return next;
}
- static inline struct btrfs_root *btrfs_get_fs_root(struct btrfs_root *root,
- u64 root_id)
- {
- struct btrfs_key root_key;
-
- if (root->objectid == root_id)
- return root;
-
- root_key.objectid = root_id;
- root_key.type = BTRFS_ROOT_ITEM_KEY;
- root_key.offset = (u64)-1;
- return btrfs_read_fs_root_no_name(root->fs_info, &root_key);
- }
-
static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct btrfs_delayed_item *item)
* btrfs_readdir_delayed_dir_index - read dir info stored in the delayed tree
*
*/
-int btrfs_readdir_delayed_dir_index(struct file *filp, void *dirent,
- filldir_t filldir,
+int btrfs_readdir_delayed_dir_index(struct dir_context *ctx,
struct list_head *ins_list)
{
struct btrfs_dir_item *di;
list_for_each_entry_safe(curr, next, ins_list, readdir_list) {
list_del(&curr->readdir_list);
- if (curr->key.offset < filp->f_pos) {
+ if (curr->key.offset < ctx->pos) {
if (atomic_dec_and_test(&curr->refs))
kfree(curr);
continue;
}
- filp->f_pos = curr->key.offset;
+ ctx->pos = curr->key.offset;
di = (struct btrfs_dir_item *)curr->data;
name = (char *)(di + 1);
d_type = btrfs_filetype_table[di->type];
btrfs_disk_key_to_cpu(&location, &di->location);
- over = filldir(dirent, name, name_len, curr->key.offset,
+ over = !dir_emit(ctx, name, name_len,
location.objectid, d_type);
if (atomic_dec_and_test(&curr->refs))
return try_release_extent_buffer(page);
}
-static void btree_invalidatepage(struct page *page, unsigned long offset)
+static void btree_invalidatepage(struct page *page, unsigned int offset,
+ unsigned int length)
{
struct extent_io_tree *tree;
tree = &BTRFS_I(page->mapping->host)->io_tree;
root->objectid = objectid;
root->last_trans = 0;
root->highest_objectid = 0;
+ root->nr_delalloc_inodes = 0;
+ root->nr_ordered_extents = 0;
root->name = NULL;
root->inode_tree = RB_ROOT;
INIT_RADIX_TREE(&root->delayed_nodes_tree, GFP_ATOMIC);
INIT_LIST_HEAD(&root->dirty_list);
INIT_LIST_HEAD(&root->root_list);
+ INIT_LIST_HEAD(&root->delalloc_inodes);
+ INIT_LIST_HEAD(&root->delalloc_root);
+ INIT_LIST_HEAD(&root->ordered_extents);
+ INIT_LIST_HEAD(&root->ordered_root);
INIT_LIST_HEAD(&root->logged_list[0]);
INIT_LIST_HEAD(&root->logged_list[1]);
spin_lock_init(&root->orphan_lock);
spin_lock_init(&root->inode_lock);
+ spin_lock_init(&root->delalloc_lock);
+ spin_lock_init(&root->ordered_extent_lock);
spin_lock_init(&root->accounting_lock);
spin_lock_init(&root->log_extents_lock[0]);
spin_lock_init(&root->log_extents_lock[1]);
atomic_set(&root->log_writers, 0);
atomic_set(&root->log_batch, 0);
atomic_set(&root->orphan_inodes, 0);
+ atomic_set(&root->refs, 1);
root->log_transid = 0;
root->last_log_commit = 0;
extent_io_tree_init(&root->dirty_log_pages,
spin_lock_init(&root->root_item_lock);
}
- static int __must_check find_and_setup_root(struct btrfs_root *tree_root,
- struct btrfs_fs_info *fs_info,
- u64 objectid,
- struct btrfs_root *root)
- {
- int ret;
- u32 blocksize;
- u64 generation;
-
- __setup_root(tree_root->nodesize, tree_root->leafsize,
- tree_root->sectorsize, tree_root->stripesize,
- root, fs_info, objectid);
- ret = btrfs_find_last_root(tree_root, objectid,
- &root->root_item, &root->root_key);
- if (ret > 0)
- return -ENOENT;
- else if (ret < 0)
- return ret;
-
- generation = btrfs_root_generation(&root->root_item);
- blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item));
- root->commit_root = NULL;
- root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item),
- blocksize, generation);
- if (!root->node || !btrfs_buffer_uptodate(root->node, generation, 0)) {
- free_extent_buffer(root->node);
- root->node = NULL;
- return -EIO;
- }
- root->commit_root = btrfs_root_node(root);
- return 0;
- }
-
static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info)
{
struct btrfs_root *root = kzalloc(sizeof(*root), GFP_NOFS);
return 0;
}
- struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root,
- struct btrfs_key *location)
+ struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
+ struct btrfs_key *key)
{
struct btrfs_root *root;
struct btrfs_fs_info *fs_info = tree_root->fs_info;
struct btrfs_path *path;
- struct extent_buffer *l;
u64 generation;
u32 blocksize;
- int ret = 0;
- int slot;
+ int ret;
- root = btrfs_alloc_root(fs_info);
- if (!root)
+ path = btrfs_alloc_path();
+ if (!path)
return ERR_PTR(-ENOMEM);
- if (location->offset == (u64)-1) {
- ret = find_and_setup_root(tree_root, fs_info,
- location->objectid, root);
- if (ret) {
- kfree(root);
- return ERR_PTR(ret);
- }
- goto out;
+
+ root = btrfs_alloc_root(fs_info);
+ if (!root) {
+ ret = -ENOMEM;
+ goto alloc_fail;
}
__setup_root(tree_root->nodesize, tree_root->leafsize,
tree_root->sectorsize, tree_root->stripesize,
- root, fs_info, location->objectid);
+ root, fs_info, key->objectid);
- path = btrfs_alloc_path();
- if (!path) {
- kfree(root);
- return ERR_PTR(-ENOMEM);
- }
- ret = btrfs_search_slot(NULL, tree_root, location, path, 0, 0);
- if (ret == 0) {
- l = path->nodes[0];
- slot = path->slots[0];
- btrfs_read_root_item(l, slot, &root->root_item);
- memcpy(&root->root_key, location, sizeof(*location));
- }
- btrfs_free_path(path);
+ ret = btrfs_find_root(tree_root, key, path,
+ &root->root_item, &root->root_key);
if (ret) {
- kfree(root);
if (ret > 0)
ret = -ENOENT;
- return ERR_PTR(ret);
+ goto find_fail;
}
generation = btrfs_root_generation(&root->root_item);
blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item));
root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item),
blocksize, generation);
- if (!root->node || !extent_buffer_uptodate(root->node)) {
- ret = (!root->node) ? -ENOMEM : -EIO;
-
- free_extent_buffer(root->node);
- kfree(root);
- return ERR_PTR(ret);
+ if (!root->node) {
+ ret = -ENOMEM;
+ goto find_fail;
+ } else if (!btrfs_buffer_uptodate(root->node, generation, 0)) {
+ ret = -EIO;
+ goto read_fail;
}
-
root->commit_root = btrfs_root_node(root);
out:
- if (location->objectid != BTRFS_TREE_LOG_OBJECTID) {
+ btrfs_free_path(path);
+ return root;
+
+ read_fail:
+ free_extent_buffer(root->node);
+ find_fail:
+ kfree(root);
+ alloc_fail:
+ root = ERR_PTR(ret);
+ goto out;
+ }
+
+ struct btrfs_root *btrfs_read_fs_root(struct btrfs_root *tree_root,
+ struct btrfs_key *location)
+ {
+ struct btrfs_root *root;
+
+ root = btrfs_read_tree_root(tree_root, location);
+ if (IS_ERR(root))
+ return root;
+
+ if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
root->ref_cows = 1;
btrfs_check_and_init_root_item(&root->root_item);
}
return root;
}
+ int btrfs_init_fs_root(struct btrfs_root *root)
+ {
+ int ret;
+
+ root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS);
+ root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned),
+ GFP_NOFS);
+ if (!root->free_ino_pinned || !root->free_ino_ctl) {
+ ret = -ENOMEM;
+ goto fail;
+ }
+
+ btrfs_init_free_ino_ctl(root);
+ mutex_init(&root->fs_commit_mutex);
+ spin_lock_init(&root->cache_lock);
+ init_waitqueue_head(&root->cache_wait);
+
+ ret = get_anon_bdev(&root->anon_dev);
+ if (ret)
+ goto fail;
+ return 0;
+ fail:
+ kfree(root->free_ino_ctl);
+ kfree(root->free_ino_pinned);
+ return ret;
+ }
+
+ struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
+ u64 root_id)
+ {
+ struct btrfs_root *root;
+
+ spin_lock(&fs_info->fs_roots_radix_lock);
+ root = radix_tree_lookup(&fs_info->fs_roots_radix,
+ (unsigned long)root_id);
+ spin_unlock(&fs_info->fs_roots_radix_lock);
+ return root;
+ }
+
+ int btrfs_insert_fs_root(struct btrfs_fs_info *fs_info,
+ struct btrfs_root *root)
+ {
+ int ret;
+
+ ret = radix_tree_preload(GFP_NOFS & ~__GFP_HIGHMEM);
+ if (ret)
+ return ret;
+
+ spin_lock(&fs_info->fs_roots_radix_lock);
+ ret = radix_tree_insert(&fs_info->fs_roots_radix,
+ (unsigned long)root->root_key.objectid,
+ root);
+ if (ret == 0)
+ root->in_radix = 1;
+ spin_unlock(&fs_info->fs_roots_radix_lock);
+ radix_tree_preload_end();
+
+ return ret;
+ }
+
struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info,
struct btrfs_key *location)
{
return fs_info->quota_root ? fs_info->quota_root :
ERR_PTR(-ENOENT);
again:
- spin_lock(&fs_info->fs_roots_radix_lock);
- root = radix_tree_lookup(&fs_info->fs_roots_radix,
- (unsigned long)location->objectid);
- spin_unlock(&fs_info->fs_roots_radix_lock);
+ root = btrfs_lookup_fs_root(fs_info, location->objectid);
if (root)
return root;
- root = btrfs_read_fs_root_no_radix(fs_info->tree_root, location);
+ root = btrfs_read_fs_root(fs_info->tree_root, location);
if (IS_ERR(root))
return root;
- root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS);
- root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned),
- GFP_NOFS);
- if (!root->free_ino_pinned || !root->free_ino_ctl) {
- ret = -ENOMEM;
+ if (btrfs_root_refs(&root->root_item) == 0) {
+ ret = -ENOENT;
goto fail;
}
- btrfs_init_free_ino_ctl(root);
- mutex_init(&root->fs_commit_mutex);
- spin_lock_init(&root->cache_lock);
- init_waitqueue_head(&root->cache_wait);
-
- ret = get_anon_bdev(&root->anon_dev);
+ ret = btrfs_init_fs_root(root);
if (ret)
goto fail;
- if (btrfs_root_refs(&root->root_item) == 0) {
- ret = -ENOENT;
- goto fail;
- }
-
ret = btrfs_find_orphan_item(fs_info->tree_root, location->objectid);
if (ret < 0)
goto fail;
if (ret == 0)
root->orphan_item_inserted = 1;
- ret = radix_tree_preload(GFP_NOFS & ~__GFP_HIGHMEM);
- if (ret)
- goto fail;
-
- spin_lock(&fs_info->fs_roots_radix_lock);
- ret = radix_tree_insert(&fs_info->fs_roots_radix,
- (unsigned long)root->root_key.objectid,
- root);
- if (ret == 0)
- root->in_radix = 1;
-
- spin_unlock(&fs_info->fs_roots_radix_lock);
- radix_tree_preload_end();
+ ret = btrfs_insert_fs_root(fs_info, root);
if (ret) {
if (ret == -EEXIST) {
free_fs_root(root);
}
goto fail;
}
-
- ret = btrfs_find_dead_roots(fs_info->tree_root,
- root->root_key.objectid);
- WARN_ON(ret);
return root;
fail:
free_fs_root(root);
static int cleaner_kthread(void *arg)
{
struct btrfs_root *root = arg;
+ int again;
do {
- int again = 0;
-
- if (!(root->fs_info->sb->s_flags & MS_RDONLY) &&
- down_read_trylock(&root->fs_info->sb->s_umount)) {
- if (mutex_trylock(&root->fs_info->cleaner_mutex)) {
- btrfs_run_delayed_iputs(root);
- again = btrfs_clean_one_deleted_snapshot(root);
- mutex_unlock(&root->fs_info->cleaner_mutex);
- }
- btrfs_run_defrag_inodes(root->fs_info);
- up_read(&root->fs_info->sb->s_umount);
+ again = 0;
+
+ /* Make the cleaner go to sleep early. */
+ if (btrfs_need_cleaner_sleep(root))
+ goto sleep;
+
+ if (!mutex_trylock(&root->fs_info->cleaner_mutex))
+ goto sleep;
+
+ /*
+ * Avoid the problem that we change the status of the fs
+ * during the above check and trylock.
