Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs

Pull btrfs updates from Chris Mason:
 "This is a pretty big pull, and most of these changes have been
  floating in btrfs-next for a long time.  Filipe's properties work is a
  cool building block for inheriting attributes like compression down on
  a per inode basis.

  Jeff Mahoney kicked in code to export filesystem info into sysfs.

  Otherwise, lots of performance improvements, cleanups and bug fixes.

  Looks like there are still a few other small pending incrementals, but
  I wanted to get the bulk of this in first"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (149 commits)
  Btrfs: fix spin_unlock in check_ref_cleanup
  Btrfs: setup inode location during btrfs_init_inode_locked
  Btrfs: don't use ram_bytes for uncompressed inline items
  Btrfs: fix btrfs_search_slot_for_read backwards iteration
  Btrfs: do not export ulist functions
  Btrfs: rework ulist with list+rb_tree
  Btrfs: fix memory leaks on walking backrefs failure
  Btrfs: fix send file hole detection leading to data corruption
  Btrfs: add a reschedule point in btrfs_find_all_roots()
  Btrfs: make send's file extent item search more efficient
  Btrfs: fix to catch all errors when resolving indirect ref
  Btrfs: fix protection between walking backrefs and root deletion
  btrfs: fix warning while merging two adjacent extents
  Btrfs: fix infinite path build loops in incremental send
  btrfs: undo sysfs when open_ctree() fails
  Btrfs: fix snprintf usage by send's gen_unique_name
  btrfs: fix defrag 32-bit integer overflow
  btrfs: sysfs: list the NO_HOLES feature
  btrfs: sysfs: don't show reserved incompat feature
  btrfs: call permission checks earlier in ioctls and return EPERM
  ...

Documentation/filesystems/btrfs.txt

@@ -38,7 +38,7 @@ Mount Options
 =============
 
 When mounting a btrfs filesystem, the following option are accepted.
-Unless otherwise specified, all options default to off.
+Options with (*) are default options and will not show in the mount options.
 
   alloc_start=<bytes>
 	Debugging option to force all block allocations above a certain
@@ -46,10 +46,12 @@ Unless otherwise specified, all options default to off.
 	bytes, optionally with a K, M, or G suffix, case insensitive.
 	Default is 1MB.
 
+  noautodefrag(*)
   autodefrag
-	Detect small random writes into files and queue them up for the
-	defrag process.  Works best for small files; Not well suited for
-	large database workloads.
+	Disable/enable auto defragmentation.
+	Auto defragmentation detects small random writes into files and queue
+	them up for the defrag process.  Works best for small files;
+	Not well suited for large database workloads.
 
   check_int
   check_int_data
@@ -96,21 +98,26 @@ Unless otherwise specified, all options default to off.
 	can be avoided.  Especially useful when trying to mount a multi-device
 	setup as root.  May be specified multiple times for multiple devices.
 
+  nodiscard(*)
   discard
-	Issue frequent commands to let the block device reclaim space freed by
-	the filesystem.  This is useful for SSD devices, thinly provisioned
+	Disable/enable discard mount option.
+	Discard issues frequent commands to let the block device reclaim space
+	freed by the filesystem.
+	This is useful for SSD devices, thinly provisioned
 	LUNs and virtual machine images, but may have a significant
 	performance impact.  (The fstrim command is also available to
 	initiate batch trims from userspace).
 
+  noenospc_debug(*)
   enospc_debug
-	Debugging option to be more verbose in some ENOSPC conditions.
+	Disable/enable debugging option to be more verbose in some ENOSPC conditions.
 
   fatal_errors=<action>
 	Action to take when encountering a fatal error: 
 	  "bug" - BUG() on a fatal error.  This is the default.
 	  "panic" - panic() on a fatal error.
 
+  noflushoncommit(*)
   flushoncommit
 	The 'flushoncommit' mount option forces any data dirtied by a write in a
 	prior transaction to commit as part of the current commit.  This makes
@@ -134,26 +141,32 @@ Unless otherwise specified, all options default to off.
 	Specify that 1 metadata chunk should be allocated after every <value>
 	data chunks.  Off by default.
 
+  acl(*)
   noacl
-	Disable support for Posix Access Control Lists (ACLs).  See the
+	Enable/disable support for Posix Access Control Lists (ACLs).  See the
 	acl(5) manual page for more information about ACLs.
 
+  barrier(*)
   nobarrier
-        Disables the use of block layer write barriers.  Write barriers ensure
-	that certain IOs make it through the device cache and are on persistent
-	storage.  If used on a device with a volatile (non-battery-backed)
-	write-back cache, this option will lead to filesystem corruption on a
-	system crash or power loss.
+        Enable/disable the use of block layer write barriers.  Write barriers
+	ensure that certain IOs make it through the device cache and are on
+	persistent storage. If disabled on a device with a volatile
+	(non-battery-backed) write-back cache, nobarrier option will lead to
+	filesystem corruption on a system crash or power loss.
 
+  datacow(*)
   nodatacow
-	Disable data copy-on-write for newly created files.  Implies nodatasum,
-	and disables all compression.
+	Enable/disable data copy-on-write for newly created files.
+	Nodatacow implies nodatasum, and disables all compression.
 
+  datasum(*)
   nodatasum
-	Disable data checksumming for newly created files.
+	Enable/disable data checksumming for newly created files.
+	Datasum implies datacow.
 
+  treelog(*)
   notreelog
-	Disable the tree logging used for fsync and O_SYNC writes.
+	Enable/disable the tree logging used for fsync and O_SYNC writes.
 
   recovery
 	Enable autorecovery attempts if a bad tree root is found at mount time.

fs/btrfs/Kconfig

@@ -1,6 +1,7 @@
 config BTRFS_FS
 	tristate "Btrfs filesystem support"
-	select LIBCRC32C
+	select CRYPTO
+	select CRYPTO_CRC32C
 	select ZLIB_INFLATE
 	select ZLIB_DEFLATE
 	select LZO_COMPRESS

fs/btrfs/Makefile

@@ -9,7 +9,7 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \
 	   export.o tree-log.o free-space-cache.o zlib.o lzo.o \
 	   compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o \
 	   reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o \
-	   uuid-tree.o
+	   uuid-tree.o props.o hash.o
 
 btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o
 btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o

fs/btrfs/backref.c

@@ -66,6 +66,16 @@ static int check_extent_in_eb(struct btrfs_key *key, struct extent_buffer *eb,
 	return 0;
 }
 
+static void free_inode_elem_list(struct extent_inode_elem *eie)
+{
+	struct extent_inode_elem *eie_next;
+
+	for (; eie; eie = eie_next) {
+		eie_next = eie->next;
+		kfree(eie);
+	}
+}
+
 static int find_extent_in_eb(struct extent_buffer *eb, u64 wanted_disk_byte,
 				u64 extent_item_pos,
 				struct extent_inode_elem **eie)
@@ -209,18 +219,19 @@ static int __add_prelim_ref(struct list_head *head, u64 root_id,
 }
 
 static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path,
-				struct ulist *parents, int level,
-				struct btrfs_key *key_for_search, u64 time_seq,
-				u64 wanted_disk_byte,
-				const u64 *extent_item_pos)
+			   struct ulist *parents, struct __prelim_ref *ref,
+			   int level, u64 time_seq, const u64 *extent_item_pos)
 {
 	int ret = 0;
 	int slot;
 	struct extent_buffer *eb;
 	struct btrfs_key key;
+	struct btrfs_key *key_for_search = &ref->key_for_search;
 	struct btrfs_file_extent_item *fi;
 	struct extent_inode_elem *eie = NULL, *old = NULL;
 	u64 disk_byte;
+	u64 wanted_disk_byte = ref->wanted_disk_byte;
+	u64 count = 0;
 
 	if (level != 0) {
 		eb = path->nodes[level];
@@ -238,7 +249,7 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path,
 	if (path->slots[0] >= btrfs_header_nritems(path->nodes[0]))
 		ret = btrfs_next_old_leaf(root, path, time_seq);
 
-	while (!ret) {
+	while (!ret && count < ref->count) {
 		eb = path->nodes[0];
 		slot = path->slots[0];
 
@@ -254,6 +265,7 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path,
 		if (disk_byte == wanted_disk_byte) {
 			eie = NULL;
 			old = NULL;
+			count++;
 			if (extent_item_pos) {
 				ret = check_extent_in_eb(&key, eb, fi,
 						*extent_item_pos,
@@ -273,6 +285,7 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path,
 					old = old->next;
 				old->next = eie;
 			}
+			eie = NULL;
 		}
 next:
 		ret = btrfs_next_old_item(root, path, time_seq);
@@ -280,6 +293,8 @@ next:
 
 	if (ret > 0)
 		ret = 0;
+	else if (ret < 0)
+		free_inode_elem_list(eie);
 	return ret;
 }
 
@@ -299,23 +314,34 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
 	int ret = 0;
 	int root_level;
 	int level = ref->level;
+	int index;
 
 	root_key.objectid = ref->root_id;
 	root_key.type = BTRFS_ROOT_ITEM_KEY;
 	root_key.offset = (u64)-1;
+
+	index = srcu_read_lock(&fs_info->subvol_srcu);
+
 	root = btrfs_read_fs_root_no_name(fs_info, &root_key);
 	if (IS_ERR(root)) {
+		srcu_read_unlock(&fs_info->subvol_srcu, index);
 		ret = PTR_ERR(root);
 		goto out;
 	}
 
 	root_level = btrfs_old_root_level(root, time_seq);
 
-	if (root_level + 1 == level)
+	if (root_level + 1 == level) {
+		srcu_read_unlock(&fs_info->subvol_srcu, index);
 		goto out;
+	}
 
 	path->lowest_level = level;
 	ret = btrfs_search_old_slot(root, &ref->key_for_search, path, time_seq);
+
+	/* root node has been locked, we can release @subvol_srcu safely here */
+	srcu_read_unlock(&fs_info->subvol_srcu, index);
+
 	pr_debug("search slot in root %llu (level %d, ref count %d) returned "
 		 "%d for key (%llu %u %llu)\n",
 		 ref->root_id, level, ref->count, ret,
@@ -334,9 +360,8 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
 		eb = path->nodes[level];
 	}
 
-	ret = add_all_parents(root, path, parents, level, &ref->key_for_search,
-				time_seq, ref->wanted_disk_byte,
-				extent_item_pos);
+	ret = add_all_parents(root, path, parents, ref, level, time_seq,
+			      extent_item_pos);
 out:
 	path->lowest_level = 0;
 	btrfs_release_path(path);
@@ -376,10 +401,16 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 			continue;
 		err = __resolve_indirect_ref(fs_info, path, time_seq, ref,
 					     parents, extent_item_pos);
-		if (err == -ENOMEM)
-			goto out;
-		if (err)
+		/*
+		 * we can only tolerate ENOENT,otherwise,we should catch error
+		 * and return directly.
+		 */
+		if (err == -ENOENT) {
 			continue;
+		} else if (err) {
+			ret = err;
+			goto out;
+		}
 
 		/* we put the first parent into the ref at hand */
 		ULIST_ITER_INIT(&uiter);
@@ -538,14 +569,13 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 	if (extent_op && extent_op->update_key)
 		btrfs_disk_key_to_cpu(&op_key, &extent_op->key);
 
-	while ((n = rb_prev(n))) {
+	spin_lock(&head->lock);
+	n = rb_first(&head->ref_root);
+	while (n) {
 		struct btrfs_delayed_ref_node *node;
 		node = rb_entry(n, struct btrfs_delayed_ref_node,
 				rb_node);
-		if (node->bytenr != head->node.bytenr)
-			break;
-		WARN_ON(node->is_head);
-
+		n = rb_next(n);
 		if (node->seq > seq)
 			continue;
 
@@ -612,10 +642,10 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			WARN_ON(1);
 		}
 		if (ret)
-			return ret;
+			break;
 	}
-
-	return 0;
+	spin_unlock(&head->lock);
+	return ret;
 }
 
 /*
@@ -828,6 +858,7 @@ static int find_parent_nodes(struct btrfs_trans_handle *trans,
 	struct list_head prefs_delayed;
 	struct list_head prefs;
 	struct __prelim_ref *ref;
+	struct extent_inode_elem *eie = NULL;
 
 	INIT_LIST_HEAD(&prefs);
 	INIT_LIST_HEAD(&prefs_delayed);
@@ -882,15 +913,15 @@ again:
 				btrfs_put_delayed_ref(&head->node);
 				goto again;
 			}
+			spin_unlock(&delayed_refs->lock);
 			ret = __add_delayed_refs(head, time_seq,
 						 &prefs_delayed);
 			mutex_unlock(&head->mutex);
-			if (ret) {
-				spin_unlock(&delayed_refs->lock);
+			if (ret)
 				goto out;
-			}
+		} else {
+			spin_unlock(&delayed_refs->lock);
 		}
-		spin_unlock(&delayed_refs->lock);
 	}
 
 	if (path->slots[0]) {
@@ -941,7 +972,6 @@ again:
 				goto out;
 		}
 		if (ref->count && ref->parent) {
-			struct extent_inode_elem *eie = NULL;
 			if (extent_item_pos && !ref->inode_list) {
 				u32 bsz;
 				struct extent_buffer *eb;
@@ -976,6 +1006,7 @@ again:
 					eie = eie->next;
 				eie->next = ref->inode_list;
 			}
+			eie = NULL;
 		}
 		list_del(&ref->list);
 		kmem_cache_free(btrfs_prelim_ref_cache, ref);
@@ -994,7 +1025,8 @@ out:
 		list_del(&ref->list);
 		kmem_cache_free(btrfs_prelim_ref_cache, ref);
 	}
-
+	if (ret < 0)
+		free_inode_elem_list(eie);
 	return ret;
 }
 
@@ -1002,7 +1034,6 @@ static void free_leaf_list(struct ulist *blocks)
 {
 	struct ulist_node *node = NULL;
 	struct extent_inode_elem *eie;
-	struct extent_inode_elem *eie_next;
 	struct ulist_iterator uiter;
 
 	ULIST_ITER_INIT(&uiter);
@@ -1010,10 +1041,7 @@ static void free_leaf_list(struct ulist *blocks)
 		if (!node->aux)
 			continue;
 		eie = (struct extent_inode_elem *)(uintptr_t)node->aux;
-		for (; eie; eie = eie_next) {
-			eie_next = eie->next;
-			kfree(eie);
-		}
+		free_inode_elem_list(eie);
 		node->aux = 0;
 	}
 
@@ -1101,44 +1129,13 @@ int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
 		if (!node)
 			break;
 		bytenr = node->val;
+		cond_resched();
 	}
 
 	ulist_free(tmp);
 	return 0;
 }
 
-
-static int __inode_info(u64 inum, u64 ioff, u8 key_type,
-			struct btrfs_root *fs_root, struct btrfs_path *path,
-			struct btrfs_key *found_key)
-{
-	int ret;
-	struct btrfs_key key;
-	struct extent_buffer *eb;
-
-	key.type = key_type;
-	key.objectid = inum;
-	key.offset = ioff;
-
-	ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0);
-	if (ret < 0)
-		return ret;
-
-	eb = path->nodes[0];
-	if (ret && path->slots[0] >= btrfs_header_nritems(eb)) {
-		ret = btrfs_next_leaf(fs_root, path);
-		if (ret)
-			return ret;
-		eb = path->nodes[0];
-	}
-
-	btrfs_item_key_to_cpu(eb, found_key, path->slots[0]);
-	if (found_key->type != key.type || found_key->objectid != key.objectid)
-		return 1;
-
-	return 0;
-}
-
 /*
  * this makes the path point to (inum INODE_ITEM ioff)
  */
@@ -1146,16 +1143,16 @@ int inode_item_info(u64 inum, u64 ioff, struct btrfs_root *fs_root,
 			struct btrfs_path *path)
 {
 	struct btrfs_key key;
-	return __inode_info(inum, ioff, BTRFS_INODE_ITEM_KEY, fs_root, path,
-				&key);
+	return btrfs_find_item(fs_root, path, inum, ioff,
+			BTRFS_INODE_ITEM_KEY, &key);
 }
 
 static int inode_ref_info(u64 inum, u64 ioff, struct btrfs_root *fs_root,
 				struct btrfs_path *path,
 				struct btrfs_key *found_key)
 {
-	return __inode_info(inum, ioff, BTRFS_INODE_REF_KEY, fs_root, path,
-				found_key);
+	return btrfs_find_item(fs_root, path, inum, ioff,
+			BTRFS_INODE_REF_KEY, found_key);
 }
 
 int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
@@ -1335,20 +1332,45 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
 	ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
 	if (ret < 0)
 		return ret;
-	ret = btrfs_previous_item(fs_info->extent_root, path,
-					0, BTRFS_EXTENT_ITEM_KEY);
-	if (ret < 0)
-		return ret;
 
-	btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]);
+	while (1) {
+		u32 nritems;
+		if (path->slots[0] == 0) {
+			btrfs_set_path_blocking(path);
+			ret = btrfs_prev_leaf(fs_info->extent_root, path);
+			if (ret != 0) {
+				if (ret > 0) {
+					pr_debug("logical %llu is not within "
+						 "any extent\n", logical);
+					ret = -ENOENT;
+				}
+				return ret;
+			}
+		} else {
+			path->slots[0]--;
+		}
+		nritems = btrfs_header_nritems(path->nodes[0]);
+		if (nritems == 0) {
+			pr_debug("logical %llu is not within any extent\n",
+				 logical);
+			return -ENOENT;
+		}
+		if (path->slots[0] == nritems)
+			path->slots[0]--;
+
+		btrfs_item_key_to_cpu(path->nodes[0], found_key,
+				      path->slots[0]);
+		if (found_key->type == BTRFS_EXTENT_ITEM_KEY ||
+		    found_key->type == BTRFS_METADATA_ITEM_KEY)
+			break;
+	}
+
 	if (found_key->type == BTRFS_METADATA_ITEM_KEY)
 		size = fs_info->extent_root->leafsize;
 	else if (found_key->type == BTRFS_EXTENT_ITEM_KEY)
 		size = found_key->offset;
 
-	if ((found_key->type != BTRFS_EXTENT_ITEM_KEY &&
-	     found_key->type != BTRFS_METADATA_ITEM_KEY) ||
-	    found_key->objectid > logical ||
+	if (found_key->objectid > logical ||
 	    found_key->objectid + size <= logical) {
 		pr_debug("logical %llu is not within any extent\n", logical);
 		return -ENOENT;
@@ -1601,7 +1623,6 @@ static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root,
 	struct btrfs_key found_key;
 
 	while (!ret) {
-		path->leave_spinning = 1;
 		ret = inode_ref_info(inum, parent ? parent+1 : 0, fs_root, path,
 				     &found_key);
 		if (ret < 0)
@@ -1614,9 +1635,12 @@ static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root,
 
 		parent = found_key.offset;
 		slot = path->slots[0];
-		eb = path->nodes[0];
-		/* make sure we can use eb after releasing the path */
-		atomic_inc(&eb->refs);
+		eb = btrfs_clone_extent_buffer(path->nodes[0]);
+		if (!eb) {
+			ret = -ENOMEM;
+			break;
+		}
+		extent_buffer_get(eb);
 		btrfs_tree_read_lock(eb);
 		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
 		btrfs_release_path(path);
@@ -1674,17 +1698,20 @@ static int iterate_inode_extrefs(u64 inum, struct btrfs_root *fs_root,
 		++found;
 
 		slot = path->slots[0];
-		eb = path->nodes[0];
-		/* make sure we can use eb after releasing the path */
-		atomic_inc(&eb->refs);
+		eb = btrfs_clone_extent_buffer(path->nodes[0]);
+		if (!eb) {
+			ret = -ENOMEM;
+			break;
+		}
+		extent_buffer_get(eb);
 
 		btrfs_tree_read_lock(eb);
 		btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
 		btrfs_release_path(path);
 
 		leaf = path->nodes[0];
-		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
-		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
+		item_size = btrfs_item_size_nr(leaf, slot);
+		ptr = btrfs_item_ptr_offset(leaf, slot);
 		cur_offset = 0;
 
 		while (cur_offset < item_size) {

fs/btrfs/btrfs_inode.h

@@ -43,6 +43,7 @@
 #define BTRFS_INODE_COPY_EVERYTHING		8
 #define BTRFS_INODE_IN_DELALLOC_LIST		9
 #define BTRFS_INODE_READDIO_NEED_LOCK		10
+#define BTRFS_INODE_HAS_PROPS		        11
 
 /* in memory btrfs inode */
 struct btrfs_inode {
@@ -135,6 +136,9 @@ struct btrfs_inode {
 	 */
 	u64 index_cnt;
 
+	/* Cache the directory index number to speed the dir/file remove */
+	u64 dir_index;
+
 	/* the fsync log has some corner cases that mean we have to check
 	 * directories to see if any unlinks have been done before
 	 * the directory was logged.  See tree-log.c for all the

fs/btrfs/check-integrity.c

@@ -1456,10 +1456,14 @@ static int btrfsic_handle_extent_data(
 	btrfsic_read_from_block_data(block_ctx, &file_extent_item,
 				     file_extent_item_offset,
 				     sizeof(struct btrfs_file_extent_item));
-	next_bytenr = btrfs_stack_file_extent_disk_bytenr(&file_extent_item) +
-		      btrfs_stack_file_extent_offset(&file_extent_item);
-	generation = btrfs_stack_file_extent_generation(&file_extent_item);
-	num_bytes = btrfs_stack_file_extent_num_bytes(&file_extent_item);
+	next_bytenr = btrfs_stack_file_extent_disk_bytenr(&file_extent_item);
+	if (btrfs_stack_file_extent_compression(&file_extent_item) ==
+	    BTRFS_COMPRESS_NONE) {
+		next_bytenr += btrfs_stack_file_extent_offset(&file_extent_item);
+		num_bytes = btrfs_stack_file_extent_num_bytes(&file_extent_item);
+	} else {
+		num_bytes = btrfs_stack_file_extent_disk_num_bytes(&file_extent_item);
+	}
 	generation = btrfs_stack_file_extent_generation(&file_extent_item);
 
 	if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)

fs/btrfs/compression.c

@@ -128,11 +128,10 @@ static int check_compressed_csum(struct inode *inode,
 		kunmap_atomic(kaddr);
 
 		if (csum != *cb_sum) {
-			printk(KERN_INFO "btrfs csum failed ino %llu "
-			       "extent %llu csum %u "
-			       "wanted %u mirror %d\n",
-			       btrfs_ino(inode), disk_start, csum, *cb_sum,
-			       cb->mirror_num);
+			btrfs_info(BTRFS_I(inode)->root->fs_info,
+			   "csum failed ino %llu extent %llu csum %u wanted %u mirror %d",
+			   btrfs_ino(inode), disk_start, csum, *cb_sum,
+			   cb->mirror_num);
 			ret = -EIO;
 			goto fail;
 		}
@@ -411,7 +410,8 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start,
 			bio_add_page(bio, page, PAGE_CACHE_SIZE, 0);
 		}
 		if (bytes_left < PAGE_CACHE_SIZE) {
-			printk("bytes left %lu compress len %lu nr %lu\n",
+			btrfs_info(BTRFS_I(inode)->root->fs_info,
+					"bytes left %lu compress len %lu nr %lu",
 			       bytes_left, cb->compressed_len, cb->nr_pages);
 		}
 		bytes_left -= PAGE_CACHE_SIZE;

fs/btrfs/ctree.c

@@ -39,9 +39,8 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
 			      struct extent_buffer *src_buf);
 static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
 		    int level, int slot);
-static void tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
+static int tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
 				 struct extent_buffer *eb);
-static int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path);
 
 struct btrfs_path *btrfs_alloc_path(void)
 {
@@ -475,6 +474,8 @@ void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,
  * the index is the shifted logical of the *new* root node for root replace
  * operations, or the shifted logical of the affected block for all other
  * operations.
+ *
+ * Note: must be called with write lock (tree_mod_log_write_lock).
  */
 static noinline int
 __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
@@ -483,24 +484,9 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
 	struct rb_node **new;
 	struct rb_node *parent = NULL;
 	struct tree_mod_elem *cur;
-	int ret = 0;
 
 	BUG_ON(!tm);
 
-	tree_mod_log_write_lock(fs_info);
-	if (list_empty(&fs_info->tree_mod_seq_list)) {
-		tree_mod_log_write_unlock(fs_info);
-		/*
-		 * Ok we no longer care about logging modifications, free up tm
-		 * and return 0.  Any callers shouldn't be using tm after
-		 * calling tree_mod_log_insert, but if they do we can just
-		 * change this to return a special error code to let the callers
-		 * do their own thing.
-		 */
-		kfree(tm);
-		return 0;
-	}
-
 	spin_lock(&fs_info->tree_mod_seq_lock);
 	tm->seq = btrfs_inc_tree_mod_seq_minor(fs_info);
 	spin_unlock(&fs_info->tree_mod_seq_lock);
@@ -518,18 +504,13 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
 			new = &((*new)->rb_left);
 		else if (cur->seq > tm->seq)
 			new = &((*new)->rb_right);
-		else {
-			ret = -EEXIST;
-			kfree(tm);
-			goto out;
-		}
+		else
+			return -EEXIST;
 	}
 
 	rb_link_node(&tm->node, parent, new);
 	rb_insert_color(&tm->node, tm_root);
-out:
-	tree_mod_log_write_unlock(fs_info);
-	return ret;
+	return 0;
 }
 
 /*
@@ -545,19 +526,38 @@ static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info,
 		return 1;
 	if (eb && btrfs_header_level(eb) == 0)
 		return 1;
+
+	tree_mod_log_write_lock(fs_info);
+	if (list_empty(&(fs_info)->tree_mod_seq_list)) {
+		tree_mod_log_write_unlock(fs_info);
+		return 1;
+	}
+
 	return 0;
 }
 
-static inline int
-__tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
-			  struct extent_buffer *eb, int slot,
-			  enum mod_log_op op, gfp_t flags)
+/* Similar to tree_mod_dont_log, but doesn't acquire any locks. */
+static inline int tree_mod_need_log(const struct btrfs_fs_info *fs_info,
+				    struct extent_buffer *eb)
+{
+	smp_mb();
+	if (list_empty(&(fs_info)->tree_mod_seq_list))
+		return 0;
+	if (eb && btrfs_header_level(eb) == 0)
+		return 0;
+
+	return 1;
+}
+
+static struct tree_mod_elem *
+alloc_tree_mod_elem(struct extent_buffer *eb, int slot,
+		    enum mod_log_op op, gfp_t flags)
 {
 	struct tree_mod_elem *tm;
 
 	tm = kzalloc(sizeof(*tm), flags);
 	if (!tm)
-		return -ENOMEM;
+		return NULL;
 
 	tm->index = eb->start >> PAGE_CACHE_SHIFT;
 	if (op != MOD_LOG_KEY_ADD) {
@@ -567,8 +567,9 @@ __tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
 	tm->op = op;
 	tm->slot = slot;
 	tm->generation = btrfs_node_ptr_generation(eb, slot);
+	RB_CLEAR_NODE(&tm->node);
 
-	return __tree_mod_log_insert(fs_info, tm);
+	return tm;
 }
 
 static noinline int
@@ -576,10 +577,27 @@ tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
 			struct extent_buffer *eb, int slot,
 			enum mod_log_op op, gfp_t flags)
 {
-	if (tree_mod_dont_log(fs_info, eb))
+	struct tree_mod_elem *tm;
+	int ret;
+
+	if (!tree_mod_need_log(fs_info, eb))
 		return 0;
 
-	return __tree_mod_log_insert_key(fs_info, eb, slot, op, flags);
+	tm = alloc_tree_mod_elem(eb, slot, op, flags);
+	if (!tm)
+		return -ENOMEM;
+
+	if (tree_mod_dont_log(fs_info, eb)) {
+		kfree(tm);
+		return 0;
+	}
+
+	ret = __tree_mod_log_insert(fs_info, tm);
+	tree_mod_log_write_unlock(fs_info);
+	if (ret)
+		kfree(tm);
+
+	return ret;
 }
 
 static noinline int
@@ -587,53 +605,95 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
 			 struct extent_buffer *eb, int dst_slot, int src_slot,
 			 int nr_items, gfp_t flags)
 {
-	struct tree_mod_elem *tm;
-	int ret;
+	struct tree_mod_elem *tm = NULL;
+	struct tree_mod_elem **tm_list = NULL;
+	int ret = 0;
 	int i;
+	int locked = 0;
 
-	if (tree_mod_dont_log(fs_info, eb))
+	if (!tree_mod_need_log(fs_info, eb))
 		return 0;
 
+	tm_list = kzalloc(nr_items * sizeof(struct tree_mod_elem *), flags);
+	if (!tm_list)
+		return -ENOMEM;
+
+	tm = kzalloc(sizeof(*tm), flags);
+	if (!tm) {
+		ret = -ENOMEM;
+		goto free_tms;
+	}
+
+	tm->index = eb->start >> PAGE_CACHE_SHIFT;
+	tm->slot = src_slot;
+	tm->move.dst_slot = dst_slot;
+	tm->move.nr_items = nr_items;
+	tm->op = MOD_LOG_MOVE_KEYS;
+
+	for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
+		tm_list[i] = alloc_tree_mod_elem(eb, i + dst_slot,
+		    MOD_LOG_KEY_REMOVE_WHILE_MOVING, flags);
+		if (!tm_list[i]) {
+			ret = -ENOMEM;
+			goto free_tms;
+		}
+	}
+
+	if (tree_mod_dont_log(fs_info, eb))
+		goto free_tms;
+	locked = 1;
+
 	/*
 	 * When we override something during the move, we log these removals.
 	 * This can only happen when we move towards the beginning of the
 	 * buffer, i.e. dst_slot < src_slot.
 	 */
 	for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
-		ret = __tree_mod_log_insert_key(fs_info, eb, i + dst_slot,
-				MOD_LOG_KEY_REMOVE_WHILE_MOVING, GFP_NOFS);
-		BUG_ON(ret < 0);
+		ret = __tree_mod_log_insert(fs_info, tm_list[i]);
+		if (ret)
+			goto free_tms;
 	}
 
-	tm = kzalloc(sizeof(*tm), flags);
-	if (!tm)
-		return -ENOMEM;
+	ret = __tree_mod_log_insert(fs_info, tm);
+	if (ret)
+		goto free_tms;
+	tree_mod_log_write_unlock(fs_info);
+	kfree(tm_list);
 
-	tm->index = eb->start >> PAGE_CACHE_SHIFT;
-	tm->slot = src_slot;
-	tm->move.dst_slot = dst_slot;
-	tm->move.nr_items = nr_items;
-	tm->op = MOD_LOG_MOVE_KEYS;
+	return 0;
+free_tms:
+	for (i = 0; i < nr_items; i++) {
+		if (tm_list[i] && !RB_EMPTY_NODE(&tm_list[i]->node))
+			rb_erase(&tm_list[i]->node, &fs_info->tree_mod_log);
+		kfree(tm_list[i]);
+	}
+	if (locked)
+		tree_mod_log_write_unlock(fs_info);
+	kfree(tm_list);
+	kfree(tm);
 
-	return __tree_mod_log_insert(fs_info, tm);
+	return ret;
 }
 
-static inline void
-__tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
+static inline int
+__tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
+		       struct tree_mod_elem **tm_list,
+		       int nritems)
 {
-	int i;
-	u32 nritems;
+	int i, j;
 	int ret;
 
-	if (btrfs_header_level(eb) == 0)
-		return;
-
-	nritems = btrfs_header_nritems(eb);
 	for (i = nritems - 1; i >= 0; i--) {
-		ret = __tree_mod_log_insert_key(fs_info, eb, i,
-				MOD_LOG_KEY_REMOVE_WHILE_FREEING, GFP_NOFS);
-		BUG_ON(ret < 0);
+		ret = __tree_mod_log_insert(fs_info, tm_list[i]);
+		if (ret) {
+			for (j = nritems - 1; j > i; j--)
+				rb_erase(&tm_list[j]->node,
+					 &fs_info->tree_mod_log);
+			return ret;
+		}
 	}
+
+	return 0;
 }
 
 static noinline int
@@ -642,17 +702,38 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 			 struct extent_buffer *new_root, gfp_t flags,
 			 int log_removal)
 {
-	struct tree_mod_elem *tm;
+	struct tree_mod_elem *tm = NULL;
+	struct tree_mod_elem **tm_list = NULL;
+	int nritems = 0;
+	int ret = 0;
+	int i;
 
-	if (tree_mod_dont_log(fs_info, NULL))
+	if (!tree_mod_need_log(fs_info, NULL))
 		return 0;
 
-	if (log_removal)
-		__tree_mod_log_free_eb(fs_info, old_root);
+	if (log_removal && btrfs_header_level(old_root) > 0) {
+		nritems = btrfs_header_nritems(old_root);
+		tm_list = kzalloc(nritems * sizeof(struct tree_mod_elem *),
+				  flags);
+		if (!tm_list) {
+			ret = -ENOMEM;
+			goto free_tms;
+		}
+		for (i = 0; i < nritems; i++) {
+			tm_list[i] = alloc_tree_mod_elem(old_root, i,
+			    MOD_LOG_KEY_REMOVE_WHILE_FREEING, flags);
+			if (!tm_list[i]) {
+				ret = -ENOMEM;
+				goto free_tms;
+			}
+		}
+	}
 
 	tm = kzalloc(sizeof(*tm), flags);
-	if (!tm)
-		return -ENOMEM;
+	if (!tm) {
+		ret = -ENOMEM;
+		goto free_tms;
+	}
 
 	tm->index = new_root->start >> PAGE_CACHE_SHIFT;
 	tm->old_root.logical = old_root->start;
@@ -660,7 +741,30 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 	tm->generation = btrfs_header_generation(old_root);
 	tm->op = MOD_LOG_ROOT_REPLACE;
 
-	return __tree_mod_log_insert(fs_info, tm);
+	if (tree_mod_dont_log(fs_info, NULL))
+		goto free_tms;
+
+	if (tm_list)
+		ret = __tree_mod_log_free_eb(fs_info, tm_list, nritems);
+	if (!ret)
+		ret = __tree_mod_log_insert(fs_info, tm);
+
+	tree_mod_log_write_unlock(fs_info);
+	if (ret)
+		goto free_tms;
+	kfree(tm_list);
+
+	return ret;
+
+free_tms:
+	if (tm_list) {
+		for (i = 0; i < nritems; i++)
+			kfree(tm_list[i]);
+		kfree(tm_list);
+	}
+	kfree(tm);
+
+	return ret;
 }
 
 static struct tree_mod_elem *
@@ -729,31 +833,75 @@ tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq)
 	return __tree_mod_log_search(fs_info, start, min_seq, 0);
 }
 
-static noinline void
+static noinline int
 tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
 		     struct extent_buffer *src, unsigned long dst_offset,
 		     unsigned long src_offset, int nr_items)
 {
-	int ret;
+	int ret = 0;
+	struct tree_mod_elem **tm_list = NULL;
+	struct tree_mod_elem **tm_list_add, **tm_list_rem;
 	int i;
+	int locked = 0;
 
-	if (tree_mod_dont_log(fs_info, NULL))
-		return;
+	if (!tree_mod_need_log(fs_info, NULL))
+		return 0;
 
 	if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0)
-		return;
+		return 0;
+
+	tm_list = kzalloc(nr_items * 2 * sizeof(struct tree_mod_elem *),
+			  GFP_NOFS);
+	if (!tm_list)
+		return -ENOMEM;
 
+	tm_list_add = tm_list;
+	tm_list_rem = tm_list + nr_items;
 	for (i = 0; i < nr_items; i++) {
-		ret = __tree_mod_log_insert_key(fs_info, src,
-						i + src_offset,
-						MOD_LOG_KEY_REMOVE, GFP_NOFS);
-		BUG_ON(ret < 0);
-		ret = __tree_mod_log_insert_key(fs_info, dst,
-						     i + dst_offset,
-						     MOD_LOG_KEY_ADD,
-						     GFP_NOFS);
-		BUG_ON(ret < 0);
+		tm_list_rem[i] = alloc_tree_mod_elem(src, i + src_offset,
+		    MOD_LOG_KEY_REMOVE, GFP_NOFS);
+		if (!tm_list_rem[i]) {
+			ret = -ENOMEM;
+			goto free_tms;
+		}
+
+		tm_list_add[i] = alloc_tree_mod_elem(dst, i + dst_offset,
+		    MOD_LOG_KEY_ADD, GFP_NOFS);
+		if (!tm_list_add[i]) {
+			ret = -ENOMEM;
+			goto free_tms;
+		}
 	}
+
+	if (tree_mod_dont_log(fs_info, NULL))
+		goto free_tms;
+	locked = 1;
+
+	for (i = 0; i < nr_items; i++) {
+		ret = __tree_mod_log_insert(fs_info, tm_list_rem[i]);
+		if (ret)
+			goto free_tms;
+		ret = __tree_mod_log_insert(fs_info, tm_list_add[i]);
+		if (ret)
+			goto free_tms;
+	}
+
+	tree_mod_log_write_unlock(fs_info);
+	kfree(tm_list);
+
+	return 0;
+
+free_tms:
+	for (i = 0; i < nr_items * 2; i++) {
+		if (tm_list[i] && !RB_EMPTY_NODE(&tm_list[i]->node))
+			rb_erase(&tm_list[i]->node, &fs_info->tree_mod_log);
+		kfree(tm_list[i]);
+	}
+	if (locked)
+		tree_mod_log_write_unlock(fs_info);
+	kfree(tm_list);
+
+	return ret;
 }
 
 static inline void
@@ -772,18 +920,58 @@ tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info,
 {
 	int ret;
 
-	ret = __tree_mod_log_insert_key(fs_info, eb, slot,
+	ret = tree_mod_log_insert_key(fs_info, eb, slot,
 					MOD_LOG_KEY_REPLACE,
 					atomic ? GFP_ATOMIC : GFP_NOFS);
 	BUG_ON(ret < 0);
 }
 
-static noinline void
+static noinline int
 tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 {
+	struct tree_mod_elem **tm_list = NULL;
+	int nritems = 0;
+	int i;
+	int ret = 0;
+
+	if (btrfs_header_level(eb) == 0)
+		return 0;
+
+	if (!tree_mod_need_log(fs_info, NULL))
+		return 0;
+
+	nritems = btrfs_header_nritems(eb);
+	tm_list = kzalloc(nritems * sizeof(struct tree_mod_elem *),
+			  GFP_NOFS);
+	if (!tm_list)
+		return -ENOMEM;
+
+	for (i = 0; i < nritems; i++) {
+		tm_list[i] = alloc_tree_mod_elem(eb, i,
+		    MOD_LOG_KEY_REMOVE_WHILE_FREEING, GFP_NOFS);
+		if (!tm_list[i]) {
+			ret = -ENOMEM;
+			goto free_tms;
+		}
+	}
+
 	if (tree_mod_dont_log(fs_info, eb))
-		return;
-	__tree_mod_log_free_eb(fs_info, eb);
+		goto free_tms;
+
+	ret = __tree_mod_log_free_eb(fs_info, tm_list, nritems);
+	tree_mod_log_write_unlock(fs_info);
+	if (ret)
+		goto free_tms;
+	kfree(tm_list);
+
+	return 0;
+
+free_tms:
+	for (i = 0; i < nritems; i++)
+		kfree(tm_list[i]);
+	kfree(tm_list);
+
+	return ret;
 }
 
 static noinline void
@@ -1041,8 +1229,13 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 		btrfs_set_node_ptr_generation(parent, parent_slot,
 					      trans->transid);
 		btrfs_mark_buffer_dirty(parent);
-		if (last_ref)
-			tree_mod_log_free_eb(root->fs_info, buf);
+		if (last_ref) {
+			ret = tree_mod_log_free_eb(root->fs_info, buf);
+			if (ret) {
+				btrfs_abort_transaction(trans, root, ret);
+				return ret;
+			}
+		}
 		btrfs_free_tree_block(trans, root, buf, parent_start,
 				      last_ref);
 	}
@@ -1287,8 +1480,8 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 		old = read_tree_block(root, logical, blocksize, 0);
 		if (WARN_ON(!old || !extent_buffer_uptodate(old))) {
 			free_extent_buffer(old);
-			pr_warn("btrfs: failed to read tree block %llu from get_old_root\n",
-				logical);
+			btrfs_warn(root->fs_info,
+				"failed to read tree block %llu from get_old_root", logical);
 		} else {
 			eb = btrfs_clone_extent_buffer(old);
 			free_extent_buffer(old);
@@ -2462,6 +2655,49 @@ static int key_search(struct extent_buffer *b, struct btrfs_key *key,
 	return 0;
 }
 
+int btrfs_find_item(struct btrfs_root *fs_root, struct btrfs_path *found_path,
+		u64 iobjectid, u64 ioff, u8 key_type,
+		struct btrfs_key *found_key)
+{
+	int ret;
+	struct btrfs_key key;
+	struct extent_buffer *eb;
+	struct btrfs_path *path;
+
+	key.type = key_type;
+	key.objectid = iobjectid;
+	key.offset = ioff;
+
+	if (found_path == NULL) {
+		path = btrfs_alloc_path();
+		if (!path)
+			return -ENOMEM;
+	} else
+		path = found_path;
+
+	ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0);
+	if ((ret < 0) || (found_key == NULL)) {
+		if (path != found_path)
+			btrfs_free_path(path);
+		return ret;
+	}
+
+	eb = path->nodes[0];
+	if (ret && path->slots[0] >= btrfs_header_nritems(eb)) {
+		ret = btrfs_next_leaf(fs_root, path);
+		if (ret)
+			return ret;
+		eb = path->nodes[0];
+	}
+
+	btrfs_item_key_to_cpu(eb, found_key, path->slots[0]);
+	if (found_key->type != key.type ||
+			found_key->objectid != key.objectid)
+		return 1;
+
+	return 0;
+}
+
 /*
  * look for key in the tree.  path is filled in with nodes along the way
  * if key is found, we return zero and you can find the item in the leaf
@@ -2495,6 +2731,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
 	lowest_level = p->lowest_level;
 	WARN_ON(lowest_level && ins_len > 0);
 	WARN_ON(p->nodes[0] != NULL);
+	BUG_ON(!cow && ins_len);
 
 	if (ins_len < 0) {
 		lowest_unlock = 2;
@@ -2603,8 +2840,6 @@ again:
 			}
 		}
 cow_done:
-		BUG_ON(!cow && ins_len);
-
 		p->nodes[level] = b;
 		btrfs_clear_path_blocking(p, NULL, 0);
 
@@ -2614,13 +2849,19 @@ cow_done:
 		 * It is safe to drop the lock on our parent before we
 		 * go through the expensive btree search on b.
 		 *
-		 * If cow is true, then we might be changing slot zero,
-		 * which may require changing the parent.  So, we can't
-		 * drop the lock until after we know which slot we're
-		 * operating on.
+		 * If we're inserting or deleting (ins_len != 0), then we might
+		 * be changing slot zero, which may require changing the parent.
+		 * So, we can't drop the lock until after we know which slot
+		 * we're operating on.
 		 */
-		if (!cow)
-			btrfs_unlock_up_safe(p, level + 1);
+		if (!ins_len && !p->keep_locks) {
+			int u = level + 1;
+
+			if (u < BTRFS_MAX_LEVEL && p->locks[u]) {
+				btrfs_tree_unlock_rw(p->nodes[u], p->locks[u]);
+				p->locks[u] = 0;
+			}
+		}
 
 		ret = key_search(b, key, level, &prev_cmp, &slot);
 
@@ -2648,7 +2889,7 @@ cow_done:
 			 * which means we must have a write lock
 			 * on the parent
 			 */
-			if (slot == 0 && cow &&
+			if (slot == 0 && ins_len &&
 			    write_lock_level < level + 1) {
 				write_lock_level = level + 1;
 				btrfs_release_path(p);
@@ -2901,7 +3142,9 @@ again:
 			if (ret < 0)
 				return ret;
 			if (!ret) {
-				p->slots[0] = btrfs_header_nritems(leaf) - 1;
+				leaf = p->nodes[0];
+				if (p->slots[0] == btrfs_header_nritems(leaf))
+					p->slots[0]--;
 				return 0;
 			}
 			if (!return_any)
@@ -3022,8 +3265,12 @@ static int push_node_left(struct btrfs_trans_handle *trans,
 	} else
 		push_items = min(src_nritems - 8, push_items);
 
-	tree_mod_log_eb_copy(root->fs_info, dst, src, dst_nritems, 0,
-			     push_items);
+	ret = tree_mod_log_eb_copy(root->fs_info, dst, src, dst_nritems, 0,
+				   push_items);
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		return ret;
+	}
 	copy_extent_buffer(dst, src,
 			   btrfs_node_key_ptr_offset(dst_nritems),
 			   btrfs_node_key_ptr_offset(0),
@@ -3093,8 +3340,12 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
 				      (dst_nritems) *
 				      sizeof(struct btrfs_key_ptr));
 
-	tree_mod_log_eb_copy(root->fs_info, dst, src, 0,
-			     src_nritems - push_items, push_items);
+	ret = tree_mod_log_eb_copy(root->fs_info, dst, src, 0,
+				   src_nritems - push_items, push_items);
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		return ret;
+	}
 	copy_extent_buffer(dst, src,
 			   btrfs_node_key_ptr_offset(0),
 			   btrfs_node_key_ptr_offset(src_nritems - push_items),
@@ -3295,7 +3546,12 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 			    btrfs_header_chunk_tree_uuid(split),
 			    BTRFS_UUID_SIZE);
 
-	tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid);
+	ret = tree_mod_log_eb_copy(root->fs_info, split, c, 0,
+				   mid, c_nritems - mid);
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		return ret;
+	}
 	copy_extent_buffer(split, c,
 			   btrfs_node_key_ptr_offset(0),
 			   btrfs_node_key_ptr_offset(mid),
@@ -3362,8 +3618,8 @@ noinline int btrfs_leaf_free_space(struct btrfs_root *root,
 	int ret;
 	ret = BTRFS_LEAF_DATA_SIZE(root) - leaf_space_used(leaf, 0, nritems);
 	if (ret < 0) {
-		printk(KERN_CRIT "leaf free space ret %d, leaf data size %lu, "
-		       "used %d nritems %d\n",
+		btrfs_crit(root->fs_info,
+			"leaf free space ret %d, leaf data size %lu, used %d nritems %d",
 		       ret, (unsigned long) BTRFS_LEAF_DATA_SIZE(root),
 		       leaf_space_used(leaf, 0, nritems), nritems);
 	}
@@ -3571,6 +3827,19 @@ static int push_leaf_right(struct btrfs_trans_handle *trans, struct btrfs_root
 	if (left_nritems == 0)
 		goto out_unlock;
 
+	if (path->slots[0] == left_nritems && !empty) {
+		/* Key greater than all keys in the leaf, right neighbor has
+		 * enough room for it and we're not emptying our leaf to delete
+		 * it, therefore use right neighbor to insert the new item and
+		 * no need to touch/dirty our left leaft. */
+		btrfs_tree_unlock(left);
+		free_extent_buffer(left);
+		path->nodes[0] = right;
+		path->slots[0] = 0;
+		path->slots[1]++;
+		return 0;
+	}
+
 	return __push_leaf_right(trans, root, path, min_data_size, empty,
 				right, free_space, left_nritems, min_slot);
 out_unlock:
@@ -3887,14 +4156,17 @@ static noinline int push_for_double_split(struct btrfs_trans_handle *trans,
 	int progress = 0;
 	int slot;
 	u32 nritems;
+	int space_needed = data_size;
 
 	slot = path->slots[0];
+	if (slot < btrfs_header_nritems(path->nodes[0]))
+		space_needed -= btrfs_leaf_free_space(root, path->nodes[0]);
 
 	/*
 	 * try to push all the items after our slot into the
 	 * right leaf
 	 */
-	ret = push_leaf_right(trans, root, path, 1, data_size, 0, slot);
+	ret = push_leaf_right(trans, root, path, 1, space_needed, 0, slot);
 	if (ret < 0)
 		return ret;
 
@@ -3914,7 +4186,7 @@ static noinline int push_for_double_split(struct btrfs_trans_handle *trans,
 
 	/* try to push all the items before our slot into the next leaf */
 	slot = path->slots[0];
-	ret = push_leaf_left(trans, root, path, 1, data_size, 0, slot);
+	ret = push_leaf_left(trans, root, path, 1, space_needed, 0, slot);
 	if (ret < 0)
 		return ret;
 
@@ -3958,13 +4230,18 @@ static noinline int split_leaf(struct btrfs_trans_handle *trans,
 
 	/* first try to make some room by pushing left and right */
 	if (data_size && path->nodes[1]) {
-		wret = push_leaf_right(trans, root, path, data_size,
-				       data_size, 0, 0);
+		int space_needed = data_size;
+
+		if (slot < btrfs_header_nritems(l))
+			space_needed -= btrfs_leaf_free_space(root, l);
+
+		wret = push_leaf_right(trans, root, path, space_needed,
+				       space_needed, 0, 0);
 		if (wret < 0)
 			return wret;
 		if (wret) {
-			wret = push_leaf_left(trans, root, path, data_size,
-					      data_size, 0, (u32)-1);
+			wret = push_leaf_left(trans, root, path, space_needed,
+					      space_needed, 0, (u32)-1);
 			if (wret < 0)
 				return wret;
 		}
@@ -4432,7 +4709,7 @@ void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
 	BUG_ON(slot < 0);
 	if (slot >= nritems) {
 		btrfs_print_leaf(root, leaf);
-		printk(KERN_CRIT "slot %d too large, nritems %d\n",
+		btrfs_crit(root->fs_info, "slot %d too large, nritems %d",
 		       slot, nritems);
 		BUG_ON(1);
 	}
@@ -4495,7 +4772,7 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 
 	if (btrfs_leaf_free_space(root, leaf) < total_size) {
 		btrfs_print_leaf(root, leaf);
-		printk(KERN_CRIT "not enough freespace need %u have %d\n",
+		btrfs_crit(root->fs_info, "not enough freespace need %u have %d",
 		       total_size, btrfs_leaf_free_space(root, leaf));
 		BUG();
 	}
@@ -4505,7 +4782,7 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 
 		if (old_data < data_end) {
 			btrfs_print_leaf(root, leaf);
-			printk(KERN_CRIT "slot %d old_data %d data_end %d\n",
+			btrfs_crit(root->fs_info, "slot %d old_data %d data_end %d",
 			       slot, old_data, data_end);
 			BUG_ON(1);
 		}
@@ -4817,7 +5094,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
  * This may release the path, and so you may lose any locks held at the
  * time you call it.
  */
-static int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
+int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
 {
 	struct btrfs_key key;
 	struct btrfs_disk_key found_key;
@@ -5240,7 +5517,7 @@ int btrfs_compare_trees(struct btrfs_root *left_root,
 
 			if (!left_start_ctransid || !right_start_ctransid) {
 				WARN(1, KERN_WARNING
-					"btrfs: btrfs_compare_tree detected "
+					"BTRFS: btrfs_compare_tree detected "
 					"a change in one of the trees while "
 					"iterating. This is probably a "
 					"bug.\n");
@@ -5680,3 +5957,46 @@ int btrfs_previous_item(struct btrfs_root *root,
 	}
 	return 1;
 }
+
+/*
+ * search in extent tree to find a previous Metadata/Data extent item with
+ * min objecitd.
+ *
+ * returns 0 if something is found, 1 if nothing was found and < 0 on error
+ */
+int btrfs_previous_extent_item(struct btrfs_root *root,
+			struct btrfs_path *path, u64 min_objectid)
+{
+	struct btrfs_key found_key;
+	struct extent_buffer *leaf;
+	u32 nritems;
+	int ret;
+
+	while (1) {
+		if (path->slots[0] == 0) {
+			btrfs_set_path_blocking(path);
+			ret = btrfs_prev_leaf(root, path);
+			if (ret != 0)
+				return ret;
+		} else {
+			path->slots[0]--;
+		}
+		leaf = path->nodes[0];
+		nritems = btrfs_header_nritems(leaf);
+		if (nritems == 0)
+			return 1;
+		if (path->slots[0] == nritems)
+			path->slots[0]--;
+
+		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
+		if (found_key.objectid < min_objectid)
+			break;
+		if (found_key.type == BTRFS_EXTENT_ITEM_KEY ||
+		    found_key.type == BTRFS_METADATA_ITEM_KEY)
+			return 0;
+		if (found_key.objectid == min_objectid &&
+		    found_key.type < BTRFS_EXTENT_ITEM_KEY)
+			break;
+	}
+	return 1;
+}

fs/btrfs/ctree.h

@@ -521,9 +521,15 @@ struct btrfs_super_block {
 #define BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF	(1ULL << 6)
 #define BTRFS_FEATURE_INCOMPAT_RAID56		(1ULL << 7)
 #define BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA	(1ULL << 8)
+#define BTRFS_FEATURE_INCOMPAT_NO_HOLES		(1ULL << 9)
 
 #define BTRFS_FEATURE_COMPAT_SUPP		0ULL
+#define BTRFS_FEATURE_COMPAT_SAFE_SET		0ULL
+#define BTRFS_FEATURE_COMPAT_SAFE_CLEAR		0ULL
 #define BTRFS_FEATURE_COMPAT_RO_SUPP		0ULL
+#define BTRFS_FEATURE_COMPAT_RO_SAFE_SET	0ULL
+#define BTRFS_FEATURE_COMPAT_RO_SAFE_CLEAR	0ULL
+
 #define BTRFS_FEATURE_INCOMPAT_SUPP			\
 	(BTRFS_FEATURE_INCOMPAT_MIXED_BACKREF |		\
 	 BTRFS_FEATURE_INCOMPAT_DEFAULT_SUBVOL |	\
@@ -532,7 +538,12 @@ struct btrfs_super_block {
 	 BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO |		\
 	 BTRFS_FEATURE_INCOMPAT_RAID56 |		\
 	 BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF |		\
-	 BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA)
+	 BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA |	\
+	 BTRFS_FEATURE_INCOMPAT_NO_HOLES)
+
+#define BTRFS_FEATURE_INCOMPAT_SAFE_SET			\
+	(BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF)
+#define BTRFS_FEATURE_INCOMPAT_SAFE_CLEAR		0ULL
 
 /*
  * A leaf is full of items. offset and size tell us where to find
@@ -1094,7 +1105,7 @@ struct btrfs_qgroup_limit_item {
 } __attribute__ ((__packed__));
 
 struct btrfs_space_info {
-	u64 flags;
+	spinlock_t lock;
 
 	u64 total_bytes;	/* total bytes in the space,
 				   this doesn't take mirrors into account */
@@ -1104,14 +1115,25 @@ struct btrfs_space_info {
 				   transaction finishes */
 	u64 bytes_reserved;	/* total bytes the allocator has reserved for
 				   current allocations */
-	u64 bytes_readonly;	/* total bytes that are read only */
-
 	u64 bytes_may_use;	/* number of bytes that may be used for
 				   delalloc/allocations */
+	u64 bytes_readonly;	/* total bytes that are read only */
+
+	unsigned int full:1;	/* indicates that we cannot allocate any more
+				   chunks for this space */
+	unsigned int chunk_alloc:1;	/* set if we are allocating a chunk */
+
+	unsigned int flush:1;		/* set if we are trying to make space */
+
+	unsigned int force_alloc;	/* set if we need to force a chunk
+					   alloc for this space */
+
 	u64 disk_used;		/* total bytes used on disk */
 	u64 disk_total;		/* total bytes on disk, takes mirrors into
 				   account */
 
+	u64 flags;
+
 	/*
 	 * bytes_pinned is kept in line with what is actually pinned, as in
 	 * we've called update_block_group and dropped the bytes_used counter
@@ -1124,22 +1146,15 @@ struct btrfs_space_info {
 	 */
 	struct percpu_counter total_bytes_pinned;
 
-	unsigned int full:1;	/* indicates that we cannot allocate any more
-				   chunks for this space */
-	unsigned int chunk_alloc:1;	/* set if we are allocating a chunk */
-
-	unsigned int flush:1;		/* set if we are trying to make space */
-
-	unsigned int force_alloc;	/* set if we need to force a chunk
-					   alloc for this space */
-
 	struct list_head list;
 
+	struct rw_semaphore groups_sem;
 	/* for block groups in our same type */
 	struct list_head block_groups[BTRFS_NR_RAID_TYPES];
-	spinlock_t lock;
-	struct rw_semaphore groups_sem;
 	wait_queue_head_t wait;
+
+	struct kobject kobj;
+	struct kobject block_group_kobjs[BTRFS_NR_RAID_TYPES];
 };
 
 #define	BTRFS_BLOCK_RSV_GLOBAL		1
@@ -1346,6 +1361,7 @@ struct btrfs_fs_info {
 
 	u64 generation;
 	u64 last_trans_committed;
+	u64 avg_delayed_ref_runtime;
 
