Merge branch 'cleanups_and_fixes' into inode_numbers

Conflicts:
	fs/btrfs/tree-log.c
	fs/btrfs/volumes.c

Signed-off-by: Chris Mason <chris.mason@oracle.com>

fs/btrfs/ctree.c

@@ -3211,7 +3211,6 @@ int btrfs_truncate_item(struct btrfs_trans_handle *trans,
 			struct btrfs_path *path,
 			u32 new_size, int from_end)
 {
-	int ret = 0;
 	int slot;
 	struct extent_buffer *leaf;
 	struct btrfs_item *item;
@@ -3309,12 +3308,11 @@ int btrfs_truncate_item(struct btrfs_trans_handle *trans,
 	btrfs_set_item_size(leaf, item, new_size);
 	btrfs_mark_buffer_dirty(leaf);
 
-	ret = 0;
 	if (btrfs_leaf_free_space(root, leaf) < 0) {
 		btrfs_print_leaf(root, leaf);
 		BUG();
 	}
-	return ret;
+	return 0;
 }
 
 /*
@@ -3324,7 +3322,6 @@ int btrfs_extend_item(struct btrfs_trans_handle *trans,
 		      struct btrfs_root *root, struct btrfs_path *path,
 		      u32 data_size)
 {
-	int ret = 0;
 	int slot;
 	struct extent_buffer *leaf;
 	struct btrfs_item *item;
@@ -3389,12 +3386,11 @@ int btrfs_extend_item(struct btrfs_trans_handle *trans,
 	btrfs_set_item_size(leaf, item, old_size + data_size);
 	btrfs_mark_buffer_dirty(leaf);
 
-	ret = 0;
 	if (btrfs_leaf_free_space(root, leaf) < 0) {
 		btrfs_print_leaf(root, leaf);
 		BUG();
 	}
-	return ret;
+	return 0;
 }
 
 /*

fs/btrfs/ctree.h

@@ -1113,9 +1113,6 @@ struct btrfs_fs_info {
 struct btrfs_root {
 	struct extent_buffer *node;
 
-	/* the node lock is held while changing the node pointer */
-	spinlock_t node_lock;
-
 	struct extent_buffer *commit_root;
 	struct btrfs_root *log_root;
 	struct btrfs_root *reloc_root;

fs/btrfs/dir-item.c

@@ -50,7 +50,6 @@ static struct btrfs_dir_item *insert_with_overflow(struct btrfs_trans_handle
 		if (di)
 			return ERR_PTR(-EEXIST);
 		ret = btrfs_extend_item(trans, root, path, data_size);
-		WARN_ON(ret > 0);
 	}
 	if (ret < 0)
 		return ERR_PTR(ret);
@@ -438,7 +437,7 @@ int verify_dir_item(struct btrfs_root *root,
 		namelen = XATTR_NAME_MAX;
 
 	if (btrfs_dir_name_len(leaf, dir_item) > namelen) {
-		printk(KERN_CRIT "btrfS: invalid dir item name len: %u\n",
+		printk(KERN_CRIT "btrfs: invalid dir item name len: %u\n",
 		       (unsigned)btrfs_dir_data_len(leaf, dir_item));
 		return 1;
 	}

fs/btrfs/disk-io.c

@@ -1053,7 +1053,6 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	INIT_LIST_HEAD(&root->dirty_list);
 	INIT_LIST_HEAD(&root->orphan_list);
 	INIT_LIST_HEAD(&root->root_list);
-	spin_lock_init(&root->node_lock);
 	spin_lock_init(&root->orphan_lock);
 	spin_lock_init(&root->inode_lock);
 	spin_lock_init(&root->accounting_lock);
@@ -1363,7 +1362,8 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
 	struct btrfs_device *device;
 	struct backing_dev_info *bdi;
 
-	list_for_each_entry(device, &info->fs_devices->devices, dev_list) {
+	rcu_read_lock();
+	list_for_each_entry_rcu(device, &info->fs_devices->devices, dev_list) {
 		if (!device->bdev)
 			continue;
 		bdi = blk_get_backing_dev_info(device->bdev);
@@ -1372,6 +1372,7 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
 			break;
 		}
 	}
+	rcu_read_unlock();
 	return ret;
 }
 