+ */
+ if (btrfs_need_cleaner_sleep(root)) {
+ mutex_unlock(&root->fs_info->cleaner_mutex);
+ goto sleep;
}
+ btrfs_run_delayed_iputs(root);
+ again = btrfs_clean_one_deleted_snapshot(root);
+ mutex_unlock(&root->fs_info->cleaner_mutex);
+
+ /*
+ * The defragger has dealt with the R/O remount and umount,
+ * needn't do anything special here.
+ */
+ btrfs_run_defrag_inodes(root->fs_info);
+ sleep:
if (!try_to_freeze() && !again) {
set_current_state(TASK_INTERRUPTIBLE);
if (!kthread_should_stop())
}
now = get_seconds();
- if (!cur->blocked &&
+ if (cur->state < TRANS_STATE_BLOCKED &&
(now < cur->start_time || now - cur->start_time < 30)) {
spin_unlock(&root->fs_info->trans_lock);
delay = HZ * 5;
list_del(&gang[0]->root_list);
if (gang[0]->in_radix) {
- btrfs_free_fs_root(fs_info, gang[0]);
+ btrfs_drop_and_free_fs_root(fs_info, gang[0]);
} else {
free_extent_buffer(gang[0]->node);
free_extent_buffer(gang[0]->commit_root);
- kfree(gang[0]);
+ btrfs_put_fs_root(gang[0]);
}
}
if (!ret)
break;
for (i = 0; i < ret; i++)
- btrfs_free_fs_root(fs_info, gang[i]);
+ btrfs_drop_and_free_fs_root(fs_info, gang[i]);
}
}
int backup_index = 0;
tree_root = fs_info->tree_root = btrfs_alloc_root(fs_info);
- extent_root = fs_info->extent_root = btrfs_alloc_root(fs_info);
- csum_root = fs_info->csum_root = btrfs_alloc_root(fs_info);
chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info);
- dev_root = fs_info->dev_root = btrfs_alloc_root(fs_info);
- quota_root = fs_info->quota_root = btrfs_alloc_root(fs_info);
-
- if (!tree_root || !extent_root || !csum_root ||
- !chunk_root || !dev_root || !quota_root) {
+ if (!tree_root || !chunk_root) {
err = -ENOMEM;
goto fail;
}
INIT_LIST_HEAD(&fs_info->trans_list);
INIT_LIST_HEAD(&fs_info->dead_roots);
INIT_LIST_HEAD(&fs_info->delayed_iputs);
- INIT_LIST_HEAD(&fs_info->delalloc_inodes);
+ INIT_LIST_HEAD(&fs_info->delalloc_roots);
INIT_LIST_HEAD(&fs_info->caching_block_groups);
- spin_lock_init(&fs_info->delalloc_lock);
+ spin_lock_init(&fs_info->delalloc_root_lock);
spin_lock_init(&fs_info->trans_lock);
spin_lock_init(&fs_info->fs_roots_radix_lock);
spin_lock_init(&fs_info->delayed_iput_lock);
fs_info->max_inline = 8192 * 1024;
fs_info->metadata_ratio = 0;
fs_info->defrag_inodes = RB_ROOT;
- fs_info->trans_no_join = 0;
fs_info->free_chunk_space = 0;
fs_info->tree_mod_log = RB_ROOT;
fs_info->thread_pool_size = min_t(unsigned long,
num_online_cpus() + 2, 8);
- INIT_LIST_HEAD(&fs_info->ordered_extents);
- spin_lock_init(&fs_info->ordered_extent_lock);
+ INIT_LIST_HEAD(&fs_info->ordered_roots);
+ spin_lock_init(&fs_info->ordered_root_lock);
fs_info->delayed_root = kmalloc(sizeof(struct btrfs_delayed_root),
GFP_NOFS);
if (!fs_info->delayed_root) {
fs_info->qgroup_seq = 1;
fs_info->quota_enabled = 0;
fs_info->pending_quota_state = 0;
+ fs_info->qgroup_ulist = NULL;
mutex_init(&fs_info->qgroup_rescan_lock);
btrfs_init_free_cluster(&fs_info->meta_alloc_cluster);
btrfs_set_root_node(&tree_root->root_item, tree_root->node);
tree_root->commit_root = btrfs_root_node(tree_root);
- ret = find_and_setup_root(tree_root, fs_info,
- BTRFS_EXTENT_TREE_OBJECTID, extent_root);
- if (ret)
+ location.objectid = BTRFS_EXTENT_TREE_OBJECTID;
+ location.type = BTRFS_ROOT_ITEM_KEY;
+ location.offset = 0;
+
+ extent_root = btrfs_read_tree_root(tree_root, &location);
+ if (IS_ERR(extent_root)) {
+ ret = PTR_ERR(extent_root);
goto recovery_tree_root;
+ }
extent_root->track_dirty = 1;
+ fs_info->extent_root = extent_root;
- ret = find_and_setup_root(tree_root, fs_info,
- BTRFS_DEV_TREE_OBJECTID, dev_root);
- if (ret)
+ location.objectid = BTRFS_DEV_TREE_OBJECTID;
+ dev_root = btrfs_read_tree_root(tree_root, &location);
+ if (IS_ERR(dev_root)) {
+ ret = PTR_ERR(dev_root);
goto recovery_tree_root;
+ }
dev_root->track_dirty = 1;
+ fs_info->dev_root = dev_root;
+ btrfs_init_devices_late(fs_info);
- ret = find_and_setup_root(tree_root, fs_info,
- BTRFS_CSUM_TREE_OBJECTID, csum_root);
- if (ret)
+ location.objectid = BTRFS_CSUM_TREE_OBJECTID;
+ csum_root = btrfs_read_tree_root(tree_root, &location);
+ if (IS_ERR(csum_root)) {
+ ret = PTR_ERR(csum_root);
goto recovery_tree_root;
+ }
csum_root->track_dirty = 1;
+ fs_info->csum_root = csum_root;
- ret = find_and_setup_root(tree_root, fs_info,
- BTRFS_QUOTA_TREE_OBJECTID, quota_root);
- if (ret) {
- kfree(quota_root);
- quota_root = fs_info->quota_root = NULL;
- } else {
+ location.objectid = BTRFS_QUOTA_TREE_OBJECTID;
+ quota_root = btrfs_read_tree_root(tree_root, &location);
+ if (!IS_ERR(quota_root)) {
quota_root->track_dirty = 1;
fs_info->quota_enabled = 1;
fs_info->pending_quota_state = 1;
+ fs_info->quota_root = quota_root;
}
fs_info->generation = generation;
location.objectid = BTRFS_FS_TREE_OBJECTID;
location.type = BTRFS_ROOT_ITEM_KEY;
- location.offset = (u64)-1;
+ location.offset = 0;
fs_info->fs_root = btrfs_read_fs_root_no_name(fs_info, &location);
- if (!fs_info->fs_root)
- goto fail_qgroup;
if (IS_ERR(fs_info->fs_root)) {
err = PTR_ERR(fs_info->fs_root);
goto fail_qgroup;
return ret;
}
+ btrfs_qgroup_rescan_resume(fs_info);
+
return 0;
fail_qgroup:
BTRFS_BLOCK_GROUP_RAID10)) {
num_tolerated_disk_barrier_failures = 1;
} else if (flags &
- BTRFS_BLOCK_GROUP_RAID5) {
+ BTRFS_BLOCK_GROUP_RAID6) {
num_tolerated_disk_barrier_failures = 2;
}
}
return ret;
}
- void btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
+ /* Drop a fs root from the radix tree and free it. */
+ void btrfs_drop_and_free_fs_root(struct btrfs_fs_info *fs_info,
+ struct btrfs_root *root)
{
spin_lock(&fs_info->fs_roots_radix_lock);
radix_tree_delete(&fs_info->fs_roots_radix,
kfree(root->free_ino_ctl);
kfree(root->free_ino_pinned);
kfree(root->name);
- kfree(root);
+ btrfs_put_fs_root(root);
+ }
+
+ void btrfs_free_fs_root(struct btrfs_root *root)
+ {
+ free_fs_root(root);
}
int btrfs_cleanup_fs_roots(struct btrfs_fs_info *fs_info)
INIT_LIST_HEAD(&splice);
mutex_lock(&root->fs_info->ordered_operations_mutex);
- spin_lock(&root->fs_info->ordered_extent_lock);
+ spin_lock(&root->fs_info->ordered_root_lock);
list_splice_init(&t->ordered_operations, &splice);
while (!list_empty(&splice)) {
ordered_operations);
list_del_init(&btrfs_inode->ordered_operations);
- spin_unlock(&root->fs_info->ordered_extent_lock);
+ spin_unlock(&root->fs_info->ordered_root_lock);
btrfs_invalidate_inodes(btrfs_inode->root);
- spin_lock(&root->fs_info->ordered_extent_lock);
+ spin_lock(&root->fs_info->ordered_root_lock);
}
- spin_unlock(&root->fs_info->ordered_extent_lock);
+ spin_unlock(&root->fs_info->ordered_root_lock);
mutex_unlock(&root->fs_info->ordered_operations_mutex);
}
{
struct btrfs_ordered_extent *ordered;
- spin_lock(&root->fs_info->ordered_extent_lock);
+ spin_lock(&root->ordered_extent_lock);
/*
* This will just short circuit the ordered completion stuff which will
* make sure the ordered extent gets properly cleaned up.
*/
- list_for_each_entry(ordered, &root->fs_info->ordered_extents,
+ list_for_each_entry(ordered, &root->ordered_extents,
root_extent_list)
set_bit(BTRFS_ORDERED_IOERR, &ordered->flags);
- spin_unlock(&root->fs_info->ordered_extent_lock);
+ spin_unlock(&root->ordered_extent_lock);
+ }
+
+ static void btrfs_destroy_all_ordered_extents(struct btrfs_fs_info *fs_info)
+ {
+ struct btrfs_root *root;
+ struct list_head splice;
+
+ INIT_LIST_HEAD(&splice);
+
+ spin_lock(&fs_info->ordered_root_lock);
+ list_splice_init(&fs_info->ordered_roots, &splice);
+ while (!list_empty(&splice)) {
+ root = list_first_entry(&splice, struct btrfs_root,
+ ordered_root);
+ list_del_init(&root->ordered_root);
+
+ btrfs_destroy_ordered_extents(root);
+
+ cond_resched_lock(&fs_info->ordered_root_lock);
+ }
+ spin_unlock(&fs_info->ordered_root_lock);
}
int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
while ((node = rb_first(&delayed_refs->root)) != NULL) {
struct btrfs_delayed_ref_head *head = NULL;
+ bool pin_bytes = false;
ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
atomic_set(&ref->refs, 1);
}
if (head->must_insert_reserved)
- btrfs_pin_extent(root, ref->bytenr,
- ref->num_bytes, 1);
+ pin_bytes = true;
btrfs_free_delayed_extent_op(head->extent_op);
delayed_refs->num_heads--;
if (list_empty(&head->cluster))
ref->in_tree = 0;
rb_erase(&ref->rb_node, &delayed_refs->root);
delayed_refs->num_entries--;
- if (head)
- mutex_unlock(&head->mutex);
spin_unlock(&delayed_refs->lock);
+ if (head) {
+ if (pin_bytes)
+ btrfs_pin_extent(root, ref->bytenr,
+ ref->num_bytes, 1);
+ mutex_unlock(&head->mutex);
+ }
btrfs_put_delayed_ref(ref);
cond_resched();
INIT_LIST_HEAD(&splice);
- spin_lock(&root->fs_info->delalloc_lock);
- list_splice_init(&root->fs_info->delalloc_inodes, &splice);
+ spin_lock(&root->delalloc_lock);
+ list_splice_init(&root->delalloc_inodes, &splice);
while (!list_empty(&splice)) {
- btrfs_inode = list_entry(splice.next, struct btrfs_inode,
- delalloc_inodes);
+ btrfs_inode = list_first_entry(&splice, struct btrfs_inode,
+ delalloc_inodes);
list_del_init(&btrfs_inode->delalloc_inodes);
clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
&btrfs_inode->runtime_flags);
- spin_unlock(&root->fs_info->delalloc_lock);
+ spin_unlock(&root->delalloc_lock);
btrfs_invalidate_inodes(btrfs_inode->root);
- spin_lock(&root->fs_info->delalloc_lock);
+ spin_lock(&root->delalloc_lock);
}
- spin_unlock(&root->fs_info->delalloc_lock);
+ spin_unlock(&root->delalloc_lock);
+ }
+
+ static void btrfs_destroy_all_delalloc_inodes(struct btrfs_fs_info *fs_info)
+ {
+ struct btrfs_root *root;
+ struct list_head splice;
+
+ INIT_LIST_HEAD(&splice);
+
+ spin_lock(&fs_info->delalloc_root_lock);
+ list_splice_init(&fs_info->delalloc_roots, &splice);
+ while (!list_empty(&splice)) {
+ root = list_first_entry(&splice, struct btrfs_root,
+ delalloc_root);
+ list_del_init(&root->delalloc_root);
+ root = btrfs_grab_fs_root(root);
+ BUG_ON(!root);
+ spin_unlock(&fs_info->delalloc_root_lock);
+
+ btrfs_destroy_delalloc_inodes(root);
+ btrfs_put_fs_root(root);
+
+ spin_lock(&fs_info->delalloc_root_lock);
+ }
+ spin_unlock(&fs_info->delalloc_root_lock);
}
static int btrfs_destroy_marked_extents(struct btrfs_root *root,
btrfs_block_rsv_release(root, &root->fs_info->trans_block_rsv,
cur_trans->dirty_pages.dirty_bytes);
- /* FIXME: cleanup wait for commit */
- cur_trans->in_commit = 1;
- cur_trans->blocked = 1;
+ cur_trans->state = TRANS_STATE_COMMIT_START;
wake_up(&root->fs_info->transaction_blocked_wait);
btrfs_evict_pending_snapshots(cur_trans);
- cur_trans->blocked = 0;
+ cur_trans->state = TRANS_STATE_UNBLOCKED;
wake_up(&root->fs_info->transaction_wait);
- cur_trans->commit_done = 1;
- wake_up(&cur_trans->commit_wait);
-
btrfs_destroy_delayed_inodes(root);
btrfs_assert_delayed_root_empty(root);
btrfs_destroy_pinned_extent(root,
root->fs_info->pinned_extents);
+ cur_trans->state =TRANS_STATE_COMPLETED;
+ wake_up(&cur_trans->commit_wait);
+
/*
memset(cur_trans, 0, sizeof(*cur_trans));
kmem_cache_free(btrfs_transaction_cachep, cur_trans);
spin_lock(&root->fs_info->trans_lock);
list_splice_init(&root->fs_info->trans_list, &list);
- root->fs_info->trans_no_join = 1;
+ root->fs_info->running_transaction = NULL;
spin_unlock(&root->fs_info->trans_lock);
while (!list_empty(&list)) {
btrfs_destroy_ordered_operations(t, root);
- btrfs_destroy_ordered_extents(root);
+ btrfs_destroy_all_ordered_extents(root->fs_info);
btrfs_destroy_delayed_refs(t, root);
- /* FIXME: cleanup wait for commit */
- t->in_commit = 1;
- t->blocked = 1;
+ /*
+ * FIXME: cleanup wait for commit
+ * We needn't acquire the lock here, because we are during
+ * the umount, there is no other task which will change it.