 	/*
 	 * this is updated to the current trans every time a full commit
@@ -1448,7 +1464,6 @@ struct btrfs_fs_info {
 	spinlock_t tree_mod_seq_lock;
 	atomic64_t tree_mod_seq;
 	struct list_head tree_mod_seq_list;
-	struct seq_list tree_mod_seq_elem;
 
 	/* this protects tree_mod_log */
 	rwlock_t tree_mod_log_lock;
@@ -1515,6 +1530,8 @@ struct btrfs_fs_info {
 	int thread_pool_size;
 
 	struct kobject super_kobj;
+	struct kobject *space_info_kobj;
+	struct kobject *device_dir_kobj;
 	struct completion kobj_unregister;
 	int do_barriers;
 	int closing;
@@ -1643,6 +1660,10 @@ struct btrfs_fs_info {
 	spinlock_t reada_lock;
 	struct radix_tree_root reada_tree;
 
+	/* Extent buffer radix tree */
+	spinlock_t buffer_lock;
+	struct radix_tree_root buffer_radix;
+
 	/* next backup root to be overwritten */
 	int backup_root_index;
 
@@ -1795,6 +1816,12 @@ struct btrfs_root {
 	struct list_head ordered_extents;
 	struct list_head ordered_root;
 	u64 nr_ordered_extents;
+
+	/*
+	 * Number of currently running SEND ioctls to prevent
+	 * manipulation with the read-only status via SUBVOL_SETFLAGS
+	 */
+	int send_in_progress;
 };
 
 struct btrfs_ioctl_defrag_range_args {
@@ -1997,6 +2024,7 @@ struct btrfs_ioctl_defrag_range_args {
 #define BTRFS_MOUNT_CHECK_INTEGRITY_INCLUDING_EXTENT_DATA (1 << 21)
 #define BTRFS_MOUNT_PANIC_ON_FATAL_ERROR	(1 << 22)
 #define BTRFS_MOUNT_RESCAN_UUID_TREE	(1 << 23)
+#define	BTRFS_MOUNT_CHANGE_INODE_CACHE	(1 << 24)
 
 #define BTRFS_DEFAULT_COMMIT_INTERVAL	(30)
 
@@ -2925,6 +2953,10 @@ BTRFS_SETGET_STACK_FUNCS(stack_file_extent_generation,
 			 struct btrfs_file_extent_item, generation, 64);
 BTRFS_SETGET_STACK_FUNCS(stack_file_extent_num_bytes,
 			 struct btrfs_file_extent_item, num_bytes, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_disk_num_bytes,
+			 struct btrfs_file_extent_item, disk_num_bytes, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_compression,
+			 struct btrfs_file_extent_item, compression, 8);
 
 static inline unsigned long
 btrfs_file_extent_inline_start(struct btrfs_file_extent_item *e)
@@ -2958,15 +2990,6 @@ BTRFS_SETGET_FUNCS(file_extent_encryption, struct btrfs_file_extent_item,
 BTRFS_SETGET_FUNCS(file_extent_other_encoding, struct btrfs_file_extent_item,
 		   other_encoding, 16);
 
-/* this returns the number of file bytes represented by the inline item.
- * If an item is compressed, this is the uncompressed size
- */
-static inline u32 btrfs_file_extent_inline_len(struct extent_buffer *eb,
-					       struct btrfs_file_extent_item *e)
-{
-	return btrfs_file_extent_ram_bytes(eb, e);
-}
-
 /*
  * this returns the number of bytes used by the item on disk, minus the
  * size of any extent headers.  If a file is compressed on disk, this is
@@ -2980,6 +3003,32 @@ static inline u32 btrfs_file_extent_inline_item_len(struct extent_buffer *eb,
 	return btrfs_item_size(eb, e) - offset;
 }
 
+/* this returns the number of file bytes represented by the inline item.
+ * If an item is compressed, this is the uncompressed size
+ */
+static inline u32 btrfs_file_extent_inline_len(struct extent_buffer *eb,
+					       int slot,
+					       struct btrfs_file_extent_item *fi)
+{
+	struct btrfs_map_token token;
+
+	btrfs_init_map_token(&token);
+	/*
+	 * return the space used on disk if this item isn't
+	 * compressed or encoded
+	 */
+	if (btrfs_token_file_extent_compression(eb, fi, &token) == 0 &&
+	    btrfs_token_file_extent_encryption(eb, fi, &token) == 0 &&
+	    btrfs_token_file_extent_other_encoding(eb, fi, &token) == 0) {
+		return btrfs_file_extent_inline_item_len(eb,
+							 btrfs_item_nr(slot));
+	}
+
+	/* otherwise use the ram bytes field */
+	return btrfs_token_file_extent_ram_bytes(eb, fi, &token);
+}
+
+
 /* btrfs_dev_stats_item */
 static inline u64 btrfs_dev_stats_value(struct extent_buffer *eb,
 					struct btrfs_dev_stats_item *ptr,
@@ -3143,6 +3192,8 @@ static inline u64 btrfs_calc_trunc_metadata_size(struct btrfs_root *root,
 
 int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans,
 				       struct btrfs_root *root);
+int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans,
+				       struct btrfs_root *root);
 void btrfs_put_block_group(struct btrfs_block_group_cache *cache);
 int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root, unsigned long count);
@@ -3163,6 +3214,7 @@ struct btrfs_block_group_cache *btrfs_lookup_block_group(
 						 struct btrfs_fs_info *info,
 						 u64 bytenr);
 void btrfs_put_block_group(struct btrfs_block_group_cache *cache);
+int get_block_group_index(struct btrfs_block_group_cache *cache);
 struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
 					struct btrfs_root *root, u32 blocksize,
 					u64 parent, u64 root_objectid,
@@ -3301,6 +3353,8 @@ int btrfs_comp_cpu_keys(struct btrfs_key *k1, struct btrfs_key *k2);
 int btrfs_previous_item(struct btrfs_root *root,
 			struct btrfs_path *path, u64 min_objectid,
 			int type);
+int btrfs_previous_extent_item(struct btrfs_root *root,
+			struct btrfs_path *path, u64 min_objectid);
 void btrfs_set_item_key_safe(struct btrfs_root *root, struct btrfs_path *path,
 			     struct btrfs_key *new_key);
 struct extent_buffer *btrfs_root_node(struct btrfs_root *root);
@@ -3350,6 +3404,8 @@ int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
 			 struct btrfs_root *root,
 			 struct btrfs_path *path,
 			 struct btrfs_key *new_key);
+int btrfs_find_item(struct btrfs_root *fs_root, struct btrfs_path *path,
+		u64 inum, u64 ioff, u8 key_type, struct btrfs_key *found_key);
 int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
 		      *root, struct btrfs_key *key, struct btrfs_path *p, int
 		      ins_len, int cow);
@@ -3399,6 +3455,7 @@ static inline int btrfs_insert_empty_item(struct btrfs_trans_handle *trans,
 }
 
 int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path);
+int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path);
 int btrfs_next_old_leaf(struct btrfs_root *root, struct btrfs_path *path,
 			u64 time_seq);
 static inline int btrfs_next_old_item(struct btrfs_root *root,
@@ -3563,12 +3620,6 @@ int btrfs_del_inode_ref(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root,
 			   const char *name, int name_len,
 			   u64 inode_objectid, u64 ref_objectid, u64 *index);
-int btrfs_get_inode_ref_index(struct btrfs_trans_handle *trans,
-			      struct btrfs_root *root,
-			      struct btrfs_path *path,
-			      const char *name, int name_len,
-			      u64 inode_objectid, u64 ref_objectid, int mod,
-			      u64 *ret_index);
 int btrfs_insert_empty_inode(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root,
 			     struct btrfs_path *path, u64 objectid);
@@ -3676,7 +3727,9 @@ int btrfs_start_delalloc_roots(struct btrfs_fs_info *fs_info, int delay_iput);
 int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end,
 			      struct extent_state **cached_state);
 int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *new_root, u64 new_dirid);
+			     struct btrfs_root *new_root,
+			     struct btrfs_root *parent_root,
+			     u64 new_dirid);
 int btrfs_merge_bio_hook(int rw, struct page *page, unsigned long offset,
 			 size_t size, struct bio *bio,
 			 unsigned long bio_flags);
@@ -3745,7 +3798,10 @@ extern const struct file_operations btrfs_file_operations;
 int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
 			 struct btrfs_root *root, struct inode *inode,
 			 struct btrfs_path *path, u64 start, u64 end,
-			 u64 *drop_end, int drop_cache);
+			 u64 *drop_end, int drop_cache,
+			 int replace_extent,
+			 u32 extent_item_size,
+			 int *key_inserted);
 int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 		       struct btrfs_root *root, struct inode *inode, u64 start,
 		       u64 end, int drop_cache);
@@ -3764,6 +3820,8 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
 /* sysfs.c */
 int btrfs_init_sysfs(void);
 void btrfs_exit_sysfs(void);
+int btrfs_sysfs_add_one(struct btrfs_fs_info *fs_info);
+void btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info);
 
 /* xattr.c */
 ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size);
@@ -3796,14 +3854,20 @@ void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...)
 	btrfs_printk(fs_info, KERN_NOTICE fmt, ##args)
 #define btrfs_info(fs_info, fmt, args...) \
 	btrfs_printk(fs_info, KERN_INFO fmt, ##args)
+
+#ifdef DEBUG
 #define btrfs_debug(fs_info, fmt, args...) \
 	btrfs_printk(fs_info, KERN_DEBUG fmt, ##args)
+#else
+#define btrfs_debug(fs_info, fmt, args...) \
+    no_printk(KERN_DEBUG fmt, ##args)
+#endif
 
 #ifdef CONFIG_BTRFS_ASSERT
 
 static inline void assfail(char *expr, char *file, int line)
 {
-	printk(KERN_ERR "BTRFS assertion failed: %s, file: %s, line: %d",
+	pr_err("BTRFS: assertion failed: %s, file: %s, line: %d",
 	       expr, file, line);
 	BUG();
 }
@@ -3841,7 +3905,7 @@ static inline void __btrfs_set_fs_incompat(struct btrfs_fs_info *fs_info,
 		if (!(features & flag)) {
 			features |= flag;
 			btrfs_set_super_incompat_flags(disk_super, features);
-			printk(KERN_INFO "btrfs: setting %llu feature flag\n",
+			btrfs_info(fs_info, "setting %llu feature flag",
 					 flag);
 		}
 		spin_unlock(&fs_info->super_lock);

fs/btrfs/delayed-inode.c

@@ -55,8 +55,7 @@ static inline void btrfs_init_delayed_node(
 	delayed_node->inode_id = inode_id;
 	atomic_set(&delayed_node->refs, 0);
 	delayed_node->count = 0;
-	delayed_node->in_list = 0;
-	delayed_node->inode_dirty = 0;
+	delayed_node->flags = 0;
 	delayed_node->ins_root = RB_ROOT;
 	delayed_node->del_root = RB_ROOT;
 	mutex_init(&delayed_node->mutex);
@@ -172,7 +171,7 @@ static void btrfs_queue_delayed_node(struct btrfs_delayed_root *root,
 				     int mod)
 {
 	spin_lock(&root->lock);
-	if (node->in_list) {
+	if (test_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags)) {
 		if (!list_empty(&node->p_list))
 			list_move_tail(&node->p_list, &root->prepare_list);
 		else if (mod)
@@ -182,7 +181,7 @@ static void btrfs_queue_delayed_node(struct btrfs_delayed_root *root,
 		list_add_tail(&node->p_list, &root->prepare_list);
 		atomic_inc(&node->refs);	/* inserted into list */
 		root->nodes++;
-		node->in_list = 1;
+		set_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags);
 	}
 	spin_unlock(&root->lock);
 }
@@ -192,13 +191,13 @@ static void btrfs_dequeue_delayed_node(struct btrfs_delayed_root *root,
 				       struct btrfs_delayed_node *node)
 {
 	spin_lock(&root->lock);
-	if (node->in_list) {
+	if (test_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags)) {
 		root->nodes--;
 		atomic_dec(&node->refs);	/* not in the list */
 		list_del_init(&node->n_list);
 		if (!list_empty(&node->p_list))
 			list_del_init(&node->p_list);
-		node->in_list = 0;
+		clear_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags);
 	}
 	spin_unlock(&root->lock);
 }
@@ -231,7 +230,8 @@ static struct btrfs_delayed_node *btrfs_next_delayed_node(
 
 	delayed_root = node->root->fs_info->delayed_root;
 	spin_lock(&delayed_root->lock);
-	if (!node->in_list) {	/* not in the list */
+	if (!test_bit(BTRFS_DELAYED_NODE_IN_LIST, &node->flags)) {
+		/* not in the list */
 		if (list_empty(&delayed_root->node_list))
 			goto out;
 		p = delayed_root->node_list.next;
@@ -1004,9 +1004,10 @@ static void btrfs_release_delayed_inode(struct btrfs_delayed_node *delayed_node)
 {
 	struct btrfs_delayed_root *delayed_root;
 
-	if (delayed_node && delayed_node->inode_dirty) {
+	if (delayed_node &&
+	    test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
 		BUG_ON(!delayed_node->root);
-		delayed_node->inode_dirty = 0;
+		clear_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags);
 		delayed_node->count--;
 
 		delayed_root = delayed_node->root->fs_info->delayed_root;
@@ -1014,6 +1015,18 @@ static void btrfs_release_delayed_inode(struct btrfs_delayed_node *delayed_node)
 	}
 }
 
+static void btrfs_release_delayed_iref(struct btrfs_delayed_node *delayed_node)
+{
+	struct btrfs_delayed_root *delayed_root;
+
+	ASSERT(delayed_node->root);
+	clear_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags);
+	delayed_node->count--;
+
+	delayed_root = delayed_node->root->fs_info->delayed_root;
+	finish_one_item(delayed_root);
+}
+
 static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
 					struct btrfs_root *root,
 					struct btrfs_path *path,
@@ -1022,13 +1035,19 @@ static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
 	struct btrfs_key key;
 	struct btrfs_inode_item *inode_item;
 	struct extent_buffer *leaf;
+	int mod;
 	int ret;
 
 	key.objectid = node->inode_id;
 	btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
 	key.offset = 0;
 
-	ret = btrfs_lookup_inode(trans, root, path, &key, 1);
+	if (test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &node->flags))
+		mod = -1;
+	else
+		mod = 1;
+
+	ret = btrfs_lookup_inode(trans, root, path, &key, mod);
 	if (ret > 0) {
 		btrfs_release_path(path);
 		return -ENOENT;
@@ -1036,19 +1055,58 @@ static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
 		return ret;
 	}
 
-	btrfs_unlock_up_safe(path, 1);
 	leaf = path->nodes[0];
 	inode_item = btrfs_item_ptr(leaf, path->slots[0],
 				    struct btrfs_inode_item);
 	write_extent_buffer(leaf, &node->inode_item, (unsigned long)inode_item,
 			    sizeof(struct btrfs_inode_item));
 	btrfs_mark_buffer_dirty(leaf);
-	btrfs_release_path(path);
 
+	if (!test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &node->flags))
+		goto no_iref;
+
+	path->slots[0]++;
+	if (path->slots[0] >= btrfs_header_nritems(leaf))
+		goto search;
+again:
+	btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
+	if (key.objectid != node->inode_id)
+		goto out;
+
+	if (key.type != BTRFS_INODE_REF_KEY &&
+	    key.type != BTRFS_INODE_EXTREF_KEY)
+		goto out;
+
+	/*
+	 * Delayed iref deletion is for the inode who has only one link,
+	 * so there is only one iref. The case that several irefs are
+	 * in the same item doesn't exist.
+	 */
+	btrfs_del_item(trans, root, path);
+out:
+	btrfs_release_delayed_iref(node);
+no_iref:
+	btrfs_release_path(path);
+err_out:
 	btrfs_delayed_inode_release_metadata(root, node);
 	btrfs_release_delayed_inode(node);
 
-	return 0;
+	return ret;
+
+search:
+	btrfs_release_path(path);
+
+	btrfs_set_key_type(&key, BTRFS_INODE_EXTREF_KEY);
+	key.offset = -1;
+	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
+	if (ret < 0)
+		goto err_out;
+	ASSERT(ret);
+
+	ret = 0;
+	leaf = path->nodes[0];
+	path->slots[0]--;
+	goto again;
 }
 
 static inline int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
@@ -1059,7 +1117,7 @@ static inline int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
 	int ret;
 
 	mutex_lock(&node->mutex);
-	if (!node->inode_dirty) {
+	if (!test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &node->flags)) {
 		mutex_unlock(&node->mutex);
 		return 0;
 	}
@@ -1203,7 +1261,7 @@ int btrfs_commit_inode_delayed_inode(struct inode *inode)
 		return 0;
 
 	mutex_lock(&delayed_node->mutex);
-	if (!delayed_node->inode_dirty) {
+	if (!test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
 		mutex_unlock(&delayed_node->mutex);
 		btrfs_release_delayed_node(delayed_node);
 		return 0;
@@ -1227,7 +1285,7 @@ int btrfs_commit_inode_delayed_inode(struct inode *inode)
 	trans->block_rsv = &delayed_node->root->fs_info->delayed_block_rsv;
 
 	mutex_lock(&delayed_node->mutex);
-	if (delayed_node->inode_dirty)
+	if (test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags))
 		ret = __btrfs_update_delayed_inode(trans, delayed_node->root,
 						   path, delayed_node);
 	else
@@ -1300,36 +1358,9 @@ again:
 	trans->block_rsv = &root->fs_info->delayed_block_rsv;
 
 	__btrfs_commit_inode_delayed_items(trans, path, delayed_node);
-	/*
-	 * Maybe new delayed items have been inserted, so we need requeue
-	 * the work. Besides that, we must dequeue the empty delayed nodes
-	 * to avoid the race between delayed items balance and the worker.
-	 * The race like this:
-	 * 	Task1				Worker thread
-	 * 					count == 0, needn't requeue
-	 * 					  also needn't insert the
-	 * 					  delayed node into prepare
-	 * 					  list again.
-	 * 	add lots of delayed items
-	 * 	queue the delayed node
-	 * 	  already in the list,
-	 * 	  and not in the prepare
-	 * 	  list, it means the delayed
-	 * 	  node is being dealt with
-	 * 	  by the worker.
-	 * 	do delayed items balance
-	 * 	  the delayed node is being
-	 * 	  dealt with by the worker
-	 * 	  now, just wait.
-	 * 	  				the worker goto idle.
-	 * Task1 will sleep until the transaction is commited.
-	 */
-	mutex_lock(&delayed_node->mutex);
-	btrfs_dequeue_delayed_node(root->fs_info->delayed_root, delayed_node);
-	mutex_unlock(&delayed_node->mutex);
 
 	trans->block_rsv = block_rsv;
-	btrfs_end_transaction_dmeta(trans, root);
+	btrfs_end_transaction(trans, root);
 	btrfs_btree_balance_dirty_nodelay(root);
 
 release_path:
@@ -1376,52 +1407,41 @@ void btrfs_assert_delayed_root_empty(struct btrfs_root *root)
 	WARN_ON(btrfs_first_delayed_node(delayed_root));
 }
 
-static int refs_newer(struct btrfs_delayed_root *delayed_root,
-		      int seq, int count)
+static int could_end_wait(struct btrfs_delayed_root *delayed_root, int seq)
 {
 	int val = atomic_read(&delayed_root->items_seq);
 
-	if (val < seq || val >= seq + count)
+	if (val < seq || val >= seq + BTRFS_DELAYED_BATCH)
+		return 1;
+
+	if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND)
 		return 1;
+
 	return 0;
 }
 
 void btrfs_balance_delayed_items(struct btrfs_root *root)
 {
 	struct btrfs_delayed_root *delayed_root;
-	int seq;
 
 	delayed_root = btrfs_get_delayed_root(root);
 
 	if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND)
 		return;
 
-	seq = atomic_read(&delayed_root->items_seq);
-
 	if (atomic_read(&delayed_root->items) >= BTRFS_DELAYED_WRITEBACK) {
+		int seq;
 		int ret;
-		DEFINE_WAIT(__wait);
+
+		seq = atomic_read(&delayed_root->items_seq);
 
 		ret = btrfs_wq_run_delayed_node(delayed_root, root, 0);
 		if (ret)
 			return;
 
-		while (1) {
-			prepare_to_wait(&delayed_root->wait, &__wait,
-					TASK_INTERRUPTIBLE);
-
-			if (refs_newer(delayed_root, seq,
-				       BTRFS_DELAYED_BATCH) ||
-			    atomic_read(&delayed_root->items) <
-			    BTRFS_DELAYED_BACKGROUND) {
-				break;
-			}
-			if (!signal_pending(current))
-				schedule();
-			else
-				break;
-		}
-		finish_wait(&delayed_root->wait, &__wait);
+		wait_event_interruptible(delayed_root->wait,
+					 could_end_wait(delayed_root, seq));
+		return;
 	}
 
 	btrfs_wq_run_delayed_node(delayed_root, root, BTRFS_DELAYED_BATCH);
@@ -1472,9 +1492,9 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans,
 	mutex_lock(&delayed_node->mutex);
 	ret = __btrfs_add_delayed_insertion_item(delayed_node, delayed_item);
 	if (unlikely(ret)) {
-		printk(KERN_ERR "err add delayed dir index item(name: %.*s) "
+		btrfs_err(root->fs_info, "err add delayed dir index item(name: %.*s) "
 				"into the insertion tree of the delayed node"
-				"(root id: %llu, inode id: %llu, errno: %d)\n",
+				"(root id: %llu, inode id: %llu, errno: %d)",
 				name_len, name, delayed_node->root->objectid,
 				delayed_node->inode_id, ret);
 		BUG();
@@ -1544,9 +1564,9 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans,
 	mutex_lock(&node->mutex);
 	ret = __btrfs_add_delayed_deletion_item(node, item);
 	if (unlikely(ret)) {
-		printk(KERN_ERR "err add delayed dir index item(index: %llu) "
+		btrfs_err(root->fs_info, "err add delayed dir index item(index: %llu) "
 				"into the deletion tree of the delayed node"
-				"(root id: %llu, inode id: %llu, errno: %d)\n",
+				"(root id: %llu, inode id: %llu, errno: %d)",
 				index, node->root->objectid, node->inode_id,
 				ret);
 		BUG();
@@ -1759,7 +1779,7 @@ int btrfs_fill_inode(struct inode *inode, u32 *rdev)
 		return -ENOENT;
 
 	mutex_lock(&delayed_node->mutex);
-	if (!delayed_node->inode_dirty) {
+	if (!test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
 		mutex_unlock(&delayed_node->mutex);
 		btrfs_release_delayed_node(delayed_node);
 		return -ENOENT;
@@ -1810,7 +1830,7 @@ int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans,
 		return PTR_ERR(delayed_node);
 
 	mutex_lock(&delayed_node->mutex);
-	if (delayed_node->inode_dirty) {
+	if (test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
 		fill_stack_inode_item(trans, &delayed_node->inode_item, inode);
 		goto release_node;
 	}
@@ -1821,7 +1841,7 @@ int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans,
 		goto release_node;
 
 	fill_stack_inode_item(trans, &delayed_node->inode_item, inode);
-	delayed_node->inode_dirty = 1;
+	set_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags);
 	delayed_node->count++;
 	atomic_inc(&root->fs_info->delayed_root->items);
 release_node:
@@ -1830,6 +1850,41 @@ release_node:
 	return ret;
 }
 
+int btrfs_delayed_delete_inode_ref(struct inode *inode)
+{
+	struct btrfs_delayed_node *delayed_node;
+
+	delayed_node = btrfs_get_or_create_delayed_node(inode);
+	if (IS_ERR(delayed_node))
+		return PTR_ERR(delayed_node);
+
+	/*
+	 * We don't reserve space for inode ref deletion is because:
+	 * - We ONLY do async inode ref deletion for the inode who has only
+	 *   one link(i_nlink == 1), it means there is only one inode ref.
+	 *   And in most case, the inode ref and the inode item are in the
+	 *   same leaf, and we will deal with them at the same time.
+	 *   Since we are sure we will reserve the space for the inode item,
+	 *   it is unnecessary to reserve space for inode ref deletion.
+	 * - If the inode ref and the inode item are not in the same leaf,
+	 *   We also needn't worry about enospc problem, because we reserve
+	 *   much more space for the inode update than it needs.
+	 * - At the worst, we can steal some space from the global reservation.
+	 *   It is very rare.
+	 */
+	mutex_lock(&delayed_node->mutex);
+	if (test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags))
+		goto release_node;
+
+	set_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags);
+	delayed_node->count++;
+	atomic_inc(&BTRFS_I(inode)->root->fs_info->delayed_root->items);
+release_node:
+	mutex_unlock(&delayed_node->mutex);
+	btrfs_release_delayed_node(delayed_node);
+	return 0;
+}
+
 static void __btrfs_kill_delayed_node(struct btrfs_delayed_node *delayed_node)
 {
 	struct btrfs_root *root = delayed_node->root;
@@ -1852,7 +1907,10 @@ static void __btrfs_kill_delayed_node(struct btrfs_delayed_node *delayed_node)
 		btrfs_release_delayed_item(prev_item);
 	}
 
-	if (delayed_node->inode_dirty) {
+	if (test_bit(BTRFS_DELAYED_NODE_DEL_IREF, &delayed_node->flags))
+		btrfs_release_delayed_iref(delayed_node);
+
+	if (test_bit(BTRFS_DELAYED_NODE_INODE_DIRTY, &delayed_node->flags)) {
 		btrfs_delayed_inode_release_metadata(root, delayed_node);
 		btrfs_release_delayed_inode(delayed_node);
 	}

fs/btrfs/delayed-inode.h

@@ -48,6 +48,10 @@ struct btrfs_delayed_root {
 	wait_queue_head_t wait;
 };
 
+#define BTRFS_DELAYED_NODE_IN_LIST	0
+#define BTRFS_DELAYED_NODE_INODE_DIRTY	1
+#define BTRFS_DELAYED_NODE_DEL_IREF	2
+
 struct btrfs_delayed_node {
 	u64 inode_id;
 	u64 bytes_reserved;
@@ -65,8 +69,7 @@ struct btrfs_delayed_node {
 	struct btrfs_inode_item inode_item;
 	atomic_t refs;
 	u64 index_cnt;
-	bool in_list;
-	bool inode_dirty;
+	unsigned long flags;
 	int count;
 };
 
@@ -125,6 +128,7 @@ int btrfs_commit_inode_delayed_inode(struct inode *inode);
 int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root, struct inode *inode);
 int btrfs_fill_inode(struct inode *inode, u32 *rdev);
+int btrfs_delayed_delete_inode_ref(struct inode *inode);
 
 /* Used for drop dead root */
 void btrfs_kill_all_delayed_nodes(struct btrfs_root *root);

fs/btrfs/delayed-ref.c

@@ -161,35 +161,61 @@ static struct btrfs_delayed_ref_node *tree_insert(struct rb_root *root,
 	return NULL;
 }
 
+/* insert a new ref to head ref rbtree */
+static struct btrfs_delayed_ref_head *htree_insert(struct rb_root *root,
+						   struct rb_node *node)
+{
+	struct rb_node **p = &root->rb_node;
+	struct rb_node *parent_node = NULL;
+	struct btrfs_delayed_ref_head *entry;
+	struct btrfs_delayed_ref_head *ins;
+	u64 bytenr;
+
+	ins = rb_entry(node, struct btrfs_delayed_ref_head, href_node);
+	bytenr = ins->node.bytenr;
+	while (*p) {
+		parent_node = *p;
+		entry = rb_entry(parent_node, struct btrfs_delayed_ref_head,
+				 href_node);
+
+		if (bytenr < entry->node.bytenr)
+			p = &(*p)->rb_left;
+		else if (bytenr > entry->node.bytenr)
+			p = &(*p)->rb_right;
+		else
+			return entry;
+	}
+
+	rb_link_node(node, parent_node, p);
+	rb_insert_color(node, root);
+	return NULL;
+}
+
 /*
  * find an head entry based on bytenr. This returns the delayed ref
  * head if it was able to find one, or NULL if nothing was in that spot.
  * If return_bigger is given, the next bigger entry is returned if no exact
  * match is found.
  */
-static struct btrfs_delayed_ref_node *find_ref_head(struct rb_root *root,
-				  u64 bytenr,
-				  struct btrfs_delayed_ref_node **last,
-				  int return_bigger)
+static struct btrfs_delayed_ref_head *
+find_ref_head(struct rb_root *root, u64 bytenr,
+	      struct btrfs_delayed_ref_head **last, int return_bigger)
 {
 	struct rb_node *n;
-	struct btrfs_delayed_ref_node *entry;
+	struct btrfs_delayed_ref_head *entry;
 	int cmp = 0;
 
 again:
 	n = root->rb_node;
 	entry = NULL;
 	while (n) {
-		entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
-		WARN_ON(!entry->in_tree);
+		entry = rb_entry(n, struct btrfs_delayed_ref_head, href_node);
 		if (last)
 			*last = entry;
 
-		if (bytenr < entry->bytenr)
+		if (bytenr < entry->node.bytenr)
 			cmp = -1;
-		else if (bytenr > entry->bytenr)
-			cmp = 1;
-		else if (!btrfs_delayed_ref_is_head(entry))
+		else if (bytenr > entry->node.bytenr)
 			cmp = 1;
 		else
 			cmp = 0;
@@ -203,12 +229,12 @@ again:
 	}
 	if (entry && return_bigger) {
 		if (cmp > 0) {
-			n = rb_next(&entry->rb_node);
+			n = rb_next(&entry->href_node);
 			if (!n)
 				n = rb_first(root);
-			entry = rb_entry(n, struct btrfs_delayed_ref_node,
-					 rb_node);
-			bytenr = entry->bytenr;
+			entry = rb_entry(n, struct btrfs_delayed_ref_head,
+					 href_node);
+			bytenr = entry->node.bytenr;
 			return_bigger = 0;
 			goto again;
 		}
@@ -243,33 +269,38 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans,
 
 static inline void drop_delayed_ref(struct btrfs_trans_handle *trans,
 				    struct btrfs_delayed_ref_root *delayed_refs,
+				    struct btrfs_delayed_ref_head *head,
 				    struct btrfs_delayed_ref_node *ref)
 {
-	rb_erase(&ref->rb_node, &delayed_refs->root);
+	if (btrfs_delayed_ref_is_head(ref)) {
+		head = btrfs_delayed_node_to_head(ref);
+		rb_erase(&head->href_node, &delayed_refs->href_root);
+	} else {
+		assert_spin_locked(&head->lock);
+		rb_erase(&ref->rb_node, &head->ref_root);
+	}
 	ref->in_tree = 0;
 	btrfs_put_delayed_ref(ref);
-	delayed_refs->num_entries--;
+	atomic_dec(&delayed_refs->num_entries);
 	if (trans->delayed_ref_updates)
 		trans->delayed_ref_updates--;
 }
 
 static int merge_ref(struct btrfs_trans_handle *trans,
 		     struct btrfs_delayed_ref_root *delayed_refs,
+		     struct btrfs_delayed_ref_head *head,
 		     struct btrfs_delayed_ref_node *ref, u64 seq)
 {
 	struct rb_node *node;
-	int merged = 0;
 	int mod = 0;
 	int done = 0;
 
-	node = rb_prev(&ref->rb_node);
-	while (node) {
+	node = rb_next(&ref->rb_node);
+	while (!done && node) {
 		struct btrfs_delayed_ref_node *next;
 
 		next = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-		node = rb_prev(node);
-		if (next->bytenr != ref->bytenr)
-			break;
+		node = rb_next(node);
 		if (seq && next->seq >= seq)
 			break;
 		if (comp_entry(ref, next, 0))
@@ -289,12 +320,11 @@ static int merge_ref(struct btrfs_trans_handle *trans,
 			mod = -next->ref_mod;
 		}
 
-		merged++;
-		drop_delayed_ref(trans, delayed_refs, next);
+		drop_delayed_ref(trans, delayed_refs, head, next);
 		ref->ref_mod += mod;
 		if (ref->ref_mod == 0) {
-			drop_delayed_ref(trans, delayed_refs, ref);
-			break;
+			drop_delayed_ref(trans, delayed_refs, head, ref);
+			done = 1;
 		} else {
 			/*
 			 * You can't have multiples of the same ref on a tree
@@ -303,13 +333,8 @@ static int merge_ref(struct btrfs_trans_handle *trans,
 			WARN_ON(ref->type == BTRFS_TREE_BLOCK_REF_KEY ||
 				ref->type == BTRFS_SHARED_BLOCK_REF_KEY);
 		}
-
-		if (done)
-			break;
-		node = rb_prev(&ref->rb_node);
 	}
-
-	return merged;
+	return done;
 }
 
 void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans,
@@ -320,6 +345,14 @@ void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans,
 	struct rb_node *node;
 	u64 seq = 0;
 
+	assert_spin_locked(&head->lock);
+	/*
+	 * We don't have too much refs to merge in the case of delayed data
+	 * refs.
+	 */
+	if (head->is_data)
+		return;
+
 	spin_lock(&fs_info->tree_mod_seq_lock);
 	if (!list_empty(&fs_info->tree_mod_seq_list)) {
 		struct seq_list *elem;
@@ -330,22 +363,19 @@ void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans,
 	}
 	spin_unlock(&fs_info->tree_mod_seq_lock);
 
-	node = rb_prev(&head->node.rb_node);
+	node = rb_first(&head->ref_root);
 	while (node) {
 		struct btrfs_delayed_ref_node *ref;
 
 		ref = rb_entry(node, struct btrfs_delayed_ref_node,
 			       rb_node);
-		if (ref->bytenr != head->node.bytenr)
-			break;
-
 		/* We can't merge refs that are outside of our seq count */
 		if (seq && ref->seq >= seq)
 			break;
-		if (merge_ref(trans, delayed_refs, ref, seq))
-			node = rb_prev(&head->node.rb_node);
+		if (merge_ref(trans, delayed_refs, head, ref, seq))
+			node = rb_first(&head->ref_root);
 		else
-			node = rb_prev(node);
+			node = rb_next(&ref->rb_node);
 	}
 }
 
@@ -373,71 +403,52 @@ int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,
 	return ret;
 }
 
-int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans,
-			   struct list_head *cluster, u64 start)
+struct btrfs_delayed_ref_head *
+btrfs_select_ref_head(struct btrfs_trans_handle *trans)
 {
-	int count = 0;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	struct rb_node *node;
-	struct btrfs_delayed_ref_node *ref;
 	struct btrfs_delayed_ref_head *head;
+	u64 start;
+	bool loop = false;
 
 	delayed_refs = &trans->transaction->delayed_refs;
-	if (start == 0) {
-		node = rb_first(&delayed_refs->root);
-	} else {
-		ref = NULL;
-		find_ref_head(&delayed_refs->root, start + 1, &ref, 1);
-		if (ref) {
-			node = &ref->rb_node;
-		} else
-			node = rb_first(&delayed_refs->root);
-	}
+
 again:
-	while (node && count < 32) {
-		ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-		if (btrfs_delayed_ref_is_head(ref)) {
-			head = btrfs_delayed_node_to_head(ref);
-			if (list_empty(&head->cluster)) {
-				list_add_tail(&head->cluster, cluster);
-				delayed_refs->run_delayed_start =
-					head->node.bytenr;
-				count++;
-
-				WARN_ON(delayed_refs->num_heads_ready == 0);
-				delayed_refs->num_heads_ready--;
-			} else if (count) {
-				/* the goal of the clustering is to find extents
-				 * that are likely to end up in the same extent
-				 * leaf on disk.  So, we don't want them spread
-				 * all over the tree.  Stop now if we've hit
-				 * a head that was already in use
-				 */
-				break;
-			}
-		}
-		node = rb_next(node);
-	}
-	if (count) {
-		return 0;
-	} else if (start) {
-		/*
-		 * we've gone to the end of the rbtree without finding any
-		 * clusters.  start from the beginning and try again
-		 */
+	start = delayed_refs->run_delayed_start;
+	head = find_ref_head(&delayed_refs->href_root, start, NULL, 1);
+	if (!head && !loop) {
+		delayed_refs->run_delayed_start = 0;
 		start = 0;
-		node = rb_first(&delayed_refs->root);
-		goto again;
+		loop = true;
+		head = find_ref_head(&delayed_refs->href_root, start, NULL, 1);
+		if (!head)
+			return NULL;
+	} else if (!head && loop) {
+		return NULL;
 	}
-	return 1;
-}
 
-void btrfs_release_ref_cluster(struct list_head *cluster)
-{
-	struct list_head *pos, *q;
+	while (head->processing) {
+		struct rb_node *node;
+
+		node = rb_next(&head->href_node);
+		if (!node) {
+			if (loop)
+				return NULL;
+			delayed_refs->run_delayed_start = 0;
+			start = 0;
+			loop = true;
+			goto again;
+		}
+		head = rb_entry(node, struct btrfs_delayed_ref_head,
+				href_node);
+	}
 
-	list_for_each_safe(pos, q, cluster)
-		list_del_init(pos);
+	head->processing = 1;
+	WARN_ON(delayed_refs->num_heads_ready == 0);
+	delayed_refs->num_heads_ready--;
+	delayed_refs->run_delayed_start = head->node.bytenr +
+		head->node.num_bytes;
+	return head;
 }
 
 /*
@@ -451,6 +462,7 @@ void btrfs_release_ref_cluster(struct list_head *cluster)
 static noinline void
 update_existing_ref(struct btrfs_trans_handle *trans,
 		    struct btrfs_delayed_ref_root *delayed_refs,
+		    struct btrfs_delayed_ref_head *head,
 		    struct btrfs_delayed_ref_node *existing,
 		    struct btrfs_delayed_ref_node *update)
 {
@@ -463,7 +475,7 @@ update_existing_ref(struct btrfs_trans_handle *trans,
 		 */
 		existing->ref_mod--;
 		if (existing->ref_mod == 0)
-			drop_delayed_ref(trans, delayed_refs, existing);
+			drop_delayed_ref(trans, delayed_refs, head, existing);
 		else
 			WARN_ON(existing->type == BTRFS_TREE_BLOCK_REF_KEY ||
 				existing->type == BTRFS_SHARED_BLOCK_REF_KEY);
@@ -533,9 +545,13 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing,
 		}
 	}
 	/*
-	 * update the reference mod on the head to reflect this new operation
+	 * update the reference mod on the head to reflect this new operation,
+	 * only need the lock for this case cause we could be processing it
+	 * currently, for refs we just added we know we're a-ok.
 	 */
+	spin_lock(&existing_ref->lock);
 	existing->ref_mod += update->ref_mod;
+	spin_unlock(&existing_ref->lock);
 }
 
 /*
@@ -543,13 +559,13 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing,
  * this does all the dirty work in terms of maintaining the correct
  * overall modification count.
  */
-static noinline void add_delayed_ref_head(struct btrfs_fs_info *fs_info,
-					struct btrfs_trans_handle *trans,
-					struct btrfs_delayed_ref_node *ref,
-					u64 bytenr, u64 num_bytes,
-					int action, int is_data)
+static noinline struct btrfs_delayed_ref_head *
+add_delayed_ref_head(struct btrfs_fs_info *fs_info,
+		     struct btrfs_trans_handle *trans,
+		     struct btrfs_delayed_ref_node *ref, u64 bytenr,
+		     u64 num_bytes, int action, int is_data)
 {
-	struct btrfs_delayed_ref_node *existing;
+	struct btrfs_delayed_ref_head *existing;
 	struct btrfs_delayed_ref_head *head_ref = NULL;
 	struct btrfs_delayed_ref_root *delayed_refs;
 	int count_mod = 1;
@@ -596,38 +612,43 @@ static noinline void add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	head_ref = btrfs_delayed_node_to_head(ref);
 	head_ref->must_insert_reserved = must_insert_reserved;
 	head_ref->is_data = is_data;
+	head_ref->ref_root = RB_ROOT;
+	head_ref->processing = 0;
 
-	INIT_LIST_HEAD(&head_ref->cluster);
+	spin_lock_init(&head_ref->lock);
 	mutex_init(&head_ref->mutex);
 
 	trace_add_delayed_ref_head(ref, head_ref, action);
 
-	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
-
+	existing = htree_insert(&delayed_refs->href_root,
+				&head_ref->href_node);
 	if (existing) {
-		update_existing_head_ref(existing, ref);
+		update_existing_head_ref(&existing->node, ref);
 		/*
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
 		kmem_cache_free(btrfs_delayed_ref_head_cachep, head_ref);
+		head_ref = existing;
 	} else {
 		delayed_refs->num_heads++;
 		delayed_refs->num_heads_ready++;
-		delayed_refs->num_entries++;
+		atomic_inc(&delayed_refs->num_entries);
 		trans->delayed_ref_updates++;
 	}
+	return head_ref;
 }
 
 /*
  * helper to insert a delayed tree ref into the rbtree.
  */
-static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
-					 struct btrfs_trans_handle *trans,
-					 struct btrfs_delayed_ref_node *ref,
-					 u64 bytenr, u64 num_bytes, u64 parent,
-					 u64 ref_root, int level, int action,
-					 int for_cow)
+static noinline void
+add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
+		     struct btrfs_trans_handle *trans,
+		     struct btrfs_delayed_ref_head *head_ref,
+		     struct btrfs_delayed_ref_node *ref, u64 bytenr,
+		     u64 num_bytes, u64 parent, u64 ref_root, int level,
+		     int action, int for_cow)
 {
 	struct btrfs_delayed_ref_node *existing;
 	struct btrfs_delayed_tree_ref *full_ref;
@@ -663,30 +684,33 @@ static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 
 	trace_add_delayed_tree_ref(ref, full_ref, action);
 
-	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
-
+	spin_lock(&head_ref->lock);
+	existing = tree_insert(&head_ref->ref_root, &ref->rb_node);
 	if (existing) {
-		update_existing_ref(trans, delayed_refs, existing, ref);
+		update_existing_ref(trans, delayed_refs, head_ref, existing,
+				    ref);
 		/*
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
 		kmem_cache_free(btrfs_delayed_tree_ref_cachep, full_ref);
 	} else {
-		delayed_refs->num_entries++;
+		atomic_inc(&delayed_refs->num_entries);
 		trans->delayed_ref_updates++;
 	}
+	spin_unlock(&head_ref->lock);
 }
 
 /*
  * helper to insert a delayed data ref into the rbtree.
  */
-static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info,
-					 struct btrfs_trans_handle *trans,
-					 struct btrfs_delayed_ref_node *ref,
-					 u64 bytenr, u64 num_bytes, u64 parent,
-					 u64 ref_root, u64 owner, u64 offset,
-					 int action, int for_cow)
+static noinline void
+add_delayed_data_ref(struct btrfs_fs_info *fs_info,
+		     struct btrfs_trans_handle *trans,
+		     struct btrfs_delayed_ref_head *head_ref,
+		     struct btrfs_delayed_ref_node *ref, u64 bytenr,
+		     u64 num_bytes, u64 parent, u64 ref_root, u64 owner,
+		     u64 offset, int action, int for_cow)
 {
 	struct btrfs_delayed_ref_node *existing;
 	struct btrfs_delayed_data_ref *full_ref;
@@ -724,19 +748,21 @@ static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 
 	trace_add_delayed_data_ref(ref, full_ref, action);
 
-	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
-
+	spin_lock(&head_ref->lock);
+	existing = tree_insert(&head_ref->ref_root, &ref->rb_node);
 	if (existing) {
-		update_existing_ref(trans, delayed_refs, existing, ref);
+		update_existing_ref(trans, delayed_refs, head_ref, existing,
+				    ref);
 		/*
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
 		kmem_cache_free(btrfs_delayed_data_ref_cachep, full_ref);
 	} else {
-		delayed_refs->num_entries++;
+		atomic_inc(&delayed_refs->num_entries);
 		trans->delayed_ref_updates++;
 	}
+	spin_unlock(&head_ref->lock);
 }
 
 /*
@@ -775,10 +801,10 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 	 * insert both the head node and the new ref without dropping
 	 * the spin lock
 	 */
-	add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
-				   num_bytes, action, 0);
+	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node,
+					bytenr, num_bytes, action, 0);
 
-	add_delayed_tree_ref(fs_info, trans, &ref->node, bytenr,
+	add_delayed_tree_ref(fs_info, trans, head_ref, &ref->node, bytenr,
 				   num_bytes, parent, ref_root, level, action,
 				   for_cow);
 	spin_unlock(&delayed_refs->lock);
@@ -823,10 +849,10 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	 * insert both the head node and the new ref without dropping
 	 * the spin lock
 	 */
-	add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
-				   num_bytes, action, 1);
+	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node,
+					bytenr, num_bytes, action, 1);
 
-	add_delayed_data_ref(fs_info, trans, &ref->node, bytenr,
+	add_delayed_data_ref(fs_info, trans, head_ref, &ref->node, bytenr,
 				   num_bytes, parent, ref_root, owner, offset,
 				   action, for_cow);
 	spin_unlock(&delayed_refs->lock);
@@ -869,14 +895,10 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 struct btrfs_delayed_ref_head *
 btrfs_find_delayed_ref_head(struct btrfs_trans_handle *trans, u64 bytenr)
 {
-	struct btrfs_delayed_ref_node *ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
 
 	delayed_refs = &trans->transaction->delayed_refs;
-	ref = find_ref_head(&delayed_refs->root, bytenr, NULL, 0);
-	if (ref)
-		return btrfs_delayed_node_to_head(ref);
-	return NULL;
+	return find_ref_head(&delayed_refs->href_root, bytenr, NULL, 0);
 }
 
 void btrfs_delayed_ref_exit(void)

fs/btrfs/delayed-ref.h

@@ -81,7 +81,10 @@ struct btrfs_delayed_ref_head {
 	 */
 	struct mutex mutex;
 
-	struct list_head cluster;
+	spinlock_t lock;
+	struct rb_root ref_root;
+
+	struct rb_node href_node;
 
 	struct btrfs_delayed_extent_op *extent_op;
 	/*
@@ -98,6 +101,7 @@ struct btrfs_delayed_ref_head {
 	 */
 	unsigned int must_insert_reserved:1;
 	unsigned int is_data:1;
+	unsigned int processing:1;
 };
 
 struct btrfs_delayed_tree_ref {
@@ -116,7 +120,8 @@ struct btrfs_delayed_data_ref {
 };
 
 struct btrfs_delayed_ref_root {
-	struct rb_root root;
+	/* head ref rbtree */
+	struct rb_root href_root;
 
 	/* this spin lock protects the rbtree and the entries inside */
 	spinlock_t lock;
@@ -124,7 +129,7 @@ struct btrfs_delayed_ref_root {
 	/* how many delayed ref updates we've queued, used by the
 	 * throttling code
 	 */
-	unsigned long num_entries;
+	atomic_t num_entries;
 
 	/* total number of head nodes in tree */
 	unsigned long num_heads;
@@ -132,15 +137,6 @@ struct btrfs_delayed_ref_root {
 	/* total number of head nodes ready for processing */
 	unsigned long num_heads_ready;
 
-	/*
-	 * bumped when someone is making progress on the delayed
-	 * refs, so that other procs know they are just adding to
-	 * contention intead of helping
-	 */
-	atomic_t procs_running_refs;
-	atomic_t ref_seq;
-	wait_queue_head_t wait;
-
 	/*
 	 * set when the tree is flushing before a transaction commit,
 	 * used by the throttling code to decide if new updates need
@@ -226,9 +222,9 @@ static inline void btrfs_delayed_ref_unlock(struct btrfs_delayed_ref_head *head)
 	mutex_unlock(&head->mutex);
 }
 
-int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans,
-			   struct list_head *cluster, u64 search_start);
-void btrfs_release_ref_cluster(struct list_head *cluster);
+
+struct btrfs_delayed_ref_head *
+btrfs_select_ref_head(struct btrfs_trans_handle *trans);
 
 int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,
 			    struct btrfs_delayed_ref_root *delayed_refs,

fs/btrfs/dev-replace.c

@@ -102,7 +102,8 @@ no_valid_dev_replace_entry_found:
 	ptr = btrfs_item_ptr(eb, slot, struct btrfs_dev_replace_item);
 
 	if (item_size != sizeof(struct btrfs_dev_replace_item)) {
-		pr_warn("btrfs: dev_replace entry found has unexpected size, ignore entry\n");
+		btrfs_warn(fs_info,
+			"dev_replace entry found has unexpected size, ignore entry");
 		goto no_valid_dev_replace_entry_found;
 	}
 
@@ -145,13 +146,19 @@ no_valid_dev_replace_entry_found:
 		if (!dev_replace->srcdev &&
 		    !btrfs_test_opt(dev_root, DEGRADED)) {
 			ret = -EIO;
-			pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "srcdev (devid %llu) is missing, need to run 'btrfs dev scan'?\n",
-				src_devid);
+			btrfs_warn(fs_info,
+			   "cannot mount because device replace operation is ongoing and");
+			btrfs_warn(fs_info,
+			   "srcdev (devid %llu) is missing, need to run 'btrfs dev scan'?",
+			   src_devid);
 		}
 		if (!dev_replace->tgtdev &&
 		    !btrfs_test_opt(dev_root, DEGRADED)) {
 			ret = -EIO;
-			pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "tgtdev (devid %llu) is missing, need to run btrfs dev scan?\n",
+			btrfs_warn(fs_info,
+			   "cannot mount because device replace operation is ongoing and");
+			btrfs_warn(fs_info,
+			   "tgtdev (devid %llu) is missing, need to run 'btrfs dev scan'?",
 				BTRFS_DEV_REPLACE_DEVID);
 		}
 		if (dev_replace->tgtdev) {
@@ -210,7 +217,7 @@ int btrfs_run_dev_replace(struct btrfs_trans_handle *trans,
 	}
 	ret = btrfs_search_slot(trans, dev_root, &key, path, -1, 1);
 	if (ret < 0) {
-		pr_warn("btrfs: error %d while searching for dev_replace item!\n",
+		btrfs_warn(fs_info, "error %d while searching for dev_replace item!",
 			ret);
 		goto out;
 	}
@@ -230,7 +237,7 @@ int btrfs_run_dev_replace(struct btrfs_trans_handle *trans,
 		 */
 		ret = btrfs_del_item(trans, dev_root, path);
 		if (ret != 0) {
-			pr_warn("btrfs: delete too small dev_replace item failed %d!\n",
+			btrfs_warn(fs_info, "delete too small dev_replace item failed %d!",
 				ret);
 			goto out;
 		}
@@ -243,7 +250,7 @@ int btrfs_run_dev_replace(struct btrfs_trans_handle *trans,
 		ret = btrfs_insert_empty_item(trans, dev_root, path,
 					      &key, sizeof(*ptr));
 		if (ret < 0) {
-			pr_warn("btrfs: insert dev_replace item failed %d!\n",
+			btrfs_warn(fs_info, "insert dev_replace item failed %d!",
 				ret);
 			goto out;
 		}
@@ -305,7 +312,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	struct btrfs_device *src_device = NULL;
 
 	if (btrfs_fs_incompat(fs_info, RAID56)) {
-		pr_warn("btrfs: dev_replace cannot yet handle RAID5/RAID6\n");
+		btrfs_warn(fs_info, "dev_replace cannot yet handle RAID5/RAID6");
 		return -EINVAL;
 	}
 
@@ -325,7 +332,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	ret = btrfs_init_dev_replace_tgtdev(root, args->start.tgtdev_name,
 					    &tgt_device);
 	if (ret) {
-		pr_err("btrfs: target device %s is invalid!\n",
+		btrfs_err(fs_info, "target device %s is invalid!",
 		       args->start.tgtdev_name);
 		mutex_unlock(&fs_info->volume_mutex);
 		return -EINVAL;
@@ -341,7 +348,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	}
 
 	if (tgt_device->total_bytes < src_device->total_bytes) {
-		pr_err("btrfs: target device is smaller than source device!\n");
+		btrfs_err(fs_info, "target device is smaller than source device!");
 		ret = -EINVAL;
 		goto leave_no_lock;
 	}
@@ -366,7 +373,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	dev_replace->tgtdev = tgt_device;
 
 	printk_in_rcu(KERN_INFO
-		      "btrfs: dev_replace from %s (devid %llu) to %s started\n",
+		      "BTRFS: dev_replace from %s (devid %llu) to %s started\n",
 		      src_device->missing ? "<missing disk>" :
 		        rcu_str_deref(src_device->name),
 		      src_device->devid,
@@ -489,7 +496,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 
 	if (scrub_ret) {
 		printk_in_rcu(KERN_ERR
-			      "btrfs: btrfs_scrub_dev(%s, %llu, %s) failed %d\n",
+			      "BTRFS: btrfs_scrub_dev(%s, %llu, %s) failed %d\n",
 			      src_device->missing ? "<missing disk>" :
 			        rcu_str_deref(src_device->name),
 			      src_device->devid,
@@ -504,7 +511,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	}
 
 	printk_in_rcu(KERN_INFO
-		      "btrfs: dev_replace from %s (devid %llu) to %s) finished\n",
+		      "BTRFS: dev_replace from %s (devid %llu) to %s) finished\n",
 		      src_device->missing ? "<missing disk>" :
 		        rcu_str_deref(src_device->name),
 		      src_device->devid,
@@ -699,7 +706,7 @@ void btrfs_dev_replace_suspend_for_unmount(struct btrfs_fs_info *fs_info)
 			BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED;
 		dev_replace->time_stopped = get_seconds();
 		dev_replace->item_needs_writeback = 1;
-		pr_info("btrfs: suspending dev_replace for unmount\n");
+		btrfs_info(fs_info, "suspending dev_replace for unmount");
 		break;
 	}
 
@@ -728,8 +735,9 @@ int btrfs_resume_dev_replace_async(struct btrfs_fs_info *fs_info)
 		break;
 	}
 	if (!dev_replace->tgtdev || !dev_replace->tgtdev->bdev) {
-		pr_info("btrfs: cannot continue dev_replace, tgtdev is missing\n"
-			"btrfs: you may cancel the operation after 'mount -o degraded'\n");
+		btrfs_info(fs_info, "cannot continue dev_replace, tgtdev is missing");
+		btrfs_info(fs_info,
+			"you may cancel the operation after 'mount -o degraded'");
 		btrfs_dev_replace_unlock(dev_replace);
 		return 0;
 	}
@@ -755,14 +763,14 @@ static int btrfs_dev_replace_kthread(void *data)
 		kfree(status_args);
 		do_div(progress, 10);
 		printk_in_rcu(KERN_INFO
-			      "btrfs: continuing dev_replace from %s (devid %llu) to %s @%u%%\n",
-			      dev_replace->srcdev->missing ? "<missing disk>" :
-				rcu_str_deref(dev_replace->srcdev->name),
-			      dev_replace->srcdev->devid,
-			      dev_replace->tgtdev ?
-				rcu_str_deref(dev_replace->tgtdev->name) :
-				"<missing target disk>",
-			      (unsigned int)progress);
+			"BTRFS: continuing dev_replace from %s (devid %llu) to %s @%u%%\n",
+			dev_replace->srcdev->missing ? "<missing disk>" :
+			rcu_str_deref(dev_replace->srcdev->name),
+			dev_replace->srcdev->devid,
+			dev_replace->tgtdev ?
+			rcu_str_deref(dev_replace->tgtdev->name) :
+			"<missing target disk>",
+			(unsigned int)progress);
 	}
 	btrfs_dev_replace_continue_on_mount(fs_info);
 	atomic_set(&fs_info->mutually_exclusive_operation_running, 0);

fs/btrfs/dir-item.c

@@ -261,7 +261,7 @@ int btrfs_check_dir_item_collision(struct btrfs_root *root, u64 dir,
 	 * see if there is room in the item to insert this
 	 * name
 	 */
-	data_size = sizeof(*di) + name_len + sizeof(struct btrfs_item);
+	data_size = sizeof(*di) + name_len;
 	leaf = path->nodes[0];
 	slot = path->slots[0];
 	if (data_size + btrfs_item_size_nr(leaf, slot) +
@@ -459,7 +459,7 @@ int verify_dir_item(struct btrfs_root *root,
 	u8 type = btrfs_dir_type(leaf, dir_item);
 
 	if (type >= BTRFS_FT_MAX) {
-		printk(KERN_CRIT "btrfs: invalid dir item type: %d\n",
+		btrfs_crit(root->fs_info, "invalid dir item type: %d",
 		       (int)type);
 		return 1;
 	}
@@ -468,7 +468,7 @@ 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",
+		btrfs_crit(root->fs_info, "invalid dir item name len: %u",
 		       (unsigned)btrfs_dir_data_len(leaf, dir_item));
 		return 1;
 	}
@@ -476,7 +476,7 @@ int verify_dir_item(struct btrfs_root *root,
 	/* BTRFS_MAX_XATTR_SIZE is the same for all dir items */
 	if ((btrfs_dir_data_len(leaf, dir_item) +
 	     btrfs_dir_name_len(leaf, dir_item)) > BTRFS_MAX_XATTR_SIZE(root)) {
-		printk(KERN_CRIT "btrfs: invalid dir item name + data len: %u + %u\n",
+		btrfs_crit(root->fs_info, "invalid dir item name + data len: %u + %u",
 		       (unsigned)btrfs_dir_name_len(leaf, dir_item),
 		       (unsigned)btrfs_dir_data_len(leaf, dir_item));
 		return 1;

fs/btrfs/disk-io.c

@@ -48,6 +48,7 @@
 #include "rcu-string.h"
 #include "dev-replace.h"
 #include "raid56.h"
+#include "sysfs.h"
 