@@ -2305,9 +2306,9 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
 	sb = &root->fs_info->super_for_commit;
 	dev_item = &sb->dev_item;
 
-	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+	rcu_read_lock();
 	head = &root->fs_info->fs_devices->devices;
-	list_for_each_entry(dev, head, dev_list) {
+	list_for_each_entry_rcu(dev, head, dev_list) {
 		if (!dev->bdev) {
 			total_errors++;
 			continue;
@@ -2340,7 +2341,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
 	}
 
 	total_errors = 0;
-	list_for_each_entry(dev, head, dev_list) {
+	list_for_each_entry_rcu(dev, head, dev_list) {
 		if (!dev->bdev)
 			continue;
 		if (!dev->in_fs_metadata || !dev->writeable)
@@ -2350,7 +2351,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
 		if (ret)
 			total_errors++;
 	}
-	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+	rcu_read_unlock();
 	if (total_errors > max_errors) {
 		printk(KERN_ERR "btrfs: %d errors while writing supers\n",
 		       total_errors);

fs/btrfs/extent-tree.c

@@ -952,7 +952,6 @@ static int convert_extent_item_v0(struct btrfs_trans_handle *trans,
 	BUG_ON(ret);
 
 	ret = btrfs_extend_item(trans, root, path, new_size);
-	BUG_ON(ret);
 
 	leaf = path->nodes[0];
 	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
@@ -1560,7 +1559,6 @@ int setup_inline_extent_backref(struct btrfs_trans_handle *trans,
 	size = btrfs_extent_inline_ref_size(type);
 
 	ret = btrfs_extend_item(trans, root, path, size);
-	BUG_ON(ret);
 
 	ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
 	refs = btrfs_extent_refs(leaf, ei);
@@ -1689,7 +1687,6 @@ int update_inline_extent_backref(struct btrfs_trans_handle *trans,
 					      end - ptr - size);
 		item_size -= size;
 		ret = btrfs_truncate_item(trans, root, path, item_size, 1);
-		BUG_ON(ret);
 	}
 	btrfs_mark_buffer_dirty(leaf);
 	return 0;
@@ -3307,6 +3304,10 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans,
 	if (reserved == 0)
 		return 0;
 
+	/* nothing to shrink - nothing to reclaim */
+	if (root->fs_info->delalloc_bytes == 0)
+		return 0;
+
 	max_reclaim = min(reserved, to_reclaim);
 
 	while (loops < 1024) {

fs/btrfs/extent_io.c

@@ -439,6 +439,15 @@ static int clear_state_bit(struct extent_io_tree *tree,
 	return ret;
 }
 
+static struct extent_state *
+alloc_extent_state_atomic(struct extent_state *prealloc)
+{
+	if (!prealloc)
+		prealloc = alloc_extent_state(GFP_ATOMIC);
+
+	return prealloc;
+}
+
 /*
  * clear some bits on a range in the tree.  This may require splitting
  * or inserting elements in the tree, so the gfp mask is used to
@@ -476,8 +485,7 @@ int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 again:
 	if (!prealloc && (mask & __GFP_WAIT)) {
 		prealloc = alloc_extent_state(mask);
-		if (!prealloc)
-			return -ENOMEM;
+		BUG_ON(!prealloc);
 	}
 
 	spin_lock(&tree->lock);
@@ -529,8 +537,8 @@ hit_next:
 	 */
 
 	if (state->start < start) {
-		if (!prealloc)
-			prealloc = alloc_extent_state(GFP_ATOMIC);
+		prealloc = alloc_extent_state_atomic(prealloc);
+		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, start);
 		BUG_ON(err == -EEXIST);
 		prealloc = NULL;
@@ -551,8 +559,8 @@ hit_next:
 	 * on the first half
 	 */
 	if (state->start <= end && state->end > end) {
-		if (!prealloc)
-			prealloc = alloc_extent_state(GFP_ATOMIC);
+		prealloc = alloc_extent_state_atomic(prealloc);
+		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, end + 1);
 		BUG_ON(err == -EEXIST);
 		if (wake)
@@ -725,8 +733,7 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 again:
 	if (!prealloc && (mask & __GFP_WAIT)) {
 		prealloc = alloc_extent_state(mask);
-		if (!prealloc)
-			return -ENOMEM;
+		BUG_ON(!prealloc);
 	}
 