+ */
+ t->state = TRANS_STATE_COMMIT_START;
smp_mb();
if (waitqueue_active(&root->fs_info->transaction_blocked_wait))
wake_up(&root->fs_info->transaction_blocked_wait);
btrfs_evict_pending_snapshots(t);
- t->blocked = 0;
+ t->state = TRANS_STATE_UNBLOCKED;
smp_mb();
if (waitqueue_active(&root->fs_info->transaction_wait))
wake_up(&root->fs_info->transaction_wait);
- t->commit_done = 1;
- smp_mb();
- if (waitqueue_active(&t->commit_wait))
- wake_up(&t->commit_wait);
-
btrfs_destroy_delayed_inodes(root);
btrfs_assert_delayed_root_empty(root);
- btrfs_destroy_delalloc_inodes(root);
-
- spin_lock(&root->fs_info->trans_lock);
- root->fs_info->running_transaction = NULL;
- spin_unlock(&root->fs_info->trans_lock);
+ btrfs_destroy_all_delalloc_inodes(root->fs_info);
btrfs_destroy_marked_extents(root, &t->dirty_pages,
EXTENT_DIRTY);
btrfs_destroy_pinned_extent(root,
root->fs_info->pinned_extents);
+ t->state = TRANS_STATE_COMPLETED;
+ smp_mb();
+ if (waitqueue_active(&t->commit_wait))
+ wake_up(&t->commit_wait);
+
atomic_set(&t->use_count, 0);
list_del_init(&t->list);
memset(t, 0, sizeof(*t));
kmem_cache_free(btrfs_transaction_cachep, t);
}
- spin_lock(&root->fs_info->trans_lock);
- root->fs_info->trans_no_join = 0;
- spin_unlock(&root->fs_info->trans_lock);
mutex_unlock(&root->fs_info->transaction_kthread_mutex);
return 0;
kmem_cache_free(extent_buffer_cache, eb);
}
}
+
+ #define btrfs_debug_check_extent_io_range(inode, start, end) \
+ __btrfs_debug_check_extent_io_range(__func__, (inode), (start), (end))
+ static inline void __btrfs_debug_check_extent_io_range(const char *caller,
+ struct inode *inode, u64 start, u64 end)
+ {
+ u64 isize = i_size_read(inode);
+
+ if (end >= PAGE_SIZE && (end % 2) == 0 && end != isize - 1) {
+ printk_ratelimited(KERN_DEBUG
+ "btrfs: %s: ino %llu isize %llu odd range [%llu,%llu]\n",
+ caller,
+ (unsigned long long)btrfs_ino(inode),
+ (unsigned long long)isize,
+ (unsigned long long)start,
+ (unsigned long long)end);
+ }
+ }
#else
#define btrfs_leak_debug_add(new, head) do {} while (0)
#define btrfs_leak_debug_del(entry) do {} while (0)
#define btrfs_leak_debug_check() do {} while (0)
+ #define btrfs_debug_check_extent_io_range(c, s, e) do {} while (0)
#endif
#define BUFFER_LRU_MAX 64
int err;
int clear = 0;
+ btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+
+ if (bits & EXTENT_DELALLOC)
+ bits |= EXTENT_NORESERVE;
+
if (delete)
bits |= ~EXTENT_CTLBITS;
bits |= EXTENT_FIRST_DELALLOC;
struct extent_state *state;
struct rb_node *node;
+ btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+
spin_lock(&tree->lock);
again:
while (1) {
u64 last_start;
u64 last_end;
+ btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+
bits |= EXTENT_FIRST_DELALLOC;
again:
if (!prealloc && (mask & __GFP_WAIT)) {
u64 last_start;
u64 last_end;
+ btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+
again:
if (!prealloc && (mask & __GFP_WAIT)) {
prealloc = alloc_extent_state(mask);
struct extent_state *cached = NULL;
struct extent_state *state;
struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
+ struct inode *inode = page->mapping->host;
pr_debug("end_bio_extent_readpage: bi_sector=%llu, err=%d, "
"mirror=%lu\n", (u64)bio->bi_sector, err,
io_bio->mirror_num);
- tree = &BTRFS_I(page->mapping->host)->io_tree;
+ tree = &BTRFS_I(inode)->io_tree;
/* We always issue full-page reads, but if some block
* in a page fails to read, blk_update_request() will
unlock_extent_cached(tree, start, end, &cached, GFP_ATOMIC);
if (uptodate) {
+ loff_t i_size = i_size_read(inode);
+ pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
+ unsigned offset;
+
+ /* Zero out the end if this page straddles i_size */
+ offset = i_size & (PAGE_CACHE_SIZE-1);
+ if (page->index == end_index && offset)
+ zero_user_segment(page, offset, PAGE_CACHE_SIZE);
SetPageUptodate(page);
} else {
ClearPageUptodate(page);
if (old_compressed)
contig = bio->bi_sector == sector;
else
- contig = bio->bi_sector + (bio->bi_size >> 9) ==
- sector;
+ contig = bio_end_sector(bio) == sector;
if (prev_bio_flags != bio_flags || !contig ||
merge_bio(rw, tree, page, offset, page_size, bio, bio_flags) ||
pg_offset = i_size & (PAGE_CACHE_SIZE - 1);
if (page->index > end_index ||
(page->index == end_index && !pg_offset)) {
- page->mapping->a_ops->invalidatepage(page, 0);
+ page->mapping->a_ops->invalidatepage(page, 0, PAGE_CACHE_SIZE);
unlock_page(page);
return 0;
}
#include <linux/string.h>
#include <linux/backing-dev.h>
#include <linux/mpage.h>
+#include <linux/aio.h>
#include <linux/falloc.h>
#include <linux/swap.h>
#include <linux/writeback.h>
ret = PTR_ERR(inode_root);
goto cleanup;
}
- if (btrfs_root_refs(&inode_root->root_item) == 0) {
- ret = -ENOENT;
- goto cleanup;
- }
key.objectid = defrag->ino;
btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
}
+ static noinline int check_can_nocow(struct inode *inode, loff_t pos,
+ size_t *write_bytes)
+ {
+ struct btrfs_trans_handle *trans;
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+ struct btrfs_ordered_extent *ordered;
+ u64 lockstart, lockend;
+ u64 num_bytes;
+ int ret;
+
+ lockstart = round_down(pos, root->sectorsize);
+ lockend = lockstart + round_up(*write_bytes, root->sectorsize) - 1;
+
+ while (1) {
+ lock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+ ordered = btrfs_lookup_ordered_range(inode, lockstart,
+ lockend - lockstart + 1);
+ if (!ordered) {
+ break;
+ }
+ unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+ btrfs_start_ordered_extent(inode, ordered, 1);
+ btrfs_put_ordered_extent(ordered);
+ }
+
+ trans = btrfs_join_transaction(root);
+ if (IS_ERR(trans)) {
+ unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+ return PTR_ERR(trans);
+ }
+
+ num_bytes = lockend - lockstart + 1;
+ ret = can_nocow_extent(trans, inode, lockstart, &num_bytes, NULL, NULL,
+ NULL);
+ btrfs_end_transaction(trans, root);
+ if (ret <= 0) {
+ ret = 0;
+ } else {
+ clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend,
+ EXTENT_DIRTY | EXTENT_DELALLOC |
+ EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 0, 0,
+ NULL, GFP_NOFS);
+ *write_bytes = min_t(size_t, *write_bytes, num_bytes);
+ }
+
+ unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+
+ return ret;
+ }
+
static noinline ssize_t __btrfs_buffered_write(struct file *file,
struct iov_iter *i,
loff_t pos)
struct inode *inode = file_inode(file);
struct btrfs_root *root = BTRFS_I(inode)->root;
struct page **pages = NULL;
+ u64 release_bytes = 0;
unsigned long first_index;
size_t num_written = 0;
int nrptrs;
int ret = 0;
+ bool only_release_metadata = false;
bool force_page_uptodate = false;
nrptrs = min((iov_iter_count(i) + PAGE_CACHE_SIZE - 1) /
offset);
size_t num_pages = (write_bytes + offset +
PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+ size_t reserve_bytes;
size_t dirty_pages;
size_t copied;
break;
}
- ret = btrfs_delalloc_reserve_space(inode,
- num_pages << PAGE_CACHE_SHIFT);
+ reserve_bytes = num_pages << PAGE_CACHE_SHIFT;
+ ret = btrfs_check_data_free_space(inode, reserve_bytes);
+ if (ret == -ENOSPC &&
+ (BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
+ BTRFS_INODE_PREALLOC))) {
+ ret = check_can_nocow(inode, pos, &write_bytes);
+ if (ret > 0) {
+ only_release_metadata = true;
+ /*
+ * our prealloc extent may be smaller than
+ * write_bytes, so scale down.
+ */
+ num_pages = (write_bytes + offset +
+ PAGE_CACHE_SIZE - 1) >>
+ PAGE_CACHE_SHIFT;
+ reserve_bytes = num_pages << PAGE_CACHE_SHIFT;
+ ret = 0;
+ } else {
+ ret = -ENOSPC;
+ }
+ }
+
if (ret)
break;
+ ret = btrfs_delalloc_reserve_metadata(inode, reserve_bytes);
+ if (ret) {
+ if (!only_release_metadata)
+ btrfs_free_reserved_data_space(inode,
+ reserve_bytes);
+ break;
+ }
+
+ release_bytes = reserve_bytes;
+
/*
* This is going to setup the pages array with the number of
* pages we want, so we don't really need to worry about the
ret = prepare_pages(root, file, pages, num_pages,
pos, first_index, write_bytes,
force_page_uptodate);
- if (ret) {
- btrfs_delalloc_release_space(inode,
- num_pages << PAGE_CACHE_SHIFT);
+ if (ret)
break;
- }
copied = btrfs_copy_from_user(pos, num_pages,
write_bytes, pages, i);
* managed to copy.