 #ifdef CONFIG_X86
 #include <asm/cpufeature.h>
@@ -299,11 +300,11 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
 			memcpy(&found, result, csum_size);
 
 			read_extent_buffer(buf, &val, 0, csum_size);
-			printk_ratelimited(KERN_INFO "btrfs: %s checksum verify "
-				       "failed on %llu wanted %X found %X "
-				       "level %d\n",
-				       root->fs_info->sb->s_id, buf->start,
-				       val, found, btrfs_header_level(buf));
+			printk_ratelimited(KERN_INFO
+				"BTRFS: %s checksum verify failed on %llu wanted %X found %X "
+				"level %d\n",
+				root->fs_info->sb->s_id, buf->start,
+				val, found, btrfs_header_level(buf));
 			if (result != (char *)&inline_result)
 				kfree(result);
 			return 1;
@@ -382,13 +383,14 @@ static int btrfs_check_super_csum(char *raw_disk_sb)
 			ret = 1;
 
 		if (ret && btrfs_super_generation(disk_sb) < 10) {
-			printk(KERN_WARNING "btrfs: super block crcs don't match, older mkfs detected\n");
+			printk(KERN_WARNING
+				"BTRFS: super block crcs don't match, older mkfs detected\n");
 			ret = 0;
 		}
 	}
 
 	if (csum_type >= ARRAY_SIZE(btrfs_csum_sizes)) {
-		printk(KERN_ERR "btrfs: unsupported checksum algorithm %u\n",
+		printk(KERN_ERR "BTRFS: unsupported checksum algorithm %u\n",
 				csum_type);
 		ret = 1;
 	}
@@ -464,13 +466,10 @@ static int btree_read_extent_buffer_pages(struct btrfs_root *root,
 
 static int csum_dirty_buffer(struct btrfs_root *root, struct page *page)
 {
-	struct extent_io_tree *tree;
 	u64 start = page_offset(page);
 	u64 found_start;
 	struct extent_buffer *eb;
 
-	tree = &BTRFS_I(page->mapping->host)->io_tree;
-
 	eb = (struct extent_buffer *)page->private;
 	if (page != eb->pages[0])
 		return 0;
@@ -500,8 +499,8 @@ static int check_tree_block_fsid(struct btrfs_root *root,
 }
 
 #define CORRUPT(reason, eb, root, slot)				\
-	printk(KERN_CRIT "btrfs: corrupt leaf, %s: block=%llu,"	\
-	       "root=%llu, slot=%d\n", reason,			\
+	btrfs_crit(root->fs_info, "corrupt leaf, %s: block=%llu,"	\
+		   "root=%llu, slot=%d", reason,			\
 	       btrfs_header_bytenr(eb),	root->objectid, slot)
 
 static noinline int check_leaf(struct btrfs_root *root,
@@ -569,7 +568,6 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 				      u64 phy_offset, struct page *page,
 				      u64 start, u64 end, int mirror)
 {
-	struct extent_io_tree *tree;
 	u64 found_start;
 	int found_level;
 	struct extent_buffer *eb;
@@ -580,7 +578,6 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 	if (!page->private)
 		goto out;
 
-	tree = &BTRFS_I(page->mapping->host)->io_tree;
 	eb = (struct extent_buffer *)page->private;
 
 	/* the pending IO might have been the only thing that kept this buffer
@@ -600,21 +597,21 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 
 	found_start = btrfs_header_bytenr(eb);
 	if (found_start != eb->start) {
-		printk_ratelimited(KERN_INFO "btrfs bad tree block start "
+		printk_ratelimited(KERN_INFO "BTRFS: bad tree block start "
 			       "%llu %llu\n",
 			       found_start, eb->start);
 		ret = -EIO;
 		goto err;
 	}
 	if (check_tree_block_fsid(root, eb)) {
-		printk_ratelimited(KERN_INFO "btrfs bad fsid on block %llu\n",
+		printk_ratelimited(KERN_INFO "BTRFS: bad fsid on block %llu\n",
 			       eb->start);
 		ret = -EIO;
 		goto err;
 	}
 	found_level = btrfs_header_level(eb);
 	if (found_level >= BTRFS_MAX_LEVEL) {
-		btrfs_info(root->fs_info, "bad tree block level %d\n",
+		btrfs_info(root->fs_info, "bad tree block level %d",
 			   (int)btrfs_header_level(eb));
 		ret = -EIO;
 		goto err;
@@ -964,11 +961,9 @@ static int btree_migratepage(struct address_space *mapping,
 static int btree_writepages(struct address_space *mapping,
 			    struct writeback_control *wbc)
 {
-	struct extent_io_tree *tree;
 	struct btrfs_fs_info *fs_info;
 	int ret;
 
-	tree = &BTRFS_I(mapping->host)->io_tree;
 	if (wbc->sync_mode == WB_SYNC_NONE) {
 
 		if (wbc->for_kupdate)
@@ -1007,8 +1002,9 @@ static void btree_invalidatepage(struct page *page, unsigned int offset,
 	extent_invalidatepage(tree, page, offset);
 	btree_releasepage(page, GFP_NOFS);
 	if (PagePrivate(page)) {
-		printk(KERN_WARNING "btrfs warning page private not zero "
-		       "on page %llu\n", (unsigned long long)page_offset(page));
+		btrfs_warn(BTRFS_I(page->mapping->host)->root->fs_info,
+			   "page private not zero on page %llu",
+			   (unsigned long long)page_offset(page));
 		ClearPagePrivate(page);
 		set_page_private(page, 0);
 		page_cache_release(page);
@@ -1092,21 +1088,13 @@ int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize,
 struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
 					    u64 bytenr, u32 blocksize)
 {
-	struct inode *btree_inode = root->fs_info->btree_inode;
-	struct extent_buffer *eb;
-	eb = find_extent_buffer(&BTRFS_I(btree_inode)->io_tree, bytenr);
-	return eb;
+	return find_extent_buffer(root->fs_info, bytenr);
 }
 
 struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root,
 						 u64 bytenr, u32 blocksize)
 {
-	struct inode *btree_inode = root->fs_info->btree_inode;
-	struct extent_buffer *eb;
-
-	eb = alloc_extent_buffer(&BTRFS_I(btree_inode)->io_tree,
-				 bytenr, blocksize);
-	return eb;
+	return alloc_extent_buffer(root->fs_info, bytenr, blocksize);
 }
 
 
@@ -1270,7 +1258,6 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 	struct btrfs_root *root;
 	struct btrfs_key key;
 	int ret = 0;
-	u64 bytenr;
 	uuid_le uuid;
 
 	root = btrfs_alloc_root(fs_info);
@@ -1292,7 +1279,6 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 		goto fail;
 	}
 
-	bytenr = leaf->start;
 	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
 	btrfs_set_header_bytenr(leaf, leaf->start);
 	btrfs_set_header_generation(leaf, trans->transid);
@@ -1613,7 +1599,8 @@ again:
 	if (ret)
 		goto fail;
 
-	ret = btrfs_find_orphan_item(fs_info->tree_root, location->objectid);
+	ret = btrfs_find_item(fs_info->tree_root, NULL, BTRFS_ORPHAN_OBJECTID,
+			location->objectid, BTRFS_ORPHAN_ITEM_KEY, NULL);
 	if (ret < 0)
 		goto fail;
 	if (ret == 0)
@@ -1681,12 +1668,10 @@ static void end_workqueue_fn(struct btrfs_work *work)
 {
 	struct bio *bio;
 	struct end_io_wq *end_io_wq;
-	struct btrfs_fs_info *fs_info;
 	int error;
 
 	end_io_wq = container_of(work, struct end_io_wq, work);
 	bio = end_io_wq->bio;
-	fs_info = end_io_wq->info;
 
 	error = end_io_wq->error;
 	bio->bi_private = end_io_wq->private;
@@ -2077,6 +2062,12 @@ static void del_fs_roots(struct btrfs_fs_info *fs_info)
 		for (i = 0; i < ret; i++)
 			btrfs_drop_and_free_fs_root(fs_info, gang[i]);
 	}
+
+	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) {
+		btrfs_free_log_root_tree(NULL, fs_info);
+		btrfs_destroy_pinned_extent(fs_info->tree_root,
+					    fs_info->pinned_extents);
+	}
 }
 
 int open_ctree(struct super_block *sb,
@@ -2151,6 +2142,7 @@ int open_ctree(struct super_block *sb,
 	mapping_set_gfp_mask(fs_info->btree_inode->i_mapping, GFP_NOFS);
 
 	INIT_RADIX_TREE(&fs_info->fs_roots_radix, GFP_ATOMIC);
+	INIT_RADIX_TREE(&fs_info->buffer_radix, GFP_ATOMIC);
 	INIT_LIST_HEAD(&fs_info->trans_list);
 	INIT_LIST_HEAD(&fs_info->dead_roots);
 	INIT_LIST_HEAD(&fs_info->delayed_iputs);
@@ -2164,6 +2156,7 @@ int open_ctree(struct super_block *sb,
 	spin_lock_init(&fs_info->free_chunk_lock);
 	spin_lock_init(&fs_info->tree_mod_seq_lock);
 	spin_lock_init(&fs_info->super_lock);
+	spin_lock_init(&fs_info->buffer_lock);
 	rwlock_init(&fs_info->tree_mod_log_lock);
 	mutex_init(&fs_info->reloc_mutex);
 	seqlock_init(&fs_info->profiles_lock);
@@ -2195,7 +2188,7 @@ int open_ctree(struct super_block *sb,
 	fs_info->free_chunk_space = 0;
 	fs_info->tree_mod_log = RB_ROOT;
 	fs_info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
-
+	fs_info->avg_delayed_ref_runtime = div64_u64(NSEC_PER_SEC, 64);
 	/* readahead state */
 	INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT);
 	spin_lock_init(&fs_info->reada_lock);
@@ -2334,7 +2327,7 @@ int open_ctree(struct super_block *sb,
 	 * Pass the whole disk block of size BTRFS_SUPER_INFO_SIZE (4k).
 	 */
 	if (btrfs_check_super_csum(bh->b_data)) {
-		printk(KERN_ERR "btrfs: superblock checksum mismatch\n");
+		printk(KERN_ERR "BTRFS: superblock checksum mismatch\n");
 		err = -EINVAL;
 		goto fail_alloc;
 	}
@@ -2353,7 +2346,7 @@ int open_ctree(struct super_block *sb,
 
 	ret = btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY);
 	if (ret) {
-		printk(KERN_ERR "btrfs: superblock contains fatal errors\n");
+		printk(KERN_ERR "BTRFS: superblock contains fatal errors\n");
 		err = -EINVAL;
 		goto fail_alloc;
 	}
@@ -2418,7 +2411,7 @@ int open_ctree(struct super_block *sb,
 		features |= BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO;
 
 	if (features & BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA)
-		printk(KERN_ERR "btrfs: has skinny extents\n");
+		printk(KERN_ERR "BTRFS: has skinny extents\n");
 
 	/*
 	 * flag our filesystem as having big metadata blocks if
@@ -2426,7 +2419,7 @@ int open_ctree(struct super_block *sb,
 	 */
 	if (btrfs_super_leafsize(disk_super) > PAGE_CACHE_SIZE) {
 		if (!(features & BTRFS_FEATURE_INCOMPAT_BIG_METADATA))
-			printk(KERN_INFO "btrfs flagging fs with big metadata feature\n");
+			printk(KERN_INFO "BTRFS: flagging fs with big metadata feature\n");
 		features |= BTRFS_FEATURE_INCOMPAT_BIG_METADATA;
 	}
 
@@ -2443,7 +2436,7 @@ int open_ctree(struct super_block *sb,
 	 */
 	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
 	    (sectorsize != leafsize)) {
-		printk(KERN_WARNING "btrfs: unequal leaf/node/sector sizes "
+		printk(KERN_WARNING "BTRFS: unequal leaf/node/sector sizes "
 				"are not allowed for mixed block groups on %s\n",
 				sb->s_id);
 		goto fail_alloc;
@@ -2580,12 +2573,12 @@ int open_ctree(struct super_block *sb,
 	sb->s_blocksize_bits = blksize_bits(sectorsize);
 
 	if (btrfs_super_magic(disk_super) != BTRFS_MAGIC) {
-		printk(KERN_INFO "btrfs: valid FS not found on %s\n", sb->s_id);
+		printk(KERN_INFO "BTRFS: valid FS not found on %s\n", sb->s_id);
 		goto fail_sb_buffer;
 	}
 
 	if (sectorsize != PAGE_SIZE) {
-		printk(KERN_WARNING "btrfs: Incompatible sector size(%lu) "
+		printk(KERN_WARNING "BTRFS: Incompatible sector size(%lu) "
 		       "found on %s\n", (unsigned long)sectorsize, sb->s_id);
 		goto fail_sb_buffer;
 	}
@@ -2594,7 +2587,7 @@ int open_ctree(struct super_block *sb,
 	ret = btrfs_read_sys_array(tree_root);
 	mutex_unlock(&fs_info->chunk_mutex);
 	if (ret) {
-		printk(KERN_WARNING "btrfs: failed to read the system "
+		printk(KERN_WARNING "BTRFS: failed to read the system "
 		       "array on %s\n", sb->s_id);
 		goto fail_sb_buffer;
 	}
@@ -2611,7 +2604,7 @@ int open_ctree(struct super_block *sb,
 					   blocksize, generation);
 	if (!chunk_root->node ||
 	    !test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) {
-		printk(KERN_WARNING "btrfs: failed to read chunk root on %s\n",
+		printk(KERN_WARNING "BTRFS: failed to read chunk root on %s\n",
 		       sb->s_id);
 		goto fail_tree_roots;
 	}
@@ -2623,7 +2616,7 @@ int open_ctree(struct super_block *sb,
 
 	ret = btrfs_read_chunk_tree(chunk_root);
 	if (ret) {
-		printk(KERN_WARNING "btrfs: failed to read chunk tree on %s\n",
+		printk(KERN_WARNING "BTRFS: failed to read chunk tree on %s\n",
 		       sb->s_id);
 		goto fail_tree_roots;
 	}
@@ -2635,7 +2628,7 @@ int open_ctree(struct super_block *sb,
 	btrfs_close_extra_devices(fs_info, fs_devices, 0);
 
 	if (!fs_devices->latest_bdev) {
-		printk(KERN_CRIT "btrfs: failed to read devices on %s\n",
+		printk(KERN_CRIT "BTRFS: failed to read devices on %s\n",
 		       sb->s_id);
 		goto fail_tree_roots;
 	}
@@ -2650,7 +2643,7 @@ retry_root_backup:
 					  blocksize, generation);
 	if (!tree_root->node ||
 	    !test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) {
-		printk(KERN_WARNING "btrfs: failed to read tree root on %s\n",
+		printk(KERN_WARNING "BTRFS: failed to read tree root on %s\n",
 		       sb->s_id);
 
 		goto recovery_tree_root;
@@ -2721,50 +2714,56 @@ retry_root_backup:
 
 	ret = btrfs_recover_balance(fs_info);
 	if (ret) {
-		printk(KERN_WARNING "btrfs: failed to recover balance\n");
+		printk(KERN_WARNING "BTRFS: failed to recover balance\n");
 		goto fail_block_groups;
 	}
 
 	ret = btrfs_init_dev_stats(fs_info);
 	if (ret) {
-		printk(KERN_ERR "btrfs: failed to init dev_stats: %d\n",
+		printk(KERN_ERR "BTRFS: failed to init dev_stats: %d\n",
 		       ret);
 		goto fail_block_groups;
 	}
 
 	ret = btrfs_init_dev_replace(fs_info);
 	if (ret) {
-		pr_err("btrfs: failed to init dev_replace: %d\n", ret);
+		pr_err("BTRFS: failed to init dev_replace: %d\n", ret);
 		goto fail_block_groups;
 	}
 
 	btrfs_close_extra_devices(fs_info, fs_devices, 1);
 
-	ret = btrfs_init_space_info(fs_info);
+	ret = btrfs_sysfs_add_one(fs_info);
 	if (ret) {
-		printk(KERN_ERR "Failed to initial space info: %d\n", ret);
+		pr_err("BTRFS: failed to init sysfs interface: %d\n", ret);
 		goto fail_block_groups;
 	}
 
+	ret = btrfs_init_space_info(fs_info);
+	if (ret) {
+		printk(KERN_ERR "BTRFS: Failed to initial space info: %d\n", ret);
+		goto fail_sysfs;
+	}
+
 	ret = btrfs_read_block_groups(extent_root);
 	if (ret) {
-		printk(KERN_ERR "Failed to read block groups: %d\n", ret);
-		goto fail_block_groups;
+		printk(KERN_ERR "BTRFS: Failed to read block groups: %d\n", ret);
+		goto fail_sysfs;
 	}
 	fs_info->num_tolerated_disk_barrier_failures =
 		btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
 	if (fs_info->fs_devices->missing_devices >
 	     fs_info->num_tolerated_disk_barrier_failures &&
 	    !(sb->s_flags & MS_RDONLY)) {
-		printk(KERN_WARNING
-		       "Btrfs: too many missing devices, writeable mount is not allowed\n");
-		goto fail_block_groups;
+		printk(KERN_WARNING "BTRFS: "
+			"too many missing devices, writeable mount is not allowed\n");
+		goto fail_sysfs;
 	}
 
 	fs_info->cleaner_kthread = kthread_run(cleaner_kthread, tree_root,
 					       "btrfs-cleaner");
 	if (IS_ERR(fs_info->cleaner_kthread))
-		goto fail_block_groups;
+		goto fail_sysfs;
 
 	fs_info->transaction_kthread = kthread_run(transaction_kthread,
 						   tree_root,
@@ -2775,11 +2774,15 @@ retry_root_backup:
 	if (!btrfs_test_opt(tree_root, SSD) &&
 	    !btrfs_test_opt(tree_root, NOSSD) &&
 	    !fs_info->fs_devices->rotating) {
-		printk(KERN_INFO "Btrfs detected SSD devices, enabling SSD "
+		printk(KERN_INFO "BTRFS: detected SSD devices, enabling SSD "
 		       "mode\n");
 		btrfs_set_opt(fs_info->mount_opt, SSD);
 	}
 
+	/* Set the real inode map cache flag */
+	if (btrfs_test_opt(tree_root, CHANGE_INODE_CACHE))
+		btrfs_set_opt(tree_root->fs_info->mount_opt, INODE_MAP_CACHE);
+
 #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
 	if (btrfs_test_opt(tree_root, CHECK_INTEGRITY)) {
 		ret = btrfsic_mount(tree_root, fs_devices,
@@ -2788,7 +2791,7 @@ retry_root_backup:
 				    1 : 0,
 				    fs_info->check_integrity_print_mask);
 		if (ret)
-			printk(KERN_WARNING "btrfs: failed to initialize"
+			printk(KERN_WARNING "BTRFS: failed to initialize"
 			       " integrity check module %s\n", sb->s_id);
 	}
 #endif
@@ -2801,7 +2804,7 @@ retry_root_backup:
 		u64 bytenr = btrfs_super_log_root(disk_super);
 
 		if (fs_devices->rw_devices == 0) {
-			printk(KERN_WARNING "Btrfs log replay required "
+			printk(KERN_WARNING "BTRFS: log replay required "
 			       "on RO media\n");
 			err = -EIO;
 			goto fail_qgroup;
@@ -2824,7 +2827,7 @@ retry_root_backup:
 						      generation + 1);
 		if (!log_tree_root->node ||
 		    !extent_buffer_uptodate(log_tree_root->node)) {
-			printk(KERN_ERR "btrfs: failed to read log tree\n");
+			printk(KERN_ERR "BTRFS: failed to read log tree\n");
 			free_extent_buffer(log_tree_root->node);
 			kfree(log_tree_root);
 			goto fail_trans_kthread;
@@ -2858,7 +2861,7 @@ retry_root_backup:
 		ret = btrfs_recover_relocation(tree_root);
 		if (ret < 0) {
 			printk(KERN_WARNING
-			       "btrfs: failed to recover relocation\n");
+			       "BTRFS: failed to recover relocation\n");
 			err = -EINVAL;
 			goto fail_qgroup;
 		}
@@ -2888,14 +2891,14 @@ retry_root_backup:
 
 	ret = btrfs_resume_balance_async(fs_info);
 	if (ret) {
-		printk(KERN_WARNING "btrfs: failed to resume balance\n");
+		printk(KERN_WARNING "BTRFS: failed to resume balance\n");
 		close_ctree(tree_root);
 		return ret;
 	}
 
 	ret = btrfs_resume_dev_replace_async(fs_info);
 	if (ret) {
-		pr_warn("btrfs: failed to resume dev_replace\n");
+		pr_warn("BTRFS: failed to resume dev_replace\n");
 		close_ctree(tree_root);
 		return ret;
 	}
@@ -2903,20 +2906,20 @@ retry_root_backup:
 	btrfs_qgroup_rescan_resume(fs_info);
 
 	if (create_uuid_tree) {
-		pr_info("btrfs: creating UUID tree\n");
+		pr_info("BTRFS: creating UUID tree\n");
 		ret = btrfs_create_uuid_tree(fs_info);
 		if (ret) {
-			pr_warn("btrfs: failed to create the UUID tree %d\n",
+			pr_warn("BTRFS: failed to create the UUID tree %d\n",
 				ret);
 			close_ctree(tree_root);
 			return ret;
 		}
 	} else if (check_uuid_tree ||
 		   btrfs_test_opt(tree_root, RESCAN_UUID_TREE)) {
-		pr_info("btrfs: checking UUID tree\n");
+		pr_info("BTRFS: checking UUID tree\n");
 		ret = btrfs_check_uuid_tree(fs_info);
 		if (ret) {
-			pr_warn("btrfs: failed to check the UUID tree %d\n",
+			pr_warn("BTRFS: failed to check the UUID tree %d\n",
 				ret);
 			close_ctree(tree_root);
 			return ret;
@@ -2942,6 +2945,9 @@ fail_cleaner:
 	 */
 	filemap_write_and_wait(fs_info->btree_inode->i_mapping);
 
+fail_sysfs:
+	btrfs_sysfs_remove_one(fs_info);
+
 fail_block_groups:
 	btrfs_put_block_group_cache(fs_info);
 	btrfs_free_block_groups(fs_info);
@@ -2997,7 +3003,7 @@ static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate)
 		struct btrfs_device *device = (struct btrfs_device *)
 			bh->b_private;
 
-		printk_ratelimited_in_rcu(KERN_WARNING "lost page write due to "
+		printk_ratelimited_in_rcu(KERN_WARNING "BTRFS: lost page write due to "
 					  "I/O error on %s\n",
 					  rcu_str_deref(device->name));
 		/* note, we dont' set_buffer_write_io_error because we have
@@ -3116,7 +3122,7 @@ static int write_dev_supers(struct btrfs_device *device,
 			bh = __getblk(device->bdev, bytenr / 4096,
 				      BTRFS_SUPER_INFO_SIZE);
 			if (!bh) {
-				printk(KERN_ERR "btrfs: couldn't get super "
+				printk(KERN_ERR "BTRFS: couldn't get super "
 				       "buffer head for bytenr %Lu\n", bytenr);
 				errors++;
 				continue;
@@ -3137,7 +3143,10 @@ static int write_dev_supers(struct btrfs_device *device,
 		 * we fua the first super.  The others we allow
 		 * to go down lazy.
 		 */
-		ret = btrfsic_submit_bh(WRITE_FUA, bh);
+		if (i == 0)
+			ret = btrfsic_submit_bh(WRITE_FUA, bh);
+		else
+			ret = btrfsic_submit_bh(WRITE_SYNC, bh);
 		if (ret)
 			errors++;
 	}
@@ -3183,7 +3192,7 @@ static int write_dev_flush(struct btrfs_device *device, int wait)
 		wait_for_completion(&device->flush_wait);
 
 		if (bio_flagged(bio, BIO_EOPNOTSUPP)) {
-			printk_in_rcu("btrfs: disabling barriers on dev %s\n",
+			printk_in_rcu("BTRFS: disabling barriers on dev %s\n",
 				      rcu_str_deref(device->name));
 			device->nobarriers = 1;
 		} else if (!bio_flagged(bio, BIO_UPTODATE)) {
@@ -3404,7 +3413,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 			total_errors++;
 	}
 	if (total_errors > max_errors) {
-		printk(KERN_ERR "btrfs: %d errors while writing supers\n",
+		btrfs_err(root->fs_info, "%d errors while writing supers",
 		       total_errors);
 		mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
@@ -3452,10 +3461,8 @@ void btrfs_drop_and_free_fs_root(struct btrfs_fs_info *fs_info,
 	if (btrfs_root_refs(&root->root_item) == 0)
 		synchronize_srcu(&fs_info->subvol_srcu);
 
-	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) {
+	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state))
 		btrfs_free_log(NULL, root);
-		btrfs_free_log_root_tree(NULL, fs_info);
-	}
 
 	__btrfs_remove_free_space_cache(root->free_ino_pinned);
 	__btrfs_remove_free_space_cache(root->free_ino_ctl);
@@ -3560,14 +3567,12 @@ int close_ctree(struct btrfs_root *root)
 	if (!(fs_info->sb->s_flags & MS_RDONLY)) {
 		ret = btrfs_commit_super(root);
 		if (ret)
-			printk(KERN_ERR "btrfs: commit super ret %d\n", ret);
+			btrfs_err(root->fs_info, "commit super ret %d", ret);
 	}
 
 	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state))
 		btrfs_error_commit_super(root);
 
-	btrfs_put_block_group_cache(fs_info);
-
 	kthread_stop(fs_info->transaction_kthread);
 	kthread_stop(fs_info->cleaner_kthread);
 
@@ -3577,12 +3582,16 @@ int close_ctree(struct btrfs_root *root)
 	btrfs_free_qgroup_config(root->fs_info);
 
 	if (percpu_counter_sum(&fs_info->delalloc_bytes)) {
-		printk(KERN_INFO "btrfs: at unmount delalloc count %lld\n",
+		btrfs_info(root->fs_info, "at unmount delalloc count %lld",
 		       percpu_counter_sum(&fs_info->delalloc_bytes));
 	}
 
+	btrfs_sysfs_remove_one(fs_info);
+
 	del_fs_roots(fs_info);
 
+	btrfs_put_block_group_cache(fs_info);
+
 	btrfs_free_block_groups(fs_info);
 
 	btrfs_stop_all_workers(fs_info);
@@ -3800,55 +3809,55 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 	delayed_refs = &trans->delayed_refs;
 
 	spin_lock(&delayed_refs->lock);
-	if (delayed_refs->num_entries == 0) {
+	if (atomic_read(&delayed_refs->num_entries) == 0) {
 		spin_unlock(&delayed_refs->lock);
-		printk(KERN_INFO "delayed_refs has NO entry\n");
+		btrfs_info(root->fs_info, "delayed_refs has NO entry");
 		return ret;
 	}
 
-	while ((node = rb_first(&delayed_refs->root)) != NULL) {
-		struct btrfs_delayed_ref_head *head = NULL;
+	while ((node = rb_first(&delayed_refs->href_root)) != NULL) {
+		struct btrfs_delayed_ref_head *head;
 		bool pin_bytes = false;
 
-		ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-		atomic_set(&ref->refs, 1);
-		if (btrfs_delayed_ref_is_head(ref)) {
-
-			head = btrfs_delayed_node_to_head(ref);
-			if (!mutex_trylock(&head->mutex)) {
-				atomic_inc(&ref->refs);
-				spin_unlock(&delayed_refs->lock);
-
-				/* Need to wait for the delayed ref to run */
-				mutex_lock(&head->mutex);
-				mutex_unlock(&head->mutex);
-				btrfs_put_delayed_ref(ref);
-
-				spin_lock(&delayed_refs->lock);
-				continue;
-			}
-
-			if (head->must_insert_reserved)
-				pin_bytes = true;
-			btrfs_free_delayed_extent_op(head->extent_op);
-			delayed_refs->num_heads--;
-			if (list_empty(&head->cluster))
-				delayed_refs->num_heads_ready--;
-			list_del_init(&head->cluster);
-		}
+		head = rb_entry(node, struct btrfs_delayed_ref_head,
+				href_node);
+		if (!mutex_trylock(&head->mutex)) {
+			atomic_inc(&head->node.refs);
+			spin_unlock(&delayed_refs->lock);
 
-		ref->in_tree = 0;
-		rb_erase(&ref->rb_node, &delayed_refs->root);
-		delayed_refs->num_entries--;
-		spin_unlock(&delayed_refs->lock);
-		if (head) {
-			if (pin_bytes)
-				btrfs_pin_extent(root, ref->bytenr,
-						 ref->num_bytes, 1);
+			mutex_lock(&head->mutex);
 			mutex_unlock(&head->mutex);
+			btrfs_put_delayed_ref(&head->node);
+			spin_lock(&delayed_refs->lock);
+			continue;
+		}
+		spin_lock(&head->lock);
+		while ((node = rb_first(&head->ref_root)) != NULL) {
+			ref = rb_entry(node, struct btrfs_delayed_ref_node,
+				       rb_node);
+			ref->in_tree = 0;
+			rb_erase(&ref->rb_node, &head->ref_root);
+			atomic_dec(&delayed_refs->num_entries);
+			btrfs_put_delayed_ref(ref);
+			cond_resched_lock(&head->lock);
 		}
-		btrfs_put_delayed_ref(ref);
+		if (head->must_insert_reserved)
+			pin_bytes = true;
+		btrfs_free_delayed_extent_op(head->extent_op);
+		delayed_refs->num_heads--;
+		if (head->processing == 0)
+			delayed_refs->num_heads_ready--;
+		atomic_dec(&delayed_refs->num_entries);
+		head->node.in_tree = 0;
+		rb_erase(&head->href_node, &delayed_refs->href_root);
+		spin_unlock(&head->lock);
+		spin_unlock(&delayed_refs->lock);
+		mutex_unlock(&head->mutex);
 
+		if (pin_bytes)
+			btrfs_pin_extent(root, head->node.bytenr,
+					 head->node.num_bytes, 1);
+		btrfs_put_delayed_ref(&head->node);
 		cond_resched();
 		spin_lock(&delayed_refs->lock);
 	}

fs/btrfs/extent-tree.c

@@ -35,6 +35,7 @@
 #include "locking.h"
 #include "free-space-cache.h"
 #include "math.h"
+#include "sysfs.h"
 
 #undef SCRAMBLE_DELAYED_REFS
 
@@ -441,7 +442,8 @@ next:
 			if (ret)
 				break;
 
-			if (need_resched()) {
+			if (need_resched() ||
+			    rwsem_is_contended(&fs_info->extent_commit_sem)) {
 				caching_ctl->progress = last;
 				btrfs_release_path(path);
 				up_read(&fs_info->extent_commit_sem);
@@ -855,12 +857,14 @@ again:
 			btrfs_put_delayed_ref(&head->node);
 			goto search_again;
 		}
+		spin_lock(&head->lock);
 		if (head->extent_op && head->extent_op->update_flags)
 			extent_flags |= head->extent_op->flags_to_set;
 		else
 			BUG_ON(num_refs == 0);
 
 		num_refs += head->node.ref_mod;
+		spin_unlock(&head->lock);
 		mutex_unlock(&head->mutex);
 	}
 	spin_unlock(&delayed_refs->lock);
@@ -1070,11 +1074,11 @@ static u64 hash_extent_data_ref(u64 root_objectid, u64 owner, u64 offset)
 	__le64 lenum;
 
 	lenum = cpu_to_le64(root_objectid);
-	high_crc = crc32c(high_crc, &lenum, sizeof(lenum));
+	high_crc = btrfs_crc32c(high_crc, &lenum, sizeof(lenum));
 	lenum = cpu_to_le64(owner);
-	low_crc = crc32c(low_crc, &lenum, sizeof(lenum));
+	low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
 	lenum = cpu_to_le64(offset);
-	low_crc = crc32c(low_crc, &lenum, sizeof(lenum));
+	low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
 
 	return ((u64)high_crc << 31) ^ (u64)low_crc;
 }
@@ -2285,64 +2289,62 @@ static noinline struct btrfs_delayed_ref_node *
 select_delayed_ref(struct btrfs_delayed_ref_head *head)
 {
 	struct rb_node *node;
-	struct btrfs_delayed_ref_node *ref;
-	int action = BTRFS_ADD_DELAYED_REF;
-again:
+	struct btrfs_delayed_ref_node *ref, *last = NULL;;
+
 	/*
 	 * select delayed ref of type BTRFS_ADD_DELAYED_REF first.
 	 * this prevents ref count from going down to zero when
 	 * there still are pending delayed ref.
 	 */
-	node = rb_prev(&head->node.rb_node);
-	while (1) {
-		if (!node)
-			break;
+	node = rb_first(&head->ref_root);
+	while (node) {
 		ref = rb_entry(node, struct btrfs_delayed_ref_node,
 				rb_node);
-		if (ref->bytenr != head->node.bytenr)
-			break;
-		if (ref->action == action)
+		if (ref->action == BTRFS_ADD_DELAYED_REF)
 			return ref;
-		node = rb_prev(node);
-	}
-	if (action == BTRFS_ADD_DELAYED_REF) {
-		action = BTRFS_DROP_DELAYED_REF;
-		goto again;
+		else if (last == NULL)
+			last = ref;
+		node = rb_next(node);
 	}
-	return NULL;
+	return last;
 }
 
 /*
  * Returns 0 on success or if called with an already aborted transaction.
  * Returns -ENOMEM or -EIO on failure and will abort the transaction.
  */
-static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
-				       struct btrfs_root *root,
-				       struct list_head *cluster)
+static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
+					     struct btrfs_root *root,
+					     unsigned long nr)
 {
 	struct btrfs_delayed_ref_root *delayed_refs;
 	struct btrfs_delayed_ref_node *ref;
 	struct btrfs_delayed_ref_head *locked_ref = NULL;
 	struct btrfs_delayed_extent_op *extent_op;
 	struct btrfs_fs_info *fs_info = root->fs_info;
+	ktime_t start = ktime_get();
 	int ret;
-	int count = 0;
+	unsigned long count = 0;
+	unsigned long actual_count = 0;
 	int must_insert_reserved = 0;
 
 	delayed_refs = &trans->transaction->delayed_refs;
 	while (1) {
 		if (!locked_ref) {
-			/* pick a new head ref from the cluster list */
-			if (list_empty(cluster))
+			if (count >= nr)
 				break;
 
-			locked_ref = list_entry(cluster->next,
-				     struct btrfs_delayed_ref_head, cluster);
+			spin_lock(&delayed_refs->lock);
+			locked_ref = btrfs_select_ref_head(trans);
+			if (!locked_ref) {
+				spin_unlock(&delayed_refs->lock);
+				break;
+			}
 
 			/* grab the lock that says we are going to process
 			 * all the refs for this head */
 			ret = btrfs_delayed_ref_lock(trans, locked_ref);
-
+			spin_unlock(&delayed_refs->lock);
 			/*
 			 * we may have dropped the spin lock to get the head
 			 * mutex lock, and that might have given someone else
@@ -2363,6 +2365,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 		 * finish.  If we merged anything we need to re-loop so we can
 		 * get a good ref.
 		 */
+		spin_lock(&locked_ref->lock);
 		btrfs_merge_delayed_refs(trans, fs_info, delayed_refs,
 					 locked_ref);
 
@@ -2374,17 +2377,14 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 
 		if (ref && ref->seq &&
 		    btrfs_check_delayed_seq(fs_info, delayed_refs, ref->seq)) {
-			/*
-			 * there are still refs with lower seq numbers in the
-			 * process of being added. Don't run this ref yet.
-			 */
-			list_del_init(&locked_ref->cluster);
+			spin_unlock(&locked_ref->lock);
 			btrfs_delayed_ref_unlock(locked_ref);
-			locked_ref = NULL;
+			spin_lock(&delayed_refs->lock);
+			locked_ref->processing = 0;
 			delayed_refs->num_heads_ready++;
 			spin_unlock(&delayed_refs->lock);
+			locked_ref = NULL;
 			cond_resched();
-			spin_lock(&delayed_refs->lock);
 			continue;
 		}
 
@@ -2399,6 +2399,8 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 		locked_ref->extent_op = NULL;
 
 		if (!ref) {
+
+
 			/* All delayed refs have been processed, Go ahead
 			 * and send the head node to run_one_delayed_ref,
 			 * so that any accounting fixes can happen
@@ -2411,8 +2413,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 			}
 
 			if (extent_op) {
-				spin_unlock(&delayed_refs->lock);
-
+				spin_unlock(&locked_ref->lock);
 				ret = run_delayed_extent_op(trans, root,
 							    ref, extent_op);
 				btrfs_free_delayed_extent_op(extent_op);
@@ -2426,19 +2427,39 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 					 */
 					if (must_insert_reserved)
 						locked_ref->must_insert_reserved = 1;
+					locked_ref->processing = 0;
 					btrfs_debug(fs_info, "run_delayed_extent_op returned %d", ret);
-					spin_lock(&delayed_refs->lock);
 					btrfs_delayed_ref_unlock(locked_ref);
 					return ret;
 				}
+				continue;
+			}
 
-				goto next;
+			/*
+			 * Need to drop our head ref lock and re-aqcuire the
+			 * delayed ref lock and then re-check to make sure
+			 * nobody got added.
+			 */
+			spin_unlock(&locked_ref->lock);
+			spin_lock(&delayed_refs->lock);
+			spin_lock(&locked_ref->lock);
+			if (rb_first(&locked_ref->ref_root)) {
+				spin_unlock(&locked_ref->lock);
+				spin_unlock(&delayed_refs->lock);
+				continue;
 			}
+			ref->in_tree = 0;
+			delayed_refs->num_heads--;
+			rb_erase(&locked_ref->href_node,
+				 &delayed_refs->href_root);
+			spin_unlock(&delayed_refs->lock);
+		} else {
+			actual_count++;
+			ref->in_tree = 0;
+			rb_erase(&ref->rb_node, &locked_ref->ref_root);
 		}
+		atomic_dec(&delayed_refs->num_entries);
 
-		ref->in_tree = 0;
-		rb_erase(&ref->rb_node, &delayed_refs->root);
-		delayed_refs->num_entries--;
 		if (!btrfs_delayed_ref_is_head(ref)) {
 			/*
 			 * when we play the delayed ref, also correct the
@@ -2455,20 +2476,18 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 			default:
 				WARN_ON(1);
 			}
-		} else {
-			list_del_init(&locked_ref->cluster);
 		}
-		spin_unlock(&delayed_refs->lock);
+		spin_unlock(&locked_ref->lock);
 
 		ret = run_one_delayed_ref(trans, root, ref, extent_op,
 					  must_insert_reserved);
 
 		btrfs_free_delayed_extent_op(extent_op);
 		if (ret) {
+			locked_ref->processing = 0;
 			btrfs_delayed_ref_unlock(locked_ref);
 			btrfs_put_delayed_ref(ref);
 			btrfs_debug(fs_info, "run_one_delayed_ref returned %d", ret);
-			spin_lock(&delayed_refs->lock);
 			return ret;
 		}
 
@@ -2484,11 +2503,29 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 		}
 		btrfs_put_delayed_ref(ref);
 		count++;
-next:
 		cond_resched();
+	}
+
+	/*
+	 * We don't want to include ref heads since we can have empty ref heads
+	 * and those will drastically skew our runtime down since we just do
+	 * accounting, no actual extent tree updates.
+	 */
+	if (actual_count > 0) {
+		u64 runtime = ktime_to_ns(ktime_sub(ktime_get(), start));
+		u64 avg;
+
+		/*
+		 * We weigh the current average higher than our current runtime
+		 * to avoid large swings in the average.
+		 */
 		spin_lock(&delayed_refs->lock);
+		avg = fs_info->avg_delayed_ref_runtime * 3 + runtime;
+		avg = div64_u64(avg, 4);
+		fs_info->avg_delayed_ref_runtime = avg;
+		spin_unlock(&delayed_refs->lock);
 	}
-	return count;
+	return 0;
 }
 
 #ifdef SCRAMBLE_DELAYED_REFS
@@ -2570,16 +2607,6 @@ int btrfs_delayed_refs_qgroup_accounting(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
-static int refs_newer(struct btrfs_delayed_ref_root *delayed_refs, int seq,
-		      int count)
-{
-	int val = atomic_read(&delayed_refs->ref_seq);
-
-	if (val < seq || val >= seq + count)
-		return 1;
-	return 0;
-}
-
 static inline u64 heads_to_leaves(struct btrfs_root *root, u64 heads)
 {
 	u64 num_bytes;
@@ -2596,7 +2623,7 @@ static inline u64 heads_to_leaves(struct btrfs_root *root, u64 heads)
 	return div64_u64(num_bytes, BTRFS_LEAF_DATA_SIZE(root));
 }
 
-int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans,
+int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans,
 				       struct btrfs_root *root)
 {
 	struct btrfs_block_rsv *global_rsv;
@@ -2625,6 +2652,22 @@ int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
+int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans,
+				       struct btrfs_root *root)
+{
+	struct btrfs_fs_info *fs_info = root->fs_info;
+	u64 num_entries =
+		atomic_read(&trans->transaction->delayed_refs.num_entries);
+	u64 avg_runtime;
+
+	smp_mb();
+	avg_runtime = fs_info->avg_delayed_ref_runtime;
+	if (num_entries * avg_runtime >= NSEC_PER_SEC)
+		return 1;
+
+	return btrfs_check_space_for_delayed_refs(trans, root);
+}
+
 /*
  * this starts processing the delayed reference count updates and
  * extent insertions we have queued up so far.  count can be
@@ -2640,13 +2683,10 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 {
 	struct rb_node *node;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	struct btrfs_delayed_ref_node *ref;
-	struct list_head cluster;
+	struct btrfs_delayed_ref_head *head;
 	int ret;
-	u64 delayed_start;
 	int run_all = count == (unsigned long)-1;
 	int run_most = 0;
-	int loops;
 
 	/* We'll clean this up in btrfs_cleanup_transaction */
 	if (trans->aborted)
@@ -2658,130 +2698,40 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 	btrfs_delayed_refs_qgroup_accounting(trans, root->fs_info);
 
 	delayed_refs = &trans->transaction->delayed_refs;
-	INIT_LIST_HEAD(&cluster);
 	if (count == 0) {
-		count = delayed_refs->num_entries * 2;
+		count = atomic_read(&delayed_refs->num_entries) * 2;
 		run_most = 1;
 	}
 
-	if (!run_all && !run_most) {
-		int old;
-		int seq = atomic_read(&delayed_refs->ref_seq);
-
-progress:
-		old = atomic_cmpxchg(&delayed_refs->procs_running_refs, 0, 1);
-		if (old) {
-			DEFINE_WAIT(__wait);
-			if (delayed_refs->flushing ||
-			    !btrfs_should_throttle_delayed_refs(trans, root))
-				return 0;
-
-			prepare_to_wait(&delayed_refs->wait, &__wait,
-					TASK_UNINTERRUPTIBLE);
-
-			old = atomic_cmpxchg(&delayed_refs->procs_running_refs, 0, 1);
-			if (old) {
-				schedule();
-				finish_wait(&delayed_refs->wait, &__wait);
-
-				if (!refs_newer(delayed_refs, seq, 256))
-					goto progress;
-				else
-					return 0;
-			} else {
-				finish_wait(&delayed_refs->wait, &__wait);
-				goto again;
-			}
-		}
-
-	} else {
-		atomic_inc(&delayed_refs->procs_running_refs);
-	}
-
 again:
-	loops = 0;
-	spin_lock(&delayed_refs->lock);
-
 #ifdef SCRAMBLE_DELAYED_REFS
 	delayed_refs->run_delayed_start = find_middle(&delayed_refs->root);
 #endif
-
-	while (1) {
-		if (!(run_all || run_most) &&
-		    !btrfs_should_throttle_delayed_refs(trans, root))
-			break;
-
-		/*
-		 * go find something we can process in the rbtree.  We start at
-		 * the beginning of the tree, and then build a cluster
-		 * of refs to process starting at the first one we are able to
-		 * lock
-		 */
-		delayed_start = delayed_refs->run_delayed_start;
-		ret = btrfs_find_ref_cluster(trans, &cluster,
-					     delayed_refs->run_delayed_start);
-		if (ret)
-			break;
-
-		ret = run_clustered_refs(trans, root, &cluster);
-		if (ret < 0) {
-			btrfs_release_ref_cluster(&cluster);
-			spin_unlock(&delayed_refs->lock);
-			btrfs_abort_transaction(trans, root, ret);
-			atomic_dec(&delayed_refs->procs_running_refs);
-			wake_up(&delayed_refs->wait);
-			return ret;
-		}
-
-		atomic_add(ret, &delayed_refs->ref_seq);
-
-		count -= min_t(unsigned long, ret, count);
-
-		if (count == 0)
-			break;
-
-		if (delayed_start >= delayed_refs->run_delayed_start) {
-			if (loops == 0) {
-				/*
-				 * btrfs_find_ref_cluster looped. let's do one
-				 * more cycle. if we don't run any delayed ref
-				 * during that cycle (because we can't because
-				 * all of them are blocked), bail out.
-				 */
-				loops = 1;
-			} else {
-				/*
-				 * no runnable refs left, stop trying
-				 */
-				BUG_ON(run_all);
-				break;
-			}
-		}
-		if (ret) {
-			/* refs were run, let's reset staleness detection */
-			loops = 0;
-		}
+	ret = __btrfs_run_delayed_refs(trans, root, count);
+	if (ret < 0) {
+		btrfs_abort_transaction(trans, root, ret);
+		return ret;
 	}
 
 	if (run_all) {
-		if (!list_empty(&trans->new_bgs)) {
-			spin_unlock(&delayed_refs->lock);
+		if (!list_empty(&trans->new_bgs))
 			btrfs_create_pending_block_groups(trans, root);
-			spin_lock(&delayed_refs->lock);
-		}
 
-		node = rb_first(&delayed_refs->root);
-		if (!node)
+		spin_lock(&delayed_refs->lock);
+		node = rb_first(&delayed_refs->href_root);
+		if (!node) {
+			spin_unlock(&delayed_refs->lock);
 			goto out;
+		}
 		count = (unsigned long)-1;
 
 		while (node) {
-			ref = rb_entry(node, struct btrfs_delayed_ref_node,
-				       rb_node);
-			if (btrfs_delayed_ref_is_head(ref)) {
-				struct btrfs_delayed_ref_head *head;
+			head = rb_entry(node, struct btrfs_delayed_ref_head,
+					href_node);
+			if (btrfs_delayed_ref_is_head(&head->node)) {
+				struct btrfs_delayed_ref_node *ref;
 
-				head = btrfs_delayed_node_to_head(ref);
+				ref = &head->node;
 				atomic_inc(&ref->refs);
 
 				spin_unlock(&delayed_refs->lock);
@@ -2795,20 +2745,16 @@ again:
 				btrfs_put_delayed_ref(ref);
 				cond_resched();
 				goto again;
+			} else {
+				WARN_ON(1);
 			}
 			node = rb_next(node);
 		}
 		spin_unlock(&delayed_refs->lock);
-		schedule_timeout(1);
+		cond_resched();
 		goto again;
 	}
 out:
-	atomic_dec(&delayed_refs->procs_running_refs);
-	smp_mb();
-	if (waitqueue_active(&delayed_refs->wait))
-		wake_up(&delayed_refs->wait);
-
-	spin_unlock(&delayed_refs->lock);
 	assert_qgroups_uptodate(trans);
 	return 0;
 }
@@ -2850,12 +2796,13 @@ static noinline int check_delayed_ref(struct btrfs_trans_handle *trans,
 	struct rb_node *node;
 	int ret = 0;
 
-	ret = -ENOENT;
 	delayed_refs = &trans->transaction->delayed_refs;
 	spin_lock(&delayed_refs->lock);
 	head = btrfs_find_delayed_ref_head(trans, bytenr);
-	if (!head)
-		goto out;
+	if (!head) {
+		spin_unlock(&delayed_refs->lock);
+		return 0;
+	}
 
 	if (!mutex_trylock(&head->mutex)) {
 		atomic_inc(&head->node.refs);
@@ -2872,40 +2819,35 @@ static noinline int check_delayed_ref(struct btrfs_trans_handle *trans,
 		btrfs_put_delayed_ref(&head->node);
 		return -EAGAIN;
 	}
+	spin_unlock(&delayed_refs->lock);
 
-	node = rb_prev(&head->node.rb_node);
-	if (!node)
-		goto out_unlock;
-
-	ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-
-	if (ref->bytenr != bytenr)
-		goto out_unlock;
-
-	ret = 1;
-	if (ref->type != BTRFS_EXTENT_DATA_REF_KEY)
-		goto out_unlock;
+	spin_lock(&head->lock);
+	node = rb_first(&head->ref_root);
+	while (node) {
+		ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
+		node = rb_next(node);
 
-	data_ref = btrfs_delayed_node_to_data_ref(ref);
+		/* If it's a shared ref we know a cross reference exists */
+		if (ref->type != BTRFS_EXTENT_DATA_REF_KEY) {
+			ret = 1;
+			break;
+		}
 
-	node = rb_prev(node);
-	if (node) {
-		int seq = ref->seq;
+		data_ref = btrfs_delayed_node_to_data_ref(ref);
 
-		ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-		if (ref->bytenr == bytenr && ref->seq == seq)
-			goto out_unlock;
+		/*
+		 * If our ref doesn't match the one we're currently looking at
+		 * then we have a cross reference.
+		 */
+		if (data_ref->root != root->root_key.objectid ||
+		    data_ref->objectid != objectid ||
+		    data_ref->offset != offset) {
+			ret = 1;
+			break;
+		}
 	}
-
-	if (data_ref->root != root->root_key.objectid ||
-	    data_ref->objectid != objectid || data_ref->offset != offset)
-		goto out_unlock;
-
-	ret = 0;
-out_unlock:
+	spin_unlock(&head->lock);
 	mutex_unlock(&head->mutex);
-out:
-	spin_unlock(&delayed_refs->lock);
 	return ret;
 }
 
@@ -3402,6 +3344,23 @@ int btrfs_extent_readonly(struct btrfs_root *root, u64 bytenr)
 	return readonly;
 }
 
+static const char *alloc_name(u64 flags)
+{
+	switch (flags) {
+	case BTRFS_BLOCK_GROUP_METADATA|BTRFS_BLOCK_GROUP_DATA:
+		return "mixed";
+	case BTRFS_BLOCK_GROUP_METADATA:
+		return "metadata";
+	case BTRFS_BLOCK_GROUP_DATA:
+		return "data";
+	case BTRFS_BLOCK_GROUP_SYSTEM:
+		return "system";
+	default:
+		WARN_ON(1);
+		return "invalid-combination";
+	};
+}
+
 static int update_space_info(struct btrfs_fs_info *info, u64 flags,
 			     u64 total_bytes, u64 bytes_used,
 			     struct btrfs_space_info **space_info)
@@ -3439,8 +3398,10 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags,
 		return ret;
 	}
 
-	for (i = 0; i < BTRFS_NR_RAID_TYPES; i++)
+	for (i = 0; i < BTRFS_NR_RAID_TYPES; i++) {
 		INIT_LIST_HEAD(&found->block_groups[i]);
+		kobject_init(&found->block_group_kobjs[i], &btrfs_raid_ktype);
+	}
 	init_rwsem(&found->groups_sem);
 	spin_lock_init(&found->lock);
 	found->flags = flags & BTRFS_BLOCK_GROUP_TYPE_MASK;
@@ -3457,11 +3418,21 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags,
 	found->chunk_alloc = 0;
 	found->flush = 0;
 	init_waitqueue_head(&found->wait);
+
+	ret = kobject_init_and_add(&found->kobj, &space_info_ktype,
+				    info->space_info_kobj, "%s",
+				    alloc_name(found->flags));
+	if (ret) {
+		kfree(found);
+		return ret;
+	}
+
 	*space_info = found;
 	list_add_rcu(&found->list, &info->space_info);
 	if (flags & BTRFS_BLOCK_GROUP_DATA)
 		info->data_sinfo = found;
-	return 0;
+
+	return ret;
 }
 
 static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags)
@@ -4637,7 +4608,7 @@ void btrfs_block_rsv_release(struct btrfs_root *root,
 			     u64 num_bytes)
 {
 	struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv;
-	if (global_rsv->full || global_rsv == block_rsv ||
+	if (global_rsv == block_rsv ||
 	    block_rsv->space_info != global_rsv->space_info)
 		global_rsv = NULL;
 	block_rsv_release_bytes(root->fs_info, block_rsv, global_rsv,
@@ -5916,24 +5887,16 @@ static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_delayed_ref_head *head;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	struct btrfs_delayed_ref_node *ref;
-	struct rb_node *node;
 	int ret = 0;
 
 	delayed_refs = &trans->transaction->delayed_refs;
 	spin_lock(&delayed_refs->lock);
 	head = btrfs_find_delayed_ref_head(trans, bytenr);
 	if (!head)
-		goto out;
+		goto out_delayed_unlock;
 
-	node = rb_prev(&head->node.rb_node);
-	if (!node)
-		goto out;
-
-	ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-
-	/* there are still entries for this ref, we can't drop it */
-	if (ref->bytenr == bytenr)
+	spin_lock(&head->lock);
+	if (rb_first(&head->ref_root))
 		goto out;
 
 	if (head->extent_op) {
@@ -5955,19 +5918,19 @@ static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans,
 	 * ahead and process it.
 	 */
 	head->node.in_tree = 0;
-	rb_erase(&head->node.rb_node, &delayed_refs->root);
+	rb_erase(&head->href_node, &delayed_refs->href_root);
 
-	delayed_refs->num_entries--;
+	atomic_dec(&delayed_refs->num_entries);
 
 	/*
 	 * we don't take a ref on the node because we're removing it from the
 	 * tree, so we just steal the ref the tree was holding.
 	 */
 	delayed_refs->num_heads--;
-	if (list_empty(&head->cluster))
+	if (head->processing == 0)
 		delayed_refs->num_heads_ready--;
-
-	list_del_init(&head->cluster);
+	head->processing = 0;
+	spin_unlock(&head->lock);
 	spin_unlock(&delayed_refs->lock);
 
 	BUG_ON(head->extent_op);
@@ -5978,6 +5941,9 @@ static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans,
 	btrfs_put_delayed_ref(&head->node);
 	return ret;
 out:
+	spin_unlock(&head->lock);
+
+out_delayed_unlock:
 	spin_unlock(&delayed_refs->lock);
 	return 0;
 }
@@ -6145,11 +6111,29 @@ int __get_raid_index(u64 flags)
 	return BTRFS_RAID_SINGLE; /* BTRFS_BLOCK_GROUP_SINGLE */
 }
 
-static int get_block_group_index(struct btrfs_block_group_cache *cache)
+int get_block_group_index(struct btrfs_block_group_cache *cache)
 {
 	return __get_raid_index(cache->flags);
 }
 
+static const char *btrfs_raid_type_names[BTRFS_NR_RAID_TYPES] = {
+	[BTRFS_RAID_RAID10]	= "raid10",
+	[BTRFS_RAID_RAID1]	= "raid1",
+	[BTRFS_RAID_DUP]	= "dup",
+	[BTRFS_RAID_RAID0]	= "raid0",
+	[BTRFS_RAID_SINGLE]	= "single",
+	[BTRFS_RAID_RAID5]	= "raid5",
+	[BTRFS_RAID_RAID6]	= "raid6",
+};
+
+static const char *get_raid_name(enum btrfs_raid_types type)
+{
+	if (type >= BTRFS_NR_RAID_TYPES)
+		return NULL;
+
+	return btrfs_raid_type_names[type];
+}
+
 enum btrfs_loop_type {
 	LOOP_CACHING_NOWAIT = 0,
 	LOOP_CACHING_WAIT = 1,
@@ -6177,7 +6161,6 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	struct btrfs_root *root = orig_root->fs_info->extent_root;
 	struct btrfs_free_cluster *last_ptr = NULL;
 	struct btrfs_block_group_cache *block_group = NULL;
-	struct btrfs_block_group_cache *used_block_group;
 	u64 search_start = 0;
 	u64 max_extent_size = 0;
 	int empty_cluster = 2 * 1024 * 1024;
@@ -6186,7 +6169,6 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	int index = __get_raid_index(flags);
 	int alloc_type = (flags & BTRFS_BLOCK_GROUP_DATA) ?
 		RESERVE_ALLOC_NO_ACCOUNT : RESERVE_ALLOC;
-	bool found_uncached_bg = false;
 	bool failed_cluster_refill = false;
 	bool failed_alloc = false;
 	bool use_cluster = true;
@@ -6239,7 +6221,6 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	if (search_start == hint_byte) {
 		block_group = btrfs_lookup_block_group(root->fs_info,
 						       search_start);
-		used_block_group = block_group;
 		/*
 		 * we don't want to use the block group if it doesn't match our
 		 * allocation bits, or if its not cached.
@@ -6276,7 +6257,6 @@ search:
 		u64 offset;
 		int cached;
 