 	spin_lock(&tree->lock);
@@ -743,6 +750,8 @@ again:
 	 */
 	node = tree_search(tree, start);
 	if (!node) {
+		prealloc = alloc_extent_state_atomic(prealloc);
+		BUG_ON(!prealloc);
 		err = insert_state(tree, prealloc, start, end, &bits);
 		prealloc = NULL;
 		BUG_ON(err == -EEXIST);
@@ -771,20 +780,18 @@ hit_next:
 		if (err)
 			goto out;
 
+		next_node = rb_next(node);
 		cache_state(state, cached_state);
 		merge_state(tree, state);
 		if (last_end == (u64)-1)
 			goto out;
 
 		start = last_end + 1;
-		if (start < end && prealloc && !need_resched()) {
-			next_node = rb_next(node);
-			if (next_node) {
-				state = rb_entry(next_node, struct extent_state,
-						 rb_node);
-				if (state->start == start)
-					goto hit_next;
-			}
+		if (next_node && start < end && prealloc && !need_resched()) {
+			state = rb_entry(next_node, struct extent_state,
+					 rb_node);
+			if (state->start == start)
+				goto hit_next;
 		}
 		goto search_again;
 	}
@@ -811,6 +818,9 @@ hit_next:
 			err = -EEXIST;
 			goto out;
 		}
+
+		prealloc = alloc_extent_state_atomic(prealloc);
+		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, start);
 		BUG_ON(err == -EEXIST);
 		prealloc = NULL;
@@ -841,14 +851,25 @@ hit_next:
 			this_end = end;
 		else
 			this_end = last_start - 1;
+
+		prealloc = alloc_extent_state_atomic(prealloc);
+		BUG_ON(!prealloc);
+
+		/*
+		 * Avoid to free 'prealloc' if it can be merged with
+		 * the later extent.
+		 */
+		atomic_inc(&prealloc->refs);
 		err = insert_state(tree, prealloc, start, this_end,
 				   &bits);
 		BUG_ON(err == -EEXIST);
 		if (err) {
+			free_extent_state(prealloc);
 			prealloc = NULL;
 			goto out;
 		}
 		cache_state(prealloc, cached_state);
+		free_extent_state(prealloc);
 		prealloc = NULL;
 		start = this_end + 1;
 		goto search_again;
@@ -865,6 +886,9 @@ hit_next:
 			err = -EEXIST;
 			goto out;
 		}
+
+		prealloc = alloc_extent_state_atomic(prealloc);
+		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, end + 1);
 		BUG_ON(err == -EEXIST);
 

fs/btrfs/file-item.c

@@ -502,7 +502,6 @@ static noinline int truncate_one_csum(struct btrfs_trans_handle *trans,
 		u32 new_size = (bytenr - key->offset) >> blocksize_bits;
 		new_size *= csum_size;
 		ret = btrfs_truncate_item(trans, root, path, new_size, 1);
-		BUG_ON(ret);
 	} else if (key->offset >= bytenr && csum_end > end_byte &&
 		   end_byte > key->offset) {
 		/*
@@ -515,7 +514,6 @@ static noinline int truncate_one_csum(struct btrfs_trans_handle *trans,
 		new_size *= csum_size;
 
 		ret = btrfs_truncate_item(trans, root, path, new_size, 0);
-		BUG_ON(ret);
 
 		key->offset = end_byte;
 		ret = btrfs_set_item_key_safe(trans, root, path, key);
@@ -558,10 +556,10 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
 		ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
 		if (ret > 0) {
 			if (path->slots[0] == 0)
-				goto out;
+				break;
 			path->slots[0]--;
 		} else if (ret < 0) {
-			goto out;
+			break;
 		}
 
 		leaf = path->nodes[0];
@@ -586,7 +584,8 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
 		/* delete the entire item, it is inside our range */
 		if (key.offset >= bytenr && csum_end <= end_byte) {
 			ret = btrfs_del_item(trans, root, path);
-			BUG_ON(ret);
+			if (ret)
+				goto out;
 			if (key.offset == bytenr)
 				break;
 		} else if (key.offset < bytenr && csum_end > end_byte) {
@@ -640,9 +639,10 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
 		}
 		btrfs_release_path(path);
 	}
+	ret = 0;
 out:
 	btrfs_free_path(path);
-	return 0;
+	return ret;
 }
 