*/
if (num_pages > dirty_pages) {
+ release_bytes = (num_pages - dirty_pages) <<
+ PAGE_CACHE_SHIFT;
if (copied > 0) {
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->outstanding_extents++;
spin_unlock(&BTRFS_I(inode)->lock);
}
- btrfs_delalloc_release_space(inode,
- (num_pages - dirty_pages) <<
- PAGE_CACHE_SHIFT);
+ if (only_release_metadata)
+ btrfs_delalloc_release_metadata(inode,
+ release_bytes);
+ else
+ btrfs_delalloc_release_space(inode,
+ release_bytes);
}
+ release_bytes = dirty_pages << PAGE_CACHE_SHIFT;
if (copied > 0) {
ret = btrfs_dirty_pages(root, inode, pages,
dirty_pages, pos, copied,
NULL);
if (ret) {
- btrfs_delalloc_release_space(inode,
- dirty_pages << PAGE_CACHE_SHIFT);
btrfs_drop_pages(pages, num_pages);
break;
}
}
+ release_bytes = 0;
btrfs_drop_pages(pages, num_pages);
+ if (only_release_metadata && copied > 0) {
+ u64 lockstart = round_down(pos, root->sectorsize);
+ u64 lockend = lockstart +
+ (dirty_pages << PAGE_CACHE_SHIFT) - 1;
+
+ set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
+ lockend, EXTENT_NORESERVE, NULL,
+ NULL, GFP_NOFS);
+ only_release_metadata = false;
+ }
+
cond_resched();
balance_dirty_pages_ratelimited(inode->i_mapping);
kfree(pages);
+ if (release_bytes) {
+ if (only_release_metadata)
+ btrfs_delalloc_release_metadata(inode, release_bytes);
+ else
+ btrfs_delalloc_release_space(inode, release_bytes);
+ }
+
return num_written ? num_written : ret;
}
size_t count, ocount;
bool sync = (file->f_flags & O_DSYNC) || IS_SYNC(file->f_mapping->host);
- sb_start_write(inode->i_sb);
-
mutex_lock(&inode->i_mutex);
err = generic_segment_checks(iov, &nr_segs, &ocount, VERIFY_READ);
if (sync)
atomic_dec(&BTRFS_I(inode)->sync_writers);
out:
- sb_end_write(inode->i_sb);
current->backing_dev_info = NULL;
return num_written ? num_written : err;
}
goto out_reserve_fail;
}
- /*
- * wait for ordered IO before we have any locks. We'll loop again
- * below with the locks held.
- */
- btrfs_wait_ordered_range(inode, alloc_start, alloc_end - alloc_start);
-
mutex_lock(&inode->i_mutex);
ret = inode_newsize_ok(inode, alloc_end);
if (ret)
alloc_start);
if (ret)
goto out;
+ } else {
+ /*
+ * If we are fallocating from the end of the file onward we
+ * need to zero out the end of the page if i_size lands in the
+ * middle of a page.
+ */
+ ret = btrfs_truncate_page(inode, inode->i_size, 0, 0);
+ if (ret)
+ goto out;
}
+ /*
+ * wait for ordered IO before we have any locks. We'll loop again
+ * below with the locks held.
+ */
+ btrfs_wait_ordered_range(inode, alloc_start, alloc_end - alloc_start);
+
locked_end = alloc_end - 1;
while (1) {
struct btrfs_ordered_extent *ordered;
}
}
- if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) {
- offset = -EINVAL;
- goto out;
- }
- if (offset > inode->i_sb->s_maxbytes) {
- offset = -EINVAL;
- goto out;
- }
-
- /* Special lock needed here? */
- if (offset != file->f_pos) {
- file->f_pos = offset;
- file->f_version = 0;
- }
+ offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes);
out:
mutex_unlock(&inode->i_mutex);
return offset;
else
ret = 0;
spin_unlock(&rsv->lock);
- return 0;
+ return ret;
}
int btrfs_truncate_free_space_cache(struct btrfs_root *root,
return 0;
}
+ #define test_msg(fmt, ...) printk(KERN_INFO "btrfs: selftest: " fmt, ##__VA_ARGS__)
+
/*
* This test just does basic sanity checking, making sure we can add an exten
* entry and remove space from either end and the middle, and make sure we can
{
int ret = 0;
- printk(KERN_ERR "Running extent only tests\n");
+ test_msg("Running extent only tests\n");
/* First just make sure we can remove an entire entry */
ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Error adding initial extents %d\n", ret);
+ test_msg("Error adding initial extents %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Error removing extent %d\n", ret);
+ test_msg("Error removing extent %d\n", ret);
return ret;
}
if (check_exists(cache, 0, 4 * 1024 * 1024)) {
- printk(KERN_ERR "Full remove left some lingering space\n");
+ test_msg("Full remove left some lingering space\n");
return -1;
}
/* Ok edge and middle cases now */
ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Error adding half extent %d\n", ret);
+ test_msg("Error adding half extent %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 1 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Error removing tail end %d\n", ret);
+ test_msg("Error removing tail end %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Error removing front end %d\n", ret);
+ test_msg("Error removing front end %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 2 * 1024 * 1024, 4096);
if (ret) {
- printk(KERN_ERR "Error removing middle piece %d\n", ret);
- test_msg("Error removing middle peice %d\n", ret);
++ test_msg("Error removing middle piece %d\n", ret);
return ret;
}
if (check_exists(cache, 0, 1 * 1024 * 1024)) {
- printk(KERN_ERR "Still have space at the front\n");
+ test_msg("Still have space at the front\n");
return -1;
}
if (check_exists(cache, 2 * 1024 * 1024, 4096)) {
- printk(KERN_ERR "Still have space in the middle\n");
+ test_msg("Still have space in the middle\n");
return -1;
}
if (check_exists(cache, 3 * 1024 * 1024, 1 * 1024 * 1024)) {
- printk(KERN_ERR "Still have space at the end\n");
+ test_msg("Still have space at the end\n");
return -1;
}
u64 next_bitmap_offset;
int ret;
- printk(KERN_ERR "Running bitmap only tests\n");
+ test_msg("Running bitmap only tests\n");
ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't create a bitmap entry %d\n", ret);
+ test_msg("Couldn't create a bitmap entry %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Error removing bitmap full range %d\n", ret);
+ test_msg("Error removing bitmap full range %d\n", ret);
return ret;
}
if (check_exists(cache, 0, 4 * 1024 * 1024)) {
- printk(KERN_ERR "Left some space in bitmap\n");
+ test_msg("Left some space in bitmap\n");
return -1;
}
ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't add to our bitmap entry %d\n", ret);
+ test_msg("Couldn't add to our bitmap entry %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 2 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Couldn't remove middle chunk %d\n", ret);
+ test_msg("Couldn't remove middle chunk %d\n", ret);
return ret;
}
ret = add_free_space_entry(cache, next_bitmap_offset -
(2 * 1024 * 1024), 4 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't add space that straddles two bitmaps"
- " %d\n", ret);
+ test_msg("Couldn't add space that straddles two bitmaps %d\n",
+ ret);
return ret;
}
ret = btrfs_remove_free_space(cache, next_bitmap_offset -
(1 * 1024 * 1024), 2 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Couldn't remove overlapping space %d\n", ret);
+ test_msg("Couldn't remove overlapping space %d\n", ret);
return ret;
}
if (check_exists(cache, next_bitmap_offset - (1 * 1024 * 1024),
2 * 1024 * 1024)) {
- printk(KERN_ERR "Left some space when removing overlapping\n");
+ test_msg("Left some space when removing overlapping\n");
return -1;
}
u64 bitmap_offset = (u64)(BITS_PER_BITMAP * 4096);
int ret;
- printk(KERN_ERR "Running bitmap and extent tests\n");
+ test_msg("Running bitmap and extent tests\n");
/*
* First let's do something simple, an extent at the same offset as the
*/
ret = add_free_space_entry(cache, 4 * 1024 * 1024, 1 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't create bitmap entry %d\n", ret);
+ test_msg("Couldn't create bitmap entry %d\n", ret);
return ret;
}
ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
if (ret) {
- printk(KERN_ERR "Couldn't add extent entry %d\n", ret);
+ test_msg("Couldn't add extent entry %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Couldn't remove extent entry %d\n", ret);
+ test_msg("Couldn't remove extent entry %d\n", ret);
return ret;
}
if (check_exists(cache, 0, 1 * 1024 * 1024)) {
- printk(KERN_ERR "Left remnants after our remove\n");
+ test_msg("Left remnants after our remove\n");
return -1;
}
/* Now to add back the extent entry and remove from the bitmap */
ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
if (ret) {
- printk(KERN_ERR "Couldn't re-add extent entry %d\n", ret);
+ test_msg("Couldn't re-add extent entry %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 4 * 1024 * 1024, 1 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Couldn't remove from bitmap %d\n", ret);
+ test_msg("Couldn't remove from bitmap %d\n", ret);
return ret;
}
if (check_exists(cache, 4 * 1024 * 1024, 1 * 1024 * 1024)) {
- printk(KERN_ERR "Left remnants in the bitmap\n");
+ test_msg("Left remnants in the bitmap\n");
return -1;
}
*/
ret = add_free_space_entry(cache, 1 * 1024 * 1024, 4 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't add to a bitmap %d\n", ret);
+ test_msg("Couldn't add to a bitmap %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 512 * 1024, 3 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Couldn't remove overlapping space %d\n", ret);
+ test_msg("Couldn't remove overlapping space %d\n", ret);
return ret;
}
if (check_exists(cache, 512 * 1024, 3 * 1024 * 1024)) {
- printk(KERN_ERR "Left over peices after removing "
- "overlapping\n");
+ test_msg("Left over peices after removing overlapping\n");
return -1;
}
/* Now with the extent entry offset into the bitmap */
ret = add_free_space_entry(cache, 4 * 1024 * 1024, 4 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't add space to the bitmap %d\n", ret);
+ test_msg("Couldn't add space to the bitmap %d\n", ret);
return ret;
}
ret = add_free_space_entry(cache, 2 * 1024 * 1024, 2 * 1024 * 1024, 0);
if (ret) {
- printk(KERN_ERR "Couldn't add extent to the cache %d\n", ret);
+ test_msg("Couldn't add extent to the cache %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 4 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Problem removing overlapping space %d\n", ret);
+ test_msg("Problem removing overlapping space %d\n", ret);
return ret;
}
if (check_exists(cache, 3 * 1024 * 1024, 4 * 1024 * 1024)) {
- printk(KERN_ERR "Left something behind when removing space");
+ test_msg("Left something behind when removing space");
return -1;
}
ret = add_free_space_entry(cache, bitmap_offset + 4 * 1024 * 1024,
4 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't add bitmap %d\n", ret);
+ test_msg("Couldn't add bitmap %d\n", ret);
return ret;
}
ret = add_free_space_entry(cache, bitmap_offset - 1 * 1024 * 1024,
5 * 1024 * 1024, 0);
if (ret) {
- printk(KERN_ERR "Couldn't add extent entry %d\n", ret);
+ test_msg("Couldn't add extent entry %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, bitmap_offset + 1 * 1024 * 1024,
5 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Failed to free our space %d\n", ret);
+ test_msg("Failed to free our space %d\n", ret);
return ret;
}
if (check_exists(cache, bitmap_offset + 1 * 1024 * 1024,
5 * 1024 * 1024)) {
- printk(KERN_ERR "Left stuff over\n");