-		used_block_group = block_group;
 		btrfs_get_block_group(block_group);
 		search_start = block_group->key.objectid;
 
@@ -6304,7 +6284,6 @@ search:
 have_block_group:
 		cached = block_group_cache_done(block_group);
 		if (unlikely(!cached)) {
-			found_uncached_bg = true;
 			ret = cache_block_group(block_group, 0);
 			BUG_ON(ret < 0);
 			ret = 0;
@@ -6320,6 +6299,7 @@ have_block_group:
 		 * lets look there
 		 */
 		if (last_ptr) {
+			struct btrfs_block_group_cache *used_block_group;
 			unsigned long aligned_cluster;
 			/*
 			 * the refill lock keeps out other
@@ -6330,10 +6310,8 @@ have_block_group:
 			if (used_block_group != block_group &&
 			    (!used_block_group ||
 			     used_block_group->ro ||
-			     !block_group_bits(used_block_group, flags))) {
-				used_block_group = block_group;
+			     !block_group_bits(used_block_group, flags)))
 				goto refill_cluster;
-			}
 
 			if (used_block_group != block_group)
 				btrfs_get_block_group(used_block_group);
@@ -6347,17 +6325,19 @@ have_block_group:
 				/* we have a block, we're done */
 				spin_unlock(&last_ptr->refill_lock);
 				trace_btrfs_reserve_extent_cluster(root,
-					block_group, search_start, num_bytes);
+						used_block_group,
+						search_start, num_bytes);
+				if (used_block_group != block_group) {
+					btrfs_put_block_group(block_group);
+					block_group = used_block_group;
+				}
 				goto checks;
 			}
 
 			WARN_ON(last_ptr->block_group != used_block_group);
-			if (used_block_group != block_group) {
+			if (used_block_group != block_group)
 				btrfs_put_block_group(used_block_group);
-				used_block_group = block_group;
-			}
 refill_cluster:
-			BUG_ON(used_block_group != block_group);
 			/* If we are on LOOP_NO_EMPTY_SIZE, we can't
 			 * set up a new clusters, so lets just skip it
 			 * and let the allocator find whatever block
@@ -6476,25 +6456,25 @@ unclustered_alloc:
 			goto loop;
 		}
 checks:
-		search_start = stripe_align(root, used_block_group,
+		search_start = stripe_align(root, block_group,
 					    offset, num_bytes);
 
 		/* move on to the next group */
 		if (search_start + num_bytes >
-		    used_block_group->key.objectid + used_block_group->key.offset) {
-			btrfs_add_free_space(used_block_group, offset, num_bytes);
+		    block_group->key.objectid + block_group->key.offset) {
+			btrfs_add_free_space(block_group, offset, num_bytes);
 			goto loop;
 		}
 
 		if (offset < search_start)
-			btrfs_add_free_space(used_block_group, offset,
+			btrfs_add_free_space(block_group, offset,
 					     search_start - offset);
 		BUG_ON(offset > search_start);
 
-		ret = btrfs_update_reserved_bytes(used_block_group, num_bytes,
+		ret = btrfs_update_reserved_bytes(block_group, num_bytes,
 						  alloc_type);
 		if (ret == -EAGAIN) {
-			btrfs_add_free_space(used_block_group, offset, num_bytes);
+			btrfs_add_free_space(block_group, offset, num_bytes);
 			goto loop;
 		}
 
@@ -6504,16 +6484,12 @@ checks:
 
 		trace_btrfs_reserve_extent(orig_root, block_group,
 					   search_start, num_bytes);
-		if (used_block_group != block_group)
-			btrfs_put_block_group(used_block_group);
 		btrfs_put_block_group(block_group);
 		break;
 loop:
 		failed_cluster_refill = false;
 		failed_alloc = false;
 		BUG_ON(index != get_block_group_index(block_group));
-		if (used_block_group != block_group)
-			btrfs_put_block_group(used_block_group);
 		btrfs_put_block_group(block_group);
 	}
 	up_read(&space_info->groups_sem);
@@ -6584,12 +6560,12 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes,
 	int index = 0;
 
 	spin_lock(&info->lock);
-	printk(KERN_INFO "space_info %llu has %llu free, is %sfull\n",
+	printk(KERN_INFO "BTRFS: space_info %llu has %llu free, is %sfull\n",
 	       info->flags,
 	       info->total_bytes - info->bytes_used - info->bytes_pinned -
 	       info->bytes_reserved - info->bytes_readonly,
 	       (info->full) ? "" : "not ");
-	printk(KERN_INFO "space_info total=%llu, used=%llu, pinned=%llu, "
+	printk(KERN_INFO "BTRFS: space_info total=%llu, used=%llu, pinned=%llu, "
 	       "reserved=%llu, may_use=%llu, readonly=%llu\n",
 	       info->total_bytes, info->bytes_used, info->bytes_pinned,
 	       info->bytes_reserved, info->bytes_may_use,
@@ -6603,7 +6579,9 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes,
 again:
 	list_for_each_entry(cache, &info->block_groups[index], list) {
 		spin_lock(&cache->lock);
-		printk(KERN_INFO "block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %s\n",
+		printk(KERN_INFO "BTRFS: "
+			   "block group %llu has %llu bytes, "
+			   "%llu used %llu pinned %llu reserved %s\n",
 		       cache->key.objectid, cache->key.offset,
 		       btrfs_block_group_used(&cache->item), cache->pinned,
 		       cache->reserved, cache->ro ? "[readonly]" : "");
@@ -6966,7 +6944,7 @@ again:
 				/*DEFAULT_RATELIMIT_BURST*/ 1);
 		if (__ratelimit(&_rs))
 			WARN(1, KERN_DEBUG
-				"btrfs: block rsv returned %d\n", ret);
+				"BTRFS: block rsv returned %d\n", ret);
 	}
 try_reserve:
 	ret = reserve_metadata_bytes(root, block_rsv, blocksize,
@@ -7714,7 +7692,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 
 			btrfs_end_transaction_throttle(trans, tree_root);
 			if (!for_reloc && btrfs_need_cleaner_sleep(root)) {
-				pr_debug("btrfs: drop snapshot early exit\n");
+				pr_debug("BTRFS: drop snapshot early exit\n");
 				err = -EAGAIN;
 				goto out_free;
 			}
@@ -7779,7 +7757,7 @@ out:
 	 */
 	if (!for_reloc && root_dropped == false)
 		btrfs_add_dead_root(root);
-	if (err)
+	if (err && err != -EAGAIN)
 		btrfs_std_error(root->fs_info, err);
 	return err;
 }
@@ -8333,6 +8311,8 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 	release_global_block_rsv(info);
 
 	while (!list_empty(&info->space_info)) {
+		int i;
+
 		space_info = list_entry(info->space_info.next,
 					struct btrfs_space_info,
 					list);
@@ -8343,9 +8323,17 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 				dump_space_info(space_info, 0, 0);
 			}
 		}
-		percpu_counter_destroy(&space_info->total_bytes_pinned);
 		list_del(&space_info->list);
-		kfree(space_info);
+		for (i = 0; i < BTRFS_NR_RAID_TYPES; i++) {
+			struct kobject *kobj;
+			kobj = &space_info->block_group_kobjs[i];
+			if (kobj->parent) {
+				kobject_del(kobj);
+				kobject_put(kobj);
+			}
+		}
+		kobject_del(&space_info->kobj);
+		kobject_put(&space_info->kobj);
 	}
 	return 0;
 }
@@ -8356,10 +8344,57 @@ static void __link_block_group(struct btrfs_space_info *space_info,
 	int index = get_block_group_index(cache);
 
 	down_write(&space_info->groups_sem);
+	if (list_empty(&space_info->block_groups[index])) {
+		struct kobject *kobj = &space_info->block_group_kobjs[index];
+		int ret;
+
+		kobject_get(&space_info->kobj); /* put in release */
+		ret = kobject_add(kobj, &space_info->kobj, "%s",
+				  get_raid_name(index));
+		if (ret) {
+			pr_warn("BTRFS: failed to add kobject for block cache. ignoring.\n");
+			kobject_put(&space_info->kobj);
+		}
+	}
 	list_add_tail(&cache->list, &space_info->block_groups[index]);
 	up_write(&space_info->groups_sem);
 }
 
+static struct btrfs_block_group_cache *
+btrfs_create_block_group_cache(struct btrfs_root *root, u64 start, u64 size)
+{
+	struct btrfs_block_group_cache *cache;
+
+	cache = kzalloc(sizeof(*cache), GFP_NOFS);
+	if (!cache)
+		return NULL;
+
+	cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
+					GFP_NOFS);
+	if (!cache->free_space_ctl) {
+		kfree(cache);
+		return NULL;
+	}
+
+	cache->key.objectid = start;
+	cache->key.offset = size;
+	cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
+
+	cache->sectorsize = root->sectorsize;
+	cache->fs_info = root->fs_info;
+	cache->full_stripe_len = btrfs_full_stripe_len(root,
+					       &root->fs_info->mapping_tree,
+					       start);
+	atomic_set(&cache->count, 1);
+	spin_lock_init(&cache->lock);
+	INIT_LIST_HEAD(&cache->list);
+	INIT_LIST_HEAD(&cache->cluster_list);
+	INIT_LIST_HEAD(&cache->new_bg_list);
+	btrfs_init_free_space_ctl(cache);
+
+	return cache;
+}
+
 int btrfs_read_block_groups(struct btrfs_root *root)
 {
 	struct btrfs_path *path;
@@ -8395,26 +8430,16 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 			break;
 		if (ret != 0)
 			goto error;
+
 		leaf = path->nodes[0];
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
-		cache = kzalloc(sizeof(*cache), GFP_NOFS);
+
+		cache = btrfs_create_block_group_cache(root, found_key.objectid,
+						       found_key.offset);
 		if (!cache) {
 			ret = -ENOMEM;
 			goto error;
 		}
-		cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
-						GFP_NOFS);
-		if (!cache->free_space_ctl) {
-			kfree(cache);
-			ret = -ENOMEM;
-			goto error;
-		}
-
-		atomic_set(&cache->count, 1);
-		spin_lock_init(&cache->lock);
-		cache->fs_info = info;
-		INIT_LIST_HEAD(&cache->list);
-		INIT_LIST_HEAD(&cache->cluster_list);
 
 		if (need_clear) {
 			/*
@@ -8435,16 +8460,10 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 		read_extent_buffer(leaf, &cache->item,
 				   btrfs_item_ptr_offset(leaf, path->slots[0]),
 				   sizeof(cache->item));
-		memcpy(&cache->key, &found_key, sizeof(found_key));
+		cache->flags = btrfs_block_group_flags(&cache->item);
 
 		key.objectid = found_key.objectid + found_key.offset;
 		btrfs_release_path(path);
-		cache->flags = btrfs_block_group_flags(&cache->item);
-		cache->sectorsize = root->sectorsize;
-		cache->full_stripe_len = btrfs_full_stripe_len(root,
-					       &root->fs_info->mapping_tree,
-					       found_key.objectid);
-		btrfs_init_free_space_ctl(cache);
 
 		/*
 		 * We need to exclude the super stripes now so that the space
@@ -8458,8 +8477,7 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 			 * case.
 			 */
 			free_excluded_extents(root, cache);
-			kfree(cache->free_space_ctl);
-			kfree(cache);
+			btrfs_put_block_group(cache);
 			goto error;
 		}
 
@@ -8590,38 +8608,15 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans,
 
 	root->fs_info->last_trans_log_full_commit = trans->transid;
 
-	cache = kzalloc(sizeof(*cache), GFP_NOFS);
+	cache = btrfs_create_block_group_cache(root, chunk_offset, size);
 	if (!cache)
 		return -ENOMEM;
-	cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
-					GFP_NOFS);
-	if (!cache->free_space_ctl) {
-		kfree(cache);
-		return -ENOMEM;
-	}
-
-	cache->key.objectid = chunk_offset;
-	cache->key.offset = size;
-	cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
-	cache->sectorsize = root->sectorsize;
-	cache->fs_info = root->fs_info;
-	cache->full_stripe_len = btrfs_full_stripe_len(root,
-					       &root->fs_info->mapping_tree,
-					       chunk_offset);
-
-	atomic_set(&cache->count, 1);
-	spin_lock_init(&cache->lock);
-	INIT_LIST_HEAD(&cache->list);
-	INIT_LIST_HEAD(&cache->cluster_list);
-	INIT_LIST_HEAD(&cache->new_bg_list);
-
-	btrfs_init_free_space_ctl(cache);
 
 	btrfs_set_block_group_used(&cache->item, bytes_used);
 	btrfs_set_block_group_chunk_objectid(&cache->item, chunk_objectid);
-	cache->flags = type;
 	btrfs_set_block_group_flags(&cache->item, type);
 
+	cache->flags = type;
 	cache->last_byte_to_unpin = (u64)-1;
 	cache->cached = BTRFS_CACHE_FINISHED;
 	ret = exclude_super_stripes(root, cache);
@@ -8631,8 +8626,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans,
 		 * case.
 		 */
 		free_excluded_extents(root, cache);
-		kfree(cache->free_space_ctl);
-		kfree(cache);
+		btrfs_put_block_group(cache);
 		return ret;
 	}
 
@@ -8796,8 +8790,11 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 	 * are still on the list after taking the semaphore
 	 */
 	list_del_init(&block_group->list);
-	if (list_empty(&block_group->space_info->block_groups[index]))
+	if (list_empty(&block_group->space_info->block_groups[index])) {
+		kobject_del(&block_group->space_info->block_group_kobjs[index]);
+		kobject_put(&block_group->space_info->block_group_kobjs[index]);
 		clear_avail_alloc_bits(root->fs_info, block_group->flags);
+	}
 	up_write(&block_group->space_info->groups_sem);
 
 	if (block_group->cached == BTRFS_CACHE_STARTED)

fs/btrfs/extent_io.c

@@ -59,7 +59,7 @@ void btrfs_leak_debug_check(void)
 
 	while (!list_empty(&states)) {
 		state = list_entry(states.next, struct extent_state, leak_list);
-		printk(KERN_ERR "btrfs state leak: start %llu end %llu "
+		printk(KERN_ERR "BTRFS: state leak: start %llu end %llu "
 		       "state %lu in tree %p refs %d\n",
 		       state->start, state->end, state->state, state->tree,
 		       atomic_read(&state->refs));
@@ -69,7 +69,7 @@ void btrfs_leak_debug_check(void)
 
 	while (!list_empty(&buffers)) {
 		eb = list_entry(buffers.next, struct extent_buffer, leak_list);
-		printk(KERN_ERR "btrfs buffer leak start %llu len %lu "
+		printk(KERN_ERR "BTRFS: buffer leak start %llu len %lu "
 		       "refs %d\n",
 		       eb->start, eb->len, atomic_read(&eb->refs));
 		list_del(&eb->leak_list);
@@ -77,16 +77,22 @@ void btrfs_leak_debug_check(void)
 	}
 }
 
-#define btrfs_debug_check_extent_io_range(inode, start, end)		\
-	__btrfs_debug_check_extent_io_range(__func__, (inode), (start), (end))
+#define btrfs_debug_check_extent_io_range(tree, start, end)		\
+	__btrfs_debug_check_extent_io_range(__func__, (tree), (start), (end))
 static inline void __btrfs_debug_check_extent_io_range(const char *caller,
-		struct inode *inode, u64 start, u64 end)
+		struct extent_io_tree *tree, u64 start, u64 end)
 {
-	u64 isize = i_size_read(inode);
+	struct inode *inode;
+	u64 isize;
+
+	if (!tree->mapping)
+		return;
 
+	inode = tree->mapping->host;
+	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",
+		    "BTRFS: %s: ino %llu isize %llu odd range [%llu,%llu]\n",
 				caller, btrfs_ino(inode), isize, start, end);
 	}
 }
@@ -124,6 +130,8 @@ static noinline void flush_write_bio(void *data);
 static inline struct btrfs_fs_info *
 tree_fs_info(struct extent_io_tree *tree)
 {
+	if (!tree->mapping)
+		return NULL;
 	return btrfs_sb(tree->mapping->host->i_sb);
 }
 
@@ -186,11 +194,9 @@ void extent_io_tree_init(struct extent_io_tree *tree,
 			 struct address_space *mapping)
 {
 	tree->state = RB_ROOT;
-	INIT_RADIX_TREE(&tree->buffer, GFP_ATOMIC);
 	tree->ops = NULL;
 	tree->dirty_bytes = 0;
 	spin_lock_init(&tree->lock);
-	spin_lock_init(&tree->buffer_lock);
 	tree->mapping = mapping;
 }
 
@@ -224,12 +230,20 @@ void free_extent_state(struct extent_state *state)
 }
 
 static struct rb_node *tree_insert(struct rb_root *root, u64 offset,
-				   struct rb_node *node)
+				   struct rb_node *node,
+				   struct rb_node ***p_in,
+				   struct rb_node **parent_in)
 {
 	struct rb_node **p = &root->rb_node;
 	struct rb_node *parent = NULL;
 	struct tree_entry *entry;
 
+	if (p_in && parent_in) {
+		p = *p_in;
+		parent = *parent_in;
+		goto do_insert;
+	}
+
 	while (*p) {
 		parent = *p;
 		entry = rb_entry(parent, struct tree_entry, rb_node);
@@ -242,35 +256,43 @@ static struct rb_node *tree_insert(struct rb_root *root, u64 offset,
 			return parent;
 	}
 
+do_insert:
 	rb_link_node(node, parent, p);
 	rb_insert_color(node, root);
 	return NULL;
 }
 
 static struct rb_node *__etree_search(struct extent_io_tree *tree, u64 offset,
-				     struct rb_node **prev_ret,
-				     struct rb_node **next_ret)
+				      struct rb_node **prev_ret,
+				      struct rb_node **next_ret,
+				      struct rb_node ***p_ret,
+				      struct rb_node **parent_ret)
 {
 	struct rb_root *root = &tree->state;
-	struct rb_node *n = root->rb_node;
+	struct rb_node **n = &root->rb_node;
 	struct rb_node *prev = NULL;
 	struct rb_node *orig_prev = NULL;
 	struct tree_entry *entry;
 	struct tree_entry *prev_entry = NULL;
 
-	while (n) {
-		entry = rb_entry(n, struct tree_entry, rb_node);
-		prev = n;
+	while (*n) {
+		prev = *n;
+		entry = rb_entry(prev, struct tree_entry, rb_node);
 		prev_entry = entry;
 
 		if (offset < entry->start)
-			n = n->rb_left;
+			n = &(*n)->rb_left;
 		else if (offset > entry->end)
-			n = n->rb_right;
+			n = &(*n)->rb_right;
 		else
-			return n;
+			return *n;
 	}
 
+	if (p_ret)
+		*p_ret = n;
+	if (parent_ret)
+		*parent_ret = prev;
+
 	if (prev_ret) {
 		orig_prev = prev;
 		while (prev && offset > prev_entry->end) {
@@ -292,18 +314,27 @@ static struct rb_node *__etree_search(struct extent_io_tree *tree, u64 offset,
 	return NULL;
 }
 
-static inline struct rb_node *tree_search(struct extent_io_tree *tree,
-					  u64 offset)
+static inline struct rb_node *
+tree_search_for_insert(struct extent_io_tree *tree,
+		       u64 offset,
+		       struct rb_node ***p_ret,
+		       struct rb_node **parent_ret)
 {
 	struct rb_node *prev = NULL;
 	struct rb_node *ret;
 
-	ret = __etree_search(tree, offset, &prev, NULL);
+	ret = __etree_search(tree, offset, &prev, NULL, p_ret, parent_ret);
 	if (!ret)
 		return prev;
 	return ret;
 }
 
+static inline struct rb_node *tree_search(struct extent_io_tree *tree,
+					  u64 offset)
+{
+	return tree_search_for_insert(tree, offset, NULL, NULL);
+}
+
 static void merge_cb(struct extent_io_tree *tree, struct extent_state *new,
 		     struct extent_state *other)
 {
@@ -385,23 +416,25 @@ static void set_state_bits(struct extent_io_tree *tree,
  */
 static int insert_state(struct extent_io_tree *tree,
 			struct extent_state *state, u64 start, u64 end,
+			struct rb_node ***p,
+			struct rb_node **parent,
 			unsigned long *bits)
 {
 	struct rb_node *node;
 
 	if (end < start)
-		WARN(1, KERN_ERR "btrfs end < start %llu %llu\n",
+		WARN(1, KERN_ERR "BTRFS: end < start %llu %llu\n",
 		       end, start);
 	state->start = start;
 	state->end = end;
 
 	set_state_bits(tree, state, bits);
 
-	node = tree_insert(&tree->state, end, &state->rb_node);
+	node = tree_insert(&tree->state, end, &state->rb_node, p, parent);
 	if (node) {
 		struct extent_state *found;
 		found = rb_entry(node, struct extent_state, rb_node);
-		printk(KERN_ERR "btrfs found node %llu %llu on insert of "
+		printk(KERN_ERR "BTRFS: found node %llu %llu on insert of "
 		       "%llu %llu\n",
 		       found->start, found->end, start, end);
 		return -EEXIST;
@@ -444,7 +477,8 @@ static int split_state(struct extent_io_tree *tree, struct extent_state *orig,
 	prealloc->state = orig->state;
 	orig->start = split;
 
-	node = tree_insert(&tree->state, prealloc->end, &prealloc->rb_node);
+	node = tree_insert(&tree->state, prealloc->end, &prealloc->rb_node,
+			   NULL, NULL);
 	if (node) {
 		free_extent_state(prealloc);
 		return -EEXIST;
@@ -542,7 +576,7 @@ int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 	int err;
 	int clear = 0;
 
-	btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+	btrfs_debug_check_extent_io_range(tree, start, end);
 
 	if (bits & EXTENT_DELALLOC)
 		bits |= EXTENT_NORESERVE;
@@ -702,7 +736,7 @@ static void wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 	struct extent_state *state;
 	struct rb_node *node;
 
-	btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+	btrfs_debug_check_extent_io_range(tree, start, end);
 
 	spin_lock(&tree->lock);
 again:
@@ -783,11 +817,13 @@ __set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 	struct extent_state *state;
 	struct extent_state *prealloc = NULL;
 	struct rb_node *node;
+	struct rb_node **p;
+	struct rb_node *parent;
 	int err = 0;
 	u64 last_start;
 	u64 last_end;
 
-	btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+	btrfs_debug_check_extent_io_range(tree, start, end);
 
 	bits |= EXTENT_FIRST_DELALLOC;
 again:
@@ -809,14 +845,16 @@ again:
 	 * this search will find all the extents that end after
 	 * our range starts.
 	 */
-	node = tree_search(tree, start);
+	node = tree_search_for_insert(tree, start, &p, &parent);
 	if (!node) {
 		prealloc = alloc_extent_state_atomic(prealloc);
 		BUG_ON(!prealloc);
-		err = insert_state(tree, prealloc, start, end, &bits);
+		err = insert_state(tree, prealloc, start, end,
+				   &p, &parent, &bits);
 		if (err)
 			extent_io_tree_panic(tree, err);
 
+		cache_state(prealloc, cached_state);
 		prealloc = NULL;
 		goto out;
 	}
@@ -919,7 +957,7 @@ hit_next:
 		 * the later extent.
 		 */
 		err = insert_state(tree, prealloc, start, this_end,
-				   &bits);
+				   NULL, NULL, &bits);
 		if (err)
 			extent_io_tree_panic(tree, err);
 
@@ -1005,11 +1043,13 @@ int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 	struct extent_state *state;
 	struct extent_state *prealloc = NULL;
 	struct rb_node *node;
+	struct rb_node **p;
+	struct rb_node *parent;
 	int err = 0;
 	u64 last_start;
 	u64 last_end;
 
-	btrfs_debug_check_extent_io_range(tree->mapping->host, start, end);
+	btrfs_debug_check_extent_io_range(tree, start, end);
 
 again:
 	if (!prealloc && (mask & __GFP_WAIT)) {
@@ -1032,17 +1072,19 @@ again:
 	 * this search will find all the extents that end after
 	 * our range starts.
 	 */
-	node = tree_search(tree, start);
+	node = tree_search_for_insert(tree, start, &p, &parent);
 	if (!node) {
 		prealloc = alloc_extent_state_atomic(prealloc);
 		if (!prealloc) {
 			err = -ENOMEM;
 			goto out;
 		}
-		err = insert_state(tree, prealloc, start, end, &bits);
-		prealloc = NULL;
+		err = insert_state(tree, prealloc, start, end,
+				   &p, &parent, &bits);
 		if (err)
 			extent_io_tree_panic(tree, err);
+		cache_state(prealloc, cached_state);
+		prealloc = NULL;
 		goto out;
 	}
 	state = rb_entry(node, struct extent_state, rb_node);
@@ -1135,7 +1177,7 @@ hit_next:
 		 * the later extent.
 		 */
 		err = insert_state(tree, prealloc, start, this_end,
-				   &bits);
+				   NULL, NULL, &bits);
 		if (err)
 			extent_io_tree_panic(tree, err);
 		cache_state(prealloc, cached_state);
@@ -2012,9 +2054,10 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start,
 		return -EIO;
 	}
 
-	printk_ratelimited_in_rcu(KERN_INFO "btrfs read error corrected: ino %lu off %llu "
-		      "(dev %s sector %llu)\n", page->mapping->host->i_ino,
-		      start, rcu_str_deref(dev->name), sector);
+	printk_ratelimited_in_rcu(KERN_INFO
+			"BTRFS: read error corrected: ino %lu off %llu "
+		    "(dev %s sector %llu)\n", page->mapping->host->i_ino,
+		    start, rcu_str_deref(dev->name), sector);
 
 	bio_put(bio);
 	return 0;
@@ -2156,7 +2199,7 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 			return -EIO;
 		}
 
-		if (em->start > start || em->start + em->len < start) {
+		if (em->start > start || em->start + em->len <= start) {
 			free_extent_map(em);
 			em = NULL;
 		}
@@ -2333,25 +2376,29 @@ int end_extent_writepage(struct page *page, int err, u64 start, u64 end)
 static void end_bio_extent_writepage(struct bio *bio, int err)
 {
 	struct bio_vec *bvec;
-	struct extent_io_tree *tree;
 	u64 start;
 	u64 end;
 	int i;
 
 	bio_for_each_segment_all(bvec, bio, i) {
 		struct page *page = bvec->bv_page;
-		tree = &BTRFS_I(page->mapping->host)->io_tree;
 
 		/* We always issue full-page reads, but if some block
 		 * in a page fails to read, blk_update_request() will
 		 * advance bv_offset and adjust bv_len to compensate.
 		 * Print a warning for nonzero offsets, and an error
 		 * if they don't add up to a full page.  */
-		if (bvec->bv_offset || bvec->bv_len != PAGE_CACHE_SIZE)
-			printk("%s page write in btrfs with offset %u and length %u\n",
-			       bvec->bv_offset + bvec->bv_len != PAGE_CACHE_SIZE
-			       ? KERN_ERR "partial" : KERN_INFO "incomplete",
-			       bvec->bv_offset, bvec->bv_len);
+		if (bvec->bv_offset || bvec->bv_len != PAGE_CACHE_SIZE) {
+			if (bvec->bv_offset + bvec->bv_len != PAGE_CACHE_SIZE)
+				btrfs_err(BTRFS_I(page->mapping->host)->root->fs_info,
+				   "partial page write in btrfs with offset %u and length %u",
+					bvec->bv_offset, bvec->bv_len);
+			else
+				btrfs_info(BTRFS_I(page->mapping->host)->root->fs_info,
+				   "incomplete page write in btrfs with offset %u and "
+				   "length %u",
+					bvec->bv_offset, bvec->bv_len);
+		}
 
 		start = page_offset(page);
 		end = start + bvec->bv_offset + bvec->bv_len - 1;
@@ -2421,11 +2468,17 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 		 * advance bv_offset and adjust bv_len to compensate.
 		 * Print a warning for nonzero offsets, and an error
 		 * if they don't add up to a full page.  */
-		if (bvec->bv_offset || bvec->bv_len != PAGE_CACHE_SIZE)
-			printk("%s page read in btrfs with offset %u and length %u\n",
-			       bvec->bv_offset + bvec->bv_len != PAGE_CACHE_SIZE
-			       ? KERN_ERR "partial" : KERN_INFO "incomplete",
-			       bvec->bv_offset, bvec->bv_len);
+		if (bvec->bv_offset || bvec->bv_len != PAGE_CACHE_SIZE) {
+			if (bvec->bv_offset + bvec->bv_len != PAGE_CACHE_SIZE)
+				btrfs_err(BTRFS_I(page->mapping->host)->root->fs_info,
+				   "partial page read in btrfs with offset %u and length %u",
+					bvec->bv_offset, bvec->bv_len);
+			else
+				btrfs_info(BTRFS_I(page->mapping->host)->root->fs_info,
+				   "incomplete page read in btrfs with offset %u and "
+				   "length %u",
+					bvec->bv_offset, bvec->bv_len);
+		}
 
 		start = page_offset(page);
 		end = start + bvec->bv_offset + bvec->bv_len - 1;
@@ -3281,8 +3334,8 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc,
 
 			set_range_writeback(tree, cur, cur + iosize - 1);
 			if (!PageWriteback(page)) {
-				printk(KERN_ERR "btrfs warning page %lu not "
-				       "writeback, cur %llu end %llu\n",
+				btrfs_err(BTRFS_I(inode)->root->fs_info,
+					   "page %lu not writeback, cur %llu end %llu",
 				       page->index, cur, end);
 			}
 
@@ -3438,6 +3491,7 @@ static int write_one_eb(struct extent_buffer *eb,
 			struct extent_page_data *epd)
 {
 	struct block_device *bdev = fs_info->fs_devices->latest_bdev;
+	struct extent_io_tree *tree = &BTRFS_I(fs_info->btree_inode)->io_tree;
 	u64 offset = eb->start;
 	unsigned long i, num_pages;
 	unsigned long bio_flags = 0;
@@ -3455,7 +3509,7 @@ static int write_one_eb(struct extent_buffer *eb,
 
 		clear_page_dirty_for_io(p);
 		set_page_writeback(p);
-		ret = submit_extent_page(rw, eb->tree, p, offset >> 9,
+		ret = submit_extent_page(rw, tree, p, offset >> 9,
 					 PAGE_CACHE_SIZE, 0, bdev, &epd->bio,
 					 -1, end_bio_extent_buffer_writepage,
 					 0, epd->bio_flags, bio_flags);
@@ -4073,12 +4127,10 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	struct extent_map *em = NULL;
 	struct extent_state *cached_state = NULL;
 	struct btrfs_path *path;
-	struct btrfs_file_extent_item *item;
 	int end = 0;
 	u64 em_start = 0;
 	u64 em_len = 0;
 	u64 em_end = 0;
-	unsigned long emflags;
 
 	if (len == 0)
 		return -EINVAL;
@@ -4103,8 +4155,6 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	}
 	WARN_ON(!ret);
 	path->slots[0]--;
-	item = btrfs_item_ptr(path->nodes[0], path->slots[0],
-			      struct btrfs_file_extent_item);
 	btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]);
 	found_type = btrfs_key_type(&found_key);
 
@@ -4172,7 +4222,6 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 			offset_in_extent = em_start - em->start;
 		em_end = extent_map_end(em);
 		em_len = em_end - em_start;
-		emflags = em->flags;
 		disko = 0;
 		flags = 0;
 
@@ -4324,10 +4373,9 @@ static inline void btrfs_release_extent_buffer(struct extent_buffer *eb)
 	__free_extent_buffer(eb);
 }
 
-static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree,
-						   u64 start,
-						   unsigned long len,
-						   gfp_t mask)
+static struct extent_buffer *
+__alloc_extent_buffer(struct btrfs_fs_info *fs_info, u64 start,
+		      unsigned long len, gfp_t mask)
 {
 	struct extent_buffer *eb = NULL;
 
@@ -4336,7 +4384,7 @@ static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree,
 		return NULL;
 	eb->start = start;
 	eb->len = len;
-	eb->tree = tree;
+	eb->fs_info = fs_info;
 	eb->bflags = 0;
 	rwlock_init(&eb->lock);
 	atomic_set(&eb->write_locks, 0);
@@ -4468,13 +4516,14 @@ static void mark_extent_buffer_accessed(struct extent_buffer *eb)
 	}
 }
 
-struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree,
-					 		u64 start)
+struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info,
+					 u64 start)
 {
 	struct extent_buffer *eb;
 
 	rcu_read_lock();
-	eb = radix_tree_lookup(&tree->buffer, start >> PAGE_CACHE_SHIFT);
+	eb = radix_tree_lookup(&fs_info->buffer_radix,
+			       start >> PAGE_CACHE_SHIFT);
 	if (eb && atomic_inc_not_zero(&eb->refs)) {
 		rcu_read_unlock();
 		mark_extent_buffer_accessed(eb);
@@ -4485,7 +4534,7 @@ struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree,
 	return NULL;
 }
 
-struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
+struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info,
 					  u64 start, unsigned long len)
 {
 	unsigned long num_pages = num_extent_pages(start, len);
@@ -4494,16 +4543,15 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
 	struct extent_buffer *eb;
 	struct extent_buffer *exists = NULL;
 	struct page *p;
-	struct address_space *mapping = tree->mapping;
+	struct address_space *mapping = fs_info->btree_inode->i_mapping;
 	int uptodate = 1;
 	int ret;
 
-
-	eb = find_extent_buffer(tree, start);
+	eb = find_extent_buffer(fs_info, start);
 	if (eb)
 		return eb;
 
-	eb = __alloc_extent_buffer(tree, start, len, GFP_NOFS);
+	eb = __alloc_extent_buffer(fs_info, start, len, GFP_NOFS);
 	if (!eb)
 		return NULL;
 
@@ -4558,12 +4606,13 @@ again:
 	if (ret)
 		goto free_eb;
 
-	spin_lock(&tree->buffer_lock);
-	ret = radix_tree_insert(&tree->buffer, start >> PAGE_CACHE_SHIFT, eb);
-	spin_unlock(&tree->buffer_lock);
+	spin_lock(&fs_info->buffer_lock);
+	ret = radix_tree_insert(&fs_info->buffer_radix,
+				start >> PAGE_CACHE_SHIFT, eb);
+	spin_unlock(&fs_info->buffer_lock);
 	radix_tree_preload_end();
 	if (ret == -EEXIST) {
-		exists = find_extent_buffer(tree, start);
+		exists = find_extent_buffer(fs_info, start);
 		if (exists)
 			goto free_eb;
 		else
@@ -4571,6 +4620,7 @@ again:
 	}
 	/* add one reference for the tree */
 	check_buffer_tree_ref(eb);
+	set_bit(EXTENT_BUFFER_IN_TREE, &eb->bflags);
 
 	/*
 	 * there is a race where release page may have
@@ -4614,17 +4664,17 @@ static int release_extent_buffer(struct extent_buffer *eb)
 {
 	WARN_ON(atomic_read(&eb->refs) == 0);
 	if (atomic_dec_and_test(&eb->refs)) {
-		if (test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags)) {
-			spin_unlock(&eb->refs_lock);
-		} else {
-			struct extent_io_tree *tree = eb->tree;
+		if (test_and_clear_bit(EXTENT_BUFFER_IN_TREE, &eb->bflags)) {
+			struct btrfs_fs_info *fs_info = eb->fs_info;
 
 			spin_unlock(&eb->refs_lock);
 
-			spin_lock(&tree->buffer_lock);
-			radix_tree_delete(&tree->buffer,
+			spin_lock(&fs_info->buffer_lock);
+			radix_tree_delete(&fs_info->buffer_radix,
 					  eb->start >> PAGE_CACHE_SHIFT);
-			spin_unlock(&tree->buffer_lock);
+			spin_unlock(&fs_info->buffer_lock);
+		} else {
+			spin_unlock(&eb->refs_lock);
 		}
 
 		/* Should be safe to release our pages at this point */
@@ -5103,12 +5153,12 @@ void memcpy_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 	unsigned long src_i;
 
 	if (src_offset + len > dst->len) {
-		printk(KERN_ERR "btrfs memmove bogus src_offset %lu move "
+		printk(KERN_ERR "BTRFS: memmove bogus src_offset %lu move "
 		       "len %lu dst len %lu\n", src_offset, len, dst->len);
 		BUG_ON(1);
 	}
 	if (dst_offset + len > dst->len) {
-		printk(KERN_ERR "btrfs memmove bogus dst_offset %lu move "
+		printk(KERN_ERR "BTRFS: memmove bogus dst_offset %lu move "
 		       "len %lu dst len %lu\n", dst_offset, len, dst->len);
 		BUG_ON(1);
 	}
@@ -5150,12 +5200,12 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 	unsigned long src_i;
 
 	if (src_offset + len > dst->len) {
-		printk(KERN_ERR "btrfs memmove bogus src_offset %lu move "
+		printk(KERN_ERR "BTRFS: memmove bogus src_offset %lu move "
 		       "len %lu len %lu\n", src_offset, len, dst->len);
 		BUG_ON(1);
 	}
 	if (dst_offset + len > dst->len) {
-		printk(KERN_ERR "btrfs memmove bogus dst_offset %lu move "
+		printk(KERN_ERR "BTRFS: memmove bogus dst_offset %lu move "
 		       "len %lu len %lu\n", dst_offset, len, dst->len);
 		BUG_ON(1);
 	}

fs/btrfs/extent_io.h

@@ -43,6 +43,7 @@
 #define EXTENT_BUFFER_WRITEBACK 7
 #define EXTENT_BUFFER_IOERR 8
 #define EXTENT_BUFFER_DUMMY 9
+#define EXTENT_BUFFER_IN_TREE 10
 
 /* these are flags for extent_clear_unlock_delalloc */
 #define PAGE_UNLOCK		(1 << 0)
@@ -94,12 +95,10 @@ struct extent_io_ops {
 
 struct extent_io_tree {
 	struct rb_root state;
-	struct radix_tree_root buffer;
 	struct address_space *mapping;
 	u64 dirty_bytes;
 	int track_uptodate;
 	spinlock_t lock;
-	spinlock_t buffer_lock;
 	struct extent_io_ops *ops;
 };
 
@@ -130,7 +129,7 @@ struct extent_buffer {
 	unsigned long map_start;
 	unsigned long map_len;
 	unsigned long bflags;
-	struct extent_io_tree *tree;
+	struct btrfs_fs_info *fs_info;
 	spinlock_t refs_lock;
 	atomic_t refs;
 	atomic_t io_pages;
@@ -266,11 +265,11 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private);
 void set_page_extent_mapped(struct page *page);
 
-struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
+struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info,
 					  u64 start, unsigned long len);
 struct extent_buffer *alloc_dummy_extent_buffer(u64 start, unsigned long len);
 struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src);
-struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree,
+struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info,
 					 u64 start);
 void free_extent_buffer(struct extent_buffer *eb);
 void free_extent_buffer_stale(struct extent_buffer *eb);

fs/btrfs/extent_map.c

@@ -79,12 +79,21 @@ void free_extent_map(struct extent_map *em)
 	}
 }
 
-static struct rb_node *tree_insert(struct rb_root *root, u64 offset,
-				   struct rb_node *node)
+/* simple helper to do math around the end of an extent, handling wrap */
+static u64 range_end(u64 start, u64 len)
+{
+	if (start + len < start)
+		return (u64)-1;
+	return start + len;
+}
+
+static int tree_insert(struct rb_root *root, struct extent_map *em)
 {
 	struct rb_node **p = &root->rb_node;
 	struct rb_node *parent = NULL;
-	struct extent_map *entry;
+	struct extent_map *entry = NULL;
+	struct rb_node *orig_parent = NULL;
+	u64 end = range_end(em->start, em->len);
 
 	while (*p) {
 		parent = *p;
@@ -92,19 +101,37 @@ static struct rb_node *tree_insert(struct rb_root *root, u64 offset,
 
 		WARN_ON(!entry->in_tree);
 
-		if (offset < entry->start)
+		if (em->start < entry->start)
 			p = &(*p)->rb_left;
-		else if (offset >= extent_map_end(entry))
+		else if (em->start >= extent_map_end(entry))
 			p = &(*p)->rb_right;
 		else
-			return parent;
+			return -EEXIST;
 	}
 
-	entry = rb_entry(node, struct extent_map, rb_node);
-	entry->in_tree = 1;
-	rb_link_node(node, parent, p);
-	rb_insert_color(node, root);
-	return NULL;
+	orig_parent = parent;
+	while (parent && em->start >= extent_map_end(entry)) {
+		parent = rb_next(parent);
+		entry = rb_entry(parent, struct extent_map, rb_node);
+	}
+	if (parent)
+		if (end > entry->start && em->start < extent_map_end(entry))
+			return -EEXIST;
+
+	parent = orig_parent;
+	entry = rb_entry(parent, struct extent_map, rb_node);
+	while (parent && em->start < entry->start) {
+		parent = rb_prev(parent);
+		entry = rb_entry(parent, struct extent_map, rb_node);
+	}
+	if (parent)
+		if (end > entry->start && em->start < extent_map_end(entry))
+			return -EEXIST;
+
+	em->in_tree = 1;
+	rb_link_node(&em->rb_node, orig_parent, p);
+	rb_insert_color(&em->rb_node, root);
+	return 0;
 }
 
 /*
@@ -228,7 +255,7 @@ static void try_merge_map(struct extent_map_tree *tree, struct extent_map *em)
 		merge = rb_entry(rb, struct extent_map, rb_node);
 	if (rb && mergable_maps(em, merge)) {
 		em->len += merge->len;
-		em->block_len += merge->len;
+		em->block_len += merge->block_len;
 		rb_erase(&merge->rb_node, &tree->map);
 		merge->in_tree = 0;
 		em->mod_len = (merge->mod_start + merge->mod_len) - em->mod_start;
@@ -310,20 +337,11 @@ int add_extent_mapping(struct extent_map_tree *tree,
 		       struct extent_map *em, int modified)
 {
 	int ret = 0;
-	struct rb_node *rb;
-	struct extent_map *exist;
 
-	exist = lookup_extent_mapping(tree, em->start, em->len);
-	if (exist) {
-		free_extent_map(exist);
-		ret = -EEXIST;
-		goto out;
-	}
-	rb = tree_insert(&tree->map, em->start, &em->rb_node);
-	if (rb) {
-		ret = -EEXIST;
+	ret = tree_insert(&tree->map, em);
+	if (ret)
 		goto out;
-	}
+
 	atomic_inc(&em->refs);
 
 	em->mod_start = em->start;
@@ -337,14 +355,6 @@ out:
 	return ret;
 }
 
-/* simple helper to do math around the end of an extent, handling wrap */
-static u64 range_end(u64 start, u64 len)
-{
-	if (start + len < start)
-		return (u64)-1;
-	return start + len;
-}
-
 static struct extent_map *
 __lookup_extent_mapping(struct extent_map_tree *tree,
 			u64 start, u64 len, int strict)

fs/btrfs/file-item.c

@@ -246,8 +246,8 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 						offset + bvec->bv_len - 1,
 						EXTENT_NODATASUM, GFP_NOFS);
 				} else {
-					printk(KERN_INFO "btrfs no csum found "
-					       "for inode %llu start %llu\n",
+					btrfs_info(BTRFS_I(inode)->root->fs_info,
+						   "no csum found for inode %llu start %llu",
 					       btrfs_ino(inode), offset);
 				}
 				item = NULL;

fs/btrfs/file.c

@@ -692,7 +692,10 @@ next:
 int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
 			 struct btrfs_root *root, struct inode *inode,
 			 struct btrfs_path *path, u64 start, u64 end,
-			 u64 *drop_end, int drop_cache)
+			 u64 *drop_end, int drop_cache,
+			 int replace_extent,
+			 u32 extent_item_size,
+			 int *key_inserted)
 {
 	struct extent_buffer *leaf;
 	struct btrfs_file_extent_item *fi;
@@ -712,6 +715,7 @@ int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
 	int modify_tree = -1;
 	int update_refs = (root->ref_cows || root == root->fs_info->tree_root);
 	int found = 0;
+	int leafs_visited = 0;
 
 	if (drop_cache)
 		btrfs_drop_extent_cache(inode, start, end - 1, 0);
@@ -733,6 +737,7 @@ int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
 				path->slots[0]--;
 		}
 		ret = 0;
+		leafs_visited++;
 next_slot:
 		leaf = path->nodes[0];
 		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
@@ -744,6 +749,7 @@ next_slot:
 				ret = 0;
 				break;
 			}
+			leafs_visited++;
 			leaf = path->nodes[0];
 			recow = 1;
 		}
@@ -766,7 +772,8 @@ next_slot:
 				btrfs_file_extent_num_bytes(leaf, fi);
 		} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 			extent_end = key.offset +
-				btrfs_file_extent_inline_len(leaf, fi);
+				btrfs_file_extent_inline_len(leaf,
+						     path->slots[0], fi);
 		} else {
 			WARN_ON(1);
 			extent_end = search_start;
@@ -927,14 +934,44 @@ next_slot:
 	}
 
 	if (!ret && del_nr > 0) {
+		/*
+		 * Set path->slots[0] to first slot, so that after the delete
+		 * if items are move off from our leaf to its immediate left or
+		 * right neighbor leafs, we end up with a correct and adjusted
+		 * path->slots[0] for our insertion.
+		 */
+		path->slots[0] = del_slot;
 		ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
 		if (ret)
 			btrfs_abort_transaction(trans, root, ret);
+
+		leaf = path->nodes[0];
+		/*
+		 * leaf eb has flag EXTENT_BUFFER_STALE if it was deleted (that
+		 * is, its contents got pushed to its neighbors), in which case
+		 * it means path->locks[0] == 0
+		 */
+		if (!ret && replace_extent && leafs_visited == 1 &&
+		    path->locks[0] &&
+		    btrfs_leaf_free_space(root, leaf) >=
+		    sizeof(struct btrfs_item) + extent_item_size) {
+
+			key.objectid = ino;
+			key.type = BTRFS_EXTENT_DATA_KEY;
+			key.offset = start;
+			setup_items_for_insert(root, path, &key,
+					       &extent_item_size,
+					       extent_item_size,
+					       sizeof(struct btrfs_item) +
+					       extent_item_size, 1);
+			*key_inserted = 1;
+		}
 	}
 
+	if (!replace_extent || !(*key_inserted))
+		btrfs_release_path(path);
 	if (drop_end)
 		*drop_end = found ? min(end, extent_end) : end;
-	btrfs_release_path(path);
 	return ret;
 }
 
@@ -949,7 +986,7 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 	if (!path)
 		return -ENOMEM;
 	ret = __btrfs_drop_extents(trans, root, inode, path, start, end, NULL,
-				   drop_cache);
+				   drop_cache, 0, 0, NULL);
 	btrfs_free_path(path);
 	return ret;
 }
@@ -1235,29 +1272,18 @@ static int prepare_uptodate_page(struct page *page, u64 pos,
 }
 
 /*
- * this gets pages into the page cache and locks them down, it also properly
- * waits for data=ordered extents to finish before allowing the pages to be
- * modified.
+ * this just gets pages into the page cache and locks them down.
  */
-static noinline int prepare_pages(struct btrfs_root *root, struct file *file,
-			 struct page **pages, size_t num_pages,
-			 loff_t pos, unsigned long first_index,
-			 size_t write_bytes, bool force_uptodate)
+static noinline int prepare_pages(struct inode *inode, struct page **pages,
+				  size_t num_pages, loff_t pos,
+				  size_t write_bytes, bool force_uptodate)
 {
-	struct extent_state *cached_state = NULL;
 	int i;
 	unsigned long index = pos >> PAGE_CACHE_SHIFT;
-	struct inode *inode = file_inode(file);
 	gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
 	int err = 0;
-	int faili = 0;
-	u64 start_pos;
-	u64 last_pos;
-
-	start_pos = pos & ~((u64)root->sectorsize - 1);
-	last_pos = ((u64)index + num_pages) << PAGE_CACHE_SHIFT;
+	int faili;
 
-again:
 	for (i = 0; i < num_pages; i++) {
 		pages[i] = find_or_create_page(inode->i_mapping, index + i,
 					       mask | __GFP_WRITE);
@@ -1280,57 +1306,85 @@ again:
 		}
 		wait_on_page_writeback(pages[i]);
 	}
-	faili = num_pages - 1;
-	err = 0;
+
+	return 0;
+fail:
+	while (faili >= 0) {
+		unlock_page(pages[faili]);
+		page_cache_release(pages[faili]);
+		faili--;
+	}
+	return err;
+
+}
+
+/*
+ * This function locks the extent and properly waits for data=ordered extents
+ * to finish before allowing the pages to be modified if need.
+ *
+ * The return value:
+ * 1 - the extent is locked
+ * 0 - the extent is not locked, and everything is OK
+ * -EAGAIN - need re-prepare the pages
+ * the other < 0 number - Something wrong happens
+ */
+static noinline int
+lock_and_cleanup_extent_if_need(struct inode *inode, struct page **pages,
+				size_t num_pages, loff_t pos,
+				u64 *lockstart, u64 *lockend,
+				struct extent_state **cached_state)
+{
+	u64 start_pos;
+	u64 last_pos;
+	int i;
+	int ret = 0;
+
+	start_pos = pos & ~((u64)PAGE_CACHE_SIZE - 1);
+	last_pos = start_pos + ((u64)num_pages << PAGE_CACHE_SHIFT) - 1;
+
 	if (start_pos < inode->i_size) {
 		struct btrfs_ordered_extent *ordered;
 		lock_extent_bits(&BTRFS_I(inode)->io_tree,
-				 start_pos, last_pos - 1, 0, &cached_state);
-		ordered = btrfs_lookup_first_ordered_extent(inode,
-							    last_pos - 1);
+				 start_pos, last_pos, 0, cached_state);
+		ordered = btrfs_lookup_first_ordered_extent(inode, last_pos);
 		if (ordered &&
 		    ordered->file_offset + ordered->len > start_pos &&
-		    ordered->file_offset < last_pos) {
+		    ordered->file_offset <= last_pos) {
 			btrfs_put_ordered_extent(ordered);
 			unlock_extent_cached(&BTRFS_I(inode)->io_tree,
-					     start_pos, last_pos - 1,
-					     &cached_state, GFP_NOFS);
+					     start_pos, last_pos,
+					     cached_state, GFP_NOFS);
 			for (i = 0; i < num_pages; i++) {
 				unlock_page(pages[i]);
 				page_cache_release(pages[i]);
 			}
-			err = btrfs_wait_ordered_range(inode, start_pos,
-						       last_pos - start_pos);
-			if (err)
-				goto fail;
-			goto again;
+			ret = btrfs_wait_ordered_range(inode, start_pos,
+						last_pos - start_pos + 1);
+			if (ret)
+				return ret;
+			else
+				return -EAGAIN;
 		}
 		if (ordered)
 			btrfs_put_ordered_extent(ordered);
 
 		clear_extent_bit(&BTRFS_I(inode)->io_tree, start_pos,
-				  last_pos - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
+				  last_pos, EXTENT_DIRTY | EXTENT_DELALLOC |
 				  EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG,
-				  0, 0, &cached_state, GFP_NOFS);
-		unlock_extent_cached(&BTRFS_I(inode)->io_tree,
-				     start_pos, last_pos - 1, &cached_state,
-				     GFP_NOFS);
+				  0, 0, cached_state, GFP_NOFS);
+		*lockstart = start_pos;
+		*lockend = last_pos;
+		ret = 1;
 	}
+
 	for (i = 0; i < num_pages; i++) {
 		if (clear_page_dirty_for_io(pages[i]))
 			account_page_redirty(pages[i]);
 		set_page_extent_mapped(pages[i]);
 		WARN_ON(!PageLocked(pages[i]));
 	}
-	return 0;
-fail:
-	while (faili >= 0) {
-		unlock_page(pages[faili]);
-		page_cache_release(pages[faili]);
-		faili--;
-	}
-	return err;
 
+	return ret;
 }
 
 static noinline int check_can_nocow(struct inode *inode, loff_t pos,
@@ -1381,13 +1435,17 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 	struct inode *inode = file_inode(file);
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct page **pages = NULL;
+	struct extent_state *cached_state = NULL;
 	u64 release_bytes = 0;
+	u64 lockstart;
+	u64 lockend;
 	unsigned long first_index;
 	size_t num_written = 0;
 	int nrptrs;
 	int ret = 0;
 	bool only_release_metadata = false;
 	bool force_page_uptodate = false;
+	bool need_unlock;
 
 	nrptrs = min((iov_iter_count(i) + PAGE_CACHE_SIZE - 1) /
 		     PAGE_CACHE_SIZE, PAGE_CACHE_SIZE /
@@ -1456,18 +1514,31 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 		}
 
 		release_bytes = reserve_bytes;
-
+		need_unlock = false;
+again:
 		/*
 		 * 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
 		 * contents of pages from loop to loop
 		 */
-		ret = prepare_pages(root, file, pages, num_pages,
-				    pos, first_index, write_bytes,
+		ret = prepare_pages(inode, pages, num_pages,
+				    pos, write_bytes,
 				    force_page_uptodate);
 		if (ret)
 			break;
 
+		ret = lock_and_cleanup_extent_if_need(inode, pages, num_pages,
+						      pos, &lockstart, &lockend,
+						      &cached_state);
+		if (ret < 0) {
+			if (ret == -EAGAIN)
+				goto again;
+			break;
+		} else if (ret > 0) {
+			need_unlock = true;
+			ret = 0;
+		}
+
 		copied = btrfs_copy_from_user(pos, num_pages,
 					   write_bytes, pages, i);
 
@@ -1512,19 +1583,21 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 		}
 
 		release_bytes = dirty_pages << PAGE_CACHE_SHIFT;
-		if (copied > 0) {
+
+		if (copied > 0)
 			ret = btrfs_dirty_pages(root, inode, pages,
 						dirty_pages, pos, copied,
 						NULL);
-			if (ret) {
-				btrfs_drop_pages(pages, num_pages);
-				break;
-			}
+		if (need_unlock)
+			unlock_extent_cached(&BTRFS_I(inode)->io_tree,
+					     lockstart, lockend, &cached_state,
+					     GFP_NOFS);
+		if (ret) {
+			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 +
@@ -1536,6 +1609,8 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 			only_release_metadata = false;
 		}
 
+		btrfs_drop_pages(pages, num_pages);
+
 		cond_resched();
 
 		balance_dirty_pages_ratelimited(inode->i_mapping);
@@ -1857,12 +1932,24 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	if (file->private_data)
 		btrfs_ioctl_trans_end(file);
 
+	/*
+	 * We use start here because we will need to wait on the IO to complete
+	 * in btrfs_sync_log, which could require joining a transaction (for
+	 * example checking cross references in the nocow path).  If we use join
+	 * here we could get into a situation where we're waiting on IO to
+	 * happen that is blocked on a transaction trying to commit.  With start
+	 * we inc the extwriter counter, so we wait for all extwriters to exit
+	 * before we start blocking join'ers.  This comment is to keep somebody
+	 * from thinking they are super smart and changing this to
+	 * btrfs_join_transaction *cough*Josef*cough*.
+	 */
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
 		mutex_unlock(&inode->i_mutex);
 		goto out;
 	}
+	trans->sync = true;
 
 	ret = btrfs_log_dentry_safe(trans, root, dentry);
 	if (ret < 0) {
@@ -1963,11 +2050,13 @@ static int fill_holes(struct btrfs_trans_handle *trans, struct inode *inode,
 	struct btrfs_key key;
 	int ret;
 
+	if (btrfs_fs_incompat(root->fs_info, NO_HOLES))
+		goto out;
+
 	key.objectid = btrfs_ino(inode);
 	key.type = BTRFS_EXTENT_DATA_KEY;
 	key.offset = offset;
 