 int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
@@ -768,7 +768,6 @@ again:
 			goto insert;
 
 		ret = btrfs_extend_item(trans, root, path, diff);
-		BUG_ON(ret);
 		goto csum;
 	}
 

fs/btrfs/inode-item.c

@@ -130,7 +130,6 @@ int btrfs_del_inode_ref(struct btrfs_trans_handle *trans,
 			      item_size - (ptr + sub_item_len - item_start));
 	ret = btrfs_truncate_item(trans, root, path,
 				  item_size - sub_item_len, 1);
-	BUG_ON(ret);
 out:
 	btrfs_free_path(path);
 	return ret;
@@ -167,7 +166,6 @@ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans,
 
 		old_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]);
 		ret = btrfs_extend_item(trans, root, path, ins_len);
-		BUG_ON(ret);
 		ref = btrfs_item_ptr(path->nodes[0], path->slots[0],
 				     struct btrfs_inode_ref);
 		ref = (struct btrfs_inode_ref *)((unsigned long)ref + old_size);

fs/btrfs/inode.c

@@ -1323,7 +1323,7 @@ static int btrfs_set_bit_hook(struct inode *inode,
 
 	/*
 	 * set_bit and clear bit hooks normally require _irqsave/restore
-	 * but in this case, we are only testeing for the DELALLOC
+	 * but in this case, we are only testing for the DELALLOC
 	 * bit, which is only set or cleared with irqs on
 	 */
 	if (!(state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) {
@@ -1356,7 +1356,7 @@ static int btrfs_clear_bit_hook(struct inode *inode,
 {
 	/*
 	 * set_bit and clear bit hooks normally require _irqsave/restore
-	 * but in this case, we are only testeing for the DELALLOC
+	 * but in this case, we are only testing for the DELALLOC
 	 * bit, which is only set or cleared with irqs on
 	 */
 	if ((state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) {
@@ -3286,7 +3286,6 @@ search_again:
 				    btrfs_file_extent_calc_inline_size(size);
 				ret = btrfs_truncate_item(trans, root, path,
 							  size, 1);
-				BUG_ON(ret);
 			} else if (root->ref_cows) {
 				inode_sub_bytes(inode, item_end + 1 -
 						found_key.offset);
@@ -7167,6 +7166,7 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 				      datasize);
 	if (err) {
 		drop_inode = 1;
+		btrfs_free_path(path);
 		goto out_unlock;
 	}
 	leaf = path->nodes[0];

fs/btrfs/ioctl.c

@@ -282,8 +282,9 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	mutex_lock(&fs_info->fs_devices->device_list_mutex);
-	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {
+	rcu_read_lock();
+	list_for_each_entry_rcu(device, &fs_info->fs_devices->devices,
+				dev_list) {
 		if (!device->bdev)
 			continue;
 		q = bdev_get_queue(device->bdev);
@@ -293,7 +294,7 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
 				     minlen);
 		}
 	}
-	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
+	rcu_read_unlock();
 	if (!num_devices)
 		return -EOPNOTSUPP;
 
@@ -1279,7 +1280,6 @@ static noinline int copy_to_sk(struct btrfs_root *root,
 	int nritems;
 	int i;
 	int slot;
-	int found = 0;
 	int ret = 0;
 
 	leaf = path->nodes[0];
@@ -1326,7 +1326,7 @@ static noinline int copy_to_sk(struct btrfs_root *root,
 					   item_off, item_len);
 			*sk_offset += item_len;
 		}
-		found++;
+		(*num_found)++;
 
 		if (*num_found >= sk->nr_items)
 			break;
@@ -1345,7 +1345,6 @@ advance_key:
 	} else
 		ret = 1;
 overflow:
-	*num_found += found;
 	return ret;
 }
 

fs/btrfs/root-tree.c

@@ -338,7 +338,10 @@ again:
 		*sequence = btrfs_root_ref_sequence(leaf, ref);
 
 		ret = btrfs_del_item(trans, tree_root, path);
-		BUG_ON(ret);
+		if (ret) {
+			err = ret;
+			goto out;
+		}
 	} else
 		err = -ENOENT;
 