+ test_msg("Left stuff over\n");
return -1;
}
*/
ret = add_free_space_entry(cache, 1 * 1024 * 1024, 2 * 1024 * 1024, 1);
if (ret) {
- printk(KERN_ERR "Couldn't add bitmap entry %d\n", ret);
+ test_msg("Couldn't add bitmap entry %d\n", ret);
return ret;
}
ret = add_free_space_entry(cache, 3 * 1024 * 1024, 1 * 1024 * 1024, 0);
if (ret) {
- printk(KERN_ERR "Couldn't add extent entry %d\n", ret);
+ test_msg("Couldn't add extent entry %d\n", ret);
return ret;
}
ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 3 * 1024 * 1024);
if (ret) {
- printk(KERN_ERR "Error removing bitmap and extent "
- "overlapping %d\n", ret);
+ test_msg("Error removing bitmap and extent overlapping %d\n", ret);
return ret;
}
{
struct btrfs_block_group_cache *cache;
- printk(KERN_ERR "Running btrfs free space cache tests\n");
+ test_msg("Running btrfs free space cache tests\n");
cache = init_test_block_group();
if (!cache) {
- printk(KERN_ERR "Couldn't run the tests\n");
+ test_msg("Couldn't run the tests\n");
return;
}
__btrfs_remove_free_space_cache(cache->free_space_ctl);
kfree(cache->free_space_ctl);
kfree(cache);
- printk(KERN_ERR "Free space cache tests finished\n");
+ test_msg("Free space cache tests finished\n");
}
- #endif /* CONFIG_BTRFS_FS_RUN_SANITY_TESTS */
+ #undef test_msg
+ #else /* !CONFIG_BTRFS_FS_RUN_SANITY_TESTS */
+ void btrfs_test_free_space_cache(void) {}
+ #endif /* !CONFIG_BTRFS_FS_RUN_SANITY_TESTS */
#include <linux/writeback.h>
#include <linux/statfs.h>
#include <linux/compat.h>
+#include <linux/aio.h>
#include <linux/bit_spinlock.h>
#include <linux/xattr.h>
#include <linux/posix_acl.h>
#include <linux/mount.h>
#include <linux/btrfs.h>
#include <linux/blkdev.h>
+ #include <linux/posix_acl_xattr.h>
#include "compat.h"
#include "ctree.h"
#include "disk-io.h"
#include "free-space-cache.h"
#include "inode-map.h"
#include "backref.h"
+ #include "hash.h"
struct btrfs_iget_args {
u64 ino;
async_extent->nr_pages = 0;
async_extent->pages = NULL;
- if (ret == -ENOSPC)
+ if (ret == -ENOSPC) {
+ unlock_extent(io_tree, async_extent->start,
+ async_extent->start +
+ async_extent->ram_size - 1);
goto retry;
+ }
goto out_free;
}
spin_unlock(&BTRFS_I(inode)->lock);
}
+ static void btrfs_add_delalloc_inodes(struct btrfs_root *root,
+ struct inode *inode)
+ {
+ spin_lock(&root->delalloc_lock);
+ if (list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
+ list_add_tail(&BTRFS_I(inode)->delalloc_inodes,
+ &root->delalloc_inodes);
+ set_bit(BTRFS_INODE_IN_DELALLOC_LIST,
+ &BTRFS_I(inode)->runtime_flags);
+ root->nr_delalloc_inodes++;
+ if (root->nr_delalloc_inodes == 1) {
+ spin_lock(&root->fs_info->delalloc_root_lock);
+ BUG_ON(!list_empty(&root->delalloc_root));
+ list_add_tail(&root->delalloc_root,
+ &root->fs_info->delalloc_roots);
+ spin_unlock(&root->fs_info->delalloc_root_lock);
+ }
+ }
+ spin_unlock(&root->delalloc_lock);
+ }
+
+ static void btrfs_del_delalloc_inode(struct btrfs_root *root,
+ struct inode *inode)
+ {
+ spin_lock(&root->delalloc_lock);
+ if (!list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
+ list_del_init(&BTRFS_I(inode)->delalloc_inodes);
+ clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
+ &BTRFS_I(inode)->runtime_flags);
+ root->nr_delalloc_inodes--;
+ if (!root->nr_delalloc_inodes) {
+ spin_lock(&root->fs_info->delalloc_root_lock);
+ BUG_ON(list_empty(&root->delalloc_root));
+ list_del_init(&root->delalloc_root);
+ spin_unlock(&root->fs_info->delalloc_root_lock);
+ }
+ }
+ spin_unlock(&root->delalloc_lock);
+ }
+
/*
* extent_io.c set_bit_hook, used to track delayed allocation
* bytes in this file, and to maintain the list of inodes that
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->delalloc_bytes += len;
if (do_list && !test_bit(BTRFS_INODE_IN_DELALLOC_LIST,
- &BTRFS_I(inode)->runtime_flags)) {
- spin_lock(&root->fs_info->delalloc_lock);
- if (list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
- list_add_tail(&BTRFS_I(inode)->delalloc_inodes,
- &root->fs_info->delalloc_inodes);
- set_bit(BTRFS_INODE_IN_DELALLOC_LIST,
- &BTRFS_I(inode)->runtime_flags);
- }
- spin_unlock(&root->fs_info->delalloc_lock);
- }
+ &BTRFS_I(inode)->runtime_flags))
+ btrfs_add_delalloc_inodes(root, inode);
spin_unlock(&BTRFS_I(inode)->lock);
}
}
btrfs_delalloc_release_metadata(inode, len);
if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID
- && do_list)
+ && do_list && !(state->state & EXTENT_NORESERVE))
btrfs_free_reserved_data_space(inode, len);
__percpu_counter_add(&root->fs_info->delalloc_bytes, -len,
BTRFS_I(inode)->delalloc_bytes -= len;
if (do_list && BTRFS_I(inode)->delalloc_bytes == 0 &&
test_bit(BTRFS_INODE_IN_DELALLOC_LIST,
- &BTRFS_I(inode)->runtime_flags)) {
- spin_lock(&root->fs_info->delalloc_lock);
- if (!list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
- list_del_init(&BTRFS_I(inode)->delalloc_inodes);
- clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
- &BTRFS_I(inode)->runtime_flags);
- }
- spin_unlock(&root->fs_info->delalloc_lock);
- }
+ &BTRFS_I(inode)->runtime_flags))
+ btrfs_del_delalloc_inode(root, inode);
spin_unlock(&BTRFS_I(inode)->lock);
}
}
return 0;
return PTR_ERR(root);
}
- if (btrfs_root_refs(&root->root_item) == 0) {
- srcu_read_unlock(&fs_info->subvol_srcu, index);
- /* parse ENOENT to 0 */
- return 0;
- }
/* step 2: get inode */
key.objectid = backref->inum;
/* 1 for the orphan item deletion. */
trans = btrfs_start_transaction(root, 1);
if (IS_ERR(trans)) {
+ iput(inode);
ret = PTR_ERR(trans);
goto out;
}
ret = btrfs_orphan_add(trans, inode);
btrfs_end_transaction(trans, root);
- if (ret)
+ if (ret) {
+ iput(inode);
goto out;
+ }
ret = btrfs_truncate(inode);
if (ret)
{
u32 nritems = btrfs_header_nritems(leaf);
struct btrfs_key found_key;
+ static u64 xattr_access = 0;
+ static u64 xattr_default = 0;
int scanned = 0;
+ if (!xattr_access) {
+ xattr_access = btrfs_name_hash(POSIX_ACL_XATTR_ACCESS,
+ strlen(POSIX_ACL_XATTR_ACCESS));
+ xattr_default = btrfs_name_hash(POSIX_ACL_XATTR_DEFAULT,
+ strlen(POSIX_ACL_XATTR_DEFAULT));
+ }
+
slot++;
while (slot < nritems) {
btrfs_item_key_to_cpu(leaf, &found_key, slot);
return 0;
/* we found an xattr, assume we've got an acl */
- if (found_key.type == BTRFS_XATTR_ITEM_KEY)
- return 1;
+ if (found_key.type == BTRFS_XATTR_ITEM_KEY) {
+ if (found_key.offset == xattr_access ||
+ found_key.offset == xattr_default)
+ return 1;
+ }
/*
* we found a key greater than an xattr key, there can't
}
return ret;
}
-
-
- /* helper to check if there is any shared block in the path */
- static int check_path_shared(struct btrfs_root *root,
- struct btrfs_path *path)
- {
- struct extent_buffer *eb;
- int level;
- u64 refs = 1;
-
- for (level = 0; level < BTRFS_MAX_LEVEL; level++) {
- int ret;
-
- if (!path->nodes[level])
- break;
- eb = path->nodes[level];
- if (!btrfs_block_can_be_shared(root, eb))
- continue;
- ret = btrfs_lookup_extent_info(NULL, root, eb->start, level, 1,
- &refs, NULL);
- if (refs > 1)
- return 1;
- }
- return 0;
- }
/*
* helper to start transaction for unlink and rmdir.
*
- * unlink and rmdir are special in btrfs, they do not always free space.
- * so in enospc case, we should make sure they will free space before
- * allowing them to use the global metadata reservation.
+ * unlink and rmdir are special in btrfs, they do not always free space, so
+ * if we cannot make our reservations the normal way try and see if there is
+ * plenty of slack room in the global reserve to migrate, otherwise we cannot
+ * allow the unlink to occur.
*/
- static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir,
- struct dentry *dentry)
+ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir)
{
struct btrfs_trans_handle *trans;
struct btrfs_root *root = BTRFS_I(dir)->root;
- struct btrfs_path *path;
- struct btrfs_dir_item *di;
- struct inode *inode = dentry->d_inode;
- u64 index;
- int check_link = 1;
- int err = -ENOSPC;
int ret;
- u64 ino = btrfs_ino(inode);
- u64 dir_ino = btrfs_ino(dir);
/*
* 1 for the possible orphan item
if (!IS_ERR(trans) || PTR_ERR(trans) != -ENOSPC)
return trans;
- if (ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)
- return ERR_PTR(-ENOSPC);
-
- /* check if there is someone else holds reference */
- if (S_ISDIR(inode->i_mode) && atomic_read(&inode->i_count) > 1)
- return ERR_PTR(-ENOSPC);
-
- if (atomic_read(&inode->i_count) > 2)
- return ERR_PTR(-ENOSPC);
-
- if (xchg(&root->fs_info->enospc_unlink, 1))
- return ERR_PTR(-ENOSPC);
-
- path = btrfs_alloc_path();
- if (!path) {
- root->fs_info->enospc_unlink = 0;
- return ERR_PTR(-ENOMEM);
- }
+ if (PTR_ERR(trans) == -ENOSPC) {
+ u64 num_bytes = btrfs_calc_trans_metadata_size(root, 5);
- /* 1 for the orphan item */
- trans = btrfs_start_transaction(root, 1);
- if (IS_ERR(trans)) {
- btrfs_free_path(path);
- root->fs_info->enospc_unlink = 0;
- return trans;
- }
-
- path->skip_locking = 1;
- path->search_commit_root = 1;
-
- ret = btrfs_lookup_inode(trans, root, path,
- &BTRFS_I(dir)->location, 0);
- if (ret < 0) {
- err = ret;
- goto out;
- }
- if (ret == 0) {
- if (check_path_shared(root, path))
- goto out;
- } else {
- check_link = 0;
- }
- btrfs_release_path(path);
-
- ret = btrfs_lookup_inode(trans, root, path,
- &BTRFS_I(inode)->location, 0);
- if (ret < 0) {
- err = ret;
- goto out;
- }
- if (ret == 0) {
- if (check_path_shared(root, path))
- goto out;
- } else {
- check_link = 0;
- }
- btrfs_release_path(path);
-
- if (ret == 0 && S_ISREG(inode->i_mode)) {
- ret = btrfs_lookup_file_extent(trans, root, path,
- ino, (u64)-1, 0);
- if (ret < 0) {
- err = ret;
- goto out;
+ trans = btrfs_start_transaction(root, 0);
+ if (IS_ERR(trans))
+ return trans;
+ ret = btrfs_cond_migrate_bytes(root->fs_info,
+ &root->fs_info->trans_block_rsv,
+ num_bytes, 5);
+ if (ret) {
+ btrfs_end_transaction(trans, root);
+ return ERR_PTR(ret);
}
- BUG_ON(ret == 0); /* Corruption */
- if (check_path_shared(root, path))
- goto out;
- btrfs_release_path(path);
- }
-
- if (!check_link) {
- err = 0;
- goto out;
- }
-
- di = btrfs_lookup_dir_item(trans, root, path, dir_ino,
- dentry->d_name.name, dentry->d_name.len, 0);
- if (IS_ERR(di)) {
- err = PTR_ERR(di);
- goto out;
- }
- if (di) {
- if (check_path_shared(root, path))
- goto out;
- } else {
- err = 0;
- goto out;
- }
- btrfs_release_path(path);
-
- ret = btrfs_get_inode_ref_index(trans, root, path, dentry->d_name.name,
- dentry->d_name.len, ino, dir_ino, 0,
- &index);
- if (ret) {
- err = ret;
- goto out;
- }
-
- if (check_path_shared(root, path))
- goto out;
-
- btrfs_release_path(path);
-
- /*
- * This is a commit root search, if we can lookup inode item and other
- * relative items in the commit root, it means the transaction of
- * dir/file creation has been committed, and the dir index item that we
- * delay to insert has also been inserted into the commit root. So
- * we needn't worry about the delayed insertion of the dir index item
- * here.