-
 	ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
 	if (ret < 0)
 		return ret;
@@ -2064,8 +2153,10 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	u64 drop_end;
 	int ret = 0;
 	int err = 0;
+	int rsv_count;
 	bool same_page = ((offset >> PAGE_CACHE_SHIFT) ==
 			  ((offset + len - 1) >> PAGE_CACHE_SHIFT));
+	bool no_holes = btrfs_fs_incompat(root->fs_info, NO_HOLES);
 
 	ret = btrfs_wait_ordered_range(inode, offset, len);
 	if (ret)
@@ -2125,7 +2216,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 		 * we need to try again.
 		 */
 		if ((!ordered ||
-		    (ordered->file_offset + ordered->len < lockstart ||
+		    (ordered->file_offset + ordered->len <= lockstart ||
 		     ordered->file_offset > lockend)) &&
 		     !test_range_bit(&BTRFS_I(inode)->io_tree, lockstart,
 				     lockend, EXTENT_UPTODATE, 0,
@@ -2163,9 +2254,10 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	/*
 	 * 1 - update the inode
 	 * 1 - removing the extents in the range
-	 * 1 - adding the hole extent
+	 * 1 - adding the hole extent if no_holes isn't set
 	 */
-	trans = btrfs_start_transaction(root, 3);
+	rsv_count = no_holes ? 2 : 3;
+	trans = btrfs_start_transaction(root, rsv_count);
 	if (IS_ERR(trans)) {
 		err = PTR_ERR(trans);
 		goto out_free;
@@ -2179,7 +2271,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	while (cur_offset < lockend) {
 		ret = __btrfs_drop_extents(trans, root, inode, path,
 					   cur_offset, lockend + 1,
-					   &drop_end, 1);
+					   &drop_end, 1, 0, 0, NULL);
 		if (ret != -ENOSPC)
 			break;
 
@@ -2202,7 +2294,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 		btrfs_end_transaction(trans, root);
 		btrfs_btree_balance_dirty(root);
 
-		trans = btrfs_start_transaction(root, 3);
+		trans = btrfs_start_transaction(root, rsv_count);
 		if (IS_ERR(trans)) {
 			ret = PTR_ERR(trans);
 			trans = NULL;

fs/btrfs/free-space-cache.c

@@ -347,8 +347,8 @@ static int io_ctl_prepare_pages(struct io_ctl *io_ctl, struct inode *inode,
 			btrfs_readpage(NULL, page);
 			lock_page(page);
 			if (!PageUptodate(page)) {
-				printk(KERN_ERR "btrfs: error reading free "
-				       "space cache\n");
+				btrfs_err(BTRFS_I(inode)->root->fs_info,
+					   "error reading free space cache");
 				io_ctl_drop_pages(io_ctl);
 				return -EIO;
 			}
@@ -405,7 +405,7 @@ static int io_ctl_check_generation(struct io_ctl *io_ctl, u64 generation)
 
 	gen = io_ctl->cur;
 	if (le64_to_cpu(*gen) != generation) {
-		printk_ratelimited(KERN_ERR "btrfs: space cache generation "
+		printk_ratelimited(KERN_ERR "BTRFS: space cache generation "
 				   "(%Lu) does not match inode (%Lu)\n", *gen,
 				   generation);
 		io_ctl_unmap_page(io_ctl);
@@ -463,7 +463,7 @@ static int io_ctl_check_crc(struct io_ctl *io_ctl, int index)
 			      PAGE_CACHE_SIZE - offset);
 	btrfs_csum_final(crc, (char *)&crc);
 	if (val != crc) {
-		printk_ratelimited(KERN_ERR "btrfs: csum mismatch on free "
+		printk_ratelimited(KERN_ERR "BTRFS: csum mismatch on free "
 				   "space cache\n");
 		io_ctl_unmap_page(io_ctl);
 		return -EIO;
@@ -1902,7 +1902,7 @@ out:
 	spin_unlock(&ctl->tree_lock);
 
 	if (ret) {
-		printk(KERN_CRIT "btrfs: unable to add free space :%d\n", ret);
+		printk(KERN_CRIT "BTRFS: unable to add free space :%d\n", ret);
 		ASSERT(ret != -EEXIST);
 	}
 
@@ -2011,14 +2011,15 @@ void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
 		info = rb_entry(n, struct btrfs_free_space, offset_index);
 		if (info->bytes >= bytes && !block_group->ro)
 			count++;
-		printk(KERN_CRIT "entry offset %llu, bytes %llu, bitmap %s\n",
-		       info->offset, info->bytes,
+		btrfs_crit(block_group->fs_info,
+			   "entry offset %llu, bytes %llu, bitmap %s",
+			   info->offset, info->bytes,
 		       (info->bitmap) ? "yes" : "no");
 	}
-	printk(KERN_INFO "block group has cluster?: %s\n",
+	btrfs_info(block_group->fs_info, "block group has cluster?: %s",
 	       list_empty(&block_group->cluster_list) ? "no" : "yes");
-	printk(KERN_INFO "%d blocks of free space at or bigger than bytes is"
-	       "\n", count);
+	btrfs_info(block_group->fs_info,
+		   "%d blocks of free space at or bigger than bytes is", count);
 }
 
 void btrfs_init_free_space_ctl(struct btrfs_block_group_cache *block_group)
@@ -2421,7 +2422,6 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
 	struct btrfs_free_space *entry = NULL;
 	struct btrfs_free_space *last;
 	struct rb_node *node;
-	u64 window_start;
 	u64 window_free;
 	u64 max_extent;
 	u64 total_size = 0;
@@ -2443,7 +2443,6 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
 		entry = rb_entry(node, struct btrfs_free_space, offset_index);
 	}
 
-	window_start = entry->offset;
 	window_free = entry->bytes;
 	max_extent = entry->bytes;
 	first = entry;

fs/btrfs/hash.c

@@ -0,0 +1,50 @@
+/*
+ * Copyright (C) 2014 Filipe David Borba Manana <fdmanana@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#include <crypto/hash.h>
+#include <linux/err.h>
+#include "hash.h"
+
+static struct crypto_shash *tfm;
+
+int __init btrfs_hash_init(void)
+{
+	tfm = crypto_alloc_shash("crc32c", 0, 0);
+	if (IS_ERR(tfm))
+		return PTR_ERR(tfm);
+
+	return 0;
+}
+
+void btrfs_hash_exit(void)
+{
+	crypto_free_shash(tfm);
+}
+
+u32 btrfs_crc32c(u32 crc, const void *address, unsigned int length)
+{
+	struct {
+		struct shash_desc shash;
+		char ctx[crypto_shash_descsize(tfm)];
+	} desc;
+	int err;
+
+	desc.shash.tfm = tfm;
+	desc.shash.flags = 0;
+	*(u32 *)desc.ctx = crc;
+
+	err = crypto_shash_update(&desc.shash, address, length);
+	BUG_ON(err);
+
+	return *(u32 *)desc.ctx;
+}

fs/btrfs/hash.h

@@ -19,10 +19,15 @@
 #ifndef __HASH__
 #define __HASH__
 
-#include <linux/crc32c.h>
+int __init btrfs_hash_init(void);
+
+void btrfs_hash_exit(void);
+
+u32 btrfs_crc32c(u32 crc, const void *address, unsigned int length);
+
 static inline u64 btrfs_name_hash(const char *name, int len)
 {
-	return crc32c((u32)~1, name, len);
+	return btrfs_crc32c((u32)~1, name, len);
 }
 
 /*
@@ -31,7 +36,7 @@ static inline u64 btrfs_name_hash(const char *name, int len)
 static inline u64 btrfs_extref_hash(u64 parent_objectid, const char *name,
 				    int len)
 {
-	return (u64) crc32c(parent_objectid, name, len);
+	return (u64) btrfs_crc32c(parent_objectid, name, len);
 }
 
 #endif

fs/btrfs/inode-item.c

@@ -91,32 +91,6 @@ int btrfs_find_name_in_ext_backref(struct btrfs_path *path, u64 ref_objectid,
 	return 0;
 }
 
-static struct btrfs_inode_ref *
-btrfs_lookup_inode_ref(struct btrfs_trans_handle *trans,
-		       struct btrfs_root *root,
-		       struct btrfs_path *path,
-		       const char *name, int name_len,
-		       u64 inode_objectid, u64 ref_objectid, int ins_len,
-		       int cow)
-{
-	int ret;
-	struct btrfs_key key;
-	struct btrfs_inode_ref *ref;
-
-	key.objectid = inode_objectid;
-	key.type = BTRFS_INODE_REF_KEY;
-	key.offset = ref_objectid;
-
-	ret = btrfs_search_slot(trans, root, &key, path, ins_len, cow);
-	if (ret < 0)
-		return ERR_PTR(ret);
-	if (ret > 0)
-		return NULL;
-	if (!find_name_in_backref(path, name, name_len, &ref))
-		return NULL;
-	return ref;
-}
-
 /* Returns NULL if no extref found */
 struct btrfs_inode_extref *
 btrfs_lookup_inode_extref(struct btrfs_trans_handle *trans,
@@ -144,45 +118,6 @@ btrfs_lookup_inode_extref(struct btrfs_trans_handle *trans,
 	return extref;
 }
 
-int btrfs_get_inode_ref_index(struct btrfs_trans_handle *trans,
-			      struct btrfs_root *root,
-			      struct btrfs_path *path,
-			      const char *name, int name_len,
-			      u64 inode_objectid, u64 ref_objectid, int mod,
-			      u64 *ret_index)
-{
-	struct btrfs_inode_ref *ref;
-	struct btrfs_inode_extref *extref;
-	int ins_len = mod < 0 ? -1 : 0;
-	int cow = mod != 0;
-
-	ref = btrfs_lookup_inode_ref(trans, root, path, name, name_len,
-				     inode_objectid, ref_objectid, ins_len,
-				     cow);
-	if (IS_ERR(ref))
-		return PTR_ERR(ref);
-
-	if (ref != NULL) {
-		*ret_index = btrfs_inode_ref_index(path->nodes[0], ref);
-		return 0;
-	}
-
-	btrfs_release_path(path);
-
-	extref = btrfs_lookup_inode_extref(trans, root, path, name,
-					   name_len, inode_objectid,
-					   ref_objectid, ins_len, cow);
-	if (IS_ERR(extref))
-		return PTR_ERR(extref);
-
-	if (extref) {
-		*ret_index = btrfs_inode_extref_index(path->nodes[0], extref);
-		return 0;
-	}
-
-	return -ENOENT;
-}
-
 static int btrfs_del_inode_extref(struct btrfs_trans_handle *trans,
 				  struct btrfs_root *root,
 				  const char *name, int name_len,

fs/btrfs/inode.c

@@ -58,9 +58,10 @@
 #include "inode-map.h"
 #include "backref.h"
 #include "hash.h"
+#include "props.h"
 
 struct btrfs_iget_args {
-	u64 ino;
+	struct btrfs_key *location;
 	struct btrfs_root *root;
 };
 
@@ -125,13 +126,12 @@ static int btrfs_init_inode_security(struct btrfs_trans_handle *trans,
  * no overlapping inline items exist in the btree
  */
 static noinline int insert_inline_extent(struct btrfs_trans_handle *trans,
+				struct btrfs_path *path, int extent_inserted,
 				struct btrfs_root *root, struct inode *inode,
 				u64 start, size_t size, size_t compressed_size,
 				int compress_type,
 				struct page **compressed_pages)
 {
-	struct btrfs_key key;
-	struct btrfs_path *path;
 	struct extent_buffer *leaf;
 	struct page *page = NULL;
 	char *kaddr;
@@ -140,29 +140,29 @@ static noinline int insert_inline_extent(struct btrfs_trans_handle *trans,
 	int err = 0;
 	int ret;
 	size_t cur_size = size;
-	size_t datasize;
 	unsigned long offset;
 
 	if (compressed_size && compressed_pages)
 		cur_size = compressed_size;
 
-	path = btrfs_alloc_path();
-	if (!path)
-		return -ENOMEM;
+	inode_add_bytes(inode, size);
 
-	path->leave_spinning = 1;
+	if (!extent_inserted) {
+		struct btrfs_key key;
+		size_t datasize;
 
-	key.objectid = btrfs_ino(inode);
-	key.offset = start;
-	btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
-	datasize = btrfs_file_extent_calc_inline_size(cur_size);
+		key.objectid = btrfs_ino(inode);
+		key.offset = start;
+		btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
 
-	inode_add_bytes(inode, size);
-	ret = btrfs_insert_empty_item(trans, root, path, &key,
-				      datasize);
-	if (ret) {
-		err = ret;
-		goto fail;
+		datasize = btrfs_file_extent_calc_inline_size(cur_size);
+		path->leave_spinning = 1;
+		ret = btrfs_insert_empty_item(trans, root, path, &key,
+					      datasize);
+		if (ret) {
+			err = ret;
+			goto fail;
+		}
 	}
 	leaf = path->nodes[0];
 	ei = btrfs_item_ptr(leaf, path->slots[0],
@@ -203,7 +203,7 @@ static noinline int insert_inline_extent(struct btrfs_trans_handle *trans,
 		page_cache_release(page);
 	}
 	btrfs_mark_buffer_dirty(leaf);
-	btrfs_free_path(path);
+	btrfs_release_path(path);
 
 	/*
 	 * we're an inline extent, so nobody can
@@ -219,7 +219,6 @@ static noinline int insert_inline_extent(struct btrfs_trans_handle *trans,
 
 	return ret;
 fail:
-	btrfs_free_path(path);
 	return err;
 }
 
@@ -242,6 +241,9 @@ static noinline int cow_file_range_inline(struct btrfs_root *root,
 	u64 aligned_end = ALIGN(end, root->sectorsize);
 	u64 data_len = inline_len;
 	int ret;
+	struct btrfs_path *path;
+	int extent_inserted = 0;
+	u32 extent_item_size;
 
 	if (compressed_size)
 		data_len = compressed_size;
@@ -256,12 +258,27 @@ static noinline int cow_file_range_inline(struct btrfs_root *root,
 		return 1;
 	}
 
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+
 	trans = btrfs_join_transaction(root);
-	if (IS_ERR(trans))
+	if (IS_ERR(trans)) {
+		btrfs_free_path(path);
 		return PTR_ERR(trans);
+	}
 	trans->block_rsv = &root->fs_info->delalloc_block_rsv;
 
-	ret = btrfs_drop_extents(trans, root, inode, start, aligned_end, 1);
+	if (compressed_size && compressed_pages)
+		extent_item_size = btrfs_file_extent_calc_inline_size(
+		   compressed_size);
+	else
+		extent_item_size = btrfs_file_extent_calc_inline_size(
+		    inline_len);
+
+	ret = __btrfs_drop_extents(trans, root, inode, path,
+				   start, aligned_end, NULL,
+				   1, 1, extent_item_size, &extent_inserted);
 	if (ret) {
 		btrfs_abort_transaction(trans, root, ret);
 		goto out;
@@ -269,7 +286,8 @@ static noinline int cow_file_range_inline(struct btrfs_root *root,
 
 	if (isize > actual_end)
 		inline_len = min_t(u64, isize, actual_end);
-	ret = insert_inline_extent(trans, root, inode, start,
+	ret = insert_inline_extent(trans, path, extent_inserted,
+				   root, inode, start,
 				   inline_len, compressed_size,
 				   compress_type, compressed_pages);
 	if (ret && ret != -ENOSPC) {
@@ -284,6 +302,7 @@ static noinline int cow_file_range_inline(struct btrfs_root *root,
 	btrfs_delalloc_release_metadata(inode, end + 1 - start);
 	btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
 out:
+	btrfs_free_path(path);
 	btrfs_end_transaction(trans, root);
 	return ret;
 }
@@ -1262,7 +1281,8 @@ next_slot:
 			nocow = 1;
 		} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 			extent_end = found_key.offset +
-				btrfs_file_extent_inline_len(leaf, fi);
+				btrfs_file_extent_inline_len(leaf,
+						     path->slots[0], fi);
 			extent_end = ALIGN(extent_end, root->sectorsize);
 		} else {
 			BUG_ON(1);
@@ -1841,14 +1861,13 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
 	struct btrfs_path *path;
 	struct extent_buffer *leaf;
 	struct btrfs_key ins;
+	int extent_inserted = 0;
 	int ret;
 
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
 
-	path->leave_spinning = 1;
-
 	/*
 	 * we may be replacing one extent in the tree with another.
 	 * The new extent is pinned in the extent map, and we don't want
@@ -1858,17 +1877,23 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
 	 * the caller is expected to unpin it and allow it to be merged
 	 * with the others.
 	 */
-	ret = btrfs_drop_extents(trans, root, inode, file_pos,
-				 file_pos + num_bytes, 0);
+	ret = __btrfs_drop_extents(trans, root, inode, path, file_pos,
+				   file_pos + num_bytes, NULL, 0,
+				   1, sizeof(*fi), &extent_inserted);
 	if (ret)
 		goto out;
 
-	ins.objectid = btrfs_ino(inode);
-	ins.offset = file_pos;
-	ins.type = BTRFS_EXTENT_DATA_KEY;
-	ret = btrfs_insert_empty_item(trans, root, path, &ins, sizeof(*fi));
-	if (ret)
-		goto out;
+	if (!extent_inserted) {
+		ins.objectid = btrfs_ino(inode);
+		ins.offset = file_pos;
+		ins.type = BTRFS_EXTENT_DATA_KEY;
+
+		path->leave_spinning = 1;
+		ret = btrfs_insert_empty_item(trans, root, path, &ins,
+					      sizeof(*fi));
+		if (ret)
+			goto out;
+	}
 	leaf = path->nodes[0];
 	fi = btrfs_item_ptr(leaf, path->slots[0],
 			    struct btrfs_file_extent_item);
@@ -2290,7 +2315,7 @@ again:
 		u64 extent_len;
 		struct btrfs_key found_key;
 
-		ret = btrfs_search_slot(trans, root, &key, path, 1, 1);
+		ret = btrfs_search_slot(trans, root, &key, path, 0, 1);
 		if (ret < 0)
 			goto out_free_path;
 
@@ -2543,12 +2568,6 @@ out_kfree:
 	return NULL;
 }
 
-/*
- * helper function for btrfs_finish_ordered_io, this
- * just reads in some of the csum leaves to prime them into ram
- * before we start the transaction.  It limits the amount of btree
- * reads required while inside the transaction.
- */
 /* as ordered data IO finishes, this gets called so we can finish
  * an ordered extent if the range of bytes in the file it covers are
  * fully written.
@@ -3248,7 +3267,8 @@ out:
  * slot is the slot the inode is in, objectid is the objectid of the inode
  */
 static noinline int acls_after_inode_item(struct extent_buffer *leaf,
-					  int slot, u64 objectid)
+					  int slot, u64 objectid,
+					  int *first_xattr_slot)
 {
 	u32 nritems = btrfs_header_nritems(leaf);
 	struct btrfs_key found_key;
@@ -3264,6 +3284,7 @@ static noinline int acls_after_inode_item(struct extent_buffer *leaf,
 	}
 
 	slot++;
+	*first_xattr_slot = -1;
 	while (slot < nritems) {
 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
 
@@ -3273,6 +3294,8 @@ static noinline int acls_after_inode_item(struct extent_buffer *leaf,
 
 		/* we found an xattr, assume we've got an acl */
 		if (found_key.type == BTRFS_XATTR_ITEM_KEY) {
+			if (*first_xattr_slot == -1)
+				*first_xattr_slot = slot;
 			if (found_key.offset == xattr_access ||
 			    found_key.offset == xattr_default)
 				return 1;
@@ -3301,6 +3324,8 @@ static noinline int acls_after_inode_item(struct extent_buffer *leaf,
 	 * something larger than an xattr.  We have to assume the inode
 	 * has acls
 	 */
+	if (*first_xattr_slot == -1)
+		*first_xattr_slot = slot;
 	return 1;
 }
 
@@ -3315,10 +3340,12 @@ static void btrfs_read_locked_inode(struct inode *inode)
 	struct btrfs_timespec *tspec;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_key location;
+	unsigned long ptr;
 	int maybe_acls;
 	u32 rdev;
 	int ret;
 	bool filled = false;
+	int first_xattr_slot;
 
 	ret = btrfs_fill_inode(inode, &rdev);
 	if (!ret)
@@ -3328,7 +3355,6 @@ static void btrfs_read_locked_inode(struct inode *inode)
 	if (!path)
 		goto make_bad;
 
-	path->leave_spinning = 1;
 	memcpy(&location, &BTRFS_I(inode)->location, sizeof(location));
 
 	ret = btrfs_lookup_inode(NULL, root, path, &location, 0);
@@ -3338,7 +3364,7 @@ static void btrfs_read_locked_inode(struct inode *inode)
 	leaf = path->nodes[0];
 
 	if (filled)
-		goto cache_acl;
+		goto cache_index;
 
 	inode_item = btrfs_item_ptr(leaf, path->slots[0],
 				    struct btrfs_inode_item);
@@ -3381,18 +3407,51 @@ static void btrfs_read_locked_inode(struct inode *inode)
 
 	BTRFS_I(inode)->index_cnt = (u64)-1;
 	BTRFS_I(inode)->flags = btrfs_inode_flags(leaf, inode_item);
+
+cache_index:
+	path->slots[0]++;
+	if (inode->i_nlink != 1 ||
+	    path->slots[0] >= btrfs_header_nritems(leaf))
+		goto cache_acl;
+
+	btrfs_item_key_to_cpu(leaf, &location, path->slots[0]);
+	if (location.objectid != btrfs_ino(inode))
+		goto cache_acl;
+
+	ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
+	if (location.type == BTRFS_INODE_REF_KEY) {
+		struct btrfs_inode_ref *ref;
+
+		ref = (struct btrfs_inode_ref *)ptr;
+		BTRFS_I(inode)->dir_index = btrfs_inode_ref_index(leaf, ref);
+	} else if (location.type == BTRFS_INODE_EXTREF_KEY) {
+		struct btrfs_inode_extref *extref;
+
+		extref = (struct btrfs_inode_extref *)ptr;
+		BTRFS_I(inode)->dir_index = btrfs_inode_extref_index(leaf,
+								     extref);
+	}
 cache_acl:
 	/*
 	 * try to precache a NULL acl entry for files that don't have
 	 * any xattrs or acls
 	 */
 	maybe_acls = acls_after_inode_item(leaf, path->slots[0],
-					   btrfs_ino(inode));
+					   btrfs_ino(inode), &first_xattr_slot);
+	if (first_xattr_slot != -1) {
+		path->slots[0] = first_xattr_slot;
+		ret = btrfs_load_inode_props(inode, path);
+		if (ret)
+			btrfs_err(root->fs_info,
+				  "error loading props for ino %llu (root %llu): %d\n",
+				  btrfs_ino(inode),
+				  root->root_key.objectid, ret);
+	}
+	btrfs_free_path(path);
+
 	if (!maybe_acls)
 		cache_no_acl(inode);
 
-	btrfs_free_path(path);
-
 	switch (inode->i_mode & S_IFMT) {
 	case S_IFREG:
 		inode->i_mapping->a_ops = &btrfs_aops;
@@ -3496,7 +3555,6 @@ static noinline int btrfs_update_inode_item(struct btrfs_trans_handle *trans,
 		goto failed;
 	}
 
-	btrfs_unlock_up_safe(path, 1);
 	leaf = path->nodes[0];
 	inode_item = btrfs_item_ptr(leaf, path->slots[0],
 				    struct btrfs_inode_item);
@@ -3593,6 +3651,24 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans,
 		goto err;
 	btrfs_release_path(path);
 
+	/*
+	 * If we don't have dir index, we have to get it by looking up
+	 * the inode ref, since we get the inode ref, remove it directly,
+	 * it is unnecessary to do delayed deletion.
+	 *
+	 * But if we have dir index, needn't search inode ref to get it.
+	 * Since the inode ref is close to the inode item, it is better
+	 * that we delay to delete it, and just do this deletion when
+	 * we update the inode item.
+	 */
+	if (BTRFS_I(inode)->dir_index) {
+		ret = btrfs_delayed_delete_inode_ref(inode);
+		if (!ret) {
+			index = BTRFS_I(inode)->dir_index;
+			goto skip_backref;
+		}
+	}
+
 	ret = btrfs_del_inode_ref(trans, root, name, name_len, ino,
 				  dir_ino, &index);
 	if (ret) {
@@ -3602,7 +3678,7 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans,
 		btrfs_abort_transaction(trans, root, ret);
 		goto err;
 	}
-
+skip_backref:
 	ret = btrfs_delete_delayed_dir_index(trans, root, dir, index);
 	if (ret) {
 		btrfs_abort_transaction(trans, root, ret);
@@ -3948,7 +4024,7 @@ search_again:
 				    btrfs_file_extent_num_bytes(leaf, fi);
 			} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
 				item_end += btrfs_file_extent_inline_len(leaf,
-									 fi);
+							 path->slots[0], fi);
 			}
 			item_end--;
 		}
@@ -4018,6 +4094,12 @@ search_again:
 					inode_sub_bytes(inode, item_end + 1 -
 							new_size);
 				}
+
+				/*
+				 * update the ram bytes to properly reflect
+				 * the new size of our item
+				 */
+				btrfs_set_file_extent_ram_bytes(leaf, fi, size);
 				size =
 				    btrfs_file_extent_calc_inline_size(size);
 				btrfs_truncate_item(root, path, size, 1);
@@ -4203,6 +4285,49 @@ out:
 	return ret;
 }
 
+static int maybe_insert_hole(struct btrfs_root *root, struct inode *inode,
+			     u64 offset, u64 len)
+{
+	struct btrfs_trans_handle *trans;
+	int ret;
+
+	/*
+	 * Still need to make sure the inode looks like it's been updated so
+	 * that any holes get logged if we fsync.
+	 */
+	if (btrfs_fs_incompat(root->fs_info, NO_HOLES)) {
+		BTRFS_I(inode)->last_trans = root->fs_info->generation;
+		BTRFS_I(inode)->last_sub_trans = root->log_transid;
+		BTRFS_I(inode)->last_log_commit = root->last_log_commit;
+		return 0;
+	}
+
+	/*
+	 * 1 - for the one we're dropping
+	 * 1 - for the one we're adding
+	 * 1 - for updating the inode.
+	 */
+	trans = btrfs_start_transaction(root, 3);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+
+	ret = btrfs_drop_extents(trans, root, inode, offset, offset + len, 1);
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		btrfs_end_transaction(trans, root);
+		return ret;
+	}
+
+	ret = btrfs_insert_file_extent(trans, root, btrfs_ino(inode), offset,
+				       0, 0, len, 0, len, 0, 0, 0);
+	if (ret)
+		btrfs_abort_transaction(trans, root, ret);
+	else
+		btrfs_update_inode(trans, root, inode);
+	btrfs_end_transaction(trans, root);
+	return ret;
+}
+
 /*
  * This function puts in dummy file extents for the area we're creating a hole
  * for.  So if we are truncating this file to a larger size we need to insert
@@ -4211,7 +4336,6 @@ out:
  */
 int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 {
-	struct btrfs_trans_handle *trans;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
 	struct extent_map *em = NULL;
@@ -4266,31 +4390,10 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 			struct extent_map *hole_em;
 			hole_size = last_byte - cur_offset;
 
-			trans = btrfs_start_transaction(root, 3);
-			if (IS_ERR(trans)) {
-				err = PTR_ERR(trans);
-				break;
-			}
-
-			err = btrfs_drop_extents(trans, root, inode,
-						 cur_offset,
-						 cur_offset + hole_size, 1);
-			if (err) {
-				btrfs_abort_transaction(trans, root, err);
-				btrfs_end_transaction(trans, root);
-				break;
-			}
-
-			err = btrfs_insert_file_extent(trans, root,
-					btrfs_ino(inode), cur_offset, 0,
-					0, hole_size, 0, hole_size,
-					0, 0, 0);
-			if (err) {
-				btrfs_abort_transaction(trans, root, err);
-				btrfs_end_transaction(trans, root);
+			err = maybe_insert_hole(root, inode, cur_offset,
+						hole_size);
+			if (err)
 				break;
-			}
-
 			btrfs_drop_extent_cache(inode, cur_offset,
 						cur_offset + hole_size - 1, 0);
 			hole_em = alloc_extent_map();
@@ -4309,7 +4412,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 			hole_em->ram_bytes = hole_size;
 			hole_em->bdev = root->fs_info->fs_devices->latest_bdev;
 			hole_em->compress_type = BTRFS_COMPRESS_NONE;
-			hole_em->generation = trans->transid;
+			hole_em->generation = root->fs_info->generation;
 
 			while (1) {
 				write_lock(&em_tree->lock);
@@ -4322,17 +4425,14 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 							hole_size - 1, 0);
 			}
 			free_extent_map(hole_em);
-next:
-			btrfs_update_inode(trans, root, inode);
-			btrfs_end_transaction(trans, root);
 		}
+next:
 		free_extent_map(em);
 		em = NULL;
 		cur_offset = last_byte;
 		if (cur_offset >= block_end)
 			break;
 	}
-
 	free_extent_map(em);
 	unlock_extent_cached(io_tree, hole_start, block_end - 1, &cached_state,
 			     GFP_NOFS);
@@ -4474,6 +4574,64 @@ static int btrfs_setattr(struct dentry *dentry, struct iattr *attr)
 	return err;
 }
 
+/*
+ * While truncating the inode pages during eviction, we get the VFS calling
+ * btrfs_invalidatepage() against each page of the inode. This is slow because
+ * the calls to btrfs_invalidatepage() result in a huge amount of calls to
+ * lock_extent_bits() and clear_extent_bit(), which keep merging and splitting
+ * extent_state structures over and over, wasting lots of time.
+ *
+ * Therefore if the inode is being evicted, let btrfs_invalidatepage() skip all
+ * those expensive operations on a per page basis and do only the ordered io
+ * finishing, while we release here the extent_map and extent_state structures,
+ * without the excessive merging and splitting.
+ */
+static void evict_inode_truncate_pages(struct inode *inode)
+{
+	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
+	struct extent_map_tree *map_tree = &BTRFS_I(inode)->extent_tree;
+	struct rb_node *node;
+
+	ASSERT(inode->i_state & I_FREEING);
+	truncate_inode_pages(&inode->i_data, 0);
+
+	write_lock(&map_tree->lock);
+	while (!RB_EMPTY_ROOT(&map_tree->map)) {
+		struct extent_map *em;
+
+		node = rb_first(&map_tree->map);
+		em = rb_entry(node, struct extent_map, rb_node);
+		clear_bit(EXTENT_FLAG_PINNED, &em->flags);
+		clear_bit(EXTENT_FLAG_LOGGING, &em->flags);
+		remove_extent_mapping(map_tree, em);
+		free_extent_map(em);
+	}
+	write_unlock(&map_tree->lock);
+
+	spin_lock(&io_tree->lock);
+	while (!RB_EMPTY_ROOT(&io_tree->state)) {
+		struct extent_state *state;
+		struct extent_state *cached_state = NULL;
+
+		node = rb_first(&io_tree->state);
+		state = rb_entry(node, struct extent_state, rb_node);
+		atomic_inc(&state->refs);
+		spin_unlock(&io_tree->lock);
+
+		lock_extent_bits(io_tree, state->start, state->end,
+				 0, &cached_state);
+		clear_extent_bit(io_tree, state->start, state->end,
+				 EXTENT_LOCKED | EXTENT_DIRTY |
+				 EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING |
+				 EXTENT_DEFRAG, 1, 1,
+				 &cached_state, GFP_NOFS);
+		free_extent_state(state);
+
+		spin_lock(&io_tree->lock);
+	}
+	spin_unlock(&io_tree->lock);
+}
+
 void btrfs_evict_inode(struct inode *inode)
 {
 	struct btrfs_trans_handle *trans;
@@ -4484,7 +4642,8 @@ void btrfs_evict_inode(struct inode *inode)
 
 	trace_btrfs_inode_evict(inode);
 
-	truncate_inode_pages(&inode->i_data, 0);
+	evict_inode_truncate_pages(inode);
+
 	if (inode->i_nlink &&
 	    ((btrfs_root_refs(&root->root_item) != 0 &&
 	      root->root_key.objectid != BTRFS_ROOT_TREE_OBJECTID) ||
@@ -4659,9 +4818,9 @@ static int fixup_tree_root_location(struct btrfs_root *root,
 	}
 
 	err = -ENOENT;
-	ret = btrfs_find_root_ref(root->fs_info->tree_root, path,
-				  BTRFS_I(dir)->root->root_key.objectid,
-				  location->objectid);
+	ret = btrfs_find_item(root->fs_info->tree_root, path,
+				BTRFS_I(dir)->root->root_key.objectid,
+				location->objectid, BTRFS_ROOT_REF_KEY, NULL);
 	if (ret) {
 		if (ret < 0)
 			err = ret;
@@ -4822,7 +4981,9 @@ again:
 static int btrfs_init_locked_inode(struct inode *inode, void *p)
 {
 	struct btrfs_iget_args *args = p;
-	inode->i_ino = args->ino;
+	inode->i_ino = args->location->objectid;
+	memcpy(&BTRFS_I(inode)->location, args->location,
+	       sizeof(*args->location));
 	BTRFS_I(inode)->root = args->root;
 	return 0;
 }
@@ -4830,19 +4991,19 @@ static int btrfs_init_locked_inode(struct inode *inode, void *p)
 static int btrfs_find_actor(struct inode *inode, void *opaque)
 {
 	struct btrfs_iget_args *args = opaque;
-	return args->ino == btrfs_ino(inode) &&
+	return args->location->objectid == BTRFS_I(inode)->location.objectid &&
 		args->root == BTRFS_I(inode)->root;
 }
 
 static struct inode *btrfs_iget_locked(struct super_block *s,
-				       u64 objectid,
+				       struct btrfs_key *location,
 				       struct btrfs_root *root)
 {
 	struct inode *inode;
 	struct btrfs_iget_args args;
-	unsigned long hashval = btrfs_inode_hash(objectid, root);
+	unsigned long hashval = btrfs_inode_hash(location->objectid, root);
 
-	args.ino = objectid;
+	args.location = location;
 	args.root = root;
 
 	inode = iget5_locked(s, hashval, btrfs_find_actor,
@@ -4859,13 +5020,11 @@ struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
 {
 	struct inode *inode;
 
-	inode = btrfs_iget_locked(s, location->objectid, root);
+	inode = btrfs_iget_locked(s, location, root);
 	if (!inode)
 		return ERR_PTR(-ENOMEM);
 
 	if (inode->i_state & I_NEW) {
-		BTRFS_I(inode)->root = root;
-		memcpy(&BTRFS_I(inode)->location, location, sizeof(*location));
 		btrfs_read_locked_inode(inode);
 		if (!is_bad_inode(inode)) {
 			inode_tree_add(inode);
@@ -4921,7 +5080,7 @@ struct inode *btrfs_lookup_dentry(struct inode *dir, struct dentry *dentry)
 		return ERR_PTR(ret);
 
 	if (location.objectid == 0)
-		return NULL;
+		return ERR_PTR(-ENOENT);
 
 	if (location.type == BTRFS_INODE_ITEM_KEY) {
 		inode = btrfs_iget(dir->i_sb, &location, root, NULL);
@@ -4985,10 +5144,17 @@ static void btrfs_dentry_release(struct dentry *dentry)
 static struct dentry *btrfs_lookup(struct inode *dir, struct dentry *dentry,
 				   unsigned int flags)
 {
-	struct dentry *ret;
+	struct inode *inode;
 
-	ret = d_splice_alias(btrfs_lookup_dentry(dir, dentry), dentry);
-	return ret;
+	inode = btrfs_lookup_dentry(dir, dentry);
+	if (IS_ERR(inode)) {
+		if (PTR_ERR(inode) == -ENOENT)
+			inode = NULL;
+		else
+			return ERR_CAST(inode);
+	}
+
+	return d_splice_alias(inode, dentry);
 }
 
 unsigned char btrfs_filetype_table[] = {
@@ -5358,7 +5524,6 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 	u32 sizes[2];
 	unsigned long ptr;
 	int ret;
-	int owner;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -5392,6 +5557,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 	 * number
 	 */
 	BTRFS_I(inode)->index_cnt = 2;
+	BTRFS_I(inode)->dir_index = *index;
 	BTRFS_I(inode)->root = root;
 	BTRFS_I(inode)->generation = trans->transid;
 	inode->i_generation = BTRFS_I(inode)->generation;
@@ -5404,11 +5570,6 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 	 */
 	set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &BTRFS_I(inode)->runtime_flags);
 
-	if (S_ISDIR(mode))
-		owner = 0;
-	else
-		owner = 1;
-
 	key[0].objectid = objectid;
 	btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY);
 	key[0].offset = 0;
@@ -5473,6 +5634,12 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 
 	btrfs_update_root_times(trans, root);
 
+	ret = btrfs_inode_inherit_props(trans, inode, dir);
+	if (ret)
+		btrfs_err(root->fs_info,
+			  "error inheriting props for ino %llu (root %llu): %d",
+			  btrfs_ino(inode), root->root_key.objectid, ret);
+
 	return inode;
 fail:
 	if (dir)
@@ -5741,6 +5908,8 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 		goto fail;
 	}
 
+	/* There are several dir indexes for this inode, clear the cache. */
+	BTRFS_I(inode)->dir_index = 0ULL;
 	inc_nlink(inode);
 	inode_inc_iversion(inode);
 	inode->i_ctime = CURRENT_TIME;
@@ -6004,7 +6173,7 @@ again:
 		       btrfs_file_extent_num_bytes(leaf, item);
 	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
 		size_t size;
-		size = btrfs_file_extent_inline_len(leaf, item);
+		size = btrfs_file_extent_inline_len(leaf, path->slots[0], item);
 		extent_end = ALIGN(extent_start + size, root->sectorsize);
 	}
 next:
@@ -6073,7 +6242,7 @@ next:
 			goto out;
 		}
 
-		size = btrfs_file_extent_inline_len(leaf, item);
+		size = btrfs_file_extent_inline_len(leaf, path->slots[0], item);
 		extent_offset = page_offset(page) + pg_offset - extent_start;
 		copy_size = min_t(u64, PAGE_CACHE_SIZE - pg_offset,
 				size - extent_offset);
@@ -6390,6 +6559,7 @@ noinline int can_nocow_extent(struct inode *inode, u64 offset, u64 *len,
 	int slot;
 	int found_type;
 	bool nocow = (BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW);
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -6433,6 +6603,10 @@ noinline int can_nocow_extent(struct inode *inode, u64 offset, u64 *len,
 	if (!nocow && found_type == BTRFS_FILE_EXTENT_REG)
 		goto out;
 
+	extent_end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
+	if (extent_end <= offset)
+		goto out;
+
 	disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
 	if (disk_bytenr == 0)
 		goto out;
@@ -6450,8 +6624,6 @@ noinline int can_nocow_extent(struct inode *inode, u64 offset, u64 *len,
 		*ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
 	}
 
-	extent_end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
-
 	if (btrfs_extent_readonly(root, disk_bytenr))
 		goto out;
 	btrfs_release_path(path);
@@ -6895,8 +7067,8 @@ static void btrfs_end_dio_bio(struct bio *bio, int err)
 	struct btrfs_dio_private *dip = bio->bi_private;
 
 	if (err) {
-		printk(KERN_ERR "btrfs direct IO failed ino %llu rw %lu "
-		      "sector %#Lx len %u err no %d\n",
+		btrfs_err(BTRFS_I(dip->inode)->root->fs_info,
+			  "direct IO failed ino %llu rw %lu sector %#Lx len %u err no %d",
 		      btrfs_ino(dip->inode), bio->bi_rw,
 		      (unsigned long long)bio->bi_iter.bi_sector,
 		      bio->bi_iter.bi_size, err);
@@ -7370,6 +7542,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned int offset,
 	struct extent_state *cached_state = NULL;
 	u64 page_start = page_offset(page);
 	u64 page_end = page_start + PAGE_CACHE_SIZE - 1;
+	int inode_evicting = inode->i_state & I_FREEING;
 
 	/*
 	 * we have the page locked, so new writeback can't start,
@@ -7385,17 +7558,21 @@ static void btrfs_invalidatepage(struct page *page, unsigned int offset,
 		btrfs_releasepage(page, GFP_NOFS);
 		return;
 	}
-	lock_extent_bits(tree, page_start, page_end, 0, &cached_state);
-	ordered = btrfs_lookup_ordered_extent(inode, page_offset(page));
+
+	if (!inode_evicting)
+		lock_extent_bits(tree, page_start, page_end, 0, &cached_state);
+	ordered = btrfs_lookup_ordered_extent(inode, page_start);
 	if (ordered) {
 		/*
 		 * IO on this page will never be started, so we need
 		 * to account for any ordered extents now
 		 */
-		clear_extent_bit(tree, page_start, page_end,
-				 EXTENT_DIRTY | EXTENT_DELALLOC |
-				 EXTENT_LOCKED | EXTENT_DO_ACCOUNTING |
-				 EXTENT_DEFRAG, 1, 0, &cached_state, GFP_NOFS);
+		if (!inode_evicting)
+			clear_extent_bit(tree, page_start, page_end,
+					 EXTENT_DIRTY | EXTENT_DELALLOC |
+					 EXTENT_LOCKED | EXTENT_DO_ACCOUNTING |
+					 EXTENT_DEFRAG, 1, 0, &cached_state,
+					 GFP_NOFS);
 		/*
 		 * whoever cleared the private bit is responsible
 		 * for the finish_ordered_io
@@ -7419,14 +7596,22 @@ static void btrfs_invalidatepage(struct page *page, unsigned int offset,
 				btrfs_finish_ordered_io(ordered);
 		}
 		btrfs_put_ordered_extent(ordered);
-		cached_state = NULL;
-		lock_extent_bits(tree, page_start, page_end, 0, &cached_state);
+		if (!inode_evicting) {
+			cached_state = NULL;
+			lock_extent_bits(tree, page_start, page_end, 0,
+					 &cached_state);
+		}
+	}
+
+	if (!inode_evicting) {
+		clear_extent_bit(tree, page_start, page_end,
+				 EXTENT_LOCKED | EXTENT_DIRTY |
+				 EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING |
+				 EXTENT_DEFRAG, 1, 1,
+				 &cached_state, GFP_NOFS);
+
+		__btrfs_releasepage(page, GFP_NOFS);
 	}
-	clear_extent_bit(tree, page_start, page_end,
-		 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
-		 EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 1, 1,
-		 &cached_state, GFP_NOFS);
-	__btrfs_releasepage(page, GFP_NOFS);
 
 	ClearPageChecked(page);
 	if (PagePrivate(page)) {
@@ -7736,7 +7921,9 @@ out:
  * create a new subvolume directory/inode (helper for the ioctl).
  */
 int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *new_root, u64 new_dirid)
+			     struct btrfs_root *new_root,
+			     struct btrfs_root *parent_root,
+			     u64 new_dirid)
 {
 	struct inode *inode;
 	int err;
@@ -7754,6 +7941,12 @@ int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
 	set_nlink(inode, 1);
 	btrfs_i_size_write(inode, 0);
 
+	err = btrfs_subvol_inherit_props(trans, new_root, parent_root);
+	if (err)
+		btrfs_err(new_root->fs_info,
+			  "error inheriting subvolume %llu properties: %d\n",
+			  new_root->root_key.objectid, err);
+
 	err = btrfs_update_inode(trans, new_root, inode);
 
 	iput(inode);
@@ -7779,6 +7972,7 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 	ei->flags = 0;
 	ei->csum_bytes = 0;
 	ei->index_cnt = (u64)-1;
+	ei->dir_index = 0;
 	ei->last_unlink_trans = 0;
 	ei->last_log_commit = 0;
 
@@ -8066,6 +8260,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (ret)
 		goto out_fail;
 
+	BTRFS_I(old_inode)->dir_index = 0ULL;
 	if (unlikely(old_ino == BTRFS_FIRST_FREE_OBJECTID)) {
 		/* force full log commit if subvolume involved. */
 		root->fs_info->last_trans_log_full_commit = trans->transid;
@@ -8154,6 +8349,9 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		goto out_fail;
 	}
 
+	if (old_inode->i_nlink == 1)
+		BTRFS_I(old_inode)->dir_index = index;
+
 	if (old_ino != BTRFS_FIRST_FREE_OBJECTID) {
 		struct dentry *parent = new_dentry->d_parent;
 		btrfs_log_new_name(trans, old_inode, old_dir, parent);
@@ -8289,7 +8487,7 @@ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput)
 {
 	int ret;
 
-	if (root->fs_info->sb->s_flags & MS_RDONLY)
+	if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state))
 		return -EROFS;
 
 	ret = __start_delalloc_inodes(root, delay_iput);
@@ -8315,7 +8513,7 @@ int btrfs_start_delalloc_roots(struct btrfs_fs_info *fs_info, int delay_iput)
 	struct list_head splice;
 	int ret;
 
-	if (fs_info->sb->s_flags & MS_RDONLY)
+	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state))
 		return -EROFS;
 
 	INIT_LIST_HEAD(&splice);

fs/btrfs/ioctl.c

@@ -56,6 +56,8 @@
 #include "rcu-string.h"
 #include "send.h"
 #include "dev-replace.h"
+#include "props.h"
+#include "sysfs.h"
 
 static int btrfs_clone(struct inode *src, struct inode *inode,
 		       u64 off, u64 olen, u64 olen_aligned, u64 destoff);
@@ -190,6 +192,9 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 	unsigned int i_oldflags;
 	umode_t mode;
 
+	if (!inode_owner_or_capable(inode))
+		return -EPERM;
+
 	if (btrfs_root_readonly(root))
 		return -EROFS;
 
@@ -200,9 +205,6 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 	if (ret)
 		return ret;
 
-	if (!inode_owner_or_capable(inode))
-		return -EACCES;
-
 	ret = mnt_want_write_file(file);
 	if (ret)
 		return ret;
@@ -280,9 +282,25 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 	if (flags & FS_NOCOMP_FL) {
 		ip->flags &= ~BTRFS_INODE_COMPRESS;
 		ip->flags |= BTRFS_INODE_NOCOMPRESS;
+
+		ret = btrfs_set_prop(inode, "btrfs.compression", NULL, 0, 0);
+		if (ret && ret != -ENODATA)
+			goto out_drop;
 	} else if (flags & FS_COMPR_FL) {
+		const char *comp;
+
 		ip->flags |= BTRFS_INODE_COMPRESS;
 		ip->flags &= ~BTRFS_INODE_NOCOMPRESS;
+
+		if (root->fs_info->compress_type == BTRFS_COMPRESS_LZO)
+			comp = "lzo";
+		else
+			comp = "zlib";
+		ret = btrfs_set_prop(inode, "btrfs.compression",
+				     comp, strlen(comp), 0);
+		if (ret)
+			goto out_drop;
+
 	} else {
 		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
 	}
@@ -392,6 +410,7 @@ static noinline int create_subvol(struct inode *dir,
 	struct btrfs_root *new_root;
 	struct btrfs_block_rsv block_rsv;
 	struct timespec cur_time = CURRENT_TIME;
+	struct inode *inode;
 	int ret;
 	int err;
 	u64 objectid;
@@ -417,7 +436,9 @@ static noinline int create_subvol(struct inode *dir,
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
-		goto out;
+		btrfs_subvolume_release_metadata(root, &block_rsv,
+						 qgroup_reserved);
+		return ret;
 	}
 	trans->block_rsv = &block_rsv;
 	trans->bytes_reserved = block_rsv.size;
@@ -500,7 +521,7 @@ static noinline int create_subvol(struct inode *dir,
 
 	btrfs_record_root_in_trans(trans, new_root);
 
-	ret = btrfs_create_subvol_root(trans, new_root, new_dirid);
+	ret = btrfs_create_subvol_root(trans, new_root, root, new_dirid);
 	if (ret) {
 		/* We potentially lose an unused inode item here */
 		btrfs_abort_transaction(trans, root, ret);
@@ -542,6 +563,8 @@ static noinline int create_subvol(struct inode *dir,
 fail:
 	trans->block_rsv = NULL;
 	trans->bytes_reserved = 0;
+	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
+
 	if (async_transid) {
 		*async_transid = trans->transid;
 		err = btrfs_commit_transaction_async(trans, root, 1);
@@ -553,10 +576,12 @@ fail:
 	if (err && !ret)
 		ret = err;
 
-	if (!ret)
-		d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry));
-out:
-	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
+	if (!ret) {
+		inode = btrfs_lookup_dentry(dir, dentry);
+		if (IS_ERR(inode))
+			return PTR_ERR(inode);
+		d_instantiate(dentry, inode);
+	}
 	return ret;
 }
 
@@ -642,7 +667,7 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 		ret = PTR_ERR(inode);
 		goto fail;
 	}
-	BUG_ON(!inode);
+
 	d_instantiate(dentry, inode);
 	ret = 0;
 fail:
@@ -1011,7 +1036,7 @@ out:
 static int cluster_pages_for_defrag(struct inode *inode,
 				    struct page **pages,
 				    unsigned long start_index,
-				    int num_pages)
+				    unsigned long num_pages)
 {
 	unsigned long file_end;
 	u64 isize = i_size_read(inode);
@@ -1169,8 +1194,8 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 	int defrag_count = 0;
 	int compress_type = BTRFS_COMPRESS_ZLIB;
 	int extent_thresh = range->extent_thresh;
-	int max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
-	int cluster = max_cluster;
+	unsigned long max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
+	unsigned long cluster = max_cluster;
 	u64 new_align = ~((u64)128 * 1024 - 1);
 	struct page **pages = NULL;
 
@@ -1254,7 +1279,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			break;
 
 		if (btrfs_defrag_cancelled(root->fs_info)) {
-			printk(KERN_DEBUG "btrfs: defrag_file cancelled\n");
+			printk(KERN_DEBUG "BTRFS: defrag_file cancelled\n");
 			ret = -EAGAIN;
 			break;
 		}
@@ -1416,20 +1441,20 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 			ret = -EINVAL;
 			goto out_free;
 		}
-		printk(KERN_INFO "btrfs: resizing devid %llu\n", devid);
+		btrfs_info(root->fs_info, "resizing devid %llu", devid);
 	}
 
 	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
 	if (!device) {
-		printk(KERN_INFO "btrfs: resizer unable to find device %llu\n",
+		btrfs_info(root->fs_info, "resizer unable to find device %llu",
 		       devid);
 		ret = -ENODEV;
 		goto out_free;
 	}
 
 	if (!device->writeable) {
-		printk(KERN_INFO "btrfs: resizer unable to apply on "
-		       "readonly device %llu\n",
+		btrfs_info(root->fs_info,
+			   "resizer unable to apply on readonly device %llu",
 		       devid);
 		ret = -EPERM;
 		goto out_free;
@@ -1466,6 +1491,10 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 		}
 		new_size = old_size - new_size;
 	} else if (mod > 0) {
+		if (new_size > ULLONG_MAX - old_size) {
+			ret = -EINVAL;
+			goto out_free;
+		}
 		new_size = old_size + new_size;
 	}
 
@@ -1481,7 +1510,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 	do_div(new_size, root->sectorsize);
 	new_size *= root->sectorsize;
 
-	printk_in_rcu(KERN_INFO "btrfs: new size for %s is %llu\n",
+	printk_in_rcu(KERN_INFO "BTRFS: new size for %s is %llu\n",
 		      rcu_str_deref(device->name), new_size);
 
 	if (new_size > old_size) {
@@ -1542,9 +1571,15 @@ static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
 
 		src_inode = file_inode(src.file);
 		if (src_inode->i_sb != file_inode(file)->i_sb) {
-			printk(KERN_INFO "btrfs: Snapshot src from "
-			       "another FS\n");
+			btrfs_info(BTRFS_I(src_inode)->root->fs_info,
+				   "Snapshot src from another FS");
 			ret = -EINVAL;
+		} else if (!inode_owner_or_capable(src_inode)) {
+			/*
+			 * Subvolume creation is not restricted, but snapshots
+			 * are limited to own subvolumes only
+			 */
+			ret = -EPERM;
 		} else {
 			ret = btrfs_mksubvol(&file->f_path, name, namelen,
 					     BTRFS_I(src_inode)->root,
@@ -1662,6 +1697,9 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
 	u64 flags;
 	int ret = 0;
 
+	if (!inode_owner_or_capable(inode))
+		return -EPERM;
+
 	ret = mnt_want_write_file(file);
 	if (ret)
 		goto out;
@@ -1686,11 +1724,6 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
 		goto out_drop_write;
 	}
 
-	if (!inode_owner_or_capable(inode)) {
-		ret = -EACCES;
-		goto out_drop_write;
-	}
-
 	down_write(&root->fs_info->subvol_sem);
 
 	/* nothing to do */
@@ -1698,12 +1731,28 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
 		goto out_drop_sem;
 
 	root_flags = btrfs_root_flags(&root->root_item);
-	if (flags & BTRFS_SUBVOL_RDONLY)
+	if (flags & BTRFS_SUBVOL_RDONLY) {
 		btrfs_set_root_flags(&root->root_item,
 				     root_flags | BTRFS_ROOT_SUBVOL_RDONLY);
-	else
-		btrfs_set_root_flags(&root->root_item,
+	} else {
+		/*
+		 * Block RO -> RW transition if this subvolume is involved in
+		 * send
+		 */
+		spin_lock(&root->root_item_lock);
+		if (root->send_in_progress == 0) {
+			btrfs_set_root_flags(&root->root_item,
 				     root_flags & ~BTRFS_ROOT_SUBVOL_RDONLY);
+			spin_unlock(&root->root_item_lock);
+		} else {
+			spin_unlock(&root->root_item_lock);
+			btrfs_warn(root->fs_info,
+			"Attempt to set subvolume %llu read-write during send",
+					root->root_key.objectid);
+			ret = -EPERM;
+			goto out_drop_sem;
+		}
+	}
 
 	trans = btrfs_start_transaction(root, 1);
 	if (IS_ERR(trans)) {
@@ -1910,7 +1959,7 @@ static noinline int search_ioctl(struct inode *inode,
 		key.offset = (u64)-1;
 		root = btrfs_read_fs_root_no_name(info, &key);
 		if (IS_ERR(root)) {
-			printk(KERN_ERR "could not find root %llu\n",
+			printk(KERN_ERR "BTRFS: could not find root %llu\n",
 			       sk->tree_id);
 			btrfs_free_path(path);
 			return -ENOENT;
@@ -2000,7 +2049,7 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
 	key.offset = (u64)-1;
 	root = btrfs_read_fs_root_no_name(info, &key);
 	if (IS_ERR(root)) {
-		printk(KERN_ERR "could not find root %llu\n", tree_id);
+		printk(KERN_ERR "BTRFS: could not find root %llu\n", tree_id);
 		ret = -ENOENT;
 		goto out;
 	}
@@ -2838,12 +2887,14 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 		 * note the key will change type as we walk through the
 		 * tree.
 		 */
+		path->leave_spinning = 1;
 		ret = btrfs_search_slot(NULL, BTRFS_I(src)->root, &key, path,
 				0, 0);
 		if (ret < 0)
 			goto out;
 
 		nritems = btrfs_header_nritems(path->nodes[0]);
+process_slot:
 		if (path->slots[0] >= nritems) {
 			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
 			if (ret < 0)
@@ -2870,11 +2921,6 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 			u8 comp;
 			u64 endoff;
 
-			size = btrfs_item_size_nr(leaf, slot);
-			read_extent_buffer(leaf, buf,
-					   btrfs_item_ptr_offset(leaf, slot),
-					   size);
-
 			extent = btrfs_item_ptr(leaf, slot,
 						struct btrfs_file_extent_item);
 			comp = btrfs_file_extent_compression(leaf, extent);
@@ -2893,11 +2939,20 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 				datal = btrfs_file_extent_ram_bytes(leaf,
 								    extent);
 			}
-			btrfs_release_path(path);
 
 			if (key.offset + datal <= off ||
-			    key.offset >= off + len - 1)
-				goto next;
+			    key.offset >= off + len - 1) {
+				path->slots[0]++;
+				goto process_slot;
+			}
+
+			size = btrfs_item_size_nr(leaf, slot);
+			read_extent_buffer(leaf, buf,
+					   btrfs_item_ptr_offset(leaf, slot),
+					   size);
+
+			btrfs_release_path(path);
+			path->leave_spinning = 0;
 
 			memcpy(&new_key, &key, sizeof(new_key));
 			new_key.objectid = btrfs_ino(inode);
@@ -3068,7 +3123,6 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 			}
 			ret = btrfs_end_transaction(trans, root);
 		}
-next:
 		btrfs_release_path(path);
 		key.offset++;
 	}
@@ -3196,9 +3250,17 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
 
 	unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
 out_unlock:
-	mutex_unlock(&src->i_mutex);
-	if (!same_inode)
-		mutex_unlock(&inode->i_mutex);
+	if (!same_inode) {
+		if (inode < src) {
+			mutex_unlock(&src->i_mutex);
+			mutex_unlock(&inode->i_mutex);
+		} else {
+			mutex_unlock(&inode->i_mutex);
+			mutex_unlock(&src->i_mutex);
+		}
+	} else {
+		mutex_unlock(&src->i_mutex);
+	}
 out_fput:
 	fdput(src_file);
 out_drop_write:
@@ -3321,8 +3383,8 @@ static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
 	if (IS_ERR_OR_NULL(di)) {
 		btrfs_free_path(path);
 		btrfs_end_transaction(trans, root);
-		printk(KERN_ERR "Umm, you don't have the default dir item, "
-		       "this isn't going to work\n");
+		btrfs_err(new_root->fs_info, "Umm, you don't have the default dir"
+			   "item, this isn't going to work");
 		ret = -ENOENT;
 		goto out;
 	}
@@ -3475,6 +3537,20 @@ out:
 	return ret;
 }
 