@@ -350,6 +353,7 @@ again:
 		goto again;
 	}
 
+out:
 	btrfs_free_path(path);
 	return err;
 }

fs/btrfs/super.c

@@ -507,8 +507,10 @@ static struct dentry *get_default_root(struct super_block *sb,
 	 */
 	dir_id = btrfs_super_root_dir(&root->fs_info->super_copy);
 	di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0);
-	if (IS_ERR(di))
+	if (IS_ERR(di)) {
+		btrfs_free_path(path);
 		return ERR_CAST(di);
+	}
 	if (!di) {
 		/*
 		 * Ok the default dir item isn't there.  This is weird since

fs/btrfs/tree-log.c

@@ -382,7 +382,6 @@ insert:
 		} else if (found_size < item_size) {
 			ret = btrfs_extend_item(trans, root, path,
 						item_size - found_size);
-			BUG_ON(ret);
 		}
 	} else if (ret) {
 		return ret;
@@ -590,6 +589,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 						ins.objectid, ins.offset,
 						0, root->root_key.objectid,
 						key->objectid, offset);
+				BUG_ON(ret);
 			} else {
 				/*
 				 * insert the extent pointer in the extent
@@ -678,7 +678,10 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
 	btrfs_release_path(path);
 
 	inode = read_one_inode(root, location.objectid);
-	BUG_ON(!inode);
+	if (!inode) {
+		kfree(name);
+		return -EIO;
+	}
 
 	ret = link_to_fixup_dir(trans, root, path, location.objectid);
 	BUG_ON(ret);
@@ -817,7 +820,10 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 		return -ENOENT;
 
 	inode = read_one_inode(root, key->objectid);
-	BUG_ON(!inode);
+	if (!inode) {
+		iput(dir);
+		return -EIO;
+	}
 
 	ref_ptr = btrfs_item_ptr_offset(eb, slot);
 	ref_end = ref_ptr + btrfs_item_size_nr(eb, slot);
@@ -1051,11 +1057,13 @@ static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans,
 			break;
 
 		ret = btrfs_del_item(trans, root, path);
-		BUG_ON(ret);
+		if (ret)
+			goto out;
 
 		btrfs_release_path(path);
 		inode = read_one_inode(root, key.offset);
-		BUG_ON(!inode);
+		if (!inode)
+			return -EIO;
 
 		ret = fixup_inode_link_count(trans, root, inode);
 		BUG_ON(ret);
@@ -1069,8 +1077,10 @@ static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans,
 		 */
 		key.offset = (u64)-1;
 	}
+	ret = 0;
+out:
 	btrfs_release_path(path);
-	return 0;
+	return ret;
 }
 
 
@@ -1089,7 +1099,8 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
 	struct inode *inode;
 
 	inode = read_one_inode(root, objectid);
-	BUG_ON(!inode);
+	if (!inode)
+		return -EIO;
 
 	key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
 	btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
@@ -1176,7 +1187,8 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 	int ret;
 
 	dir = read_one_inode(root, key->objectid);
-	BUG_ON(!dir);
+	if (!dir)
+		return -EIO;
 
 	name_len = btrfs_dir_name_len(eb, di);
 	name = kmalloc(name_len, GFP_NOFS);
@@ -1432,7 +1444,10 @@ again:
 			btrfs_release_path(path);
 			btrfs_release_path(log_path);
 			inode = read_one_inode(root, location.objectid);
-			BUG_ON(!inode);
+			if (!inode) {
+				kfree(name);
+				return -EIO;
+			}
 
 			ret = link_to_fixup_dir(trans, root,
 						path, location.objectid);
@@ -2589,7 +2604,8 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans,
 			break;
 
 		ret = btrfs_del_item(trans, log, path);
-		BUG_ON(ret);
+		if (ret)
+			break;
 		btrfs_release_path(path);
 	}
 	btrfs_release_path(path);