- */
- di = btrfs_lookup_dir_index_item(trans, root, path, dir_ino, index,
- dentry->d_name.name, dentry->d_name.len, 0);
- if (IS_ERR(di)) {
- err = PTR_ERR(di);
- goto out;
- }
- BUG_ON(ret == -ENOENT);
- if (check_path_shared(root, path))
- goto out;
-
- err = 0;
- out:
- btrfs_free_path(path);
- /* Migrate the orphan reservation over */
- if (!err)
- err = btrfs_block_rsv_migrate(trans->block_rsv,
- &root->fs_info->global_block_rsv,
- trans->bytes_reserved);
-
- if (err) {
- btrfs_end_transaction(trans, root);
- root->fs_info->enospc_unlink = 0;
- return ERR_PTR(err);
- }
-
- trans->block_rsv = &root->fs_info->global_block_rsv;
- return trans;
- }
-
- static void __unlink_end_trans(struct btrfs_trans_handle *trans,
- struct btrfs_root *root)
- {
- if (trans->block_rsv->type == BTRFS_BLOCK_RSV_GLOBAL) {
- btrfs_block_rsv_release(root, trans->block_rsv,
- trans->bytes_reserved);
trans->block_rsv = &root->fs_info->trans_block_rsv;
- BUG_ON(!root->fs_info->enospc_unlink);
- root->fs_info->enospc_unlink = 0;
+ trans->bytes_reserved = num_bytes;
}
- btrfs_end_transaction(trans, root);
+ return trans;
}
static int btrfs_unlink(struct inode *dir, struct dentry *dentry)
struct inode *inode = dentry->d_inode;
int ret;
- trans = __unlink_start_trans(dir, dentry);
+ trans = __unlink_start_trans(dir);
if (IS_ERR(trans))
return PTR_ERR(trans);
}
out:
- __unlink_end_trans(trans, root);
+ btrfs_end_transaction(trans, root);
btrfs_btree_balance_dirty(root);
return ret;
}
if (btrfs_ino(inode) == BTRFS_FIRST_FREE_OBJECTID)
return -EPERM;
- trans = __unlink_start_trans(dir, dentry);
+ trans = __unlink_start_trans(dir);
if (IS_ERR(trans))
return PTR_ERR(trans);
if (!err)
btrfs_i_size_write(inode, 0);
out:
- __unlink_end_trans(trans, root);
+ btrfs_end_transaction(trans, root);
btrfs_btree_balance_dirty(root);
return err;
u64 hole_size;
int err = 0;
+ /*
+ * If our size started in the middle of a page we need to zero out the
+ * rest of the page before we expand the i_size, otherwise we could
+ * expose stale data.
+ */
+ err = btrfs_truncate_page(inode, oldsize, 0, 0);
+ if (err)
+ return err;
+
if (size <= hole_start)
return 0;
goto out;
}
- if (btrfs_root_refs(&new_root->root_item) == 0) {
- err = -ENOENT;
- goto out;
- }
-
*sub_root = new_root;
location->objectid = btrfs_root_dirid(&new_root->root_item);
location->type = BTRFS_INODE_ITEM_KEY;
if (!(inode->i_sb->s_flags & MS_RDONLY))
ret = btrfs_orphan_cleanup(sub_root);
up_read(&root->fs_info->cleanup_work_sem);
- if (ret)
+ if (ret) {
+ iput(inode);
inode = ERR_PTR(ret);
+ }
}
return inode;
DT_UNKNOWN, DT_REG, DT_DIR, DT_CHR, DT_BLK, DT_FIFO, DT_SOCK, DT_LNK
};
-static int btrfs_real_readdir(struct file *filp, void *dirent,
- filldir_t filldir)
+static int btrfs_real_readdir(struct file *file, struct dir_context *ctx)
{
- struct inode *inode = file_inode(filp);
+ struct inode *inode = file_inode(file);
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_item *item;
struct btrfs_dir_item *di;
char tmp_name[32];
char *name_ptr;
int name_len;
- int is_curr = 0; /* filp->f_pos points to the current index? */
+ int is_curr = 0; /* ctx->pos points to the current index? */
/* FIXME, use a real flag for deciding about the key type */
if (root->fs_info->tree_root == root)
key_type = BTRFS_DIR_ITEM_KEY;
- /* special case for "." */
- if (filp->f_pos == 0) {
- over = filldir(dirent, ".", 1,
- filp->f_pos, btrfs_ino(inode), DT_DIR);
- if (over)
- return 0;
- filp->f_pos = 1;
- }
- /* special case for .., just use the back ref */
- if (filp->f_pos == 1) {
- u64 pino = parent_ino(filp->f_path.dentry);
- over = filldir(dirent, "..", 2,
- filp->f_pos, pino, DT_DIR);
- if (over)
- return 0;
- filp->f_pos = 2;
- }
+ if (!dir_emit_dots(file, ctx))
+ return 0;
+
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
}
btrfs_set_key_type(&key, key_type);
- key.offset = filp->f_pos;
+ key.offset = ctx->pos;
key.objectid = btrfs_ino(inode);
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
break;
if (btrfs_key_type(&found_key) != key_type)
break;
- if (found_key.offset < filp->f_pos)
+ if (found_key.offset < ctx->pos)
goto next;
if (key_type == BTRFS_DIR_INDEX_KEY &&
btrfs_should_delete_dir_index(&del_list,
found_key.offset))
goto next;
- filp->f_pos = found_key.offset;
+ ctx->pos = found_key.offset;
is_curr = 1;
di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
over = 0;
goto skip;
}
- over = filldir(dirent, name_ptr, name_len,
- found_key.offset, location.objectid,
- d_type);
+ over = !dir_emit(ctx, name_ptr, name_len,
+ location.objectid, d_type);
skip:
if (name_ptr != tmp_name)
if (key_type == BTRFS_DIR_INDEX_KEY) {
if (is_curr)
- filp->f_pos++;
- ret = btrfs_readdir_delayed_dir_index(filp, dirent, filldir,
- &ins_list);
+ ctx->pos++;
+ ret = btrfs_readdir_delayed_dir_index(ctx, &ins_list);
if (ret)
goto nopos;
}
* 32-bit glibc will use getdents64, but then strtol -
* so the last number we can serve is this.
*/
- filp->f_pos = 0x7fffffff;
+ ctx->pos = 0x7fffffff;
else
- filp->f_pos++;
+ ctx->pos++;
nopos:
ret = 0;
err:
* returns 1 when the nocow is safe, < 1 on error, 0 if the
* block must be cow'd
*/
- static noinline int can_nocow_odirect(struct btrfs_trans_handle *trans,
- struct inode *inode, u64 offset, u64 *len,
- u64 *orig_start, u64 *orig_block_len,
- u64 *ram_bytes)
+ noinline int can_nocow_extent(struct btrfs_trans_handle *trans,
+ struct inode *inode, u64 offset, u64 *len,
+ u64 *orig_start, u64 *orig_block_len,
+ u64 *ram_bytes)
{
struct btrfs_path *path;
int ret;
u64 num_bytes;
int slot;
int found_type;
-
+ bool nocow = (BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW);
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
/* not a regular extent, must cow */
goto out;
}
+
+ if (!nocow && found_type == BTRFS_FILE_EXTENT_REG)
+ goto out;
+
disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
+ if (disk_bytenr == 0)
+ goto out;
+
+ if (btrfs_file_extent_compression(leaf, fi) ||
+ btrfs_file_extent_encryption(leaf, fi) ||
+ btrfs_file_extent_other_encoding(leaf, fi))
+ goto out;
+
backref_offset = btrfs_file_extent_offset(leaf, fi);
- *orig_start = key.offset - backref_offset;
- *orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, fi);
- *ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
+ if (orig_start) {
+ *orig_start = key.offset - backref_offset;
+ *orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, fi);
+ *ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
+ }
extent_end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
- if (extent_end < offset + *len) {
- /* extent doesn't include our full range, must cow */
- goto out;
- }
if (btrfs_extent_readonly(root, disk_bytenr))
goto out;
if (IS_ERR(trans))
goto must_cow;
- if (can_nocow_odirect(trans, inode, start, &len, &orig_start,
- &orig_block_len, &ram_bytes) == 1) {
+ if (can_nocow_extent(trans, inode, start, &len, &orig_start,
+ &orig_block_len, &ram_bytes) == 1) {
if (type == BTRFS_ORDERED_PREALLOC) {
free_extent_map(em);
em = create_pinned_em(inode, start, len,
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_dio_private *dip;
- struct bio_vec *bvec = dio_bio->bi_io_vec;
struct bio *io_bio;
int skip_sum;
int write = rw & REQ_WRITE;
}
dip->private = dio_bio->bi_private;
- io_bio->bi_private = dio_bio->bi_private;
dip->inode = inode;
dip->logical_offset = file_offset;
-
- dip->bytes = 0;
- do {
- dip->bytes += bvec->bv_len;
- bvec++;
- } while (bvec <= (dio_bio->bi_io_vec + dio_bio->bi_vcnt - 1));
-
+ dip->bytes = dio_bio->bi_size;
dip->disk_bytenr = (u64)dio_bio->bi_sector << 9;
io_bio->bi_private = dip;
dip->errors = 0;
atomic_inc(&inode->i_dio_count);
smp_mb__after_atomic_inc();
+ /*
+ * The generic stuff only does filemap_write_and_wait_range, which isn't
+ * enough if we've written compressed pages to this area, so we need to
+ * call btrfs_wait_ordered_range to make absolutely sure that any
+ * outstanding dirty pages are on disk.
+ */
+ count = iov_length(iov, nr_segs);
+ btrfs_wait_ordered_range(inode, offset, count);
+
if (rw & WRITE) {
- count = iov_length(iov, nr_segs);
/*
* If the write DIO is beyond the EOF, we need update
* the isize, but it is protected by i_mutex. So we can
return __btrfs_releasepage(page, gfp_flags & GFP_NOFS);
}
-static void btrfs_invalidatepage(struct page *page, unsigned long offset)
+static void btrfs_invalidatepage(struct page *page, unsigned int offset,
+ unsigned int length)
{
struct inode *inode = page->mapping->host;
struct extent_io_tree *tree;
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_block_rsv *rsv;
- int ret;
+ int ret = 0;
int err = 0;
struct btrfs_trans_handle *trans;
u64 mask = root->sectorsize - 1;
u64 min_size = btrfs_calc_trunc_metadata_size(root, 1);
- ret = btrfs_truncate_page(inode, inode->i_size, 0, 0);
- if (ret)
- return ret;
-
btrfs_wait_ordered_range(inode, inode->i_size & (~mask), (u64)-1);
btrfs_ordered_update_i_size(inode, inode->i_size, NULL);
*/
smp_mb();
if (!list_empty(&BTRFS_I(inode)->ordered_operations)) {
- spin_lock(&root->fs_info->ordered_extent_lock);
+ spin_lock(&root->fs_info->ordered_root_lock);
list_del_init(&BTRFS_I(inode)->ordered_operations);
- spin_unlock(&root->fs_info->ordered_extent_lock);
+ spin_unlock(&root->fs_info->ordered_root_lock);
}
if (test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
* some fairly slow code that needs optimization. This walks the list
* of all the inodes with pending delalloc and forces them to disk.