+static long btrfs_ioctl_global_rsv(struct btrfs_root *root, void __user *arg)
+{
+	struct btrfs_block_rsv *block_rsv = &root->fs_info->global_block_rsv;
+	u64 reserved;
+
+	spin_lock(&block_rsv->lock);
+	reserved = block_rsv->reserved;
+	spin_unlock(&block_rsv->lock);
+
+	if (arg && copy_to_user(arg, &reserved, sizeof(reserved)))
+		return -EFAULT;
+	return 0;
+}
+
 /*
  * there are many ways the trans_start and trans_end ioctls can lead
  * to deadlocks.  They should only be used by applications that
@@ -4303,6 +4379,9 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 	int ret = 0;
 	int received_uuid_changed;
 
+	if (!inode_owner_or_capable(inode))
+		return -EPERM;
+
 	ret = mnt_want_write_file(file);
 	if (ret < 0)
 		return ret;
@@ -4319,11 +4398,6 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 		goto out;
 	}
 
-	if (!inode_owner_or_capable(inode)) {
-		ret = -EACCES;
-		goto out;
-	}
-
 	sa = memdup_user(arg, sizeof(*sa));
 	if (IS_ERR(sa)) {
 		ret = PTR_ERR(sa);
@@ -4409,8 +4483,8 @@ static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg)
 	len = strnlen(label, BTRFS_LABEL_SIZE);
 
 	if (len == BTRFS_LABEL_SIZE) {
-		pr_warn("btrfs: label is too long, return the first %zu bytes\n",
-			--len);
+		btrfs_warn(root->fs_info,
+			"label is too long, return the first %zu bytes", --len);
 	}
 
 	ret = copy_to_user(arg, label, len);
@@ -4433,7 +4507,7 @@ static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg)
 		return -EFAULT;
 
 	if (strnlen(label, BTRFS_LABEL_SIZE) == BTRFS_LABEL_SIZE) {
-		pr_err("btrfs: unable to set label with more than %d bytes\n",
+		btrfs_err(root->fs_info, "unable to set label with more than %d bytes",
 		       BTRFS_LABEL_SIZE - 1);
 		return -EINVAL;
 	}
@@ -4458,6 +4532,166 @@ out_unlock:
 	return ret;
 }
 
+#define INIT_FEATURE_FLAGS(suffix) \
+	{ .compat_flags = BTRFS_FEATURE_COMPAT_##suffix, \
+	  .compat_ro_flags = BTRFS_FEATURE_COMPAT_RO_##suffix, \
+	  .incompat_flags = BTRFS_FEATURE_INCOMPAT_##suffix }
+
+static int btrfs_ioctl_get_supported_features(struct file *file,
+					      void __user *arg)
+{
+	static struct btrfs_ioctl_feature_flags features[3] = {
+		INIT_FEATURE_FLAGS(SUPP),
+		INIT_FEATURE_FLAGS(SAFE_SET),
+		INIT_FEATURE_FLAGS(SAFE_CLEAR)
+	};
+
+	if (copy_to_user(arg, &features, sizeof(features)))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int btrfs_ioctl_get_features(struct file *file, void __user *arg)
+{
+	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
+	struct btrfs_super_block *super_block = root->fs_info->super_copy;
+	struct btrfs_ioctl_feature_flags features;
+
+	features.compat_flags = btrfs_super_compat_flags(super_block);
+	features.compat_ro_flags = btrfs_super_compat_ro_flags(super_block);
+	features.incompat_flags = btrfs_super_incompat_flags(super_block);
+
+	if (copy_to_user(arg, &features, sizeof(features)))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int check_feature_bits(struct btrfs_root *root,
+			      enum btrfs_feature_set set,
+			      u64 change_mask, u64 flags, u64 supported_flags,
+			      u64 safe_set, u64 safe_clear)
+{
+	const char *type = btrfs_feature_set_names[set];
+	char *names;
+	u64 disallowed, unsupported;
+	u64 set_mask = flags & change_mask;
+	u64 clear_mask = ~flags & change_mask;
+
+	unsupported = set_mask & ~supported_flags;
+	if (unsupported) {
+		names = btrfs_printable_features(set, unsupported);
+		if (names) {
+			btrfs_warn(root->fs_info,
+			   "this kernel does not support the %s feature bit%s",
+			   names, strchr(names, ',') ? "s" : "");
+			kfree(names);
+		} else
+			btrfs_warn(root->fs_info,
+			   "this kernel does not support %s bits 0x%llx",
+			   type, unsupported);
+		return -EOPNOTSUPP;
+	}
+
+	disallowed = set_mask & ~safe_set;
+	if (disallowed) {
+		names = btrfs_printable_features(set, disallowed);
+		if (names) {
+			btrfs_warn(root->fs_info,
+			   "can't set the %s feature bit%s while mounted",
+			   names, strchr(names, ',') ? "s" : "");
+			kfree(names);
+		} else
+			btrfs_warn(root->fs_info,
+			   "can't set %s bits 0x%llx while mounted",
+			   type, disallowed);
+		return -EPERM;
+	}
+
+	disallowed = clear_mask & ~safe_clear;
+	if (disallowed) {
+		names = btrfs_printable_features(set, disallowed);
+		if (names) {
+			btrfs_warn(root->fs_info,
+			   "can't clear the %s feature bit%s while mounted",
+			   names, strchr(names, ',') ? "s" : "");
+			kfree(names);
+		} else
+			btrfs_warn(root->fs_info,
+			   "can't clear %s bits 0x%llx while mounted",
+			   type, disallowed);
+		return -EPERM;
+	}
+
+	return 0;
+}
+
+#define check_feature(root, change_mask, flags, mask_base)	\
+check_feature_bits(root, FEAT_##mask_base, change_mask, flags,	\
+		   BTRFS_FEATURE_ ## mask_base ## _SUPP,	\
+		   BTRFS_FEATURE_ ## mask_base ## _SAFE_SET,	\
+		   BTRFS_FEATURE_ ## mask_base ## _SAFE_CLEAR)
+
+static int btrfs_ioctl_set_features(struct file *file, void __user *arg)
+{
+	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
+	struct btrfs_super_block *super_block = root->fs_info->super_copy;
+	struct btrfs_ioctl_feature_flags flags[2];
+	struct btrfs_trans_handle *trans;
+	u64 newflags;
+	int ret;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	if (copy_from_user(flags, arg, sizeof(flags)))
+		return -EFAULT;
+
+	/* Nothing to do */
+	if (!flags[0].compat_flags && !flags[0].compat_ro_flags &&
+	    !flags[0].incompat_flags)
+		return 0;
+
+	ret = check_feature(root, flags[0].compat_flags,
+			    flags[1].compat_flags, COMPAT);
+	if (ret)
+		return ret;
+
+	ret = check_feature(root, flags[0].compat_ro_flags,
+			    flags[1].compat_ro_flags, COMPAT_RO);
+	if (ret)
+		return ret;
+
+	ret = check_feature(root, flags[0].incompat_flags,
+			    flags[1].incompat_flags, INCOMPAT);
+	if (ret)
+		return ret;
+
+	trans = btrfs_start_transaction(root, 1);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+
+	spin_lock(&root->fs_info->super_lock);
+	newflags = btrfs_super_compat_flags(super_block);
+	newflags |= flags[0].compat_flags & flags[1].compat_flags;
+	newflags &= ~(flags[0].compat_flags & ~flags[1].compat_flags);
+	btrfs_set_super_compat_flags(super_block, newflags);
+
+	newflags = btrfs_super_compat_ro_flags(super_block);
+	newflags |= flags[0].compat_ro_flags & flags[1].compat_ro_flags;
+	newflags &= ~(flags[0].compat_ro_flags & ~flags[1].compat_ro_flags);
+	btrfs_set_super_compat_ro_flags(super_block, newflags);
+
+	newflags = btrfs_super_incompat_flags(super_block);
+	newflags |= flags[0].incompat_flags & flags[1].incompat_flags;
+	newflags &= ~(flags[0].incompat_flags & ~flags[1].incompat_flags);
+	btrfs_set_super_incompat_flags(super_block, newflags);
+	spin_unlock(&root->fs_info->super_lock);
+
+	return btrfs_end_transaction(trans, root);
+}
+
 long btrfs_ioctl(struct file *file, unsigned int
 		cmd, unsigned long arg)
 {
@@ -4523,6 +4757,8 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_ioctl_logical_to_ino(root, argp);
 	case BTRFS_IOC_SPACE_INFO:
 		return btrfs_ioctl_space_info(root, argp);
+	case BTRFS_IOC_GLOBAL_RSV:
+		return btrfs_ioctl_global_rsv(root, argp);
 	case BTRFS_IOC_SYNC: {
 		int ret;
 
@@ -4576,6 +4812,12 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_ioctl_set_fslabel(file, argp);
 	case BTRFS_IOC_FILE_EXTENT_SAME:
 		return btrfs_ioctl_file_extent_same(file, argp);
+	case BTRFS_IOC_GET_SUPPORTED_FEATURES:
+		return btrfs_ioctl_get_supported_features(file, argp);
+	case BTRFS_IOC_GET_FEATURES:
+		return btrfs_ioctl_get_features(file, argp);
+	case BTRFS_IOC_SET_FEATURES:
+		return btrfs_ioctl_set_features(file, argp);
 	}
 
 	return -ENOTTY;

fs/btrfs/lzo.c

@@ -141,7 +141,7 @@ static int lzo_compress_pages(struct list_head *ws,
 		ret = lzo1x_1_compress(data_in, in_len, workspace->cbuf,
 				       &out_len, workspace->mem);
 		if (ret != LZO_E_OK) {
-			printk(KERN_DEBUG "btrfs deflate in loop returned %d\n",
+			printk(KERN_DEBUG "BTRFS: deflate in loop returned %d\n",
 			       ret);
 			ret = -1;
 			goto out;
@@ -357,7 +357,7 @@ cont:
 		if (need_unmap)
 			kunmap(pages_in[page_in_index - 1]);
 		if (ret != LZO_E_OK) {
-			printk(KERN_WARNING "btrfs decompress failed\n");
+			printk(KERN_WARNING "BTRFS: decompress failed\n");
 			ret = -1;
 			break;
 		}
@@ -401,7 +401,7 @@ static int lzo_decompress(struct list_head *ws, unsigned char *data_in,
 	out_len = PAGE_CACHE_SIZE;
 	ret = lzo1x_decompress_safe(data_in, in_len, workspace->buf, &out_len);
 	if (ret != LZO_E_OK) {
-		printk(KERN_WARNING "btrfs decompress failed!\n");
+		printk(KERN_WARNING "BTRFS: decompress failed!\n");
 		ret = -1;
 		goto out;
 	}

fs/btrfs/ordered-data.c

@@ -336,13 +336,14 @@ int btrfs_dec_test_first_ordered_pending(struct inode *inode,
 		      entry->len);
 	*file_offset = dec_end;
 	if (dec_start > dec_end) {
-		printk(KERN_CRIT "bad ordering dec_start %llu end %llu\n",
-		       dec_start, dec_end);
+		btrfs_crit(BTRFS_I(inode)->root->fs_info,
+			"bad ordering dec_start %llu end %llu", dec_start, dec_end);
 	}
 	to_dec = dec_end - dec_start;
 	if (to_dec > entry->bytes_left) {
-		printk(KERN_CRIT "bad ordered accounting left %llu size %llu\n",
-		       entry->bytes_left, to_dec);
+		btrfs_crit(BTRFS_I(inode)->root->fs_info,
+			"bad ordered accounting left %llu size %llu",
+			entry->bytes_left, to_dec);
 	}
 	entry->bytes_left -= to_dec;
 	if (!uptodate)
@@ -401,7 +402,8 @@ have_entry:
 	}
 
 	if (io_size > entry->bytes_left) {
-		printk(KERN_CRIT "bad ordered accounting left %llu size %llu\n",
+		btrfs_crit(BTRFS_I(inode)->root->fs_info,
+			   "bad ordered accounting left %llu size %llu",
 		       entry->bytes_left, io_size);
 	}
 	entry->bytes_left -= io_size;
@@ -520,7 +522,8 @@ void btrfs_remove_ordered_extent(struct inode *inode,
 	spin_lock_irq(&tree->lock);
 	node = &entry->rb_node;
 	rb_erase(node, &tree->tree);
-	tree->last = NULL;
+	if (tree->last == node)
+		tree->last = NULL;
 	set_bit(BTRFS_ORDERED_COMPLETE, &entry->flags);
 	spin_unlock_irq(&tree->lock);
 

fs/btrfs/orphan.c

@@ -69,23 +69,3 @@ out:
 	btrfs_free_path(path);
 	return ret;
 }
-
-int btrfs_find_orphan_item(struct btrfs_root *root, u64 offset)
-{
-	struct btrfs_path *path;
-	struct btrfs_key key;
-	int ret;
-
-	key.objectid = BTRFS_ORPHAN_OBJECTID;
-	key.type = BTRFS_ORPHAN_ITEM_KEY;
-	key.offset = offset;
-
-	path = btrfs_alloc_path();
-	if (!path)
-		return -ENOMEM;
-
-	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
-
-	btrfs_free_path(path);
-	return ret;
-}

fs/btrfs/print-tree.c

@@ -154,7 +154,7 @@ static void print_uuid_item(struct extent_buffer *l, unsigned long offset,
 			    u32 item_size)
 {
 	if (!IS_ALIGNED(item_size, sizeof(u64))) {
-		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+		pr_warn("BTRFS: uuid item with illegal size %lu!\n",
 			(unsigned long)item_size);
 		return;
 	}
@@ -249,7 +249,7 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 			    BTRFS_FILE_EXTENT_INLINE) {
 				printk(KERN_INFO "\t\tinline extent data "
 				       "size %u\n",
-				       btrfs_file_extent_inline_len(l, fi));
+				       btrfs_file_extent_inline_len(l, i, fi));
 				break;
 			}
 			printk(KERN_INFO "\t\textent data disk bytenr %llu "

fs/btrfs/props.c

@@ -0,0 +1,427 @@
+/*
+ * Copyright (C) 2014 Filipe David Borba Manana <fdmanana@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#include <linux/hashtable.h>
+#include "props.h"
+#include "btrfs_inode.h"
+#include "hash.h"
+#include "transaction.h"
+#include "xattr.h"
+
+#define BTRFS_PROP_HANDLERS_HT_BITS 8
+static DEFINE_HASHTABLE(prop_handlers_ht, BTRFS_PROP_HANDLERS_HT_BITS);
+
+struct prop_handler {
+	struct hlist_node node;
+	const char *xattr_name;
+	int (*validate)(const char *value, size_t len);
+	int (*apply)(struct inode *inode, const char *value, size_t len);
+	const char *(*extract)(struct inode *inode);
+	int inheritable;
+};
+
+static int prop_compression_validate(const char *value, size_t len);
+static int prop_compression_apply(struct inode *inode,
+				  const char *value,
+				  size_t len);
+static const char *prop_compression_extract(struct inode *inode);
+
+static struct prop_handler prop_handlers[] = {
+	{
+		.xattr_name = XATTR_BTRFS_PREFIX "compression",
+		.validate = prop_compression_validate,
+		.apply = prop_compression_apply,
+		.extract = prop_compression_extract,
+		.inheritable = 1
+	},
+	{
+		.xattr_name = NULL
+	}
+};
+
+void __init btrfs_props_init(void)
+{
+	struct prop_handler *p;
+
+	hash_init(prop_handlers_ht);
+
+	for (p = &prop_handlers[0]; p->xattr_name; p++) {
+		u64 h = btrfs_name_hash(p->xattr_name, strlen(p->xattr_name));
+
+		hash_add(prop_handlers_ht, &p->node, h);
+	}
+}
+
+static const struct hlist_head *find_prop_handlers_by_hash(const u64 hash)
+{
+	struct hlist_head *h;
+
+	h = &prop_handlers_ht[hash_min(hash, BTRFS_PROP_HANDLERS_HT_BITS)];
+	if (hlist_empty(h))
+		return NULL;
+
+	return h;
+}
+
+static const struct prop_handler *
+find_prop_handler(const char *name,
+		  const struct hlist_head *handlers)
+{
+	struct prop_handler *h;
+
+	if (!handlers) {
+		u64 hash = btrfs_name_hash(name, strlen(name));
+
+		handlers = find_prop_handlers_by_hash(hash);
+		if (!handlers)
+			return NULL;
+	}
+
+	hlist_for_each_entry(h, handlers, node)
+		if (!strcmp(h->xattr_name, name))
+			return h;
+
+	return NULL;
+}
+
+static int __btrfs_set_prop(struct btrfs_trans_handle *trans,
+			    struct inode *inode,
+			    const char *name,
+			    const char *value,
+			    size_t value_len,
+			    int flags)
+{
+	const struct prop_handler *handler;
+	int ret;
+
+	if (strlen(name) <= XATTR_BTRFS_PREFIX_LEN)
+		return -EINVAL;
+
+	handler = find_prop_handler(name, NULL);
+	if (!handler)
+		return -EINVAL;
+
+	if (value_len == 0) {
+		ret = __btrfs_setxattr(trans, inode, handler->xattr_name,
+				       NULL, 0, flags);
+		if (ret)
+			return ret;
+
+		ret = handler->apply(inode, NULL, 0);
+		ASSERT(ret == 0);
+
+		return ret;
+	}
+
+	ret = handler->validate(value, value_len);
+	if (ret)
+		return ret;
+	ret = __btrfs_setxattr(trans, inode, handler->xattr_name,
+			       value, value_len, flags);
+	if (ret)
+		return ret;
+	ret = handler->apply(inode, value, value_len);
+	if (ret) {
+		__btrfs_setxattr(trans, inode, handler->xattr_name,
+				 NULL, 0, flags);
+		return ret;
+	}
+
+	set_bit(BTRFS_INODE_HAS_PROPS, &BTRFS_I(inode)->runtime_flags);
+
+	return 0;
+}
+
+int btrfs_set_prop(struct inode *inode,
+		   const char *name,
+		   const char *value,
+		   size_t value_len,
+		   int flags)
+{
+	return __btrfs_set_prop(NULL, inode, name, value, value_len, flags);
+}
+
+static int iterate_object_props(struct btrfs_root *root,
+				struct btrfs_path *path,
+				u64 objectid,
+				void (*iterator)(void *,
+						 const struct prop_handler *,
+						 const char *,
+						 size_t),
+				void *ctx)
+{
+	int ret;
+	char *name_buf = NULL;
+	char *value_buf = NULL;
+	int name_buf_len = 0;
+	int value_buf_len = 0;
+
+	while (1) {
+		struct btrfs_key key;
+		struct btrfs_dir_item *di;
+		struct extent_buffer *leaf;
+		u32 total_len, cur, this_len;
+		int slot;
+		const struct hlist_head *handlers;
+
+		slot = path->slots[0];
+		leaf = path->nodes[0];
+
+		if (slot >= btrfs_header_nritems(leaf)) {
+			ret = btrfs_next_leaf(root, path);
+			if (ret < 0)
+				goto out;
+			else if (ret > 0)
+				break;
+			continue;
+		}
+
+		btrfs_item_key_to_cpu(leaf, &key, slot);
+		if (key.objectid != objectid)
+			break;
+		if (key.type != BTRFS_XATTR_ITEM_KEY)
+			break;
+
+		handlers = find_prop_handlers_by_hash(key.offset);
+		if (!handlers)
+			goto next_slot;
+
+		di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
+		cur = 0;
+		total_len = btrfs_item_size_nr(leaf, slot);
+
+		while (cur < total_len) {
+			u32 name_len = btrfs_dir_name_len(leaf, di);
+			u32 data_len = btrfs_dir_data_len(leaf, di);
+			unsigned long name_ptr, data_ptr;
+			const struct prop_handler *handler;
+
+			this_len = sizeof(*di) + name_len + data_len;
+			name_ptr = (unsigned long)(di + 1);
+			data_ptr = name_ptr + name_len;
+
+			if (name_len <= XATTR_BTRFS_PREFIX_LEN ||
+			    memcmp_extent_buffer(leaf, XATTR_BTRFS_PREFIX,
+						 name_ptr,
+						 XATTR_BTRFS_PREFIX_LEN))
+				goto next_dir_item;
+
+			if (name_len >= name_buf_len) {
+				kfree(name_buf);
+				name_buf_len = name_len + 1;
+				name_buf = kmalloc(name_buf_len, GFP_NOFS);
+				if (!name_buf) {
+					ret = -ENOMEM;
+					goto out;
+				}
+			}
+			read_extent_buffer(leaf, name_buf, name_ptr, name_len);
+			name_buf[name_len] = '\0';
+
+			handler = find_prop_handler(name_buf, handlers);
+			if (!handler)
+				goto next_dir_item;
+
+			if (data_len > value_buf_len) {
+				kfree(value_buf);
+				value_buf_len = data_len;
+				value_buf = kmalloc(data_len, GFP_NOFS);
+				if (!value_buf) {
+					ret = -ENOMEM;
+					goto out;
+				}
+			}
+			read_extent_buffer(leaf, value_buf, data_ptr, data_len);
+
+			iterator(ctx, handler, value_buf, data_len);
+next_dir_item:
+			cur += this_len;
+			di = (struct btrfs_dir_item *)((char *) di + this_len);
+		}
+
+next_slot:
+		path->slots[0]++;
+	}
+
+	ret = 0;
+out:
+	btrfs_release_path(path);
+	kfree(name_buf);
+	kfree(value_buf);
+
+	return ret;
+}
+
+static void inode_prop_iterator(void *ctx,
+				const struct prop_handler *handler,
+				const char *value,
+				size_t len)
+{
+	struct inode *inode = ctx;
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	int ret;
+
+	ret = handler->apply(inode, value, len);
+	if (unlikely(ret))
+		btrfs_warn(root->fs_info,
+			   "error applying prop %s to ino %llu (root %llu): %d",
+			   handler->xattr_name, btrfs_ino(inode),
+			   root->root_key.objectid, ret);
+	else
+		set_bit(BTRFS_INODE_HAS_PROPS, &BTRFS_I(inode)->runtime_flags);
+}
+
+int btrfs_load_inode_props(struct inode *inode, struct btrfs_path *path)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	u64 ino = btrfs_ino(inode);
+	int ret;
+
+	ret = iterate_object_props(root, path, ino, inode_prop_iterator, inode);
+
+	return ret;
+}
+
+static int inherit_props(struct btrfs_trans_handle *trans,
+			 struct inode *inode,
+			 struct inode *parent)
+{
+	const struct prop_handler *h;
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	int ret;
+
+	if (!test_bit(BTRFS_INODE_HAS_PROPS,
+		      &BTRFS_I(parent)->runtime_flags))
+		return 0;
+
+	for (h = &prop_handlers[0]; h->xattr_name; h++) {
+		const char *value;
+		u64 num_bytes;
+
+		if (!h->inheritable)
+			continue;
+
+		value = h->extract(parent);
+		if (!value)
+			continue;
+
+		num_bytes = btrfs_calc_trans_metadata_size(root, 1);
+		ret = btrfs_block_rsv_add(root, trans->block_rsv,
+					  num_bytes, BTRFS_RESERVE_NO_FLUSH);
+		if (ret)
+			goto out;
+		ret = __btrfs_set_prop(trans, inode, h->xattr_name,
+				       value, strlen(value), 0);
+		btrfs_block_rsv_release(root, trans->block_rsv, num_bytes);
+		if (ret)
+			goto out;
+	}
+	ret = 0;
+out:
+	return ret;
+}
+
+int btrfs_inode_inherit_props(struct btrfs_trans_handle *trans,
+			      struct inode *inode,
+			      struct inode *dir)
+{
+	if (!dir)
+		return 0;
+
+	return inherit_props(trans, inode, dir);
+}
+
+int btrfs_subvol_inherit_props(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *root,
+			       struct btrfs_root *parent_root)
+{
+	struct btrfs_key key;
+	struct inode *parent_inode, *child_inode;
+	int ret;
+
+	key.objectid = BTRFS_FIRST_FREE_OBJECTID;
+	key.type = BTRFS_INODE_ITEM_KEY;
+	key.offset = 0;
+
+	parent_inode = btrfs_iget(parent_root->fs_info->sb, &key,
+				  parent_root, NULL);
+	if (IS_ERR(parent_inode))
+		return PTR_ERR(parent_inode);
+
+	child_inode = btrfs_iget(root->fs_info->sb, &key, root, NULL);
+	if (IS_ERR(child_inode)) {
+		iput(parent_inode);
+		return PTR_ERR(child_inode);
+	}
+
+	ret = inherit_props(trans, child_inode, parent_inode);
+	iput(child_inode);
+	iput(parent_inode);
+
+	return ret;
+}
+
+static int prop_compression_validate(const char *value, size_t len)
+{
+	if (!strncmp("lzo", value, len))
+		return 0;
+	else if (!strncmp("zlib", value, len))
+		return 0;
+
+	return -EINVAL;
+}
+
+static int prop_compression_apply(struct inode *inode,
+				  const char *value,
+				  size_t len)
+{
+	int type;
+
+	if (len == 0) {
+		BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
+		BTRFS_I(inode)->flags &= ~BTRFS_INODE_COMPRESS;
+		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
+
+		return 0;
+	}
+
+	if (!strncmp("lzo", value, len))
+		type = BTRFS_COMPRESS_LZO;
+	else if (!strncmp("zlib", value, len))
+		type = BTRFS_COMPRESS_ZLIB;
+	else
+		return -EINVAL;
+
+	BTRFS_I(inode)->flags &= ~BTRFS_INODE_NOCOMPRESS;
+	BTRFS_I(inode)->flags |= BTRFS_INODE_COMPRESS;
+	BTRFS_I(inode)->force_compress = type;
+
+	return 0;
+}
+
+static const char *prop_compression_extract(struct inode *inode)
+{
+	switch (BTRFS_I(inode)->force_compress) {
+	case BTRFS_COMPRESS_ZLIB:
+		return "zlib";
+	case BTRFS_COMPRESS_LZO:
+		return "lzo";
+	}
+
+	return NULL;
+}

fs/btrfs/props.h

@@ -0,0 +1,42 @@
+/*
+ * Copyright (C) 2014 Filipe David Borba Manana <fdmanana@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#ifndef __BTRFS_PROPS_H
+#define __BTRFS_PROPS_H
+
+#include "ctree.h"
+
+void __init btrfs_props_init(void);
+
+int btrfs_set_prop(struct inode *inode,
+		   const char *name,
+		   const char *value,
+		   size_t value_len,
+		   int flags);
+
+int btrfs_load_inode_props(struct inode *inode, struct btrfs_path *path);
+
+int btrfs_inode_inherit_props(struct btrfs_trans_handle *trans,
+			      struct inode *inode,
+			      struct inode *dir);
+
+int btrfs_subvol_inherit_props(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *root,
+			       struct btrfs_root *parent_root);
+
+#endif

fs/btrfs/qgroup.c

@@ -301,16 +301,16 @@ int btrfs_read_qgroup_config(struct btrfs_fs_info *fs_info)
 
 			if (btrfs_qgroup_status_version(l, ptr) !=
 			    BTRFS_QGROUP_STATUS_VERSION) {
-				printk(KERN_ERR
-				 "btrfs: old qgroup version, quota disabled\n");
+				btrfs_err(fs_info,
+				 "old qgroup version, quota disabled");
 				goto out;
 			}
 			if (btrfs_qgroup_status_generation(l, ptr) !=
 			    fs_info->generation) {
 				flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
-				printk(KERN_ERR
-					"btrfs: qgroup generation mismatch, "
-					"marked as inconsistent\n");
+				btrfs_err(fs_info,
+					"qgroup generation mismatch, "
+					"marked as inconsistent");
 			}
 			fs_info->qgroup_flags = btrfs_qgroup_status_flags(l,
 									  ptr);
@@ -325,7 +325,7 @@ int btrfs_read_qgroup_config(struct btrfs_fs_info *fs_info)
 		qgroup = find_qgroup_rb(fs_info, found_key.offset);
 		if ((qgroup && found_key.type == BTRFS_QGROUP_INFO_KEY) ||
 		    (!qgroup && found_key.type == BTRFS_QGROUP_LIMIT_KEY)) {
-			printk(KERN_ERR "btrfs: inconsitent qgroup config\n");
+			btrfs_err(fs_info, "inconsitent qgroup config");
 			flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
 		}
 		if (!qgroup) {
@@ -396,8 +396,8 @@ next1:
 		ret = add_relation_rb(fs_info, found_key.objectid,
 				      found_key.offset);
 		if (ret == -ENOENT) {
-			printk(KERN_WARNING
-				"btrfs: orphan qgroup relation 0x%llx->0x%llx\n",
+			btrfs_warn(fs_info,
+				"orphan qgroup relation 0x%llx->0x%llx",
 				found_key.objectid, found_key.offset);
 			ret = 0;	/* ignore the error */
 		}
@@ -644,8 +644,7 @@ static int update_qgroup_limit_item(struct btrfs_trans_handle *trans,
 
 	l = path->nodes[0];
 	slot = path->slots[0];
-	qgroup_limit = btrfs_item_ptr(l, path->slots[0],
-				      struct btrfs_qgroup_limit_item);
+	qgroup_limit = btrfs_item_ptr(l, slot, struct btrfs_qgroup_limit_item);
 	btrfs_set_qgroup_limit_flags(l, qgroup_limit, flags);
 	btrfs_set_qgroup_limit_max_rfer(l, qgroup_limit, max_rfer);
 	btrfs_set_qgroup_limit_max_excl(l, qgroup_limit, max_excl);
@@ -687,8 +686,7 @@ static int update_qgroup_info_item(struct btrfs_trans_handle *trans,
 
 	l = path->nodes[0];
 	slot = path->slots[0];
-	qgroup_info = btrfs_item_ptr(l, path->slots[0],
-				 struct btrfs_qgroup_info_item);
+	qgroup_info = btrfs_item_ptr(l, slot, struct btrfs_qgroup_info_item);
 	btrfs_set_qgroup_info_generation(l, qgroup_info, trans->transid);
 	btrfs_set_qgroup_info_rfer(l, qgroup_info, qgroup->rfer);
 	btrfs_set_qgroup_info_rfer_cmpr(l, qgroup_info, qgroup->rfer_cmpr);
@@ -1161,7 +1159,7 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans,
 				       limit->rsv_excl);
 	if (ret) {
 		fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
-		printk(KERN_INFO "unable to update quota limit for %llu\n",
+		btrfs_info(fs_info, "unable to update quota limit for %llu",
 		       qgroupid);
 	}
 
@@ -1349,7 +1347,6 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans,
 			     struct btrfs_delayed_ref_node *node,
 			     struct btrfs_delayed_extent_op *extent_op)
 {
-	struct btrfs_key ins;
 	struct btrfs_root *quota_root;
 	u64 ref_root;
 	struct btrfs_qgroup *qgroup;
@@ -1363,10 +1360,6 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans,
 
 	BUG_ON(!fs_info->quota_root);
 
-	ins.objectid = node->bytenr;
-	ins.offset = node->num_bytes;
-	ins.type = BTRFS_EXTENT_ITEM_KEY;
-
 	if (node->type == BTRFS_TREE_BLOCK_REF_KEY ||
 	    node->type == BTRFS_SHARED_BLOCK_REF_KEY) {
 		struct btrfs_delayed_tree_ref *ref;
@@ -1840,7 +1833,9 @@ void assert_qgroups_uptodate(struct btrfs_trans_handle *trans)
 {
 	if (list_empty(&trans->qgroup_ref_list) && !trans->delayed_ref_elem.seq)
 		return;
-	pr_err("btrfs: qgroups not uptodate in trans handle %p: list is%s empty, seq is %#x.%x\n",
+	btrfs_err(trans->root->fs_info,
+		"qgroups not uptodate in trans handle %p:  list is%s empty, "
+		"seq is %#x.%x",
 		trans, list_empty(&trans->qgroup_ref_list) ? "" : " not",
 		(u32)(trans->delayed_ref_elem.seq >> 32),
 		(u32)trans->delayed_ref_elem.seq);
@@ -1902,9 +1897,17 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 	mutex_unlock(&fs_info->qgroup_rescan_lock);
 
 	for (; slot < btrfs_header_nritems(scratch_leaf); ++slot) {
+		u64 num_bytes;
+
 		btrfs_item_key_to_cpu(scratch_leaf, &found, slot);
-		if (found.type != BTRFS_EXTENT_ITEM_KEY)
+		if (found.type != BTRFS_EXTENT_ITEM_KEY &&
+		    found.type != BTRFS_METADATA_ITEM_KEY)
 			continue;
+		if (found.type == BTRFS_METADATA_ITEM_KEY)
+			num_bytes = fs_info->extent_root->leafsize;
+		else
+			num_bytes = found.offset;
+
 		ret = btrfs_find_all_roots(trans, fs_info, found.objectid,
 					   tree_mod_seq_elem.seq, &roots);
 		if (ret < 0)
@@ -1949,12 +1952,12 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 			struct btrfs_qgroup_list *glist;
 
 			qg = (struct btrfs_qgroup *)(uintptr_t) unode->aux;
-			qg->rfer += found.offset;
-			qg->rfer_cmpr += found.offset;
+			qg->rfer += num_bytes;
+			qg->rfer_cmpr += num_bytes;
 			WARN_ON(qg->tag >= seq);
 			if (qg->refcnt - seq == roots->nnodes) {
-				qg->excl += found.offset;
-				qg->excl_cmpr += found.offset;
+				qg->excl += num_bytes;
+				qg->excl_cmpr += num_bytes;
 			}
 			qgroup_dirty(fs_info, qg);
 
@@ -2037,10 +2040,10 @@ out:
 	mutex_unlock(&fs_info->qgroup_rescan_lock);
 
 	if (err >= 0) {
-		pr_info("btrfs: qgroup scan completed%s\n",
+		btrfs_info(fs_info, "qgroup scan completed%s",
 			err == 2 ? " (inconsistency flag cleared)" : "");
 	} else {
-		pr_err("btrfs: qgroup scan failed with %d\n", err);
+		btrfs_err(fs_info, "qgroup scan failed with %d", err);
 	}
 
 	complete_all(&fs_info->qgroup_rescan_completion);
@@ -2096,7 +2099,7 @@ qgroup_rescan_init(struct btrfs_fs_info *fs_info, u64 progress_objectid,
 
 	if (ret) {
 err:
-		pr_info("btrfs: qgroup_rescan_init failed with %d\n", ret);
+		btrfs_info(fs_info, "qgroup_rescan_init failed with %d", ret);
 		return ret;
 	}
 

fs/btrfs/reada.c

@@ -189,8 +189,8 @@ static int __readahead_hook(struct btrfs_root *root, struct extent_buffer *eb,
 			 */
 #ifdef DEBUG
 			if (rec->generation != generation) {
-				printk(KERN_DEBUG "generation mismatch for "
-						"(%llu,%d,%llu) %llu != %llu\n",
+				btrfs_debug(root->fs_info,
+					   "generation mismatch for (%llu,%d,%llu) %llu != %llu",
 				       key.objectid, key.type, key.offset,
 				       rec->generation, generation);
 			}
@@ -365,8 +365,9 @@ static struct reada_extent *reada_find_extent(struct btrfs_root *root,
 		goto error;
 
 	if (bbio->num_stripes > BTRFS_MAX_MIRRORS) {
-		printk(KERN_ERR "btrfs readahead: more than %d copies not "
-				"supported", BTRFS_MAX_MIRRORS);
+		btrfs_err(root->fs_info,
+			   "readahead: more than %d copies not supported",
+			   BTRFS_MAX_MIRRORS);
 		goto error;
 	}
 

fs/btrfs/relocation.c

@@ -94,6 +94,7 @@ struct backref_edge {
 
 #define LOWER	0
 #define UPPER	1
+#define RELOCATION_RESERVED_NODES	256
 
 struct backref_cache {
 	/* red black tree of all backref nodes in the cache */
@@ -176,6 +177,8 @@ struct reloc_control {
 	u64 merging_rsv_size;
 	/* size of relocated tree nodes */
 	u64 nodes_relocated;
+	/* reserved size for block group relocation*/
+	u64 reserved_bytes;
 
 	u64 search_start;
 	u64 extents_found;
@@ -184,7 +187,6 @@ struct reloc_control {
 	unsigned int create_reloc_tree:1;
 	unsigned int merge_reloc_tree:1;
 	unsigned int found_file_extent:1;
-	unsigned int commit_transaction:1;
 };
 
 /* stages of data relocation */
@@ -2309,9 +2311,6 @@ void free_reloc_roots(struct list_head *list)
 		reloc_root = list_entry(list->next, struct btrfs_root,
 					root_list);
 		__del_reloc_root(reloc_root);
-		free_extent_buffer(reloc_root->node);
-		free_extent_buffer(reloc_root->commit_root);
-		kfree(reloc_root);
 	}
 }
 
@@ -2353,10 +2352,9 @@ again:
 
 			ret = merge_reloc_root(rc, root);
 			if (ret) {
-				__del_reloc_root(reloc_root);
-				free_extent_buffer(reloc_root->node);
-				free_extent_buffer(reloc_root->commit_root);
-				kfree(reloc_root);
+				if (list_empty(&reloc_root->root_list))
+					list_add_tail(&reloc_root->root_list,
+						      &reloc_roots);
 				goto out;
 			}
 		} else {
@@ -2452,7 +2450,7 @@ static noinline_for_stack
 struct btrfs_root *select_reloc_root(struct btrfs_trans_handle *trans,
 				     struct reloc_control *rc,
 				     struct backref_node *node,
-				     struct backref_edge *edges[], int *nr)
+				     struct backref_edge *edges[])
 {
 	struct backref_node *next;
 	struct btrfs_root *root;
@@ -2494,7 +2492,6 @@ struct btrfs_root *select_reloc_root(struct btrfs_trans_handle *trans,
 	if (!root)
 		return NULL;
 
-	*nr = index;
 	next = node;
 	/* setup backref node path for btrfs_reloc_cow_block */
 	while (1) {
@@ -2590,28 +2587,36 @@ static int reserve_metadata_space(struct btrfs_trans_handle *trans,
 	struct btrfs_root *root = rc->extent_root;
 	u64 num_bytes;
 	int ret;
+	u64 tmp;
 
 	num_bytes = calcu_metadata_size(rc, node, 1) * 2;
 
 	trans->block_rsv = rc->block_rsv;
-	ret = btrfs_block_rsv_add(root, rc->block_rsv, num_bytes,
-				  BTRFS_RESERVE_FLUSH_ALL);
+	rc->reserved_bytes += num_bytes;
+	ret = btrfs_block_rsv_refill(root, rc->block_rsv, num_bytes,
+				BTRFS_RESERVE_FLUSH_ALL);
 	if (ret) {
-		if (ret == -EAGAIN)
-			rc->commit_transaction = 1;
+		if (ret == -EAGAIN) {
+			tmp = rc->extent_root->nodesize *
+				RELOCATION_RESERVED_NODES;
+			while (tmp <= rc->reserved_bytes)
+				tmp <<= 1;
+			/*
+			 * only one thread can access block_rsv at this point,
+			 * so we don't need hold lock to protect block_rsv.
+			 * we expand more reservation size here to allow enough
+			 * space for relocation and we will return eailer in
+			 * enospc case.
+			 */
+			rc->block_rsv->size = tmp + rc->extent_root->nodesize *
+					      RELOCATION_RESERVED_NODES;
+		}
 		return ret;
 	}
 
 	return 0;
 }
 
-static void release_metadata_space(struct reloc_control *rc,
-				   struct backref_node *node)
-{
-	u64 num_bytes = calcu_metadata_size(rc, node, 0) * 2;
-	btrfs_block_rsv_release(rc->extent_root, rc->block_rsv, num_bytes);
-}
-
 /*
  * relocate a block tree, and then update pointers in upper level
  * blocks that reference the block to point to the new location.
@@ -2633,7 +2638,6 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 	u32 blocksize;
 	u64 bytenr;
 	u64 generation;
-	int nr;
 	int slot;
 	int ret;
 	int err = 0;
@@ -2646,7 +2650,7 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 		cond_resched();
 
 		upper = edge->node[UPPER];
-		root = select_reloc_root(trans, rc, upper, edges, &nr);
+		root = select_reloc_root(trans, rc, upper, edges);
 		BUG_ON(!root);
 
 		if (upper->eb && !upper->locked) {
@@ -2898,7 +2902,6 @@ static int relocate_tree_block(struct btrfs_trans_handle *trans,
 				struct btrfs_path *path)
 {
 	struct btrfs_root *root;
-	int release = 0;
 	int ret = 0;
 
 	if (!node)
@@ -2915,7 +2918,6 @@ static int relocate_tree_block(struct btrfs_trans_handle *trans,
 		ret = reserve_metadata_space(trans, rc, node);
 		if (ret)
 			goto out;
-		release = 1;
 	}
 
 	if (root) {
@@ -2940,11 +2942,8 @@ static int relocate_tree_block(struct btrfs_trans_handle *trans,
 		ret = do_relocation(trans, rc, node, key, path, 1);
 	}
 out:
-	if (ret || node->level == 0 || node->cowonly) {
-		if (release)
-			release_metadata_space(rc, node);
+	if (ret || node->level == 0 || node->cowonly)
 		remove_backref_node(&rc->backref_cache, node);
-	}
 	return ret;
 }
 
@@ -3867,29 +3866,20 @@ static noinline_for_stack
 int prepare_to_relocate(struct reloc_control *rc)
 {
 	struct btrfs_trans_handle *trans;
-	int ret;
 
 	rc->block_rsv = btrfs_alloc_block_rsv(rc->extent_root,
 					      BTRFS_BLOCK_RSV_TEMP);
 	if (!rc->block_rsv)
 		return -ENOMEM;
 
-	/*
-	 * reserve some space for creating reloc trees.
-	 * btrfs_init_reloc_root will use them when there
-	 * is no reservation in transaction handle.
-	 */
-	ret = btrfs_block_rsv_add(rc->extent_root, rc->block_rsv,
-				  rc->extent_root->nodesize * 256,
-				  BTRFS_RESERVE_FLUSH_ALL);
-	if (ret)
-		return ret;
-
 	memset(&rc->cluster, 0, sizeof(rc->cluster));
 	rc->search_start = rc->block_group->key.objectid;
 	rc->extents_found = 0;
 	rc->nodes_relocated = 0;
 	rc->merging_rsv_size = 0;
+	rc->reserved_bytes = 0;
+	rc->block_rsv->size = rc->extent_root->nodesize *
+			      RELOCATION_RESERVED_NODES;
 
 	rc->create_reloc_tree = 1;
 	set_reloc_control(rc);
@@ -3933,6 +3923,14 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 	}
 
 	while (1) {
+		rc->reserved_bytes = 0;
+		ret = btrfs_block_rsv_refill(rc->extent_root,
+					rc->block_rsv, rc->block_rsv->size,
+					BTRFS_RESERVE_FLUSH_ALL);
+		if (ret) {
+			err = ret;
+			break;
+		}
 		progress++;
 		trans = btrfs_start_transaction(rc->extent_root, 0);
 		if (IS_ERR(trans)) {
@@ -4011,6 +4009,12 @@ restart:
 		if (!RB_EMPTY_ROOT(&blocks)) {
 			ret = relocate_tree_blocks(trans, rc, &blocks);
 			if (ret < 0) {
+				/*
+				 * if we fail to relocate tree blocks, force to update
+				 * backref cache when committing transaction.
+				 */
+				rc->backref_cache.last_trans = trans->transid - 1;
+
 				if (ret != -EAGAIN) {
 					err = ret;
 					break;
@@ -4020,14 +4024,8 @@ restart:
 			}
 		}
 
-		if (rc->commit_transaction) {
-			rc->commit_transaction = 0;
-			ret = btrfs_commit_transaction(trans, rc->extent_root);
-			BUG_ON(ret);
-		} else {
-			btrfs_end_transaction_throttle(trans, rc->extent_root);
-			btrfs_btree_balance_dirty(rc->extent_root);
-		}
+		btrfs_end_transaction_throttle(trans, rc->extent_root);
+		btrfs_btree_balance_dirty(rc->extent_root);
 		trans = NULL;
 
 		if (rc->stage == MOVE_DATA_EXTENTS &&
@@ -4247,7 +4245,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 		goto out;
 	}
 
-	printk(KERN_INFO "btrfs: relocating block group %llu flags %llu\n",
+	btrfs_info(extent_root->fs_info, "relocating block group %llu flags %llu",
 	       rc->block_group->key.objectid, rc->block_group->flags);
 
 	ret = btrfs_start_delalloc_roots(fs_info, 0);
@@ -4269,7 +4267,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 		if (rc->extents_found == 0)
 			break;
 
-		printk(KERN_INFO "btrfs: found %llu extents\n",
+		btrfs_info(extent_root->fs_info, "found %llu extents",
 			rc->extents_found);
 
 		if (rc->stage == MOVE_DATA_EXTENTS && rc->found_file_extent) {
@@ -4285,11 +4283,6 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 		}
 	}
 
-	filemap_write_and_wait_range(fs_info->btree_inode->i_mapping,
-				     rc->block_group->key.objectid,
-				     rc->block_group->key.objectid +
-				     rc->block_group->key.offset - 1);
-
 	WARN_ON(rc->block_group->pinned > 0);
 	WARN_ON(rc->block_group->reserved > 0);
 	WARN_ON(btrfs_block_group_used(&rc->block_group->item) > 0);

fs/btrfs/root-tree.c

@@ -44,7 +44,7 @@ static void btrfs_read_root_item(struct extent_buffer *eb, int slot,
 	if (!need_reset && btrfs_root_generation(item)
 		!= btrfs_root_generation_v2(item)) {
 		if (btrfs_root_generation_v2(item) != 0) {
-			printk(KERN_WARNING "btrfs: mismatching "
+			printk(KERN_WARNING "BTRFS: mismatching "
 					"generation and generation_v2 "
 					"found in root item. This root "
 					"was probably mounted with an "
@@ -154,7 +154,7 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root
 
 	if (ret != 0) {
 		btrfs_print_leaf(root, path->nodes[0]);
-		printk(KERN_CRIT "unable to update root key %llu %u %llu\n",
+		btrfs_crit(root->fs_info, "unable to update root key %llu %u %llu",
 		       key->objectid, key->type, key->offset);
 		BUG_ON(1);
 	}
@@ -400,21 +400,6 @@ out:
 	return err;
 }
 
-int btrfs_find_root_ref(struct btrfs_root *tree_root,
-		   struct btrfs_path *path,
-		   u64 root_id, u64 ref_id)
-{
-	struct btrfs_key key;
-	int ret;
-
-	key.objectid = root_id;
-	key.type = BTRFS_ROOT_REF_KEY;
-	key.offset = ref_id;
-
-	ret = btrfs_search_slot(NULL, tree_root, &key, path, 0, 0);
-	return ret;
-}
-
 /*
  * add a btrfs_root_ref item.  type is either BTRFS_ROOT_REF_KEY
  * or BTRFS_ROOT_BACKREF_KEY.

fs/btrfs/scrub.c

@@ -256,6 +256,8 @@ static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root,
 static int copy_nocow_pages(struct scrub_ctx *sctx, u64 logical, u64 len,
 			    int mirror_num, u64 physical_for_dev_replace);
 static void copy_nocow_pages_worker(struct btrfs_work *work);
+static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info);
+static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info);
 
 
 static void scrub_pending_bio_inc(struct scrub_ctx *sctx)
@@ -269,6 +271,29 @@ static void scrub_pending_bio_dec(struct scrub_ctx *sctx)
 	wake_up(&sctx->list_wait);
 }
 
+static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info)
+{
+	while (atomic_read(&fs_info->scrub_pause_req)) {
+		mutex_unlock(&fs_info->scrub_lock);
+		wait_event(fs_info->scrub_pause_wait,
+		   atomic_read(&fs_info->scrub_pause_req) == 0);
+		mutex_lock(&fs_info->scrub_lock);
+	}
+}
+
+static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info)
+{
+	atomic_inc(&fs_info->scrubs_paused);
+	wake_up(&fs_info->scrub_pause_wait);
+
+	mutex_lock(&fs_info->scrub_lock);
+	__scrub_blocked_if_needed(fs_info);
+	atomic_dec(&fs_info->scrubs_paused);
+	mutex_unlock(&fs_info->scrub_lock);
+
+	wake_up(&fs_info->scrub_pause_wait);
+}
+
 /*
  * used for workers that require transaction commits (i.e., for the
  * NOCOW case)
@@ -480,7 +505,7 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root,
 	 * hold all of the paths here
 	 */
 	for (i = 0; i < ipath->fspath->elem_cnt; ++i)
-		printk_in_rcu(KERN_WARNING "btrfs: %s at logical %llu on dev "
+		printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev "
 			"%s, sector %llu, root %llu, inode %llu, offset %llu, "
 			"length %llu, links %u (path: %s)\n", swarn->errstr,
 			swarn->logical, rcu_str_deref(swarn->dev->name),
@@ -492,7 +517,7 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root,
 	return 0;
 
 err:
-	printk_in_rcu(KERN_WARNING "btrfs: %s at logical %llu on dev "
+	printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev "
 		"%s, sector %llu, root %llu, inode %llu, offset %llu: path "
 		"resolving failed with ret=%d\n", swarn->errstr,
 		swarn->logical, rcu_str_deref(swarn->dev->name),
@@ -555,7 +580,7 @@ static void scrub_print_warning(const char *errstr, struct scrub_block *sblock)
 			ret = tree_backref_for_extent(&ptr, eb, ei, item_size,
 							&ref_root, &ref_level);
 			printk_in_rcu(KERN_WARNING
-				"btrfs: %s at logical %llu on dev %s, "
+				"BTRFS: %s at logical %llu on dev %s, "
 				"sector %llu: metadata %s (level %d) in tree "
 				"%llu\n", errstr, swarn.logical,
 				rcu_str_deref(dev->name),
@@ -704,13 +729,11 @@ static void scrub_fixup_nodatasum(struct btrfs_work *work)
 	struct scrub_fixup_nodatasum *fixup;
 	struct scrub_ctx *sctx;
 	struct btrfs_trans_handle *trans = NULL;
-	struct btrfs_fs_info *fs_info;
 	struct btrfs_path *path;
 	int uncorrectable = 0;
 
 	fixup = container_of(work, struct scrub_fixup_nodatasum, work);
 	sctx = fixup->sctx;
-	fs_info = fixup->root->fs_info;
 
 	path = btrfs_alloc_path();
 	if (!path) {
@@ -759,8 +782,8 @@ out:
 		btrfs_dev_replace_stats_inc(
 			&sctx->dev_root->fs_info->dev_replace.
 			num_uncorrectable_read_errors);
-		printk_ratelimited_in_rcu(KERN_ERR
-			"btrfs: unable to fixup (nodatasum) error at logical %llu on dev %s\n",
+		printk_ratelimited_in_rcu(KERN_ERR "BTRFS: "
+		    "unable to fixup (nodatasum) error at logical %llu on dev %s\n",
 			fixup->logical, rcu_str_deref(fixup->dev->name));
 	}
 
@@ -1161,7 +1184,7 @@ corrected_error:
 			sctx->stat.corrected_errors++;
 			spin_unlock(&sctx->stat_lock);
 			printk_ratelimited_in_rcu(KERN_ERR
-				"btrfs: fixed up error at logical %llu on dev %s\n",
+				"BTRFS: fixed up error at logical %llu on dev %s\n",
 				logical, rcu_str_deref(dev->name));
 		}
 	} else {
@@ -1170,7 +1193,7 @@ did_not_correct_error:
 		sctx->stat.uncorrectable_errors++;
 		spin_unlock(&sctx->stat_lock);
 		printk_ratelimited_in_rcu(KERN_ERR
-			"btrfs: unable to fixup (regular) error at logical %llu on dev %s\n",
+			"BTRFS: unable to fixup (regular) error at logical %llu on dev %s\n",
 			logical, rcu_str_deref(dev->name));
 	}
 
@@ -1418,8 +1441,9 @@ static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad,
 		int ret;
 
 		if (!page_bad->dev->bdev) {
-			printk_ratelimited(KERN_WARNING
-				"btrfs: scrub_repair_page_from_good_copy(bdev == NULL) is unexpected!\n");
+			printk_ratelimited(KERN_WARNING "BTRFS: "
+				"scrub_repair_page_from_good_copy(bdev == NULL) "
+				"is unexpected!\n");
 			return -EIO;
 		}
 
@@ -1877,7 +1901,7 @@ static void scrub_submit(struct scrub_ctx *sctx)
 		 * This case is handled correctly (but _very_ slowly).
 		 */
 		printk_ratelimited(KERN_WARNING
-			"btrfs: scrub_submit(bio bdev == NULL) is unexpected!\n");
+			"BTRFS: scrub_submit(bio bdev == NULL) is unexpected!\n");
 		bio_endio(sbio->bio, -EIO);
 	} else {
 		btrfsic_submit_bio(READ, sbio->bio);
@@ -2286,8 +2310,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 
 	wait_event(sctx->list_wait,
 		   atomic_read(&sctx->bios_in_flight) == 0);
-	atomic_inc(&fs_info->scrubs_paused);
-	wake_up(&fs_info->scrub_pause_wait);
+	scrub_blocked_if_needed(fs_info);
 
 	/* FIXME it might be better to start readahead at commit root */
 	key_start.objectid = logical;
@@ -2311,16 +2334,6 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 	if (!IS_ERR(reada2))
 		btrfs_reada_wait(reada2);
 
-	mutex_lock(&fs_info->scrub_lock);
-	while (atomic_read(&fs_info->scrub_pause_req)) {
-		mutex_unlock(&fs_info->scrub_lock);
-		wait_event(fs_info->scrub_pause_wait,
-		   atomic_read(&fs_info->scrub_pause_req) == 0);
-		mutex_lock(&fs_info->scrub_lock);
-	}
-	atomic_dec(&fs_info->scrubs_paused);
-	mutex_unlock(&fs_info->scrub_lock);
-	wake_up(&fs_info->scrub_pause_wait);
 
 	/*
 	 * collect all data csums for the stripe to avoid seeking during
@@ -2357,22 +2370,14 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 			wait_event(sctx->list_wait,
 				   atomic_read(&sctx->bios_in_flight) == 0);
 			atomic_set(&sctx->wr_ctx.flush_all_writes, 0);
-			atomic_inc(&fs_info->scrubs_paused);
-			wake_up(&fs_info->scrub_pause_wait);
-			mutex_lock(&fs_info->scrub_lock);
-			while (atomic_read(&fs_info->scrub_pause_req)) {
-				mutex_unlock(&fs_info->scrub_lock);
-				wait_event(fs_info->scrub_pause_wait,
-				   atomic_read(&fs_info->scrub_pause_req) == 0);
-				mutex_lock(&fs_info->scrub_lock);
-			}
-			atomic_dec(&fs_info->scrubs_paused);
-			mutex_unlock(&fs_info->scrub_lock);
-			wake_up(&fs_info->scrub_pause_wait);
+			scrub_blocked_if_needed(fs_info);
 		}
 
+		if (btrfs_fs_incompat(fs_info, SKINNY_METADATA))
+			key.type = BTRFS_METADATA_ITEM_KEY;
+		else
+			key.type = BTRFS_EXTENT_ITEM_KEY;
 		key.objectid = logical;
-		key.type = BTRFS_EXTENT_ITEM_KEY;
 		key.offset = (u64)-1;
 