fs/btrfs/version.sh

@@ -1,43 +0,0 @@
-#!/bin/bash
-#
-# determine-version -- report a useful version for releases
-#
-# Copyright 2008, Aron Griffis <agriffis@n01se.net>
-# Copyright 2008, Oracle
-# Released under the GNU GPLv2
- 
-v="v0.16"
-
-which git &> /dev/null
-if [ $? == 0 ]; then
-    git branch >& /dev/null
-    if [ $? == 0 ]; then
-	    if head=`git rev-parse --verify HEAD 2>/dev/null`; then
-		if tag=`git describe --tags 2>/dev/null`; then
-		    v="$tag"
-		fi
-
-		# Are there uncommitted changes?
-		git update-index --refresh --unmerged > /dev/null
-		if git diff-index --name-only HEAD | \
-		    grep -v "^scripts/package" \
-		    | read dummy; then
-		    v="$v"-dirty
-		fi
-	    fi
-    fi
-fi
- 
-echo "#ifndef __BUILD_VERSION" > .build-version.h
-echo "#define __BUILD_VERSION" >> .build-version.h
-echo "#define BTRFS_BUILD_VERSION \"Btrfs $v\"" >> .build-version.h
-echo "#endif" >> .build-version.h
-
-diff -q version.h .build-version.h >& /dev/null
-
-if [ $? == 0 ]; then
-    rm .build-version.h
-    exit 0
-fi
-
-mv .build-version.h version.h

fs/btrfs/volumes.c

@@ -350,7 +350,7 @@ static noinline int device_list_add(const char *path,
 		INIT_LIST_HEAD(&device->dev_alloc_list);
 
 		mutex_lock(&fs_devices->device_list_mutex);
-		list_add(&device->dev_list, &fs_devices->devices);
+		list_add_rcu(&device->dev_list, &fs_devices->devices);
 		mutex_unlock(&fs_devices->device_list_mutex);
 
 		device->fs_devices = fs_devices;
@@ -393,7 +393,7 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 	fs_devices->latest_trans = orig->latest_trans;
 	memcpy(fs_devices->fsid, orig->fsid, sizeof(fs_devices->fsid));
 
-	mutex_lock(&orig->device_list_mutex);
+	/* We have held the volume lock, it is safe to get the devices. */
 	list_for_each_entry(orig_dev, &orig->devices, dev_list) {
 		device = kzalloc(sizeof(*device), GFP_NOFS);
 		if (!device)
@@ -416,10 +416,8 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 		device->fs_devices = fs_devices;
 		fs_devices->num_devices++;
 	}
-	mutex_unlock(&orig->device_list_mutex);
 	return fs_devices;
 error:
-	mutex_unlock(&orig->device_list_mutex);
 	free_fs_devices(fs_devices);
 	return ERR_PTR(-ENOMEM);
 }
@@ -430,7 +428,7 @@ int btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices)
 
 	mutex_lock(&uuid_mutex);
 again:
-	mutex_lock(&fs_devices->device_list_mutex);
+	/* This is the initialized path, it is safe to release the devices. */
 	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
 		if (device->in_fs_metadata)
 			continue;
@@ -450,7 +448,6 @@ again:
 		kfree(device->name);
 		kfree(device);
 	}
-	mutex_unlock(&fs_devices->device_list_mutex);
 
 	if (fs_devices->seed) {
 		fs_devices = fs_devices->seed;
@@ -461,6 +458,29 @@ again:
 	return 0;
 }
 
+static void __free_device(struct work_struct *work)
+{
+	struct btrfs_device *device;
+
+	device = container_of(work, struct btrfs_device, rcu_work);
+
+	if (device->bdev)
+		blkdev_put(device->bdev, device->mode);
+
+	kfree(device->name);
+	kfree(device);
+}
+
+static void free_device(struct rcu_head *head)
+{
+	struct btrfs_device *device;
+
+	device = container_of(head, struct btrfs_device, rcu);
+
+	INIT_WORK(&device->rcu_work, __free_device);
+	schedule_work(&device->rcu_work);
+}
+
 static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 {
 	struct btrfs_device *device;
@@ -468,20 +488,32 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 	if (--fs_devices->opened > 0)
 		return 0;
 
