcdev->disk_total_bytes;
dev_replace->tgtdev->bytes_used =
dev_replace->srcdev->bytes_used;
}
dev_replace->tgtdev->is_tgtdev_for_dev_replace = 1;
btrfs_init_dev_replace_tgtdev_for_resume(fs_info,
dev_replace->tgtdev);
}
break;
}
out:
if (path)
btrfs_free_path(path);
return ret;
}
int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
struct btrfs_root *root, int cache_only)
{
struct btrfs_path *path = NULL;
struct btrfs_key key;
int ret = 0;
int wret;
int level;
int is_extent = 0;
int next_key_ret = 0;
u64 last_ret = 0;
u64 min_trans = 0;
if (cache_only)
goto out;
if (root->fs_info->extent_root == root) {
/*
* there's recursion here right now in the tree locking,
* we can't defrag the extent root without deadlock
*/
goto out;
}
if (root->ref_cows == 0 && !is_extent)
goto out;
if (btrfs_test_opt(root, SSD))
goto out;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
level = btrfs_header_level(root->node);
if (level == 0)
goto out;
if (root->defrag_progress.objectid == 0) {
struct extent_buffer *root_node;
u32 nritems;
root_node = btrfs_lock_root_node(root);
btrfs_set_lock_blocking(root_node);
nritems = btrfs_header_nritems(root_node);
root->defrag_max.objectid = 0;
/* from above we know this is not a leaf */
btrfs_node_key_to_cpu(root_node, &root->defrag_max,
nritems - 1);
btrfs_tree_unlock(root_node);
free_extent_buffer(root_node);
memset(&key, 0, sizeof(key));
} else {
memcpy(&key, &root->defrag_progress, sizeof(key));
}
path->keep_locks = 1;
if (cache_only)
min_trans = root->defrag_trans_start;
ret = btrfs_search_forward(root, &key, NULL, path,
cache_only, min_trans);
if (ret < 0)
goto out;
if (ret > 0) {
ret = 0;
goto out;
}
btrfs_release_path(path);
wret = btrfs_search_slot(trans, root, &key, path, 0, 1);
if (wret < 0) {
ret = wret;
goto out;
}
if (!path->nodes[1]) {
ret = 0;
goto out;
}
path->slots[1] = btrfs_header_nritems(path->nodes[1]);
next_key_ret = btrfs_find_next_key(root, path, &key, 1, cache_only,
min_trans);
ret = btrfs_realloc_node(trans, root,
path->nodes[1], 0,
cache_only, &last_ret,
&root->defrag_progress);
if (ret) {
WARN_ON(ret == -EAGAIN);
goto out;
}
if (next_key_ret == 0) {
memcpy(&root->defrag_progress, &key, sizeof(key));
ret = -EAGAIN;
}
out:
if (path)
btrfs_free_path(path);
if (ret == -EAGAIN) {
if (root->defrag_max.objectid > root->defrag_progress.objectid)
goto done;
if (root->defrag_max.type > root->defrag_progress.type)
goto done;
if (root->defrag_max.offset > root->defrag_progress.offset)
goto done;
ret = 0;
}
done:
if (ret != -EAGAIN) {
memset(&root->defrag_progress, 0,
sizeof(root->defrag_progress));
root->defrag_trans_start = trans->transid;
}
return ret;
}