 		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
@@ -2380,8 +2385,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 			goto out;
 
 		if (ret > 0) {
-			ret = btrfs_previous_item(root, path, 0,
-						  BTRFS_EXTENT_ITEM_KEY);
+			ret = btrfs_previous_extent_item(root, path, 0);
 			if (ret < 0)
 				goto out;
 			if (ret > 0) {
@@ -2439,9 +2443,9 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 
 			if (key.objectid < logical &&
 			    (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)) {
-				printk(KERN_ERR
-				       "btrfs scrub: tree block %llu spanning "
-				       "stripes, ignored. logical=%llu\n",
+				btrfs_err(fs_info,
+					   "scrub: tree block %llu spanning "
+					   "stripes, ignored. logical=%llu",
 				       key.objectid, logical);
 				goto next;
 			}
@@ -2683,21 +2687,9 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 		wait_event(sctx->list_wait,
 			   atomic_read(&sctx->bios_in_flight) == 0);
 		atomic_set(&sctx->wr_ctx.flush_all_writes, 0);
-		atomic_inc(&fs_info->scrubs_paused);
-		wake_up(&fs_info->scrub_pause_wait);
 		wait_event(sctx->list_wait,
 			   atomic_read(&sctx->workers_pending) == 0);
-
-		mutex_lock(&fs_info->scrub_lock);
-		while (atomic_read(&fs_info->scrub_pause_req)) {
-			mutex_unlock(&fs_info->scrub_lock);
-			wait_event(fs_info->scrub_pause_wait,
-			   atomic_read(&fs_info->scrub_pause_req) == 0);
-			mutex_lock(&fs_info->scrub_lock);
-		}
-		atomic_dec(&fs_info->scrubs_paused);
-		mutex_unlock(&fs_info->scrub_lock);
-		wake_up(&fs_info->scrub_pause_wait);
+		scrub_blocked_if_needed(fs_info);
 
 		btrfs_put_block_group(cache);
 		if (ret)
@@ -2823,8 +2815,8 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	 * check some assumptions
 	 */
 	if (fs_info->chunk_root->nodesize != fs_info->chunk_root->leafsize) {
-		printk(KERN_ERR
-		       "btrfs_scrub: size assumption nodesize == leafsize (%d == %d) fails\n",
+		btrfs_err(fs_info,
+			   "scrub: size assumption nodesize == leafsize (%d == %d) fails",
 		       fs_info->chunk_root->nodesize,
 		       fs_info->chunk_root->leafsize);
 		return -EINVAL;
@@ -2836,16 +2828,17 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		 * the way scrub is implemented. Do not handle this
 		 * situation at all because it won't ever happen.
 		 */
-		printk(KERN_ERR
-		       "btrfs_scrub: size assumption nodesize <= BTRFS_STRIPE_LEN (%d <= %d) fails\n",
+		btrfs_err(fs_info,
+			   "scrub: size assumption nodesize <= BTRFS_STRIPE_LEN (%d <= %d) fails",
 		       fs_info->chunk_root->nodesize, BTRFS_STRIPE_LEN);
 		return -EINVAL;
 	}
 
 	if (fs_info->chunk_root->sectorsize != PAGE_SIZE) {
 		/* not supported for data w/o checksums */
-		printk(KERN_ERR
-		       "btrfs_scrub: size assumption sectorsize != PAGE_SIZE (%d != %lu) fails\n",
+		btrfs_err(fs_info,
+			   "scrub: size assumption sectorsize != PAGE_SIZE "
+			   "(%d != %lu) fails",
 		       fs_info->chunk_root->sectorsize, PAGE_SIZE);
 		return -EINVAL;
 	}
@@ -2858,7 +2851,8 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		 * would exhaust the array bounds of pagev member in
 		 * struct scrub_block
 		 */
-		pr_err("btrfs_scrub: size assumption nodesize and sectorsize <= SCRUB_MAX_PAGES_PER_BLOCK (%d <= %d && %d <= %d) fails\n",
+		btrfs_err(fs_info, "scrub: size assumption nodesize and sectorsize "
+			   "<= SCRUB_MAX_PAGES_PER_BLOCK (%d <= %d && %d <= %d) fails",
 		       fs_info->chunk_root->nodesize,
 		       SCRUB_MAX_PAGES_PER_BLOCK,
 		       fs_info->chunk_root->sectorsize,
@@ -2908,7 +2902,13 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	}
 	sctx->readonly = readonly;
 	dev->scrub_device = sctx;
+	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 
+	/*
+	 * checking @scrub_pause_req here, we can avoid
+	 * race between committing transaction and scrubbing.
+	 */
+	__scrub_blocked_if_needed(fs_info);
 	atomic_inc(&fs_info->scrubs_running);
 	mutex_unlock(&fs_info->scrub_lock);
 
@@ -2917,9 +2917,10 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 		 * by holding device list mutex, we can
 		 * kick off writing super in log tree sync.
 		 */
+		mutex_lock(&fs_info->fs_devices->device_list_mutex);
 		ret = scrub_supers(sctx, dev);
+		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 	}
-	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 
 	if (!ret)
 		ret = scrub_enumerate_chunks(sctx, dev, start, end,
@@ -3167,7 +3168,8 @@ static void copy_nocow_pages_worker(struct btrfs_work *work)
 	ret = iterate_inodes_from_logical(logical, fs_info, path,
 					  record_inode_for_nocow, nocow_ctx);
 	if (ret != 0 && ret != -ENOENT) {
-		pr_warn("iterate_inodes_from_logical() failed: log %llu, phys %llu, len %llu, mir %u, ret %d\n",
+		btrfs_warn(fs_info, "iterate_inodes_from_logical() failed: log %llu, "
+			"phys %llu, len %llu, mir %u, ret %d",
 			logical, physical_for_dev_replace, len, mirror_num,
 			ret);
 		not_written = 1;
@@ -3289,7 +3291,7 @@ static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root,
 again:
 		page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
 		if (!page) {
-			pr_err("find_or_create_page() failed\n");
+			btrfs_err(fs_info, "find_or_create_page() failed");
 			ret = -ENOMEM;
 			goto out;
 		}
@@ -3361,7 +3363,7 @@ static int write_page_nocow(struct scrub_ctx *sctx,
 		return -EIO;
 	if (!dev->bdev) {
 		printk_ratelimited(KERN_WARNING
-			"btrfs: scrub write_page_nocow(bdev == NULL) is unexpected!\n");
+			"BTRFS: scrub write_page_nocow(bdev == NULL) is unexpected!\n");
 		return -EIO;
 	}
 	bio = btrfs_io_bio_alloc(GFP_NOFS, 1);

fs/btrfs/send.c

@@ -88,8 +88,6 @@ struct send_ctx {
 	u64 cmd_send_size[BTRFS_SEND_C_MAX + 1];
 	u64 flags;	/* 'flags' member of btrfs_ioctl_send_args is u64 */
 
-	struct vfsmount *mnt;
-
 	struct btrfs_root *send_root;
 	struct btrfs_root *parent_root;
 	struct clone_root *clone_roots;
@@ -111,6 +109,7 @@ struct send_ctx {
 	int cur_inode_deleted;
 	u64 cur_inode_size;
 	u64 cur_inode_mode;
+	u64 cur_inode_last_extent;
 
 	u64 send_progress;
 
@@ -122,6 +121,74 @@ struct send_ctx {
 	int name_cache_size;
 
 	char *read_buf;
+
+	/*
+	 * We process inodes by their increasing order, so if before an
+	 * incremental send we reverse the parent/child relationship of
+	 * directories such that a directory with a lower inode number was
+	 * the parent of a directory with a higher inode number, and the one
+	 * becoming the new parent got renamed too, we can't rename/move the
+	 * directory with lower inode number when we finish processing it - we
+	 * must process the directory with higher inode number first, then
+	 * rename/move it and then rename/move the directory with lower inode
+	 * number. Example follows.
+	 *
+	 * Tree state when the first send was performed:
+	 *
+	 * .
+	 * |-- a                   (ino 257)
+	 *     |-- b               (ino 258)
+	 *         |
+	 *         |
+	 *         |-- c           (ino 259)
+	 *         |   |-- d       (ino 260)
+	 *         |
+	 *         |-- c2          (ino 261)
+	 *
+	 * Tree state when the second (incremental) send is performed:
+	 *
+	 * .
+	 * |-- a                   (ino 257)
+	 *     |-- b               (ino 258)
+	 *         |-- c2          (ino 261)
+	 *             |-- d2      (ino 260)
+	 *                 |-- cc  (ino 259)
+	 *
+	 * The sequence of steps that lead to the second state was:
+	 *
+	 * mv /a/b/c/d /a/b/c2/d2
+	 * mv /a/b/c /a/b/c2/d2/cc
+	 *
+	 * "c" has lower inode number, but we can't move it (2nd mv operation)
+	 * before we move "d", which has higher inode number.
+	 *
+	 * So we just memorize which move/rename operations must be performed
+	 * later when their respective parent is processed and moved/renamed.
+	 */
+
+	/* Indexed by parent directory inode number. */
+	struct rb_root pending_dir_moves;
+
+	/*
+	 * Reverse index, indexed by the inode number of a directory that
+	 * is waiting for the move/rename of its immediate parent before its
+	 * own move/rename can be performed.
+	 */
+	struct rb_root waiting_dir_moves;
+};
+
+struct pending_dir_move {
+	struct rb_node node;
+	struct list_head list;
+	u64 parent_ino;
+	u64 ino;
+	u64 gen;
+	struct list_head update_refs;
+};
+
+struct waiting_dir_move {
+	struct rb_node node;
+	u64 ino;
 };
 
 struct name_cache_entry {
@@ -145,6 +212,15 @@ struct name_cache_entry {
 	char name[];
 };
 
+static int is_waiting_for_move(struct send_ctx *sctx, u64 ino);
+
+static int need_send_hole(struct send_ctx *sctx)
+{
+	return (sctx->parent_root && !sctx->cur_inode_new &&
+		!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted &&
+		S_ISREG(sctx->cur_inode_mode));
+}
+
 static void fs_path_reset(struct fs_path *p)
 {
 	if (p->reversed) {
@@ -336,16 +412,6 @@ out:
 	return ret;
 }
 
-#if 0
-static void fs_path_remove(struct fs_path *p)
-{
-	BUG_ON(p->reversed);
-	while (p->start != p->end && *p->end != '/')
-		p->end--;
-	*p->end = 0;
-}
-#endif
-
 static int fs_path_copy(struct fs_path *p, struct fs_path *from)
 {
 	int ret;
@@ -436,30 +502,15 @@ static int tlv_put(struct send_ctx *sctx, u16 attr, const void *data, int len)
 	return 0;
 }
 
-#if 0
-static int tlv_put_u8(struct send_ctx *sctx, u16 attr, u8 value)
-{
-	return tlv_put(sctx, attr, &value, sizeof(value));
-}
-
-static int tlv_put_u16(struct send_ctx *sctx, u16 attr, u16 value)
-{
-	__le16 tmp = cpu_to_le16(value);
-	return tlv_put(sctx, attr, &tmp, sizeof(tmp));
-}
-
-static int tlv_put_u32(struct send_ctx *sctx, u16 attr, u32 value)
-{
-	__le32 tmp = cpu_to_le32(value);
-	return tlv_put(sctx, attr, &tmp, sizeof(tmp));
-}
-#endif
+#define TLV_PUT_DEFINE_INT(bits) \
+	static int tlv_put_u##bits(struct send_ctx *sctx,	 	\
+			u##bits attr, u##bits value)			\
+	{								\
+		__le##bits __tmp = cpu_to_le##bits(value);		\
+		return tlv_put(sctx, attr, &__tmp, sizeof(__tmp));	\
+	}
 
-static int tlv_put_u64(struct send_ctx *sctx, u16 attr, u64 value)
-{
-	__le64 tmp = cpu_to_le64(value);
-	return tlv_put(sctx, attr, &tmp, sizeof(tmp));
-}
+TLV_PUT_DEFINE_INT(64)
 
 static int tlv_put_string(struct send_ctx *sctx, u16 attr,
 			  const char *str, int len)
@@ -475,17 +526,6 @@ static int tlv_put_uuid(struct send_ctx *sctx, u16 attr,
 	return tlv_put(sctx, attr, uuid, BTRFS_UUID_SIZE);
 }
 
-#if 0
-static int tlv_put_timespec(struct send_ctx *sctx, u16 attr,
-			    struct timespec *ts)
-{
-	struct btrfs_timespec bts;
-	bts.sec = cpu_to_le64(ts->tv_sec);
-	bts.nsec = cpu_to_le32(ts->tv_nsec);
-	return tlv_put(sctx, attr, &bts, sizeof(bts));
-}
-#endif
-
 static int tlv_put_btrfs_timespec(struct send_ctx *sctx, u16 attr,
 				  struct extent_buffer *eb,
 				  struct btrfs_timespec *ts)
@@ -533,12 +573,6 @@ static int tlv_put_btrfs_timespec(struct send_ctx *sctx, u16 attr,
 		if (ret < 0) \
 			goto tlv_put_failure; \
 	} while (0)
-#define TLV_PUT_TIMESPEC(sctx, attrtype, ts) \
-	do { \
-		ret = tlv_put_timespec(sctx, attrtype, ts); \
-		if (ret < 0) \
-			goto tlv_put_failure; \
-	} while (0)
 #define TLV_PUT_BTRFS_TIMESPEC(sctx, attrtype, eb, ts) \
 	do { \
 		ret = tlv_put_btrfs_timespec(sctx, attrtype, eb, ts); \
@@ -1270,7 +1304,7 @@ static int find_extent_clone(struct send_ctx *sctx,
 	if (!backref_ctx->found_itself) {
 		/* found a bug in backref code? */
 		ret = -EIO;
-		printk(KERN_ERR "btrfs: ERROR did not find backref in "
+		btrfs_err(sctx->send_root->fs_info, "did not find backref in "
 				"send_root. inode=%llu, offset=%llu, "
 				"disk_byte=%llu found extent=%llu\n",
 				ino, data_offset, disk_byte, found_key.objectid);
@@ -1343,7 +1377,7 @@ static int read_symlink(struct btrfs_root *root,
 	BUG_ON(compression);
 
 	off = btrfs_file_extent_inline_start(ei);
-	len = btrfs_file_extent_inline_len(path->nodes[0], ei);
+	len = btrfs_file_extent_inline_len(path->nodes[0], path->slots[0], ei);
 
 	ret = fs_path_add_from_extent_buffer(dest, path->nodes[0], off, len);
 
@@ -1372,7 +1406,7 @@ static int gen_unique_name(struct send_ctx *sctx,
 		return -ENOMEM;
 
 	while (1) {
-		len = snprintf(tmp, sizeof(tmp) - 1, "o%llu-%llu-%llu",
+		len = snprintf(tmp, sizeof(tmp), "o%llu-%llu-%llu",
 				ino, gen, idx);
 		if (len >= sizeof(tmp)) {
 			/* should really not happen */
@@ -1933,6 +1967,7 @@ static void name_cache_free(struct send_ctx *sctx)
  */
 static int __get_cur_name_and_parent(struct send_ctx *sctx,
 				     u64 ino, u64 gen,
+				     int skip_name_cache,
 				     u64 *parent_ino,
 				     u64 *parent_gen,
 				     struct fs_path *dest)
@@ -1942,6 +1977,8 @@ static int __get_cur_name_and_parent(struct send_ctx *sctx,
 	struct btrfs_path *path = NULL;
 	struct name_cache_entry *nce = NULL;
 
+	if (skip_name_cache)
+		goto get_ref;
 	/*
 	 * First check if we already did a call to this function with the same
 	 * ino/gen. If yes, check if the cache entry is still up-to-date. If yes
@@ -1986,11 +2023,12 @@ static int __get_cur_name_and_parent(struct send_ctx *sctx,
 		goto out_cache;
 	}
 
+get_ref:
 	/*
 	 * Depending on whether the inode was already processed or not, use
 	 * send_root or parent_root for ref lookup.
 	 */
-	if (ino < sctx->send_progress)
+	if (ino < sctx->send_progress && !skip_name_cache)
 		ret = get_first_ref(sctx->send_root, ino,
 				    parent_ino, parent_gen, dest);
 	else
@@ -2014,6 +2052,8 @@ static int __get_cur_name_and_parent(struct send_ctx *sctx,
 			goto out;
 		ret = 1;
 	}
+	if (skip_name_cache)
+		goto out;
 
 out_cache:
 	/*
@@ -2081,6 +2121,9 @@ static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen,
 	u64 parent_inode = 0;
 	u64 parent_gen = 0;
 	int stop = 0;
+	u64 start_ino = ino;
+	u64 start_gen = gen;
+	int skip_name_cache = 0;
 
 	name = fs_path_alloc();
 	if (!name) {
@@ -2088,19 +2131,32 @@ static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen,
 		goto out;
 	}
 
+	if (is_waiting_for_move(sctx, ino))
+		skip_name_cache = 1;
+
+again:
 	dest->reversed = 1;
 	fs_path_reset(dest);
 
 	while (!stop && ino != BTRFS_FIRST_FREE_OBJECTID) {
 		fs_path_reset(name);
 
-		ret = __get_cur_name_and_parent(sctx, ino, gen,
+		ret = __get_cur_name_and_parent(sctx, ino, gen, skip_name_cache,
 				&parent_inode, &parent_gen, name);
 		if (ret < 0)
 			goto out;
 		if (ret)
 			stop = 1;
 
+		if (!skip_name_cache &&
+		    is_waiting_for_move(sctx, parent_inode)) {
+			ino = start_ino;
+			gen = start_gen;
+			stop = 0;
+			skip_name_cache = 1;
+			goto again;
+		}
+
 		ret = fs_path_add_path(dest, name);
 		if (ret < 0)
 			goto out;
@@ -2131,7 +2187,7 @@ static int send_subvol_begin(struct send_ctx *sctx)
 	char *name = NULL;
 	int namelen;
 
-	path = alloc_path_for_send();
+	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
 
@@ -2180,12 +2236,12 @@ static int send_subvol_begin(struct send_ctx *sctx)
 	TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID,
 			sctx->send_root->root_item.uuid);
 	TLV_PUT_U64(sctx, BTRFS_SEND_A_CTRANSID,
-			sctx->send_root->root_item.ctransid);
+		    le64_to_cpu(sctx->send_root->root_item.ctransid));
 	if (parent_root) {
 		TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
 				sctx->parent_root->root_item.uuid);
 		TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID,
-				sctx->parent_root->root_item.ctransid);
+			    le64_to_cpu(sctx->parent_root->root_item.ctransid));
 	}
 
 	ret = send_cmd(sctx);
@@ -2672,10 +2728,349 @@ out:
 	return ret;
 }
 
+static int is_waiting_for_move(struct send_ctx *sctx, u64 ino)
+{
+	struct rb_node *n = sctx->waiting_dir_moves.rb_node;
+	struct waiting_dir_move *entry;
+
+	while (n) {
+		entry = rb_entry(n, struct waiting_dir_move, node);
+		if (ino < entry->ino)
+			n = n->rb_left;
+		else if (ino > entry->ino)
+			n = n->rb_right;
+		else
+			return 1;
+	}
+	return 0;
+}
+
+static int add_waiting_dir_move(struct send_ctx *sctx, u64 ino)
+{
+	struct rb_node **p = &sctx->waiting_dir_moves.rb_node;
+	struct rb_node *parent = NULL;
+	struct waiting_dir_move *entry, *dm;
+
+	dm = kmalloc(sizeof(*dm), GFP_NOFS);
+	if (!dm)
+		return -ENOMEM;
+	dm->ino = ino;
+
+	while (*p) {
+		parent = *p;
+		entry = rb_entry(parent, struct waiting_dir_move, node);
+		if (ino < entry->ino) {
+			p = &(*p)->rb_left;
+		} else if (ino > entry->ino) {
+			p = &(*p)->rb_right;
+		} else {
+			kfree(dm);
+			return -EEXIST;
+		}
+	}
+
+	rb_link_node(&dm->node, parent, p);
+	rb_insert_color(&dm->node, &sctx->waiting_dir_moves);
+	return 0;
+}
+
+#ifdef CONFIG_BTRFS_ASSERT
+
+static int del_waiting_dir_move(struct send_ctx *sctx, u64 ino)
+{
+	struct rb_node *n = sctx->waiting_dir_moves.rb_node;
+	struct waiting_dir_move *entry;
+
+	while (n) {
+		entry = rb_entry(n, struct waiting_dir_move, node);
+		if (ino < entry->ino) {
+			n = n->rb_left;
+		} else if (ino > entry->ino) {
+			n = n->rb_right;
+		} else {
+			rb_erase(&entry->node, &sctx->waiting_dir_moves);
+			kfree(entry);
+			return 0;
+		}
+	}
+	return -ENOENT;
+}
+
+#endif
+
+static int add_pending_dir_move(struct send_ctx *sctx, u64 parent_ino)
+{
+	struct rb_node **p = &sctx->pending_dir_moves.rb_node;
+	struct rb_node *parent = NULL;
+	struct pending_dir_move *entry, *pm;
+	struct recorded_ref *cur;
+	int exists = 0;
+	int ret;
+
+	pm = kmalloc(sizeof(*pm), GFP_NOFS);
+	if (!pm)
+		return -ENOMEM;
+	pm->parent_ino = parent_ino;
+	pm->ino = sctx->cur_ino;
+	pm->gen = sctx->cur_inode_gen;
+	INIT_LIST_HEAD(&pm->list);
+	INIT_LIST_HEAD(&pm->update_refs);
+	RB_CLEAR_NODE(&pm->node);
+
+	while (*p) {
+		parent = *p;
+		entry = rb_entry(parent, struct pending_dir_move, node);
+		if (parent_ino < entry->parent_ino) {
+			p = &(*p)->rb_left;
+		} else if (parent_ino > entry->parent_ino) {
+			p = &(*p)->rb_right;
+		} else {
+			exists = 1;
+			break;
+		}
+	}
+
+	list_for_each_entry(cur, &sctx->deleted_refs, list) {
+		ret = dup_ref(cur, &pm->update_refs);
+		if (ret < 0)
+			goto out;
+	}
+	list_for_each_entry(cur, &sctx->new_refs, list) {
+		ret = dup_ref(cur, &pm->update_refs);
+		if (ret < 0)
+			goto out;
+	}
+
+	ret = add_waiting_dir_move(sctx, pm->ino);
+	if (ret)
+		goto out;
+
+	if (exists) {
+		list_add_tail(&pm->list, &entry->list);
+	} else {
+		rb_link_node(&pm->node, parent, p);
+		rb_insert_color(&pm->node, &sctx->pending_dir_moves);
+	}
+	ret = 0;
+out:
+	if (ret) {
+		__free_recorded_refs(&pm->update_refs);
+		kfree(pm);
+	}
+	return ret;
+}
+
+static struct pending_dir_move *get_pending_dir_moves(struct send_ctx *sctx,
+						      u64 parent_ino)
+{
+	struct rb_node *n = sctx->pending_dir_moves.rb_node;
+	struct pending_dir_move *entry;
+
+	while (n) {
+		entry = rb_entry(n, struct pending_dir_move, node);
+		if (parent_ino < entry->parent_ino)
+			n = n->rb_left;
+		else if (parent_ino > entry->parent_ino)
+			n = n->rb_right;
+		else
+			return entry;
+	}
+	return NULL;
+}
+
+static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm)
+{
+	struct fs_path *from_path = NULL;
+	struct fs_path *to_path = NULL;
+	u64 orig_progress = sctx->send_progress;
+	struct recorded_ref *cur;
+	int ret;
+
+	from_path = fs_path_alloc();
+	if (!from_path)
+		return -ENOMEM;
+
+	sctx->send_progress = pm->ino;
+	ret = get_cur_path(sctx, pm->ino, pm->gen, from_path);
+	if (ret < 0)
+		goto out;
+
+	to_path = fs_path_alloc();
+	if (!to_path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	sctx->send_progress = sctx->cur_ino + 1;
+	ASSERT(del_waiting_dir_move(sctx, pm->ino) == 0);
+	ret = get_cur_path(sctx, pm->ino, pm->gen, to_path);
+	if (ret < 0)
+		goto out;
+
+	ret = send_rename(sctx, from_path, to_path);
+	if (ret < 0)
+		goto out;
+
+	ret = send_utimes(sctx, pm->ino, pm->gen);
+	if (ret < 0)
+		goto out;
+
+	/*
+	 * After rename/move, need to update the utimes of both new parent(s)
+	 * and old parent(s).
+	 */
+	list_for_each_entry(cur, &pm->update_refs, list) {
+		ret = send_utimes(sctx, cur->dir, cur->dir_gen);
+		if (ret < 0)
+			goto out;
+	}
+
+out:
+	fs_path_free(from_path);
+	fs_path_free(to_path);
+	sctx->send_progress = orig_progress;
+
+	return ret;
+}
+
+static void free_pending_move(struct send_ctx *sctx, struct pending_dir_move *m)
+{
+	if (!list_empty(&m->list))
+		list_del(&m->list);
+	if (!RB_EMPTY_NODE(&m->node))
+		rb_erase(&m->node, &sctx->pending_dir_moves);
+	__free_recorded_refs(&m->update_refs);
+	kfree(m);
+}
+
+static void tail_append_pending_moves(struct pending_dir_move *moves,
+				      struct list_head *stack)
+{
+	if (list_empty(&moves->list)) {
+		list_add_tail(&moves->list, stack);
+	} else {
+		LIST_HEAD(list);
+		list_splice_init(&moves->list, &list);
+		list_add_tail(&moves->list, stack);
+		list_splice_tail(&list, stack);
+	}
+}
+
+static int apply_children_dir_moves(struct send_ctx *sctx)
+{
+	struct pending_dir_move *pm;
+	struct list_head stack;
+	u64 parent_ino = sctx->cur_ino;
+	int ret = 0;
+
+	pm = get_pending_dir_moves(sctx, parent_ino);
+	if (!pm)
+		return 0;
+
+	INIT_LIST_HEAD(&stack);
+	tail_append_pending_moves(pm, &stack);
+
+	while (!list_empty(&stack)) {
+		pm = list_first_entry(&stack, struct pending_dir_move, list);
+		parent_ino = pm->ino;
+		ret = apply_dir_move(sctx, pm);
+		free_pending_move(sctx, pm);
+		if (ret)
+			goto out;
+		pm = get_pending_dir_moves(sctx, parent_ino);
+		if (pm)
+			tail_append_pending_moves(pm, &stack);
+	}
+	return 0;
+
+out:
+	while (!list_empty(&stack)) {
+		pm = list_first_entry(&stack, struct pending_dir_move, list);
+		free_pending_move(sctx, pm);
+	}
+	return ret;
+}
+
+static int wait_for_parent_move(struct send_ctx *sctx,
+				struct recorded_ref *parent_ref)
+{
+	int ret;
+	u64 ino = parent_ref->dir;
+	u64 parent_ino_before, parent_ino_after;
+	u64 new_gen, old_gen;
+	struct fs_path *path_before = NULL;
+	struct fs_path *path_after = NULL;
+	int len1, len2;
+
+	if (parent_ref->dir <= sctx->cur_ino)
+		return 0;
+
+	if (is_waiting_for_move(sctx, ino))
+		return 1;
+
+	ret = get_inode_info(sctx->parent_root, ino, NULL, &old_gen,
+			     NULL, NULL, NULL, NULL);
+	if (ret == -ENOENT)
+		return 0;
+	else if (ret < 0)
+		return ret;
+
+	ret = get_inode_info(sctx->send_root, ino, NULL, &new_gen,
+			     NULL, NULL, NULL, NULL);
+	if (ret < 0)
+		return ret;
+
+	if (new_gen != old_gen)
+		return 0;
+
+	path_before = fs_path_alloc();
+	if (!path_before)
+		return -ENOMEM;
+
+	ret = get_first_ref(sctx->parent_root, ino, &parent_ino_before,
+			    NULL, path_before);
+	if (ret == -ENOENT) {
+		ret = 0;
+		goto out;
+	} else if (ret < 0) {
+		goto out;
+	}
+
+	path_after = fs_path_alloc();
+	if (!path_after) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = get_first_ref(sctx->send_root, ino, &parent_ino_after,
+			    NULL, path_after);
+	if (ret == -ENOENT) {
+		ret = 0;
+		goto out;
+	} else if (ret < 0) {
+		goto out;
+	}
+
+	len1 = fs_path_len(path_before);
+	len2 = fs_path_len(path_after);
+	if ((parent_ino_before != parent_ino_after) && (len1 != len2 ||
+	     memcmp(path_before->start, path_after->start, len1))) {
+		ret = 1;
+		goto out;
+	}
+	ret = 0;
+
+out:
+	fs_path_free(path_before);
+	fs_path_free(path_after);
+
+	return ret;
+}
+
 /*
  * This does all the move/link/unlink/rmdir magic.
  */
-static int process_recorded_refs(struct send_ctx *sctx)
+static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
 {
 	int ret = 0;
 	struct recorded_ref *cur;
@@ -2824,11 +3219,17 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 				 * dirs, we always have one new and one deleted
 				 * ref. The deleted ref is ignored later.
 				 */
-				ret = send_rename(sctx, valid_path,
-						cur->full_path);
-				if (ret < 0)
-					goto out;
-				ret = fs_path_copy(valid_path, cur->full_path);
+				if (wait_for_parent_move(sctx, cur)) {
+					ret = add_pending_dir_move(sctx,
+								   cur->dir);
+					*pending_move = 1;
+				} else {
+					ret = send_rename(sctx, valid_path,
+							  cur->full_path);
+					if (!ret)
+						ret = fs_path_copy(valid_path,
+							       cur->full_path);
+				}
 				if (ret < 0)
 					goto out;
 			} else {
@@ -3197,6 +3598,7 @@ static int process_all_refs(struct send_ctx *sctx,
 	struct extent_buffer *eb;
 	int slot;
 	iterate_inode_ref_t cb;
+	int pending_move = 0;
 
 	path = alloc_path_for_send();
 	if (!path)
@@ -3240,7 +3642,9 @@ static int process_all_refs(struct send_ctx *sctx,
 	}
 	btrfs_release_path(path);
 
-	ret = process_recorded_refs(sctx);
+	ret = process_recorded_refs(sctx, &pending_move);
+	/* Only applicable to an incremental send. */
+	ASSERT(pending_move == 0);
 
 out:
 	btrfs_free_path(path);
@@ -3706,7 +4110,7 @@ verbose_printk("btrfs: send_clone offset=%llu, len=%d, clone_root=%llu, "
 	TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
 			clone_root->root->root_item.uuid);
 	TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID,
-			clone_root->root->root_item.ctransid);
+		    le64_to_cpu(clone_root->root->root_item.ctransid));
 	TLV_PUT_PATH(sctx, BTRFS_SEND_A_CLONE_PATH, p);
 	TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_OFFSET,
 			clone_root->offset);
@@ -3752,6 +4156,39 @@ out:
 	return ret;
 }
 
+static int send_hole(struct send_ctx *sctx, u64 end)
+{
+	struct fs_path *p = NULL;
+	u64 offset = sctx->cur_inode_last_extent;
+	u64 len;
+	int ret = 0;
+
+	p = fs_path_alloc();
+	if (!p)
+		return -ENOMEM;
+	memset(sctx->read_buf, 0, BTRFS_SEND_READ_SIZE);
+	while (offset < end) {
+		len = min_t(u64, end - offset, BTRFS_SEND_READ_SIZE);
+
+		ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE);
+		if (ret < 0)
+			break;
+		ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
+		if (ret < 0)
+			break;
+		TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
+		TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
+		TLV_PUT(sctx, BTRFS_SEND_A_DATA, sctx->read_buf, len);
+		ret = send_cmd(sctx);
+		if (ret < 0)
+			break;
+		offset += len;
+	}
+tlv_put_failure:
+	fs_path_free(p);
+	return ret;
+}
+
 static int send_write_or_clone(struct send_ctx *sctx,
 			       struct btrfs_path *path,
 			       struct btrfs_key *key,
@@ -3764,12 +4201,14 @@ static int send_write_or_clone(struct send_ctx *sctx,
 	u64 len;
 	u32 l;
 	u8 type;
+	u64 bs = sctx->send_root->fs_info->sb->s_blocksize;
 
 	ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
 			struct btrfs_file_extent_item);
 	type = btrfs_file_extent_type(path->nodes[0], ei);
 	if (type == BTRFS_FILE_EXTENT_INLINE) {
-		len = btrfs_file_extent_inline_len(path->nodes[0], ei);
+		len = btrfs_file_extent_inline_len(path->nodes[0],
+						   path->slots[0], ei);
 		/*
 		 * it is possible the inline item won't cover the whole page,
 		 * but there may be items after this page.  Make
@@ -3787,7 +4226,7 @@ static int send_write_or_clone(struct send_ctx *sctx,
 		goto out;
 	}
 
-	if (clone_root) {
+	if (clone_root && IS_ALIGNED(offset + len, bs)) {
 		ret = send_clone(sctx, offset, len, clone_root);
 	} else if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA) {
 		ret = send_update_extent(sctx, offset, len);
@@ -3979,6 +4418,101 @@ out:
 	return ret;
 }
 
+static int get_last_extent(struct send_ctx *sctx, u64 offset)
+{
+	struct btrfs_path *path;
+	struct btrfs_root *root = sctx->send_root;
+	struct btrfs_file_extent_item *fi;
+	struct btrfs_key key;
+	u64 extent_end;
+	u8 type;
+	int ret;
+
+	path = alloc_path_for_send();
+	if (!path)
+		return -ENOMEM;
+
+	sctx->cur_inode_last_extent = 0;
+
+	key.objectid = sctx->cur_ino;
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = offset;
+	ret = btrfs_search_slot_for_read(root, &key, path, 0, 1);
+	if (ret < 0)
+		goto out;
+	ret = 0;
+	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
+	if (key.objectid != sctx->cur_ino || key.type != BTRFS_EXTENT_DATA_KEY)
+		goto out;
+
+	fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
+			    struct btrfs_file_extent_item);
+	type = btrfs_file_extent_type(path->nodes[0], fi);
+	if (type == BTRFS_FILE_EXTENT_INLINE) {
+		u64 size = btrfs_file_extent_inline_len(path->nodes[0],
+							path->slots[0], fi);
+		extent_end = ALIGN(key.offset + size,
+				   sctx->send_root->sectorsize);
+	} else {
+		extent_end = key.offset +
+			btrfs_file_extent_num_bytes(path->nodes[0], fi);
+	}
+	sctx->cur_inode_last_extent = extent_end;
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+static int maybe_send_hole(struct send_ctx *sctx, struct btrfs_path *path,
+			   struct btrfs_key *key)
+{
+	struct btrfs_file_extent_item *fi;
+	u64 extent_end;
+	u8 type;
+	int ret = 0;
+
+	if (sctx->cur_ino != key->objectid || !need_send_hole(sctx))
+		return 0;
+
+	if (sctx->cur_inode_last_extent == (u64)-1) {
+		ret = get_last_extent(sctx, key->offset - 1);
+		if (ret)
+			return ret;
+	}
+
+	fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
+			    struct btrfs_file_extent_item);
+	type = btrfs_file_extent_type(path->nodes[0], fi);
+	if (type == BTRFS_FILE_EXTENT_INLINE) {
+		u64 size = btrfs_file_extent_inline_len(path->nodes[0],
+							path->slots[0], fi);
+		extent_end = ALIGN(key->offset + size,
+				   sctx->send_root->sectorsize);
+	} else {
+		extent_end = key->offset +
+			btrfs_file_extent_num_bytes(path->nodes[0], fi);
+	}
+
+	if (path->slots[0] == 0 &&
+	    sctx->cur_inode_last_extent < key->offset) {
+		/*
+		 * We might have skipped entire leafs that contained only
+		 * file extent items for our current inode. These leafs have
+		 * a generation number smaller (older) than the one in the
+		 * current leaf and the leaf our last extent came from, and
+		 * are located between these 2 leafs.
+		 */
+		ret = get_last_extent(sctx, key->offset - 1);
+		if (ret)
+			return ret;
+	}
+
+	if (sctx->cur_inode_last_extent < key->offset)
+		ret = send_hole(sctx, key->offset);
+	sctx->cur_inode_last_extent = extent_end;
+	return ret;
+}
+
 static int process_extent(struct send_ctx *sctx,
 			  struct btrfs_path *path,
 			  struct btrfs_key *key)
@@ -3995,7 +4529,7 @@ static int process_extent(struct send_ctx *sctx,
 			goto out;
 		if (ret) {
 			ret = 0;
-			goto out;
+			goto out_hole;
 		}
 	} else {
 		struct btrfs_file_extent_item *ei;
@@ -4031,7 +4565,10 @@ static int process_extent(struct send_ctx *sctx,
 		goto out;
 
 	ret = send_write_or_clone(sctx, path, key, found_clone);
-
+	if (ret)
+		goto out;
+out_hole:
+	ret = maybe_send_hole(sctx, path, key);
 out:
 	return ret;
 }
@@ -4054,17 +4591,25 @@ static int process_all_extents(struct send_ctx *sctx)
 	key.objectid = sctx->cmp_key->objectid;
 	key.type = BTRFS_EXTENT_DATA_KEY;
 	key.offset = 0;
-	while (1) {
-		ret = btrfs_search_slot_for_read(root, &key, path, 1, 0);
-		if (ret < 0)
-			goto out;
-		if (ret) {
-			ret = 0;
-			goto out;
-		}
+	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	if (ret < 0)
+		goto out;
 
+	while (1) {
 		eb = path->nodes[0];
 		slot = path->slots[0];
+
+		if (slot >= btrfs_header_nritems(eb)) {
+			ret = btrfs_next_leaf(root, path);
+			if (ret < 0) {
+				goto out;
+			} else if (ret > 0) {
+				ret = 0;
+				break;
+			}
+			continue;
+		}
+
 		btrfs_item_key_to_cpu(eb, &found_key, slot);
 
 		if (found_key.objectid != key.objectid ||
@@ -4077,8 +4622,7 @@ static int process_all_extents(struct send_ctx *sctx)
 		if (ret < 0)
 			goto out;
 
-		btrfs_release_path(path);
-		key.offset = found_key.offset + 1;
+		path->slots[0]++;
 	}
 
 out:
@@ -4086,7 +4630,9 @@ out:
 	return ret;
 }
 
-static int process_recorded_refs_if_needed(struct send_ctx *sctx, int at_end)
+static int process_recorded_refs_if_needed(struct send_ctx *sctx, int at_end,
+					   int *pending_move,
+					   int *refs_processed)
 {
 	int ret = 0;
 
@@ -4098,17 +4644,11 @@ static int process_recorded_refs_if_needed(struct send_ctx *sctx, int at_end)
 	if (list_empty(&sctx->new_refs) && list_empty(&sctx->deleted_refs))
 		goto out;
 
-	ret = process_recorded_refs(sctx);
+	ret = process_recorded_refs(sctx, pending_move);
 	if (ret < 0)
 		goto out;
 
-	/*
-	 * We have processed the refs and thus need to advance send_progress.
-	 * Now, calls to get_cur_xxx will take the updated refs of the current
-	 * inode into account.
-	 */
-	sctx->send_progress = sctx->cur_ino + 1;
-
+	*refs_processed = 1;
 out:
 	return ret;
 }
@@ -4124,11 +4664,29 @@ static int finish_inode_if_needed(struct send_ctx *sctx, int at_end)
 	u64 right_gid;
 	int need_chmod = 0;
 	int need_chown = 0;
+	int pending_move = 0;
+	int refs_processed = 0;
 
-	ret = process_recorded_refs_if_needed(sctx, at_end);
+	ret = process_recorded_refs_if_needed(sctx, at_end, &pending_move,
+					      &refs_processed);
 	if (ret < 0)
 		goto out;
 
+	/*
+	 * We have processed the refs and thus need to advance send_progress.
+	 * Now, calls to get_cur_xxx will take the updated refs of the current
+	 * inode into account.
+	 *
+	 * On the other hand, if our current inode is a directory and couldn't
+	 * be moved/renamed because its parent was renamed/moved too and it has
+	 * a higher inode number, we can only move/rename our current inode
+	 * after we moved/renamed its parent. Therefore in this case operate on
+	 * the old path (pre move/rename) of our current inode, and the
+	 * move/rename will be performed later.
+	 */
+	if (refs_processed && !pending_move)
+		sctx->send_progress = sctx->cur_ino + 1;
+
 	if (sctx->cur_ino == 0 || sctx->cur_inode_deleted)
 		goto out;
 	if (!at_end && sctx->cmp_key->objectid == sctx->cur_ino)
@@ -4157,6 +4715,19 @@ static int finish_inode_if_needed(struct send_ctx *sctx, int at_end)
 	}
 
 	if (S_ISREG(sctx->cur_inode_mode)) {
+		if (need_send_hole(sctx)) {
+			if (sctx->cur_inode_last_extent == (u64)-1) {
+				ret = get_last_extent(sctx, (u64)-1);
+				if (ret)
+					goto out;
+			}
+			if (sctx->cur_inode_last_extent <
+			    sctx->cur_inode_size) {
+				ret = send_hole(sctx, sctx->cur_inode_size);
+				if (ret)
+					goto out;
+			}
+		}
 		ret = send_truncate(sctx, sctx->cur_ino, sctx->cur_inode_gen,
 				sctx->cur_inode_size);
 		if (ret < 0)
@@ -4177,9 +4748,21 @@ static int finish_inode_if_needed(struct send_ctx *sctx, int at_end)
 	}
 
 	/*
-	 * Need to send that every time, no matter if it actually changed
-	 * between the two trees as we have done changes to the inode before.
+	 * If other directory inodes depended on our current directory
+	 * inode's move/rename, now do their move/rename operations.
+	 */
+	if (!is_waiting_for_move(sctx, sctx->cur_ino)) {
+		ret = apply_children_dir_moves(sctx);
+		if (ret)
+			goto out;
+	}
+
+	/*
+	 * Need to send that every time, no matter if it actually
+	 * changed between the two trees as we have done changes to
+	 * the inode before.
 	 */
+	sctx->send_progress = sctx->cur_ino + 1;
 	ret = send_utimes(sctx, sctx->cur_ino, sctx->cur_inode_gen);
 	if (ret < 0)
 		goto out;
@@ -4200,6 +4783,7 @@ static int changed_inode(struct send_ctx *sctx,
 
 	sctx->cur_ino = key->objectid;
 	sctx->cur_inode_new_gen = 0;
+	sctx->cur_inode_last_extent = (u64)-1;
 
 	/*
 	 * Set send_progress to current inode. This will tell all get_cur_xxx
@@ -4480,14 +5064,18 @@ static int changed_cb(struct btrfs_root *left_root,
 	struct send_ctx *sctx = ctx;
 
 	if (result == BTRFS_COMPARE_TREE_SAME) {
-		if (key->type != BTRFS_INODE_REF_KEY &&
-		    key->type != BTRFS_INODE_EXTREF_KEY)
-			return 0;
-		ret = compare_refs(sctx, left_path, key);
-		if (!ret)
+		if (key->type == BTRFS_INODE_REF_KEY ||
+		    key->type == BTRFS_INODE_EXTREF_KEY) {
+			ret = compare_refs(sctx, left_path, key);
+			if (!ret)
+				return 0;
+			if (ret < 0)
+				return ret;
+		} else if (key->type == BTRFS_EXTENT_DATA_KEY) {
+			return maybe_send_hole(sctx, left_path, key);
+		} else {
 			return 0;
-		if (ret < 0)
-			return ret;
+		}
 		result = BTRFS_COMPARE_TREE_CHANGED;
 		ret = 0;
 	}
@@ -4522,7 +5110,6 @@ out:
 static int full_send_tree(struct send_ctx *sctx)
 {
 	int ret;
-	struct btrfs_trans_handle *trans = NULL;
 	struct btrfs_root *send_root = sctx->send_root;
 	struct btrfs_key key;
 	struct btrfs_key found_key;
@@ -4544,19 +5131,6 @@ static int full_send_tree(struct send_ctx *sctx)
 	key.type = BTRFS_INODE_ITEM_KEY;
 	key.offset = 0;
 
-join_trans:
-	/*
-	 * We need to make sure the transaction does not get committed
-	 * while we do anything on commit roots. Join a transaction to prevent
-	 * this.
-	 */
-	trans = btrfs_join_transaction(send_root);
-	if (IS_ERR(trans)) {
-		ret = PTR_ERR(trans);
-		trans = NULL;
-		goto out;
-	}
-
 	/*
 	 * Make sure the tree has not changed after re-joining. We detect this
 	 * by comparing start_ctransid and ctransid. They should always match.
@@ -4566,7 +5140,7 @@ join_trans:
 	spin_unlock(&send_root->root_item_lock);
 
 	if (ctransid != start_ctransid) {
-		WARN(1, KERN_WARNING "btrfs: the root that you're trying to "
+		WARN(1, KERN_WARNING "BTRFS: the root that you're trying to "
 				     "send was modified in between. This is "
 				     "probably a bug.\n");
 		ret = -EIO;
@@ -4580,19 +5154,6 @@ join_trans:
 		goto out_finish;
 
 	while (1) {
-		/*
-		 * When someone want to commit while we iterate, end the
-		 * joined transaction and rejoin.
-		 */
-		if (btrfs_should_end_transaction(trans, send_root)) {
-			ret = btrfs_end_transaction(trans, send_root);
-			trans = NULL;
-			if (ret < 0)
-				goto out;
-			btrfs_release_path(path);
-			goto join_trans;
-		}
-
 		eb = path->nodes[0];
 		slot = path->slots[0];
 		btrfs_item_key_to_cpu(eb, &found_key, slot);
@@ -4620,12 +5181,6 @@ out_finish:
 
 out:
 	btrfs_free_path(path);
-	if (trans) {
-		if (!ret)
-			ret = btrfs_end_transaction(trans, send_root);
-		else
-			btrfs_end_transaction(trans, send_root);
-	}
 	return ret;
 }
 
@@ -4662,6 +5217,21 @@ out:
 	return ret;
 }
 
+static void btrfs_root_dec_send_in_progress(struct btrfs_root* root)
+{
+	spin_lock(&root->root_item_lock);
+	root->send_in_progress--;
+	/*
+	 * Not much left to do, we don't know why it's unbalanced and
+	 * can't blindly reset it to 0.
+	 */
+	if (root->send_in_progress < 0)
+		btrfs_err(root->fs_info,
+			"send_in_progres unbalanced %d root %llu\n",
+			root->send_in_progress, root->root_key.objectid);
+	spin_unlock(&root->root_item_lock);
+}
+
 long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 {
 	int ret = 0;
@@ -4673,6 +5243,9 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 	struct send_ctx *sctx = NULL;
 	u32 i;
 	u64 *clone_sources_tmp = NULL;
+	int clone_sources_to_rollback = 0;
+	int sort_clone_roots = 0;
+	int index;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
@@ -4680,6 +5253,14 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 	send_root = BTRFS_I(file_inode(mnt_file))->root;
 	fs_info = send_root->fs_info;
 
+	/*
+	 * The subvolume must remain read-only during send, protect against
+	 * making it RW.
+	 */
+	spin_lock(&send_root->root_item_lock);
+	send_root->send_in_progress++;
+	spin_unlock(&send_root->root_item_lock);
+
 	/*
 	 * This is done when we lookup the root, it should already be complete
 	 * by the time we get here.
@@ -4687,32 +5268,12 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 	WARN_ON(send_root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE);
 
 	/*
-	 * If we just created this root we need to make sure that the orphan
-	 * cleanup has been done and committed since we search the commit root,
-	 * so check its commit root transid with our otransid and if they match
-	 * commit the transaction to make sure everything is updated.
+	 * Userspace tools do the checks and warn the user if it's
+	 * not RO.
 	 */
-	down_read(&send_root->fs_info->extent_commit_sem);
-	if (btrfs_header_generation(send_root->commit_root) ==
-	    btrfs_root_otransid(&send_root->root_item)) {
-		struct btrfs_trans_handle *trans;
-
-		up_read(&send_root->fs_info->extent_commit_sem);
-
-		trans = btrfs_attach_transaction_barrier(send_root);
-		if (IS_ERR(trans)) {
-			if (PTR_ERR(trans) != -ENOENT) {
-				ret = PTR_ERR(trans);
-				goto out;
-			}
-			/* ENOENT means theres no transaction */
-		} else {
-			ret = btrfs_commit_transaction(trans, send_root);
-			if (ret)
-				goto out;
-		}
-	} else {
-		up_read(&send_root->fs_info->extent_commit_sem);
+	if (!btrfs_root_readonly(send_root)) {
+		ret = -EPERM;
+		goto out;
 	}
 
 	arg = memdup_user(arg_, sizeof(*arg));
@@ -4753,8 +5314,6 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 		goto out;
 	}
 
-	sctx->mnt = mnt_file->f_path.mnt;
-
 	sctx->send_root = send_root;
 	sctx->clone_roots_cnt = arg->clone_sources_count;
 
@@ -4771,6 +5330,9 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 		goto out;
 	}
 
+	sctx->pending_dir_moves = RB_ROOT;
+	sctx->waiting_dir_moves = RB_ROOT;
+
 	sctx->clone_roots = vzalloc(sizeof(struct clone_root) *
 			(arg->clone_sources_count + 1));
 	if (!sctx->clone_roots) {
@@ -4798,11 +5360,27 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 			key.objectid = clone_sources_tmp[i];
 			key.type = BTRFS_ROOT_ITEM_KEY;
 			key.offset = (u64)-1;
+
+			index = srcu_read_lock(&fs_info->subvol_srcu);
+
 			clone_root = btrfs_read_fs_root_no_name(fs_info, &key);
 			if (IS_ERR(clone_root)) {
+				srcu_read_unlock(&fs_info->subvol_srcu, index);
 				ret = PTR_ERR(clone_root);
 				goto out;
 			}
+			clone_sources_to_rollback = i + 1;
+			spin_lock(&clone_root->root_item_lock);
+			clone_root->send_in_progress++;
+			if (!btrfs_root_readonly(clone_root)) {
+				spin_unlock(&clone_root->root_item_lock);
+				srcu_read_unlock(&fs_info->subvol_srcu, index);
+				ret = -EPERM;
+				goto out;
+			}
+			spin_unlock(&clone_root->root_item_lock);
+			srcu_read_unlock(&fs_info->subvol_srcu, index);
+
 			sctx->clone_roots[i].root = clone_root;
 		}
 		vfree(clone_sources_tmp);
@@ -4813,11 +5391,27 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 		key.objectid = arg->parent_root;
 		key.type = BTRFS_ROOT_ITEM_KEY;
 		key.offset = (u64)-1;
+
+		index = srcu_read_lock(&fs_info->subvol_srcu);
+
 		sctx->parent_root = btrfs_read_fs_root_no_name(fs_info, &key);
 		if (IS_ERR(sctx->parent_root)) {
+			srcu_read_unlock(&fs_info->subvol_srcu, index);
 			ret = PTR_ERR(sctx->parent_root);
 			goto out;
 		}
+
+		spin_lock(&sctx->parent_root->root_item_lock);
+		sctx->parent_root->send_in_progress++;
+		if (!btrfs_root_readonly(sctx->parent_root)) {
+			spin_unlock(&sctx->parent_root->root_item_lock);
+			srcu_read_unlock(&fs_info->subvol_srcu, index);
+			ret = -EPERM;
+			goto out;
+		}
+		spin_unlock(&sctx->parent_root->root_item_lock);
+
+		srcu_read_unlock(&fs_info->subvol_srcu, index);
 	}
 
 	/*
@@ -4831,6 +5425,7 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 	sort(sctx->clone_roots, sctx->clone_roots_cnt,
 			sizeof(*sctx->clone_roots), __clone_root_cmp_sort,
 			NULL);
+	sort_clone_roots = 1;
 
 	ret = send_subvol(sctx);
 	if (ret < 0)
@@ -4846,6 +5441,48 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 	}
 
 out:
+	WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->pending_dir_moves));
+	while (sctx && !RB_EMPTY_ROOT(&sctx->pending_dir_moves)) {
+		struct rb_node *n;
+		struct pending_dir_move *pm;
+
+		n = rb_first(&sctx->pending_dir_moves);
+		pm = rb_entry(n, struct pending_dir_move, node);
+		while (!list_empty(&pm->list)) {
+			struct pending_dir_move *pm2;
+
+			pm2 = list_first_entry(&pm->list,
+					       struct pending_dir_move, list);
+			free_pending_move(sctx, pm2);
+		}
+		free_pending_move(sctx, pm);
+	}
+
+	WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves));
+	while (sctx && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves)) {
+		struct rb_node *n;
+		struct waiting_dir_move *dm;
+
+		n = rb_first(&sctx->waiting_dir_moves);
+		dm = rb_entry(n, struct waiting_dir_move, node);
+		rb_erase(&dm->node, &sctx->waiting_dir_moves);
+		kfree(dm);
+	}
+
+	if (sort_clone_roots) {
+		for (i = 0; i < sctx->clone_roots_cnt; i++)
+			btrfs_root_dec_send_in_progress(
+					sctx->clone_roots[i].root);
+	} else {
+		for (i = 0; sctx && i < clone_sources_to_rollback; i++)
+			btrfs_root_dec_send_in_progress(
+					sctx->clone_roots[i].root);
+
+		btrfs_root_dec_send_in_progress(send_root);
+	}
+	if (sctx && !IS_ERR_OR_NULL(sctx->parent_root))
+		btrfs_root_dec_send_in_progress(sctx->parent_root);
+
 	kfree(arg);
 	vfree(clone_sources_tmp);
 

fs/btrfs/super.c

@@ -48,6 +48,8 @@
 #include "transaction.h"
 #include "btrfs_inode.h"
 #include "print-tree.h"
+#include "hash.h"
+#include "props.h"
 #include "xattr.h"
 #include "volumes.h"
 #include "export.h"
@@ -152,11 +154,12 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		vaf.fmt = fmt;
 		vaf.va = &args;
 
-		printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: errno=%d %s (%pV)\n",
+		printk(KERN_CRIT
+			"BTRFS: error (device %s) in %s:%d: errno=%d %s (%pV)\n",
 			sb->s_id, function, line, errno, errstr, &vaf);
 		va_end(args);
 	} else {
-		printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: errno=%d %s\n",
+		printk(KERN_CRIT "BTRFS: error (device %s) in %s:%d: errno=%d %s\n",
 			sb->s_id, function, line, errno, errstr);
 	}
 
@@ -250,7 +253,7 @@ void __btrfs_abort_transaction(struct btrfs_trans_handle *trans,
 	 */
 	if (!test_and_set_bit(BTRFS_FS_STATE_TRANS_ABORTED,
 				&root->fs_info->fs_state)) {
-		WARN(1, KERN_DEBUG "btrfs: Transaction aborted (error %d)\n",
+		WARN(1, KERN_DEBUG "BTRFS: Transaction aborted (error %d)\n",
 				errno);
 	}
 	trans->aborted = errno;
@@ -294,8 +297,8 @@ void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function,
 		panic(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (errno=%d %s)\n",
 			s_id, function, line, &vaf, errno, errstr);
 
-	printk(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (errno=%d %s)\n",
-	       s_id, function, line, &vaf, errno, errstr);
+	btrfs_crit(fs_info, "panic in %s:%d: %pV (errno=%d %s)",
+		   function, line, &vaf, errno, errstr);
 	va_end(args);
 	/* Caller calls BUG() */
 }
@@ -322,7 +325,9 @@ enum {
 	Opt_no_space_cache, Opt_recovery, Opt_skip_balance,
 	Opt_check_integrity, Opt_check_integrity_including_extent_data,
 	Opt_check_integrity_print_mask, Opt_fatal_errors, Opt_rescan_uuid_tree,
-	Opt_commit_interval,
+	Opt_commit_interval, Opt_barrier, Opt_nodefrag, Opt_nodiscard,
+	Opt_noenospc_debug, Opt_noflushoncommit, Opt_acl, Opt_datacow,
+	Opt_datasum, Opt_treelog, Opt_noinode_cache,
 	Opt_err,
 };
 
@@ -332,8 +337,11 @@ static match_table_t tokens = {
 	{Opt_subvolid, "subvolid=%s"},
 	{Opt_device, "device=%s"},
 	{Opt_nodatasum, "nodatasum"},
+	{Opt_datasum, "datasum"},
 	{Opt_nodatacow, "nodatacow"},
+	{Opt_datacow, "datacow"},
 	{Opt_nobarrier, "nobarrier"},
+	{Opt_barrier, "barrier"},
 	{Opt_max_inline, "max_inline=%s"},
 	{Opt_alloc_start, "alloc_start=%s"},
 	{Opt_thread_pool, "thread_pool=%d"},
@@ -344,18 +352,25 @@ static match_table_t tokens = {
 	{Opt_ssd, "ssd"},
 	{Opt_ssd_spread, "ssd_spread"},
 	{Opt_nossd, "nossd"},
+	{Opt_acl, "acl"},
 	{Opt_noacl, "noacl"},
 	{Opt_notreelog, "notreelog"},
+	{Opt_treelog, "treelog"},
 	{Opt_flushoncommit, "flushoncommit"},
+	{Opt_noflushoncommit, "noflushoncommit"},
 	{Opt_ratio, "metadata_ratio=%d"},
 	{Opt_discard, "discard"},
+	{Opt_nodiscard, "nodiscard"},
 	{Opt_space_cache, "space_cache"},
 	{Opt_clear_cache, "clear_cache"},
 	{Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"},
 	{Opt_enospc_debug, "enospc_debug"},
+	{Opt_noenospc_debug, "noenospc_debug"},
 	{Opt_subvolrootid, "subvolrootid=%d"},
 	{Opt_defrag, "autodefrag"},
+	{Opt_nodefrag, "noautodefrag"},
 	{Opt_inode_cache, "inode_cache"},
+	{Opt_noinode_cache, "noinode_cache"},
 	{Opt_no_space_cache, "nospace_cache"},
 	{Opt_recovery, "recovery"},
 	{Opt_skip_balance, "skip_balance"},
@@ -368,6 +383,20 @@ static match_table_t tokens = {
 	{Opt_err, NULL},
 };
 