+	mutex_lock(&fs_devices->device_list_mutex);
 	list_for_each_entry(device, &fs_devices->devices, dev_list) {
-		if (device->bdev) {
-			blkdev_put(device->bdev, device->mode);
+		struct btrfs_device *new_device;
+
+		if (device->bdev)
 			fs_devices->open_devices--;
-		}
+
 		if (device->writeable) {
 			list_del_init(&device->dev_alloc_list);
 			fs_devices->rw_devices--;
 		}
 
-		device->bdev = NULL;
-		device->writeable = 0;
-		device->in_fs_metadata = 0;
+		new_device = kmalloc(sizeof(*new_device), GFP_NOFS);
+		BUG_ON(!new_device);
+		memcpy(new_device, device, sizeof(*new_device));
+		new_device->name = kstrdup(device->name, GFP_NOFS);
+		BUG_ON(!new_device->name);
+		new_device->bdev = NULL;
+		new_device->writeable = 0;
+		new_device->in_fs_metadata = 0;
+		list_replace_rcu(&device->dev_list, &new_device->dev_list);
+
+		call_rcu(&device->rcu, free_device);
 	}
+	mutex_unlock(&fs_devices->device_list_mutex);
+
 	WARN_ON(fs_devices->open_devices);
 	WARN_ON(fs_devices->rw_devices);
 	fs_devices->opened = 0;
@@ -584,6 +616,7 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 			list_add(&device->dev_alloc_list,
 				 &fs_devices->alloc_list);
 		}
+		brelse(bh);
 		continue;
 
 error_brelse:
@@ -933,14 +966,14 @@ static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans,
 	if (ret > 0) {
 		ret = btrfs_previous_item(root, path, key.objectid,
 					  BTRFS_DEV_EXTENT_KEY);
-		BUG_ON(ret);
+		if (ret)
+			goto out;
 		leaf = path->nodes[0];
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 		extent = btrfs_item_ptr(leaf, path->slots[0],
 					struct btrfs_dev_extent);
 		BUG_ON(found_key.offset > start || found_key.offset +
 		       btrfs_dev_extent_length(leaf, extent) < start);
-		ret = 0;
 	} else if (ret == 0) {
 		leaf = path->nodes[0];
 		extent = btrfs_item_ptr(leaf, path->slots[0],
@@ -951,8 +984,8 @@ static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans,
 	if (device->bytes_used > 0)
 		device->bytes_used -= btrfs_dev_extent_length(leaf, extent);
 	ret = btrfs_del_item(trans, root, path);
-	BUG_ON(ret);
 
+out:
 	btrfs_free_path(path);
 	return ret;
 }
@@ -1187,11 +1220,13 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	struct block_device *bdev;
 	struct buffer_head *bh = NULL;
 	struct btrfs_super_block *disk_super;
+	struct btrfs_fs_devices *cur_devices;
 	u64 all_avail;
 	u64 devid;
 	u64 num_devices;
 	u8 *dev_uuid;
 	int ret = 0;
+	bool clear_super = false;
 
 	mutex_lock(&uuid_mutex);
 	mutex_lock(&root->fs_info->volume_mutex);
@@ -1222,14 +1257,16 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 
 		device = NULL;
 		devices = &root->fs_info->fs_devices->devices;
-		mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+		/*
+		 * It is safe to read the devices since the volume_mutex
+		 * is held.
+		 */
 		list_for_each_entry(tmp, devices, dev_list) {
 			if (tmp->in_fs_metadata && !tmp->bdev) {
 				device = tmp;
 				break;
 			}
 		}
-		mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 		bdev = NULL;
 		bh = NULL;
 		disk_super = NULL;
@@ -1271,8 +1308,11 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	}
 
 	if (device->writeable) {
+		lock_chunks(root);
 		list_del_init(&device->dev_alloc_list);
+		unlock_chunks(root);
 		root->fs_info->fs_devices->rw_devices--;
+		clear_super = true;
 	}
 
 	ret = btrfs_shrink_device(device, 0);
@@ -1291,9 +1331,10 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	 * the device list while someone else is writing out all
 	 * the device supers.
 	 */
+
+	cur_devices = device->fs_devices;
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
-	list_del_init(&device->dev_list);
-	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+	list_del_rcu(&device->dev_list);
 
 	device->fs_devices->num_devices--;
 