*/
- int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput)
+ static int __start_delalloc_inodes(struct btrfs_root *root, int delay_iput)
{
struct btrfs_inode *binode;
struct inode *inode;
struct list_head splice;
int ret = 0;
- if (root->fs_info->sb->s_flags & MS_RDONLY)
- return -EROFS;
-
INIT_LIST_HEAD(&works);
INIT_LIST_HEAD(&splice);
- spin_lock(&root->fs_info->delalloc_lock);
- list_splice_init(&root->fs_info->delalloc_inodes, &splice);
+ spin_lock(&root->delalloc_lock);
+ list_splice_init(&root->delalloc_inodes, &splice);
while (!list_empty(&splice)) {
binode = list_entry(splice.next, struct btrfs_inode,
delalloc_inodes);
- list_del_init(&binode->delalloc_inodes);
-
+ list_move_tail(&binode->delalloc_inodes,
+ &root->delalloc_inodes);
inode = igrab(&binode->vfs_inode);
if (!inode) {
- clear_bit(BTRFS_INODE_IN_DELALLOC_LIST,
- &binode->runtime_flags);
+ cond_resched_lock(&root->delalloc_lock);
continue;
}
-
- list_add_tail(&binode->delalloc_inodes,
- &root->fs_info->delalloc_inodes);
- spin_unlock(&root->fs_info->delalloc_lock);
+ spin_unlock(&root->delalloc_lock);
work = btrfs_alloc_delalloc_work(inode, 0, delay_iput);
if (unlikely(!work)) {
&work->work);
cond_resched();
- spin_lock(&root->fs_info->delalloc_lock);
+ spin_lock(&root->delalloc_lock);
}
- spin_unlock(&root->fs_info->delalloc_lock);
+ spin_unlock(&root->delalloc_lock);
list_for_each_entry_safe(work, next, &works, list) {
list_del_init(&work->list);
btrfs_wait_and_free_delalloc_work(work);
}
+ return 0;
+ out:
+ list_for_each_entry_safe(work, next, &works, list) {
+ list_del_init(&work->list);
+ btrfs_wait_and_free_delalloc_work(work);
+ }
+
+ if (!list_empty_careful(&splice)) {
+ spin_lock(&root->delalloc_lock);
+ list_splice_tail(&splice, &root->delalloc_inodes);
+ spin_unlock(&root->delalloc_lock);
+ }
+ return ret;
+ }
+
+ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput)
+ {
+ int ret;
- /* the filemap_flush will queue IO into the worker threads, but
+ if (root->fs_info->sb->s_flags & MS_RDONLY)
+ return -EROFS;
+
+ ret = __start_delalloc_inodes(root, delay_iput);
+ /*
+ * the filemap_flush will queue IO into the worker threads, but
* we have to make sure the IO is actually started and that
* ordered extents get created before we return
*/
atomic_read(&root->fs_info->async_delalloc_pages) == 0));
}
atomic_dec(&root->fs_info->async_submit_draining);
- return 0;
- out:
- list_for_each_entry_safe(work, next, &works, list) {
- list_del_init(&work->list);
- btrfs_wait_and_free_delalloc_work(work);
+ return ret;
+ }
+
+ int btrfs_start_all_delalloc_inodes(struct btrfs_fs_info *fs_info,
+ int delay_iput)
+ {
+ struct btrfs_root *root;
+ struct list_head splice;
+ int ret;
+
+ if (fs_info->sb->s_flags & MS_RDONLY)
+ return -EROFS;
+
+ INIT_LIST_HEAD(&splice);
+
+ spin_lock(&fs_info->delalloc_root_lock);
+ list_splice_init(&fs_info->delalloc_roots, &splice);
+ while (!list_empty(&splice)) {
+ root = list_first_entry(&splice, struct btrfs_root,
+ delalloc_root);
+ root = btrfs_grab_fs_root(root);
+ BUG_ON(!root);
+ list_move_tail(&root->delalloc_root,
+ &fs_info->delalloc_roots);
+ spin_unlock(&fs_info->delalloc_root_lock);
+
+ ret = __start_delalloc_inodes(root, delay_iput);
+ btrfs_put_fs_root(root);
+ if (ret)
+ goto out;
+
+ spin_lock(&fs_info->delalloc_root_lock);
}
+ spin_unlock(&fs_info->delalloc_root_lock);
+ atomic_inc(&fs_info->async_submit_draining);
+ while (atomic_read(&fs_info->nr_async_submits) ||
+ atomic_read(&fs_info->async_delalloc_pages)) {
+ wait_event(fs_info->async_submit_wait,
+ (atomic_read(&fs_info->nr_async_submits) == 0 &&
+ atomic_read(&fs_info->async_delalloc_pages) == 0));
+ }
+ atomic_dec(&fs_info->async_submit_draining);
+ return 0;
+ out:
if (!list_empty_careful(&splice)) {
- spin_lock(&root->fs_info->delalloc_lock);
- list_splice_tail(&splice, &root->fs_info->delalloc_inodes);
- spin_unlock(&root->fs_info->delalloc_lock);
+ spin_lock(&fs_info->delalloc_root_lock);
+ list_splice_tail(&splice, &fs_info->delalloc_roots);
+ spin_unlock(&fs_info->delalloc_root_lock);
}
return ret;
}
static const struct file_operations btrfs_dir_file_operations = {
.llseek = generic_file_llseek,
.read = generic_read_dir,
- .readdir = btrfs_real_readdir,
+ .iterate = btrfs_real_readdir,
.unlocked_ioctl = btrfs_ioctl,
#ifdef CONFIG_COMPAT
.compat_ioctl = btrfs_ioctl,
if (!root->ref_cows)
return -EINVAL;
+ ret = btrfs_start_delalloc_inodes(root, 0);
+ if (ret)
+ return ret;
+
+ btrfs_wait_ordered_extents(root, 0);
+
pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
if (!pending_snapshot)
return -ENOMEM;
if (ret)
return ret;
- if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
- 1)) {
- pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
- mnt_drop_write_file(file);
- return -EINVAL;
- }
-
- mutex_lock(&root->fs_info->volume_mutex);
vol_args = memdup_user(arg, sizeof(*vol_args));
if (IS_ERR(vol_args)) {
ret = PTR_ERR(vol_args);
}
vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
- ret = btrfs_rm_device(root, vol_args->name);
- kfree(vol_args);
- out:
+ if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
+ 1)) {
+ ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
+ goto out;
+ }
+
+ mutex_lock(&root->fs_info->volume_mutex);
+ ret = btrfs_rm_device(root, vol_args->name);
mutex_unlock(&root->fs_info->volume_mutex);
atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
+
+ out:
+ kfree(vol_args);
mnt_drop_write_file(file);
return ret;
}
int ret;
u64 len = olen;
u64 bs = root->fs_info->sb->s_blocksize;
+ int same_inode = 0;
/*
* TODO:
ret = -EINVAL;
if (src == inode)
- goto out_fput;
+ same_inode = 1;
/* the src must be open for reading */
if (!(src_file.file->f_mode & FMODE_READ))
}
path->reada = 2;
- if (inode < src) {
- mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
- mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
+ if (!same_inode) {
+ if (inode < src) {
+ mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
+ mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
+ } else {
+ mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
+ mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
+ }
} else {
- mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
- mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
+ mutex_lock(&src->i_mutex);
}
/* determine range to clone */
!IS_ALIGNED(destoff, bs))
goto out_unlock;
+ /* verify if ranges are overlapped within the same file */
+ if (same_inode) {
+ if (destoff + len > off && destoff < off + len)
+ goto out_unlock;
+ }
+
if (destoff > inode->i_size) {
ret = btrfs_cont_expand(inode, inode->i_size, destoff);
if (ret)
unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
out_unlock:
mutex_unlock(&src->i_mutex);
- mutex_unlock(&inode->i_mutex);
+ if (!same_inode)
+ mutex_unlock(&inode->i_mutex);
vfree(buf);
btrfs_free_path(path);
out_fput:
goto out;
}
- if (btrfs_root_refs(&new_root->root_item) == 0) {
- ret = -ENOENT;
- goto out;
- }
-
path = btrfs_alloc_path();
if (!path) {
ret = -ENOMEM;
break;
}
- if (copy_to_user(arg, sa, sizeof(*sa)))
- ret = -EFAULT;
-
err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
if (err && !ret)
ret = err;
static long btrfs_ioctl_quota_rescan(struct file *file, void __user *arg)
{
- struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+ struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
struct btrfs_ioctl_quota_rescan_args *qsa;
int ret;
static long btrfs_ioctl_quota_rescan_status(struct file *file, void __user *arg)
{
- struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+ struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
struct btrfs_ioctl_quota_rescan_args *qsa;
int ret = 0;
return ret;
}
+ static long btrfs_ioctl_quota_rescan_wait(struct file *file, void __user *arg)
+ {
+ struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ return btrfs_qgroup_wait_for_completion(root->fs_info);
+ }
+
static long btrfs_ioctl_set_received_subvol(struct file *file,
void __user *arg)
{
static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg)
{
- struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+ struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
const char *label = root->fs_info->super_copy->label;
size_t len = strnlen(label, BTRFS_LABEL_SIZE);
int ret;
static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg)
{
- struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+ struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
struct btrfs_super_block *super_block = root->fs_info->super_copy;
struct btrfs_trans_handle *trans;
char label[BTRFS_LABEL_SIZE];
return btrfs_ioctl_quota_rescan(file, argp);
case BTRFS_IOC_QUOTA_RESCAN_STATUS:
return btrfs_ioctl_quota_rescan_status(file, argp);
+ case BTRFS_IOC_QUOTA_RESCAN_WAIT:
+ return btrfs_ioctl_quota_rescan_wait(file, argp);
case BTRFS_IOC_DEV_REPLACE:
return btrfs_ioctl_dev_replace(root, argp);
case BTRFS_IOC_GET_FSLABEL:
return ret;
}
+ static int contains_pending_extent(struct btrfs_trans_handle *trans,
+ struct btrfs_device *device,
+ u64 *start, u64 len)
+ {
+ struct extent_map *em;
+ int ret = 0;
+
+ list_for_each_entry(em, &trans->transaction->pending_chunks, list) {
+ struct map_lookup *map;
+ int i;
+
+ map = (struct map_lookup *)em->bdev;
+ for (i = 0; i < map->num_stripes; i++) {
+ if (map->stripes[i].dev != device)
+ continue;
+ if (map->stripes[i].physical >= *start + len ||
+ map->stripes[i].physical + em->orig_block_len <=
+ *start)
+ continue;
+ *start = map->stripes[i].physical +
+ em->orig_block_len;
+ ret = 1;
+ }
+ }
+
+ return ret;
+ }
+
+
/*
* find_free_dev_extent - find free space in the specified device
* @device: the device which we search the free space in
* But if we don't find suitable free space, it is used to store the size of
* the max free space.
*/
- int find_free_dev_extent(struct btrfs_device *device, u64 num_bytes,
+ int find_free_dev_extent(struct btrfs_trans_handle *trans,
+ struct btrfs_device *device, u64 num_bytes,
u64 *start, u64 *len)
{
struct btrfs_key key;
*/
search_start = max(root->fs_info->alloc_start, 1024ull * 1024);
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+ again:
max_hole_start = search_start;
max_hole_size = 0;
hole_size = 0;
if (search_start >= search_end || device->is_tgtdev_for_dev_replace) {
ret = -ENOSPC;
- goto error;
+ goto out;
}
- path = btrfs_alloc_path();
- if (!path) {
- ret = -ENOMEM;
- goto error;
- }
path->reada = 2;
+ path->search_commit_root = 1;
+ path->skip_locking = 1;
key.objectid = device->devid;
key.offset = search_start;
if (key.offset > search_start) {
hole_size = key.offset - search_start;
+ /*
+ * Have to check before we set max_hole_start, otherwise
+ * we could end up sending back this offset anyway.