@@ -1307,34 +1348,36 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	if (device->bdev == root->fs_info->fs_devices->latest_bdev)
 		root->fs_info->fs_devices->latest_bdev = next_device->bdev;
 
-	if (device->bdev) {
-		blkdev_put(device->bdev, device->mode);
-		device->bdev = NULL;
+	if (device->bdev)
 		device->fs_devices->open_devices--;
-	}
+
+	call_rcu(&device->rcu, free_device);
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	num_devices = btrfs_super_num_devices(&root->fs_info->super_copy) - 1;
 	btrfs_set_super_num_devices(&root->fs_info->super_copy, num_devices);
 
-	if (device->fs_devices->open_devices == 0) {
+	if (cur_devices->open_devices == 0) {
 		struct btrfs_fs_devices *fs_devices;
 		fs_devices = root->fs_info->fs_devices;
 		while (fs_devices) {
-			if (fs_devices->seed == device->fs_devices)
+			if (fs_devices->seed == cur_devices)
 				break;
 			fs_devices = fs_devices->seed;
 		}
-		fs_devices->seed = device->fs_devices->seed;
-		device->fs_devices->seed = NULL;
-		__btrfs_close_devices(device->fs_devices);
-		free_fs_devices(device->fs_devices);
+		fs_devices->seed = cur_devices->seed;
+		cur_devices->seed = NULL;
+		lock_chunks(root);
+		__btrfs_close_devices(cur_devices);
+		unlock_chunks(root);
+		free_fs_devices(cur_devices);
 	}
 
 	/*
 	 * at this point, the device is zero sized.  We want to
 	 * remove it from the devices list and zero out the old super
 	 */
-	if (device->writeable) {
+	if (clear_super) {
 		/* make sure this device isn't detected as part of
 		 * the FS anymore
 		 */
@@ -1343,8 +1386,6 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 		sync_dirty_buffer(bh);
 	}
 
-	kfree(device->name);
-	kfree(device);
 	ret = 0;
 
 error_brelse:
@@ -1358,8 +1399,10 @@ out:
 	return ret;
 error_undo:
 	if (device->writeable) {
+		lock_chunks(root);
 		list_add(&device->dev_alloc_list,
 			 &root->fs_info->fs_devices->alloc_list);
+		unlock_chunks(root);
 		root->fs_info->fs_devices->rw_devices++;
 	}
 	goto error_brelse;
@@ -1399,7 +1442,12 @@ static int btrfs_prepare_sprout(struct btrfs_trans_handle *trans,
 	INIT_LIST_HEAD(&seed_devices->devices);
 	INIT_LIST_HEAD(&seed_devices->alloc_list);
 	mutex_init(&seed_devices->device_list_mutex);
-	list_splice_init(&fs_devices->devices, &seed_devices->devices);
+
+	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+	list_splice_init_rcu(&fs_devices->devices, &seed_devices->devices,
+			      synchronize_rcu);
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+
 	list_splice_init(&fs_devices->alloc_list, &seed_devices->alloc_list);
 	list_for_each_entry(device, &seed_devices->devices, dev_list) {
 		device->fs_devices = seed_devices;
@@ -1596,7 +1644,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	 * half setup
 	 */
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
-	list_add(&device->dev_list, &root->fs_info->fs_devices->devices);
+	list_add_rcu(&device->dev_list, &root->fs_info->fs_devices->devices);
 	list_add(&device->dev_alloc_list,
 		 &root->fs_info->fs_devices->alloc_list);
 	root->fs_info->fs_devices->num_devices++;
@@ -1754,10 +1802,9 @@ static int btrfs_free_chunk(struct btrfs_trans_handle *trans,
 	BUG_ON(ret);
 
 	ret = btrfs_del_item(trans, root, path);
-	BUG_ON(ret);
 
 	btrfs_free_path(path);
-	return 0;
+	return ret;
 }
 
 static int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64

fs/btrfs/volumes.h

@@ -89,6 +89,8 @@ struct btrfs_device {
 	struct scrub_dev *scrub_device;
 
 	struct btrfs_work work;
+	struct rcu_head rcu;
+	struct work_struct rcu_work;
 };
 
 struct btrfs_fs_devices {