+ */
+ if (contains_pending_extent(trans, device,
+ &search_start,
+ hole_size))
+ hole_size = 0;
+
if (hole_size > max_hole_size) {
max_hole_start = search_start;
max_hole_size = hole_size;
max_hole_size = hole_size;
}
+ if (contains_pending_extent(trans, device, &search_start, hole_size)) {
+ btrfs_release_path(path);
+ goto again;
+ }
+
/* See above. */
if (hole_size < num_bytes)
ret = -ENOSPC;
out:
btrfs_free_path(path);
- error:
*start = max_hole_start;
if (len)
*len = max_hole_size;
return ret;
}
- static noinline int find_next_chunk(struct btrfs_root *root,
- u64 objectid, u64 *offset)
+ static u64 find_next_chunk(struct btrfs_fs_info *fs_info)
{
- struct btrfs_path *path;
- int ret;
- struct btrfs_key key;
- struct btrfs_chunk *chunk;
- struct btrfs_key found_key;
-
- path = btrfs_alloc_path();
- if (!path)
- return -ENOMEM;
-
- key.objectid = objectid;
- key.offset = (u64)-1;
- key.type = BTRFS_CHUNK_ITEM_KEY;
-
- ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
- if (ret < 0)
- goto error;
-
- BUG_ON(ret == 0); /* Corruption */
+ struct extent_map_tree *em_tree;
+ struct extent_map *em;
+ struct rb_node *n;
+ u64 ret = 0;
- ret = btrfs_previous_item(root, path, 0, BTRFS_CHUNK_ITEM_KEY);
- if (ret) {
- *offset = 0;
- } else {
- btrfs_item_key_to_cpu(path->nodes[0], &found_key,
- path->slots[0]);
- if (found_key.objectid != objectid)
- *offset = 0;
- else {
- chunk = btrfs_item_ptr(path->nodes[0], path->slots[0],
- struct btrfs_chunk);
- *offset = found_key.offset +
- btrfs_chunk_length(path->nodes[0], chunk);
- }
+ em_tree = &fs_info->mapping_tree.map_tree;
+ read_lock(&em_tree->lock);
+ n = rb_last(&em_tree->map);
+ if (n) {
+ em = rb_entry(n, struct extent_map, rb_node);
+ ret = em->start + em->len;
}
- ret = 0;
- error:
- btrfs_free_path(path);
+ read_unlock(&em_tree->lock);
+
return ret;
}
btrfs_dev_replace_unlock(&root->fs_info->dev_replace);
if ((all_avail & BTRFS_BLOCK_GROUP_RAID10) && num_devices <= 4) {
- printk(KERN_ERR "btrfs: unable to go below four devices "
- "on raid10\n");
- ret = -EINVAL;
+ ret = BTRFS_ERROR_DEV_RAID10_MIN_NOT_MET;
goto out;
}
if ((all_avail & BTRFS_BLOCK_GROUP_RAID1) && num_devices <= 2) {
- printk(KERN_ERR "btrfs: unable to go below two "
- "devices on raid1\n");
- ret = -EINVAL;
+ ret = BTRFS_ERROR_DEV_RAID1_MIN_NOT_MET;
goto out;
}
if ((all_avail & BTRFS_BLOCK_GROUP_RAID5) &&
root->fs_info->fs_devices->rw_devices <= 2) {
- printk(KERN_ERR "btrfs: unable to go below two "
- "devices on raid5\n");
- ret = -EINVAL;
+ ret = BTRFS_ERROR_DEV_RAID5_MIN_NOT_MET;
goto out;
}
if ((all_avail & BTRFS_BLOCK_GROUP_RAID6) &&
root->fs_info->fs_devices->rw_devices <= 3) {
- printk(KERN_ERR "btrfs: unable to go below three "
- "devices on raid6\n");
- ret = -EINVAL;
+ ret = BTRFS_ERROR_DEV_RAID6_MIN_NOT_MET;
goto out;
}
bh = NULL;
disk_super = NULL;
if (!device) {
- printk(KERN_ERR "btrfs: no missing devices found to "
- "remove\n");
+ ret = BTRFS_ERROR_DEV_MISSING_NOT_FOUND;
goto out;
}
} else {
}
if (device->is_tgtdev_for_dev_replace) {
- pr_err("btrfs: unable to remove the dev_replace target dev\n");
- ret = -EINVAL;
+ ret = BTRFS_ERROR_DEV_TGT_REPLACE;
goto error_brelse;
}
if (device->writeable && root->fs_info->fs_devices->rw_devices == 1) {
- printk(KERN_ERR "btrfs: unable to remove the only writeable "
- "device\n");
- ret = -EINVAL;
+ ret = BTRFS_ERROR_DEV_ONLY_WRITABLE;
goto error_brelse;
}
}
tsk = kthread_run(balance_kthread, fs_info, "btrfs-balance");
- if (IS_ERR(tsk))
- return PTR_ERR(tsk);
-
- return 0;
+ return PTR_RET(tsk);
}
int btrfs_recover_balance(struct btrfs_fs_info *fs_info)
}
static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
- struct btrfs_root *extent_root,
- struct map_lookup **map_ret,
- u64 *num_bytes_out, u64 *stripe_size_out,
- u64 start, u64 type)
+ struct btrfs_root *extent_root, u64 start,
+ u64 type)
{
struct btrfs_fs_info *info = extent_root->fs_info;
struct btrfs_fs_devices *fs_devices = info->fs_devices;
if (total_avail == 0)
continue;
- ret = find_free_dev_extent(device,
+ ret = find_free_dev_extent(trans, device,
max_stripe_size * dev_stripes,
&dev_offset, &max_avail);
if (ret && ret != -ENOSPC)
map->type = type;
map->sub_stripes = sub_stripes;
- *map_ret = map;
num_bytes = stripe_size * data_stripes;
- *stripe_size_out = stripe_size;
- *num_bytes_out = num_bytes;
-
trace_btrfs_chunk_alloc(info->chunk_root, map, start, num_bytes);
em = alloc_extent_map();
em->len = num_bytes;
em->block_start = 0;
em->block_len = em->len;
+ em->orig_block_len = stripe_size;
em_tree = &extent_root->fs_info->mapping_tree.map_tree;
write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em, 0);
+ if (!ret) {
+ list_add_tail(&em->list, &trans->transaction->pending_chunks);
+ atomic_inc(&em->refs);
+ }
write_unlock(&em_tree->lock);
if (ret) {
free_extent_map(em);
goto error;
}
- for (i = 0; i < map->num_stripes; ++i) {
- struct btrfs_device *device;
- u64 dev_offset;
-
- device = map->stripes[i].dev;
- dev_offset = map->stripes[i].physical;
-
- ret = btrfs_alloc_dev_extent(trans, device,
- info->chunk_root->root_key.objectid,
- BTRFS_FIRST_CHUNK_TREE_OBJECTID,
- start, dev_offset, stripe_size);
- if (ret)
- goto error_dev_extent;
- }
-
ret = btrfs_make_block_group(trans, extent_root, 0, type,
BTRFS_FIRST_CHUNK_TREE_OBJECTID,
start, num_bytes);
- if (ret) {
- i = map->num_stripes - 1;
- goto error_dev_extent;
- }
+ if (ret)
+ goto error_del_extent;
free_extent_map(em);
check_raid56_incompat_flag(extent_root->fs_info, type);
kfree(devices_info);
return 0;
- error_dev_extent:
- for (; i >= 0; i--) {
- struct btrfs_device *device;
- int err;
-
- device = map->stripes[i].dev;
- err = btrfs_free_dev_extent(trans, device, start);
- if (err) {
- btrfs_abort_transaction(trans, extent_root, err);
- break;
- }
- }
+ error_del_extent:
write_lock(&em_tree->lock);
remove_extent_mapping(em_tree, em);
write_unlock(&em_tree->lock);
return ret;
}
- static int __finish_chunk_alloc(struct btrfs_trans_handle *trans,
+ int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans,
struct btrfs_root *extent_root,
- struct map_lookup *map, u64 chunk_offset,
- u64 chunk_size, u64 stripe_size)
+ u64 chunk_offset, u64 chunk_size)
{
- u64 dev_offset;
struct btrfs_key key;
struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
struct btrfs_device *device;
struct btrfs_chunk *chunk;
struct btrfs_stripe *stripe;
- size_t item_size = btrfs_chunk_item_size(map->num_stripes);
- int index = 0;
+ struct extent_map_tree *em_tree;
+ struct extent_map *em;
+ struct map_lookup *map;
+ size_t item_size;
+ u64 dev_offset;
+ u64 stripe_size;
+ int i = 0;
int ret;
+ em_tree = &extent_root->fs_info->mapping_tree.map_tree;
+ read_lock(&em_tree->lock);
+ em = lookup_extent_mapping(em_tree, chunk_offset, chunk_size);
+ read_unlock(&em_tree->lock);
+
+ if (!em) {
+ btrfs_crit(extent_root->fs_info, "unable to find logical "
+ "%Lu len %Lu", chunk_offset, chunk_size);
+ return -EINVAL;
+ }
+
+ if (em->start != chunk_offset || em->len != chunk_size) {
+ btrfs_crit(extent_root->fs_info, "found a bad mapping, wanted"
+ " %Lu-%Lu, found %Lu-%Lu\n", chunk_offset,
+ chunk_size, em->start, em->len);
+ free_extent_map(em);
+ return -EINVAL;
+ }
+
+ map = (struct map_lookup *)em->bdev;
+ item_size = btrfs_chunk_item_size(map->num_stripes);
+ stripe_size = em->orig_block_len;
+
chunk = kzalloc(item_size, GFP_NOFS);
- if (!chunk)
- return -ENOMEM;
+ if (!chunk) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ for (i = 0; i < map->num_stripes; i++) {
+ device = map->stripes[i].dev;
+ dev_offset = map->stripes[i].physical;
- index = 0;
- while (index < map->num_stripes) {
- device = map->stripes[index].dev;
device->bytes_used += stripe_size;
ret = btrfs_update_device(trans, device);
if (ret)
- goto out_free;
- index++;
+ goto out;
+ ret = btrfs_alloc_dev_extent(trans, device,
+ chunk_root->root_key.objectid,
+ BTRFS_FIRST_CHUNK_TREE_OBJECTID,
+ chunk_offset, dev_offset,
+ stripe_size);
+ if (ret)
+ goto out;
}
spin_lock(&extent_root->fs_info->free_chunk_lock);
map->num_stripes);
spin_unlock(&extent_root->fs_info->free_chunk_lock);
- index = 0;
stripe = &chunk->stripe;
- while (index < map->num_stripes) {
- device = map->stripes[index].dev;
- dev_offset = map->stripes[index].physical;
+ for (i = 0; i < map->num_stripes; i++) {
+ device = map->stripes[i].dev;
+ dev_offset = map->stripes[i].physical;
btrfs_set_stack_stripe_devid(stripe, device->devid);
btrfs_set_stack_stripe_offset(stripe, dev_offset);
memcpy(stripe->dev_uuid, device->uuid, BTRFS_UUID_SIZE);
stripe++;
- index++;
}
btrfs_set_stack_chunk_length(chunk, chunk_size);
key.offset = chunk_offset;
ret = btrfs_insert_item(trans, chunk_root, &key, chunk, item_size);
-
if (ret == 0 && map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
/*
* TODO: Cleanup of inserted chunk root in case of
item_size);
}
- out_free:
+ out:
kfree(chunk);
+ free_extent_map(em);
return ret;
}
struct btrfs_root *extent_root, u64 type)
{
u64 chunk_offset;
- u64 chunk_size;
- u64 stripe_size;
- struct map_lookup *map;
- struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
- int ret;
-
- ret = find_next_chunk(chunk_root, BTRFS_FIRST_CHUNK_TREE_OBJECTID,
- &chunk_offset);
- if (ret)
- return ret;
- ret = __btrfs_alloc_chunk(trans, extent_root, &map, &chunk_size,
- &stripe_size, chunk_offset, type);
- if (ret)
- return ret;
-
- ret = __finish_chunk_alloc(trans, extent_root, map, chunk_offset,
- chunk_size, stripe_size);
- if (ret)
- return ret;
- return 0;
+ chunk_offset = find_next_chunk(extent_root->fs_info);
+ return __btrfs_alloc_chunk(trans, extent_root, chunk_offset, type);
}
static noinline int init_first_rw_device(struct btrfs_trans_handle *trans,
{
u64 chunk_offset;
u64 sys_chunk_offset;
- u64 chunk_size;
- u64 sys_chunk_size;
- u64 stripe_size;
- u64 sys_stripe_size;
u64 alloc_profile;
- struct map_lookup *map;
- struct map_lookup *sys_map;
struct btrfs_fs_info *fs_info = root->fs_info;
struct btrfs_root *extent_root = fs_info->extent_root;
int ret;
- ret = find_next_chunk(fs_info->chunk_root,
- BTRFS_FIRST_CHUNK_TREE_OBJECTID, &chunk_offset);
- if (ret)
- return ret;
-
+ chunk_offset = find_next_chunk(fs_info);
alloc_profile = btrfs_get_alloc_profile(extent_root, 0);
- ret = __btrfs_alloc_chunk(trans, extent_root, &map, &chunk_size,
- &stripe_size, chunk_offset, alloc_profile);
+ ret = __btrfs_alloc_chunk(trans, extent_root, chunk_offset,
+ alloc_profile);
if (ret)
return ret;
- sys_chunk_offset = chunk_offset + chunk_size;
-
+ sys_chunk_offset = find_next_chunk(root->fs_info);
alloc_profile = btrfs_get_alloc_profile(fs_info->chunk_root, 0);
- ret = __btrfs_alloc_chunk(trans, extent_root, &sys_map,
- &sys_chunk_size, &sys_stripe_size,
- sys_chunk_offset, alloc_profile);
+ ret = __btrfs_alloc_chunk(trans, extent_root, sys_chunk_offset,
+ alloc_profile);
if (ret) {
btrfs_abort_transaction(trans, root, ret);
goto out;
}
ret = btrfs_add_device(trans, fs_info->chunk_root, device);
- if (ret) {
- btrfs_abort_transaction(trans, root, ret);
- goto out;
- }
-
- /*
- * Modifying chunk tree needs allocating new blocks from both
- * system block group and metadata block group. So we only can
- * do operations require modifying the chunk tree after both
- * block groups were created.
- */
- ret = __finish_chunk_alloc(trans, extent_root, map, chunk_offset,
- chunk_size, stripe_size);
- if (ret) {
- btrfs_abort_transaction(trans, root, ret);
- goto out;
- }
-
- ret = __finish_chunk_alloc(trans, extent_root, sys_map,
- sys_chunk_offset, sys_chunk_size,
- sys_stripe_size);
if (ret)
btrfs_abort_transaction(trans, root, ret);
-
out:
-
return ret;
}
map = (struct map_lookup *)em->bdev;
offset = logical - em->start;
- if (mirror_num > map->num_stripes)
- mirror_num = 0;
-
stripe_len = map->stripe_len;
stripe_nr = offset;
/*
}
prev = &bio->bi_io_vec[bio->bi_vcnt - 1];
- if ((bio->bi_size >> 9) > max_sectors)
+ if (bio_sectors(bio) > max_sectors)
return 0;
if (!q->merge_bvec_fn)
return NULL;
list_add(&device->dev_list,
&fs_devices->devices);
- device->dev_root = root->fs_info->dev_root;
device->devid = devid;
device->work.func = pending_bios_fn;
device->fs_devices = fs_devices;
}
fill_device_from_item(leaf, dev_item, device);
- device->dev_root = root->fs_info->dev_root;
device->in_fs_metadata = 1;
if (device->writeable && !device->is_tgtdev_for_dev_replace) {
device->fs_devices->total_rw_bytes += device->total_bytes;
return ret;
}
+ void btrfs_init_devices_late(struct btrfs_fs_info *fs_info)
+ {
+ struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
+ struct btrfs_device *device;
+
+ mutex_lock(&fs_devices->device_list_mutex);
+ list_for_each_entry(device, &fs_devices->devices, dev_list)
+ device->dev_root = fs_info->dev_root;
+ mutex_unlock(&fs_devices->device_list_mutex);
+ }
+
static void __btrfs_reset_dev_stats(struct btrfs_device *dev)
{
int i;