+#define btrfs_set_and_info(root, opt, fmt, args...)			\
+{									\
+	if (!btrfs_test_opt(root, opt))					\
+		btrfs_info(root->fs_info, fmt, ##args);			\
+	btrfs_set_opt(root->fs_info->mount_opt, opt);			\
+}
+
+#define btrfs_clear_and_info(root, opt, fmt, args...)			\
+{									\
+	if (btrfs_test_opt(root, opt))					\
+		btrfs_info(root->fs_info, fmt, ##args);			\
+	btrfs_clear_opt(root->fs_info->mount_opt, opt);			\
+}
+
 /*
  * Regular mount options parser.  Everything that is needed only when
  * reading in a new superblock is parsed here.
@@ -383,6 +412,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 	int ret = 0;
 	char *compress_type;
 	bool compress_force = false;
+	bool compress = false;
 
 	cache_gen = btrfs_super_cache_generation(root->fs_info->super_copy);
 	if (cache_gen)
@@ -409,7 +439,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 		token = match_token(p, tokens, args);
 		switch (token) {
 		case Opt_degraded:
-			printk(KERN_INFO "btrfs: allowing degraded mounts\n");
+			btrfs_info(root->fs_info, "allowing degraded mounts");
 			btrfs_set_opt(info->mount_opt, DEGRADED);
 			break;
 		case Opt_subvol:
@@ -422,27 +452,45 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			 */
 			break;
 		case Opt_nodatasum:
-			printk(KERN_INFO "btrfs: setting nodatasum\n");
-			btrfs_set_opt(info->mount_opt, NODATASUM);
+			btrfs_set_and_info(root, NODATASUM,
+					   "setting nodatasum");
+			break;
+		case Opt_datasum:
+			if (btrfs_test_opt(root, NODATASUM)) {
+				if (btrfs_test_opt(root, NODATACOW))
+					btrfs_info(root->fs_info, "setting datasum, datacow enabled");
+				else
+					btrfs_info(root->fs_info, "setting datasum");
+			}
+			btrfs_clear_opt(info->mount_opt, NODATACOW);
+			btrfs_clear_opt(info->mount_opt, NODATASUM);
 			break;
 		case Opt_nodatacow:
-			if (!btrfs_test_opt(root, COMPRESS) ||
-				!btrfs_test_opt(root, FORCE_COMPRESS)) {
-					printk(KERN_INFO "btrfs: setting nodatacow, compression disabled\n");
-			} else {
-				printk(KERN_INFO "btrfs: setting nodatacow\n");
+			if (!btrfs_test_opt(root, NODATACOW)) {
+				if (!btrfs_test_opt(root, COMPRESS) ||
+				    !btrfs_test_opt(root, FORCE_COMPRESS)) {
+					btrfs_info(root->fs_info,
+						   "setting nodatacow, compression disabled");
+				} else {
+					btrfs_info(root->fs_info, "setting nodatacow");
+				}
 			}
 			btrfs_clear_opt(info->mount_opt, COMPRESS);
 			btrfs_clear_opt(info->mount_opt, FORCE_COMPRESS);
 			btrfs_set_opt(info->mount_opt, NODATACOW);
 			btrfs_set_opt(info->mount_opt, NODATASUM);
 			break;
+		case Opt_datacow:
+			btrfs_clear_and_info(root, NODATACOW,
+					     "setting datacow");
+			break;
 		case Opt_compress_force:
 		case Opt_compress_force_type:
 			compress_force = true;
 			/* Fallthrough */
 		case Opt_compress:
 		case Opt_compress_type:
+			compress = true;
 			if (token == Opt_compress ||
 			    token == Opt_compress_force ||
 			    strcmp(args[0].from, "zlib") == 0) {
@@ -469,34 +517,36 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			}
 
 			if (compress_force) {
-				btrfs_set_opt(info->mount_opt, FORCE_COMPRESS);
-				pr_info("btrfs: force %s compression\n",
-					compress_type);
-			} else if (btrfs_test_opt(root, COMPRESS)) {
-				pr_info("btrfs: use %s compression\n",
-					compress_type);
+				btrfs_set_and_info(root, FORCE_COMPRESS,
+						   "force %s compression",
+						   compress_type);
+			} else if (compress) {
+				if (!btrfs_test_opt(root, COMPRESS))
+					btrfs_info(root->fs_info,
+						   "btrfs: use %s compression\n",
+						   compress_type);
 			}
 			break;
 		case Opt_ssd:
-			printk(KERN_INFO "btrfs: use ssd allocation scheme\n");
-			btrfs_set_opt(info->mount_opt, SSD);
+			btrfs_set_and_info(root, SSD,
+					   "use ssd allocation scheme");
 			break;
 		case Opt_ssd_spread:
-			printk(KERN_INFO "btrfs: use spread ssd "
-			       "allocation scheme\n");
-			btrfs_set_opt(info->mount_opt, SSD);
-			btrfs_set_opt(info->mount_opt, SSD_SPREAD);
+			btrfs_set_and_info(root, SSD_SPREAD,
+					   "use spread ssd allocation scheme");
 			break;
 		case Opt_nossd:
-			printk(KERN_INFO "btrfs: not using ssd allocation "
-			       "scheme\n");
-			btrfs_set_opt(info->mount_opt, NOSSD);
+			btrfs_clear_and_info(root, NOSSD,
+					     "not using ssd allocation scheme");
 			btrfs_clear_opt(info->mount_opt, SSD);
-			btrfs_clear_opt(info->mount_opt, SSD_SPREAD);
+			break;
+		case Opt_barrier:
+			btrfs_clear_and_info(root, NOBARRIER,
+					     "turning on barriers");
 			break;
 		case Opt_nobarrier:
-			printk(KERN_INFO "btrfs: turning off barriers\n");
-			btrfs_set_opt(info->mount_opt, NOBARRIER);
+			btrfs_set_and_info(root, NOBARRIER,
+					   "turning off barriers");
 			break;
 		case Opt_thread_pool:
 			ret = match_int(&args[0], &intarg);
@@ -520,7 +570,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 						info->max_inline,
 						root->sectorsize);
 				}
-				printk(KERN_INFO "btrfs: max_inline at %llu\n",
+				btrfs_info(root->fs_info, "max_inline at %llu",
 					info->max_inline);
 			} else {
 				ret = -ENOMEM;
@@ -534,24 +584,34 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				info->alloc_start = memparse(num, NULL);
 				mutex_unlock(&info->chunk_mutex);
 				kfree(num);
-				printk(KERN_INFO
-					"btrfs: allocations start at %llu\n",
+				btrfs_info(root->fs_info, "allocations start at %llu",
 					info->alloc_start);
 			} else {
 				ret = -ENOMEM;
 				goto out;
 			}
 			break;
+		case Opt_acl:
+			root->fs_info->sb->s_flags |= MS_POSIXACL;
+			break;
 		case Opt_noacl:
 			root->fs_info->sb->s_flags &= ~MS_POSIXACL;
 			break;
 		case Opt_notreelog:
-			printk(KERN_INFO "btrfs: disabling tree log\n");
-			btrfs_set_opt(info->mount_opt, NOTREELOG);
+			btrfs_set_and_info(root, NOTREELOG,
+					   "disabling tree log");
+			break;
+		case Opt_treelog:
+			btrfs_clear_and_info(root, NOTREELOG,
+					     "enabling tree log");
 			break;
 		case Opt_flushoncommit:
-			printk(KERN_INFO "btrfs: turning on flush-on-commit\n");
-			btrfs_set_opt(info->mount_opt, FLUSHONCOMMIT);
+			btrfs_set_and_info(root, FLUSHONCOMMIT,
+					   "turning on flush-on-commit");
+			break;
+		case Opt_noflushoncommit:
+			btrfs_clear_and_info(root, FLUSHONCOMMIT,
+					     "turning off flush-on-commit");
 			break;
 		case Opt_ratio:
 			ret = match_int(&args[0], &intarg);
@@ -559,7 +619,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				goto out;
 			} else if (intarg >= 0) {
 				info->metadata_ratio = intarg;
-				printk(KERN_INFO "btrfs: metadata ratio %d\n",
+				btrfs_info(root->fs_info, "metadata ratio %d",
 				       info->metadata_ratio);
 			} else {
 				ret = -EINVAL;
@@ -567,25 +627,35 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			}
 			break;
 		case Opt_discard:
-			btrfs_set_opt(info->mount_opt, DISCARD);
+			btrfs_set_and_info(root, DISCARD,
+					   "turning on discard");
+			break;
+		case Opt_nodiscard:
+			btrfs_clear_and_info(root, DISCARD,
+					     "turning off discard");
 			break;
 		case Opt_space_cache:
-			btrfs_set_opt(info->mount_opt, SPACE_CACHE);
+			btrfs_set_and_info(root, SPACE_CACHE,
+					   "enabling disk space caching");
 			break;
 		case Opt_rescan_uuid_tree:
 			btrfs_set_opt(info->mount_opt, RESCAN_UUID_TREE);
 			break;
 		case Opt_no_space_cache:
-			printk(KERN_INFO "btrfs: disabling disk space caching\n");
-			btrfs_clear_opt(info->mount_opt, SPACE_CACHE);
+			btrfs_clear_and_info(root, SPACE_CACHE,
+					     "disabling disk space caching");
 			break;
 		case Opt_inode_cache:
-			printk(KERN_INFO "btrfs: enabling inode map caching\n");
-			btrfs_set_opt(info->mount_opt, INODE_MAP_CACHE);
+			btrfs_set_and_info(root, CHANGE_INODE_CACHE,
+					   "enabling inode map caching");
+			break;
+		case Opt_noinode_cache:
+			btrfs_clear_and_info(root, CHANGE_INODE_CACHE,
+					     "disabling inode map caching");
 			break;
 		case Opt_clear_cache:
-			printk(KERN_INFO "btrfs: force clearing of disk cache\n");
-			btrfs_set_opt(info->mount_opt, CLEAR_CACHE);
+			btrfs_set_and_info(root, CLEAR_CACHE,
+					   "force clearing of disk cache");
 			break;
 		case Opt_user_subvol_rm_allowed:
 			btrfs_set_opt(info->mount_opt, USER_SUBVOL_RM_ALLOWED);
@@ -593,12 +663,19 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 		case Opt_enospc_debug:
 			btrfs_set_opt(info->mount_opt, ENOSPC_DEBUG);
 			break;
+		case Opt_noenospc_debug:
+			btrfs_clear_opt(info->mount_opt, ENOSPC_DEBUG);
+			break;
 		case Opt_defrag:
-			printk(KERN_INFO "btrfs: enabling auto defrag\n");
-			btrfs_set_opt(info->mount_opt, AUTO_DEFRAG);
+			btrfs_set_and_info(root, AUTO_DEFRAG,
+					   "enabling auto defrag");
+			break;
+		case Opt_nodefrag:
+			btrfs_clear_and_info(root, AUTO_DEFRAG,
+					     "disabling auto defrag");
 			break;
 		case Opt_recovery:
-			printk(KERN_INFO "btrfs: enabling auto recovery\n");
+			btrfs_info(root->fs_info, "enabling auto recovery");
 			btrfs_set_opt(info->mount_opt, RECOVERY);
 			break;
 		case Opt_skip_balance:
@@ -606,14 +683,14 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			break;
 #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
 		case Opt_check_integrity_including_extent_data:
-			printk(KERN_INFO "btrfs: enabling check integrity"
-			       " including extent data\n");
+			btrfs_info(root->fs_info,
+				   "enabling check integrity including extent data");
 			btrfs_set_opt(info->mount_opt,
 				      CHECK_INTEGRITY_INCLUDING_EXTENT_DATA);
 			btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY);
 			break;
 		case Opt_check_integrity:
-			printk(KERN_INFO "btrfs: enabling check integrity\n");
+			btrfs_info(root->fs_info, "enabling check integrity");
 			btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY);
 			break;
 		case Opt_check_integrity_print_mask:
@@ -622,8 +699,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				goto out;
 			} else if (intarg >= 0) {
 				info->check_integrity_print_mask = intarg;
-				printk(KERN_INFO "btrfs:"
-				       " check_integrity_print_mask 0x%x\n",
+				btrfs_info(root->fs_info, "check_integrity_print_mask 0x%x",
 				       info->check_integrity_print_mask);
 			} else {
 				ret = -EINVAL;
@@ -634,8 +710,8 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 		case Opt_check_integrity_including_extent_data:
 		case Opt_check_integrity:
 		case Opt_check_integrity_print_mask:
-			printk(KERN_ERR "btrfs: support for check_integrity*"
-			       " not compiled in!\n");
+			btrfs_err(root->fs_info,
+				"support for check_integrity* not compiled in!");
 			ret = -EINVAL;
 			goto out;
 #endif
@@ -655,28 +731,24 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			intarg = 0;
 			ret = match_int(&args[0], &intarg);
 			if (ret < 0) {
-				printk(KERN_ERR
-					"btrfs: invalid commit interval\n");
+				btrfs_err(root->fs_info, "invalid commit interval");
 				ret = -EINVAL;
 				goto out;
 			}
 			if (intarg > 0) {
 				if (intarg > 300) {
-					printk(KERN_WARNING
-					    "btrfs: excessive commit interval %d\n",
+					btrfs_warn(root->fs_info, "excessive commit interval %d",
 							intarg);
 				}
 				info->commit_interval = intarg;
 			} else {
-				printk(KERN_INFO
-				    "btrfs: using default commit interval %ds\n",
+				btrfs_info(root->fs_info, "using default commit interval %ds",
 				    BTRFS_DEFAULT_COMMIT_INTERVAL);
 				info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
 			}
 			break;
 		case Opt_err:
-			printk(KERN_INFO "btrfs: unrecognized mount option "
-			       "'%s'\n", p);
+			btrfs_info(root->fs_info, "unrecognized mount option '%s'", p);
 			ret = -EINVAL;
 			goto out;
 		default:
@@ -685,7 +757,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 	}
 out:
 	if (!ret && btrfs_test_opt(root, SPACE_CACHE))
-		printk(KERN_INFO "btrfs: disk space caching is enabled\n");
+		btrfs_info(root->fs_info, "disk space caching is enabled");
 	kfree(orig);
 	return ret;
 }
@@ -748,7 +820,8 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags,
 			break;
 		case Opt_subvolrootid:
 			printk(KERN_WARNING
-				"btrfs: 'subvolrootid' mount option is deprecated and has no effect\n");
+				"BTRFS: 'subvolrootid' mount option is deprecated and has "
+				"no effect\n");
 			break;
 		case Opt_device:
 			device_name = match_strdup(&args[0]);
@@ -877,7 +950,7 @@ static int btrfs_fill_super(struct super_block *sb,
 	sb->s_flags |= MS_I_VERSION;
 	err = open_ctree(sb, fs_devices, (char *)data);
 	if (err) {
-		printk("btrfs: open_ctree failed\n");
+		printk(KERN_ERR "BTRFS: open_ctree failed\n");
 		return err;
 	}
 
@@ -1115,7 +1188,7 @@ static struct dentry *mount_subvol(const char *subvol_name, int flags,
 		dput(root);
 		root = ERR_PTR(-EINVAL);
 		deactivate_locked_super(s);
-		printk(KERN_ERR "btrfs: '%s' is not a valid subvolume\n",
+		printk(KERN_ERR "BTRFS: '%s' is not a valid subvolume\n",
 				subvol_name);
 	}
 
@@ -1240,7 +1313,7 @@ static void btrfs_resize_thread_pool(struct btrfs_fs_info *fs_info,
 
 	fs_info->thread_pool_size = new_pool_size;
 
-	printk(KERN_INFO "btrfs: resize thread pool %d -> %d\n",
+	btrfs_info(fs_info, "resize thread pool %d -> %d",
 	       old_pool_size, new_pool_size);
 
 	btrfs_set_max_workers(&fs_info->generic_worker, new_pool_size);
@@ -1346,7 +1419,7 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 	} else {
 		if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) {
 			btrfs_err(fs_info,
-				"Remounting read-write after error is not allowed\n");
+				"Remounting read-write after error is not allowed");
 			ret = -EINVAL;
 			goto restore;
 		}
@@ -1358,8 +1431,8 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 		if (fs_info->fs_devices->missing_devices >
 		     fs_info->num_tolerated_disk_barrier_failures &&
 		    !(*flags & MS_RDONLY)) {
-			printk(KERN_WARNING
-			       "Btrfs: too many missing devices, writeable remount is not allowed\n");
+			btrfs_warn(fs_info,
+				"too many missing devices, writeable remount is not allowed");
 			ret = -EACCES;
 			goto restore;
 		}
@@ -1384,16 +1457,15 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 
 		ret = btrfs_resume_dev_replace_async(fs_info);
 		if (ret) {
-			pr_warn("btrfs: failed to resume dev_replace\n");
+			btrfs_warn(fs_info, "failed to resume dev_replace");
 			goto restore;
 		}
 
 		if (!fs_info->uuid_root) {
-			pr_info("btrfs: creating UUID tree\n");
+			btrfs_info(fs_info, "creating UUID tree");
 			ret = btrfs_create_uuid_tree(fs_info);
 			if (ret) {
-				pr_warn("btrfs: failed to create the uuid tree"
-					"%d\n", ret);
+				btrfs_warn(fs_info, "failed to create the UUID tree %d", ret);
 				goto restore;
 			}
 		}
@@ -1773,7 +1845,7 @@ static int btrfs_interface_init(void)
 static void btrfs_interface_exit(void)
 {
 	if (misc_deregister(&btrfs_misc) < 0)
-		printk(KERN_INFO "btrfs: misc_deregister failed for control device\n");
+		printk(KERN_INFO "BTRFS: misc_deregister failed for control device\n");
 }
 
 static void btrfs_print_info(void)
@@ -1818,10 +1890,16 @@ static int __init init_btrfs_fs(void)
 {
 	int err;
 
-	err = btrfs_init_sysfs();
+	err = btrfs_hash_init();
 	if (err)
 		return err;
 
+	btrfs_props_init();
+
+	err = btrfs_init_sysfs();
+	if (err)
+		goto free_hash;
+
 	btrfs_init_compress();
 
 	err = btrfs_init_cachep();
@@ -1895,6 +1973,8 @@ free_cachep:
 free_compress:
 	btrfs_exit_compress();
 	btrfs_exit_sysfs();
+free_hash:
+	btrfs_hash_exit();
 	return err;
 }
 
@@ -1913,6 +1993,7 @@ static void __exit exit_btrfs_fs(void)
 	btrfs_exit_sysfs();
 	btrfs_cleanup_fs_uuids();
 	btrfs_exit_compress();
+	btrfs_hash_exit();
 }
 
 module_init(init_btrfs_fs)

fs/btrfs/sysfs.c

@@ -22,24 +22,641 @@
 #include <linux/completion.h>
 #include <linux/buffer_head.h>
 #include <linux/kobject.h>
+#include <linux/bug.h>
+#include <linux/genhd.h>
 
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
+#include "sysfs.h"
+#include "volumes.h"
+
+static inline struct btrfs_fs_info *to_fs_info(struct kobject *kobj);
+
+static u64 get_features(struct btrfs_fs_info *fs_info,
+			enum btrfs_feature_set set)
+{
+	struct btrfs_super_block *disk_super = fs_info->super_copy;
+	if (set == FEAT_COMPAT)
+		return btrfs_super_compat_flags(disk_super);
+	else if (set == FEAT_COMPAT_RO)
+		return btrfs_super_compat_ro_flags(disk_super);
+	else
+		return btrfs_super_incompat_flags(disk_super);
+}
+
+static void set_features(struct btrfs_fs_info *fs_info,
+			 enum btrfs_feature_set set, u64 features)
+{
+	struct btrfs_super_block *disk_super = fs_info->super_copy;
+	if (set == FEAT_COMPAT)
+		btrfs_set_super_compat_flags(disk_super, features);
+	else if (set == FEAT_COMPAT_RO)
+		btrfs_set_super_compat_ro_flags(disk_super, features);
+	else
+		btrfs_set_super_incompat_flags(disk_super, features);
+}
+
+static int can_modify_feature(struct btrfs_feature_attr *fa)
+{
+	int val = 0;
+	u64 set, clear;
+	switch (fa->feature_set) {
+	case FEAT_COMPAT:
+		set = BTRFS_FEATURE_COMPAT_SAFE_SET;
+		clear = BTRFS_FEATURE_COMPAT_SAFE_CLEAR;
+		break;
+	case FEAT_COMPAT_RO:
+		set = BTRFS_FEATURE_COMPAT_RO_SAFE_SET;
+		clear = BTRFS_FEATURE_COMPAT_RO_SAFE_CLEAR;
+		break;
+	case FEAT_INCOMPAT:
+		set = BTRFS_FEATURE_INCOMPAT_SAFE_SET;
+		clear = BTRFS_FEATURE_INCOMPAT_SAFE_CLEAR;
+		break;
+	default:
+		printk(KERN_WARNING "btrfs: sysfs: unknown feature set %d\n",
+				fa->feature_set);
+		return 0;
+	}
+
+	if (set & fa->feature_bit)
+		val |= 1;
+	if (clear & fa->feature_bit)
+		val |= 2;
+
+	return val;
+}
+
+static ssize_t btrfs_feature_attr_show(struct kobject *kobj,
+				       struct kobj_attribute *a, char *buf)
+{
+	int val = 0;
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	struct btrfs_feature_attr *fa = to_btrfs_feature_attr(a);
+	if (fs_info) {
+		u64 features = get_features(fs_info, fa->feature_set);
+		if (features & fa->feature_bit)
+			val = 1;
+	} else
+		val = can_modify_feature(fa);
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", val);
+}
+
+static ssize_t btrfs_feature_attr_store(struct kobject *kobj,
+					struct kobj_attribute *a,
+					const char *buf, size_t count)
+{
+	struct btrfs_fs_info *fs_info;
+	struct btrfs_feature_attr *fa = to_btrfs_feature_attr(a);
+	struct btrfs_trans_handle *trans;
+	u64 features, set, clear;
+	unsigned long val;
+	int ret;
+
+	fs_info = to_fs_info(kobj);
+	if (!fs_info)
+		return -EPERM;
+
+	ret = kstrtoul(skip_spaces(buf), 0, &val);
+	if (ret)
+		return ret;
+
+	if (fa->feature_set == FEAT_COMPAT) {
+		set = BTRFS_FEATURE_COMPAT_SAFE_SET;
+		clear = BTRFS_FEATURE_COMPAT_SAFE_CLEAR;
+	} else if (fa->feature_set == FEAT_COMPAT_RO) {
+		set = BTRFS_FEATURE_COMPAT_RO_SAFE_SET;
+		clear = BTRFS_FEATURE_COMPAT_RO_SAFE_CLEAR;
+	} else {
+		set = BTRFS_FEATURE_INCOMPAT_SAFE_SET;
+		clear = BTRFS_FEATURE_INCOMPAT_SAFE_CLEAR;
+	}
+
+	features = get_features(fs_info, fa->feature_set);
+
+	/* Nothing to do */
+	if ((val && (features & fa->feature_bit)) ||
+	    (!val && !(features & fa->feature_bit)))
+		return count;
+
+	if ((val && !(set & fa->feature_bit)) ||
+	    (!val && !(clear & fa->feature_bit))) {
+		btrfs_info(fs_info,
+			"%sabling feature %s on mounted fs is not supported.",
+			val ? "En" : "Dis", fa->kobj_attr.attr.name);
+		return -EPERM;
+	}
+
+	btrfs_info(fs_info, "%s %s feature flag",
+		   val ? "Setting" : "Clearing", fa->kobj_attr.attr.name);
+
+	trans = btrfs_start_transaction(fs_info->fs_root, 0);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+
+	spin_lock(&fs_info->super_lock);
+	features = get_features(fs_info, fa->feature_set);
+	if (val)
+		features |= fa->feature_bit;
+	else
+		features &= ~fa->feature_bit;
+	set_features(fs_info, fa->feature_set, features);
+	spin_unlock(&fs_info->super_lock);
+
+	ret = btrfs_commit_transaction(trans, fs_info->fs_root);
+	if (ret)
+		return ret;
+
+	return count;
+}
+
+static umode_t btrfs_feature_visible(struct kobject *kobj,
+				     struct attribute *attr, int unused)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	umode_t mode = attr->mode;
+
+	if (fs_info) {
+		struct btrfs_feature_attr *fa;
+		u64 features;
+
+		fa = attr_to_btrfs_feature_attr(attr);
+		features = get_features(fs_info, fa->feature_set);
+
+		if (can_modify_feature(fa))
+			mode |= S_IWUSR;
+		else if (!(features & fa->feature_bit))
+			mode = 0;
+	}
+
+	return mode;
+}
+
+BTRFS_FEAT_ATTR_INCOMPAT(mixed_backref, MIXED_BACKREF);
+BTRFS_FEAT_ATTR_INCOMPAT(default_subvol, DEFAULT_SUBVOL);
+BTRFS_FEAT_ATTR_INCOMPAT(mixed_groups, MIXED_GROUPS);
+BTRFS_FEAT_ATTR_INCOMPAT(compress_lzo, COMPRESS_LZO);
+BTRFS_FEAT_ATTR_INCOMPAT(big_metadata, BIG_METADATA);
+BTRFS_FEAT_ATTR_INCOMPAT(extended_iref, EXTENDED_IREF);
+BTRFS_FEAT_ATTR_INCOMPAT(raid56, RAID56);
+BTRFS_FEAT_ATTR_INCOMPAT(skinny_metadata, SKINNY_METADATA);
+BTRFS_FEAT_ATTR_INCOMPAT(no_holes, NO_HOLES);
+
+static struct attribute *btrfs_supported_feature_attrs[] = {
+	BTRFS_FEAT_ATTR_PTR(mixed_backref),
+	BTRFS_FEAT_ATTR_PTR(default_subvol),
+	BTRFS_FEAT_ATTR_PTR(mixed_groups),
+	BTRFS_FEAT_ATTR_PTR(compress_lzo),
+	BTRFS_FEAT_ATTR_PTR(big_metadata),
+	BTRFS_FEAT_ATTR_PTR(extended_iref),
+	BTRFS_FEAT_ATTR_PTR(raid56),
+	BTRFS_FEAT_ATTR_PTR(skinny_metadata),
+	BTRFS_FEAT_ATTR_PTR(no_holes),
+	NULL
+};
+
+static const struct attribute_group btrfs_feature_attr_group = {
+	.name = "features",
+	.is_visible = btrfs_feature_visible,
+	.attrs = btrfs_supported_feature_attrs,
+};
+
+static ssize_t btrfs_show_u64(u64 *value_ptr, spinlock_t *lock, char *buf)
+{
+	u64 val;
+	if (lock)
+		spin_lock(lock);
+	val = *value_ptr;
+	if (lock)
+		spin_unlock(lock);
+	return snprintf(buf, PAGE_SIZE, "%llu\n", val);
+}
+
+static ssize_t global_rsv_size_show(struct kobject *kobj,
+				    struct kobj_attribute *ka, char *buf)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj->parent);
+	struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv;
+	return btrfs_show_u64(&block_rsv->size, &block_rsv->lock, buf);
+}
+BTRFS_ATTR(global_rsv_size, 0444, global_rsv_size_show);
+
+static ssize_t global_rsv_reserved_show(struct kobject *kobj,
+					struct kobj_attribute *a, char *buf)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj->parent);
+	struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv;
+	return btrfs_show_u64(&block_rsv->reserved, &block_rsv->lock, buf);
+}
+BTRFS_ATTR(global_rsv_reserved, 0444, global_rsv_reserved_show);
+
+#define to_space_info(_kobj) container_of(_kobj, struct btrfs_space_info, kobj)
+
+static ssize_t raid_bytes_show(struct kobject *kobj,
+			       struct kobj_attribute *attr, char *buf);
+BTRFS_RAID_ATTR(total_bytes, raid_bytes_show);
+BTRFS_RAID_ATTR(used_bytes, raid_bytes_show);
+
+static ssize_t raid_bytes_show(struct kobject *kobj,
+			       struct kobj_attribute *attr, char *buf)
+
+{
+	struct btrfs_space_info *sinfo = to_space_info(kobj->parent);
+	struct btrfs_block_group_cache *block_group;
+	int index = kobj - sinfo->block_group_kobjs;
+	u64 val = 0;
+
+	down_read(&sinfo->groups_sem);
+	list_for_each_entry(block_group, &sinfo->block_groups[index], list) {
+		if (&attr->attr == BTRFS_RAID_ATTR_PTR(total_bytes))
+			val += block_group->key.offset;
+		else
+			val += btrfs_block_group_used(&block_group->item);
+	}
+	up_read(&sinfo->groups_sem);
+	return snprintf(buf, PAGE_SIZE, "%llu\n", val);
+}
+
+static struct attribute *raid_attributes[] = {
+	BTRFS_RAID_ATTR_PTR(total_bytes),
+	BTRFS_RAID_ATTR_PTR(used_bytes),
+	NULL
+};
+
+static void release_raid_kobj(struct kobject *kobj)
+{
+	kobject_put(kobj->parent);
+}
+
+struct kobj_type btrfs_raid_ktype = {
+	.sysfs_ops = &kobj_sysfs_ops,
+	.release = release_raid_kobj,
+	.default_attrs = raid_attributes,
+};
+
+#define SPACE_INFO_ATTR(field)						\
+static ssize_t btrfs_space_info_show_##field(struct kobject *kobj,	\
+					     struct kobj_attribute *a,	\
+					     char *buf)			\
+{									\
+	struct btrfs_space_info *sinfo = to_space_info(kobj);		\
+	return btrfs_show_u64(&sinfo->field, &sinfo->lock, buf);	\
+}									\
+BTRFS_ATTR(field, 0444, btrfs_space_info_show_##field)
+
+static ssize_t btrfs_space_info_show_total_bytes_pinned(struct kobject *kobj,
+						       struct kobj_attribute *a,
+						       char *buf)
+{
+	struct btrfs_space_info *sinfo = to_space_info(kobj);
+	s64 val = percpu_counter_sum(&sinfo->total_bytes_pinned);
+	return snprintf(buf, PAGE_SIZE, "%lld\n", val);
+}
+
+SPACE_INFO_ATTR(flags);
+SPACE_INFO_ATTR(total_bytes);
+SPACE_INFO_ATTR(bytes_used);
+SPACE_INFO_ATTR(bytes_pinned);
+SPACE_INFO_ATTR(bytes_reserved);
+SPACE_INFO_ATTR(bytes_may_use);
+SPACE_INFO_ATTR(disk_used);
+SPACE_INFO_ATTR(disk_total);
+BTRFS_ATTR(total_bytes_pinned, 0444, btrfs_space_info_show_total_bytes_pinned);
+
+static struct attribute *space_info_attrs[] = {
+	BTRFS_ATTR_PTR(flags),
+	BTRFS_ATTR_PTR(total_bytes),
+	BTRFS_ATTR_PTR(bytes_used),
+	BTRFS_ATTR_PTR(bytes_pinned),
+	BTRFS_ATTR_PTR(bytes_reserved),
+	BTRFS_ATTR_PTR(bytes_may_use),
+	BTRFS_ATTR_PTR(disk_used),
+	BTRFS_ATTR_PTR(disk_total),
+	BTRFS_ATTR_PTR(total_bytes_pinned),
+	NULL,
+};
+
+static void space_info_release(struct kobject *kobj)
+{
+	struct btrfs_space_info *sinfo = to_space_info(kobj);
+	percpu_counter_destroy(&sinfo->total_bytes_pinned);
+	kfree(sinfo);
+}
+
+struct kobj_type space_info_ktype = {
+	.sysfs_ops = &kobj_sysfs_ops,
+	.release = space_info_release,
+	.default_attrs = space_info_attrs,
+};
+
+static const struct attribute *allocation_attrs[] = {
+	BTRFS_ATTR_PTR(global_rsv_reserved),
+	BTRFS_ATTR_PTR(global_rsv_size),
+	NULL,
+};
+
+static ssize_t btrfs_label_show(struct kobject *kobj,
+				struct kobj_attribute *a, char *buf)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	return snprintf(buf, PAGE_SIZE, "%s\n", fs_info->super_copy->label);
+}
+
+static ssize_t btrfs_label_store(struct kobject *kobj,
+				 struct kobj_attribute *a,
+				 const char *buf, size_t len)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	struct btrfs_trans_handle *trans;
+	struct btrfs_root *root = fs_info->fs_root;
+	int ret;
+
+	if (len >= BTRFS_LABEL_SIZE) {
+		pr_err("BTRFS: unable to set label with more than %d bytes\n",
+		       BTRFS_LABEL_SIZE - 1);
+		return -EINVAL;
+	}
+
+	trans = btrfs_start_transaction(root, 0);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+
+	spin_lock(&root->fs_info->super_lock);
+	strcpy(fs_info->super_copy->label, buf);
+	spin_unlock(&root->fs_info->super_lock);
+	ret = btrfs_commit_transaction(trans, root);
+
+	if (!ret)
+		return len;
+
+	return ret;
+}
+BTRFS_ATTR_RW(label, 0644, btrfs_label_show, btrfs_label_store);
+
+static struct attribute *btrfs_attrs[] = {
+	BTRFS_ATTR_PTR(label),
+	NULL,
+};
+
+static void btrfs_release_super_kobj(struct kobject *kobj)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	complete(&fs_info->kobj_unregister);
+}
+
+static struct kobj_type btrfs_ktype = {
+	.sysfs_ops	= &kobj_sysfs_ops,
+	.release	= btrfs_release_super_kobj,
+	.default_attrs	= btrfs_attrs,
+};
+
+static inline struct btrfs_fs_info *to_fs_info(struct kobject *kobj)
+{
+	if (kobj->ktype != &btrfs_ktype)
+		return NULL;
+	return container_of(kobj, struct btrfs_fs_info, super_kobj);
+}
+
+#define NUM_FEATURE_BITS 64
+static char btrfs_unknown_feature_names[3][NUM_FEATURE_BITS][13];
+static struct btrfs_feature_attr btrfs_feature_attrs[3][NUM_FEATURE_BITS];
+
+static u64 supported_feature_masks[3] = {
+	[FEAT_COMPAT]    = BTRFS_FEATURE_COMPAT_SUPP,
+	[FEAT_COMPAT_RO] = BTRFS_FEATURE_COMPAT_RO_SUPP,
+	[FEAT_INCOMPAT]  = BTRFS_FEATURE_INCOMPAT_SUPP,
+};
+
+static int addrm_unknown_feature_attrs(struct btrfs_fs_info *fs_info, bool add)
+{
+	int set;
+
+	for (set = 0; set < FEAT_MAX; set++) {
+		int i;
+		struct attribute *attrs[2];
+		struct attribute_group agroup = {
+			.name = "features",
+			.attrs = attrs,
+		};
+		u64 features = get_features(fs_info, set);
+		features &= ~supported_feature_masks[set];
+
+		if (!features)
+			continue;
+
+		attrs[1] = NULL;
+		for (i = 0; i < NUM_FEATURE_BITS; i++) {
+			struct btrfs_feature_attr *fa;
+
+			if (!(features & (1ULL << i)))
+				continue;
+
+			fa = &btrfs_feature_attrs[set][i];
+			attrs[0] = &fa->kobj_attr.attr;
+			if (add) {
+				int ret;
+				ret = sysfs_merge_group(&fs_info->super_kobj,
+							&agroup);
+				if (ret)
+					return ret;
+			} else
+				sysfs_unmerge_group(&fs_info->super_kobj,
+						    &agroup);
+		}
+
+	}
+	return 0;
+}
+
+static void __btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info)
+{
+	kobject_del(&fs_info->super_kobj);
+	kobject_put(&fs_info->super_kobj);
+	wait_for_completion(&fs_info->kobj_unregister);
+}
+
+void btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info)
+{
+	if (fs_info->space_info_kobj) {
+		sysfs_remove_files(fs_info->space_info_kobj, allocation_attrs);
+		kobject_del(fs_info->space_info_kobj);
+		kobject_put(fs_info->space_info_kobj);
+	}
+	kobject_del(fs_info->device_dir_kobj);
+	kobject_put(fs_info->device_dir_kobj);
+	addrm_unknown_feature_attrs(fs_info, false);
+	sysfs_remove_group(&fs_info->super_kobj, &btrfs_feature_attr_group);
+	__btrfs_sysfs_remove_one(fs_info);
+}
+
+const char * const btrfs_feature_set_names[3] = {
+	[FEAT_COMPAT]	 = "compat",
+	[FEAT_COMPAT_RO] = "compat_ro",
+	[FEAT_INCOMPAT]	 = "incompat",
+};
+
+char *btrfs_printable_features(enum btrfs_feature_set set, u64 flags)
+{
+	size_t bufsize = 4096; /* safe max, 64 names * 64 bytes */
+	int len = 0;
+	int i;
+	char *str;
+
+	str = kmalloc(bufsize, GFP_KERNEL);
+	if (!str)
+		return str;
+
+	for (i = 0; i < ARRAY_SIZE(btrfs_feature_attrs[set]); i++) {
+		const char *name;
+
+		if (!(flags & (1ULL << i)))
+			continue;
+
+		name = btrfs_feature_attrs[set][i].kobj_attr.attr.name;
+		len += snprintf(str + len, bufsize - len, "%s%s",
+				len ? "," : "", name);
+	}
+
+	return str;
+}
+
+static void init_feature_attrs(void)
+{
+	struct btrfs_feature_attr *fa;
+	int set, i;
+
+	BUILD_BUG_ON(ARRAY_SIZE(btrfs_unknown_feature_names) !=
+		     ARRAY_SIZE(btrfs_feature_attrs));
+	BUILD_BUG_ON(ARRAY_SIZE(btrfs_unknown_feature_names[0]) !=
+		     ARRAY_SIZE(btrfs_feature_attrs[0]));
+
+	memset(btrfs_feature_attrs, 0, sizeof(btrfs_feature_attrs));
+	memset(btrfs_unknown_feature_names, 0,
+	       sizeof(btrfs_unknown_feature_names));
+
+	for (i = 0; btrfs_supported_feature_attrs[i]; i++) {
+		struct btrfs_feature_attr *sfa;
+		struct attribute *a = btrfs_supported_feature_attrs[i];
+		int bit;
+		sfa = attr_to_btrfs_feature_attr(a);
+		bit = ilog2(sfa->feature_bit);
+		fa = &btrfs_feature_attrs[sfa->feature_set][bit];
+
+		fa->kobj_attr.attr.name = sfa->kobj_attr.attr.name;
+	}
+
+	for (set = 0; set < FEAT_MAX; set++) {
+		for (i = 0; i < ARRAY_SIZE(btrfs_feature_attrs[set]); i++) {
+			char *name = btrfs_unknown_feature_names[set][i];
+			fa = &btrfs_feature_attrs[set][i];
+
+			if (fa->kobj_attr.attr.name)
+				continue;
+
+			snprintf(name, 13, "%s:%u",
+				 btrfs_feature_set_names[set], i);
+
+			fa->kobj_attr.attr.name = name;
+			fa->kobj_attr.attr.mode = S_IRUGO;
+			fa->feature_set = set;
+			fa->feature_bit = 1ULL << i;
+		}
+	}
+}
+
+static int add_device_membership(struct btrfs_fs_info *fs_info)
+{
+	int error = 0;
+	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
+	struct btrfs_device *dev;
+
+	fs_info->device_dir_kobj = kobject_create_and_add("devices",
+						&fs_info->super_kobj);
+	if (!fs_info->device_dir_kobj)
+		return -ENOMEM;
+
+	list_for_each_entry(dev, &fs_devices->devices, dev_list) {
+		struct hd_struct *disk = dev->bdev->bd_part;
+		struct kobject *disk_kobj = &part_to_dev(disk)->kobj;
+
+		error = sysfs_create_link(fs_info->device_dir_kobj,
+					  disk_kobj, disk_kobj->name);
+		if (error)
+			break;
+	}
+
+	return error;
+}
 
 /* /sys/fs/btrfs/ entry */
 static struct kset *btrfs_kset;
 
+int btrfs_sysfs_add_one(struct btrfs_fs_info *fs_info)
+{
+	int error;
+
+	init_completion(&fs_info->kobj_unregister);
+	fs_info->super_kobj.kset = btrfs_kset;
+	error = kobject_init_and_add(&fs_info->super_kobj, &btrfs_ktype, NULL,
+				     "%pU", fs_info->fsid);
+	if (error)
+		return error;
+
+	error = sysfs_create_group(&fs_info->super_kobj,
+				   &btrfs_feature_attr_group);
+	if (error) {
+		__btrfs_sysfs_remove_one(fs_info);
+		return error;
+	}
+
+	error = addrm_unknown_feature_attrs(fs_info, true);
+	if (error)
+		goto failure;
+
+	error = add_device_membership(fs_info);
+	if (error)
+		goto failure;
+
+	fs_info->space_info_kobj = kobject_create_and_add("allocation",
+						  &fs_info->super_kobj);
+	if (!fs_info->space_info_kobj) {
+		error = -ENOMEM;
+		goto failure;
+	}
+
+	error = sysfs_create_files(fs_info->space_info_kobj, allocation_attrs);
+	if (error)
+		goto failure;
+
+	return 0;
+failure:
+	btrfs_sysfs_remove_one(fs_info);
+	return error;
+}
+
 int btrfs_init_sysfs(void)
 {
+	int ret;
 	btrfs_kset = kset_create_and_add("btrfs", NULL, fs_kobj);
 	if (!btrfs_kset)
 		return -ENOMEM;
+
+	init_feature_attrs();
+
+	ret = sysfs_create_group(&btrfs_kset->kobj, &btrfs_feature_attr_group);
+	if (ret) {
+		kset_unregister(btrfs_kset);
+		return ret;
+	}
+
 	return 0;
 }
 
 void btrfs_exit_sysfs(void)
 {
+	sysfs_remove_group(&btrfs_kset->kobj, &btrfs_feature_attr_group);
 	kset_unregister(btrfs_kset);
 }
 

fs/btrfs/sysfs.h

@@ -0,0 +1,64 @@
+#ifndef _BTRFS_SYSFS_H_
+#define _BTRFS_SYSFS_H_
+
+enum btrfs_feature_set {
+	FEAT_COMPAT,
+	FEAT_COMPAT_RO,
+	FEAT_INCOMPAT,
+	FEAT_MAX
+};
+
+#define __INIT_KOBJ_ATTR(_name, _mode, _show, _store)			\
+{									\
+	.attr	= { .name = __stringify(_name), .mode = _mode },	\
+	.show	= _show,						\
+	.store	= _store,						\
+}
+
+#define BTRFS_ATTR_RW(_name, _mode, _show, _store)			\
+static struct kobj_attribute btrfs_attr_##_name =			\
+			__INIT_KOBJ_ATTR(_name, _mode, _show, _store)
+#define BTRFS_ATTR(_name, _mode, _show)					\
+	BTRFS_ATTR_RW(_name, _mode, _show, NULL)
+#define BTRFS_ATTR_PTR(_name)    (&btrfs_attr_##_name.attr)
+
+#define BTRFS_RAID_ATTR(_name, _show)					\
+static struct kobj_attribute btrfs_raid_attr_##_name =			\
+			__INIT_KOBJ_ATTR(_name, 0444, _show, NULL)
+#define BTRFS_RAID_ATTR_PTR(_name)    (&btrfs_raid_attr_##_name.attr)
+
+
+struct btrfs_feature_attr {
+	struct kobj_attribute kobj_attr;
+	enum btrfs_feature_set feature_set;
+	u64 feature_bit;
+};
+
+#define BTRFS_FEAT_ATTR(_name, _feature_set, _prefix, _feature_bit)	     \
+static struct btrfs_feature_attr btrfs_attr_##_name = {			     \
+	.kobj_attr = __INIT_KOBJ_ATTR(_name, S_IRUGO,			     \
+				      btrfs_feature_attr_show,		     \
+				      btrfs_feature_attr_store),	     \
+	.feature_set	= _feature_set,					     \
+	.feature_bit	= _prefix ##_## _feature_bit,			     \
+}
+#define BTRFS_FEAT_ATTR_PTR(_name)    (&btrfs_attr_##_name.kobj_attr.attr)
+
+#define BTRFS_FEAT_ATTR_COMPAT(name, feature) \
+	BTRFS_FEAT_ATTR(name, FEAT_COMPAT, BTRFS_FEATURE_COMPAT, feature)
+#define BTRFS_FEAT_ATTR_COMPAT_RO(name, feature) \
+	BTRFS_FEAT_ATTR(name, FEAT_COMPAT_RO, BTRFS_FEATURE_COMPAT, feature)
+#define BTRFS_FEAT_ATTR_INCOMPAT(name, feature) \
+	BTRFS_FEAT_ATTR(name, FEAT_INCOMPAT, BTRFS_FEATURE_INCOMPAT, feature)
+
+/* convert from attribute */
+#define to_btrfs_feature_attr(a) \
+			container_of(a, struct btrfs_feature_attr, kobj_attr)
+#define attr_to_btrfs_attr(a) container_of(a, struct kobj_attribute, attr)
+#define attr_to_btrfs_feature_attr(a) \
+			to_btrfs_feature_attr(attr_to_btrfs_attr(a))
+char *btrfs_printable_features(enum btrfs_feature_set set, u64 flags);
+extern const char * const btrfs_feature_set_names[3];
+extern struct kobj_type space_info_ktype;
+extern struct kobj_type btrfs_raid_ktype;
+#endif /* _BTRFS_SYSFS_H_ */

fs/btrfs/tests/btrfs-tests.h

@@ -21,7 +21,7 @@
 
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 
-#define test_msg(fmt, ...) pr_info("btrfs: selftest: " fmt, ##__VA_ARGS__)
+#define test_msg(fmt, ...) pr_info("BTRFS: selftest: " fmt, ##__VA_ARGS__)
 
 int btrfs_test_free_space_cache(void);
 int btrfs_test_extent_buffer_operations(void);

fs/btrfs/transaction.c

@@ -62,7 +62,7 @@ void btrfs_put_transaction(struct btrfs_transaction *transaction)
 	WARN_ON(atomic_read(&transaction->use_count) == 0);
 	if (atomic_dec_and_test(&transaction->use_count)) {
 		BUG_ON(!list_empty(&transaction->list));
-		WARN_ON(transaction->delayed_refs.root.rb_node);
+		WARN_ON(!RB_EMPTY_ROOT(&transaction->delayed_refs.href_root));
 		while (!list_empty(&transaction->pending_chunks)) {
 			struct extent_map *em;
 
@@ -183,8 +183,8 @@ loop:
 	atomic_set(&cur_trans->use_count, 2);
 	cur_trans->start_time = get_seconds();
 
-	cur_trans->delayed_refs.root = RB_ROOT;
-	cur_trans->delayed_refs.num_entries = 0;
+	cur_trans->delayed_refs.href_root = RB_ROOT;
+	atomic_set(&cur_trans->delayed_refs.num_entries, 0);
 	cur_trans->delayed_refs.num_heads_ready = 0;
 	cur_trans->delayed_refs.num_heads = 0;
 	cur_trans->delayed_refs.flushing = 0;
@@ -196,17 +196,14 @@ loop:
 	 */
 	smp_mb();
 	if (!list_empty(&fs_info->tree_mod_seq_list))
-		WARN(1, KERN_ERR "btrfs: tree_mod_seq_list not empty when "
+		WARN(1, KERN_ERR "BTRFS: tree_mod_seq_list not empty when "
 			"creating a fresh transaction\n");
 	if (!RB_EMPTY_ROOT(&fs_info->tree_mod_log))
-		WARN(1, KERN_ERR "btrfs: tree_mod_log rb tree not empty when "
+		WARN(1, KERN_ERR "BTRFS: tree_mod_log rb tree not empty when "
 			"creating a fresh transaction\n");
 	atomic64_set(&fs_info->tree_mod_seq, 0);
 
 	spin_lock_init(&cur_trans->delayed_refs.lock);
-	atomic_set(&cur_trans->delayed_refs.procs_running_refs, 0);
-	atomic_set(&cur_trans->delayed_refs.ref_seq, 0);
-	init_waitqueue_head(&cur_trans->delayed_refs.wait);
 
 	INIT_LIST_HEAD(&cur_trans->pending_snapshots);
 	INIT_LIST_HEAD(&cur_trans->ordered_operations);
@@ -472,6 +469,7 @@ again:
 	h->type = type;
 	h->allocating_chunk = false;
 	h->reloc_reserved = false;
+	h->sync = false;
 	INIT_LIST_HEAD(&h->qgroup_ref_list);
 	INIT_LIST_HEAD(&h->new_bgs);
 
@@ -647,7 +645,7 @@ static int should_end_transaction(struct btrfs_trans_handle *trans,
 				  struct btrfs_root *root)
 {
 	if (root->fs_info->global_block_rsv.space_info->full &&
-	    btrfs_should_throttle_delayed_refs(trans, root))
+	    btrfs_check_space_for_delayed_refs(trans, root))
 		return 1;
 
 	return !!btrfs_block_rsv_check(root, &root->fs_info->global_block_rsv, 5);
@@ -711,8 +709,8 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 		btrfs_create_pending_block_groups(trans, root);
 
 	trans->delayed_ref_updates = 0;
-	if (btrfs_should_throttle_delayed_refs(trans, root)) {
-		cur = max_t(unsigned long, cur, 1);
+	if (!trans->sync && btrfs_should_throttle_delayed_refs(trans, root)) {
+		cur = max_t(unsigned long, cur, 32);
 		trans->delayed_ref_updates = 0;
 		btrfs_run_delayed_refs(trans, root, cur);
 	}
@@ -788,12 +786,6 @@ int btrfs_end_transaction_throttle(struct btrfs_trans_handle *trans,
 	return __btrfs_end_transaction(trans, root, 1);
 }
 
-int btrfs_end_transaction_dmeta(struct btrfs_trans_handle *trans,
-				struct btrfs_root *root)
-{
-	return __btrfs_end_transaction(trans, root, 1);
-}
-
 /*
  * when btree blocks are allocated, they have some corresponding bits set for
  * them in one of two extent_io trees.  This is used to make sure all of
@@ -1105,7 +1097,7 @@ int btrfs_defrag_root(struct btrfs_root *root)
 			break;
 
 		if (btrfs_defrag_cancelled(root->fs_info)) {
-			printk(KERN_DEBUG "btrfs: defrag_root cancelled\n");
+			pr_debug("BTRFS: defrag_root cancelled\n");
 			ret = -EAGAIN;
 			break;
 		}
@@ -1746,6 +1738,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 		goto cleanup_transaction;
 
 	btrfs_wait_delalloc_flush(root->fs_info);
+
+	btrfs_scrub_pause(root);
 	/*
 	 * Ok now we need to make sure to block out any other joins while we
 	 * commit the transaction.  We could have started a join before setting
@@ -1810,7 +1804,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	WARN_ON(cur_trans != trans->transaction);
 
-	btrfs_scrub_pause(root);
 	/* btrfs_commit_tree_roots is responsible for getting the
 	 * various roots consistent with each other.  Every pointer
 	 * in the tree of tree roots has to point to the most up to date
@@ -1833,6 +1826,15 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 		goto cleanup_transaction;
 	}
 
+	/*
+	 * Since the transaction is done, we should set the inode map cache flag
+	 * before any other comming transaction.
+	 */
+	if (btrfs_test_opt(root, CHANGE_INODE_CACHE))
+		btrfs_set_opt(root->fs_info->mount_opt, INODE_MAP_CACHE);
+	else
+		btrfs_clear_opt(root->fs_info->mount_opt, INODE_MAP_CACHE);
+
 	/* commit_fs_roots gets rid of all the tree log roots, it is now
 	 * safe to free the root of tree log roots
 	 */
@@ -1975,10 +1977,23 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 	}
 	root = list_first_entry(&fs_info->dead_roots,
 			struct btrfs_root, root_list);
+	/*
+	 * Make sure root is not involved in send,
+	 * if we fail with first root, we return
+	 * directly rather than continue.
+	 */
+	spin_lock(&root->root_item_lock);
+	if (root->send_in_progress) {
+		spin_unlock(&fs_info->trans_lock);
+		spin_unlock(&root->root_item_lock);
+		return 0;
+	}
+	spin_unlock(&root->root_item_lock);
+
 	list_del_init(&root->root_list);
 	spin_unlock(&fs_info->trans_lock);
 
-	pr_debug("btrfs: cleaner removing %llu\n", root->objectid);
+	pr_debug("BTRFS: cleaner removing %llu\n", root->objectid);
 
 	btrfs_kill_all_delayed_nodes(root);
 

fs/btrfs/transaction.h

@@ -93,6 +93,7 @@ struct btrfs_trans_handle {
 	short adding_csums;
 	bool allocating_chunk;
 	bool reloc_reserved;
+	bool sync;
 	unsigned int type;
 	/*
 	 * this root is only needed to validate that the root passed to
@@ -154,8 +155,6 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans,
 				   int wait_for_unblock);
 int btrfs_end_transaction_throttle(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root);
-int btrfs_end_transaction_dmeta(struct btrfs_trans_handle *trans,
-				struct btrfs_root *root);
 int btrfs_should_end_transaction(struct btrfs_trans_handle *trans,
 				 struct btrfs_root *root);
 void btrfs_throttle(struct btrfs_root *root);

fs/btrfs/tree-log.c

@@ -570,7 +570,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 		if (btrfs_file_extent_disk_bytenr(eb, item) == 0)
 			nbytes = 0;
 	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
-		size = btrfs_file_extent_inline_len(eb, item);
+		size = btrfs_file_extent_inline_len(eb, slot, item);
 		nbytes = btrfs_file_extent_ram_bytes(eb, item);
 		extent_end = ALIGN(start + size, root->sectorsize);
 	} else {
@@ -1238,7 +1238,8 @@ static int insert_orphan_item(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root, u64 offset)
 {
 	int ret;
-	ret = btrfs_find_orphan_item(root, offset);
+	ret = btrfs_find_item(root, NULL, BTRFS_ORPHAN_OBJECTID,
+			offset, BTRFS_ORPHAN_ITEM_KEY, NULL);
 	if (ret > 0)
 		ret = btrfs_insert_orphan_item(trans, root, offset);
 	return ret;
@@ -3194,7 +3195,7 @@ static int log_inode_item(struct btrfs_trans_handle *trans,
 static noinline int copy_items(struct btrfs_trans_handle *trans,
 			       struct inode *inode,
 			       struct btrfs_path *dst_path,
-			       struct extent_buffer *src,
+			       struct btrfs_path *src_path, u64 *last_extent,
 			       int start_slot, int nr, int inode_only)
 {
 	unsigned long src_offset;
@@ -3202,6 +3203,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 	struct btrfs_root *log = BTRFS_I(inode)->root->log_root;
 	struct btrfs_file_extent_item *extent;
 	struct btrfs_inode_item *inode_item;
+	struct extent_buffer *src = src_path->nodes[0];
+	struct btrfs_key first_key, last_key, key;
 	int ret;
 	struct btrfs_key *ins_keys;
 	u32 *ins_sizes;
@@ -3209,6 +3212,9 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 	int i;
 	struct list_head ordered_sums;
 	int skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
+	bool has_extents = false;
+	bool need_find_last_extent = (*last_extent == 0);
+	bool done = false;
 
 	INIT_LIST_HEAD(&ordered_sums);
 
@@ -3217,6 +3223,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 	if (!ins_data)
 		return -ENOMEM;
 
+	first_key.objectid = (u64)-1;
+
 	ins_sizes = (u32 *)ins_data;
 	ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32));
 
@@ -3237,6 +3245,9 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 
 		src_offset = btrfs_item_ptr_offset(src, start_slot + i);
 
+		if ((i == (nr - 1)))
+			last_key = ins_keys[i];
+
 		if (ins_keys[i].type == BTRFS_INODE_ITEM_KEY) {
 			inode_item = btrfs_item_ptr(dst_path->nodes[0],
 						    dst_path->slots[0],
@@ -3248,6 +3259,21 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 					   src_offset, ins_sizes[i]);
 		}
 
+		/*
+		 * We set need_find_last_extent here in case we know we were
+		 * processing other items and then walk into the first extent in
+		 * the inode.  If we don't hit an extent then nothing changes,
+		 * we'll do the last search the next time around.
+		 */
+		if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY) {
+			has_extents = true;
+			if (need_find_last_extent &&
+			    first_key.objectid == (u64)-1)
+				first_key = ins_keys[i];
+		} else {
+			need_find_last_extent = false;
+		}
+
 		/* take a reference on file data extents so that truncates
 		 * or deletes of this inode don't have to relog the inode
 		 * again
@@ -3312,6 +3338,128 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 		list_del(&sums->list);
 		kfree(sums);
 	}
+
+	if (!has_extents)
+		return ret;
+
+	/*
+	 * Because we use btrfs_search_forward we could skip leaves that were
+	 * not modified and then assume *last_extent is valid when it really
+	 * isn't.  So back up to the previous leaf and read the end of the last
+	 * extent before we go and fill in holes.
+	 */
+	if (need_find_last_extent) {
+		u64 len;
+
+		ret = btrfs_prev_leaf(BTRFS_I(inode)->root, src_path);
+		if (ret < 0)
+			return ret;
+		if (ret)
+			goto fill_holes;
+		if (src_path->slots[0])
+			src_path->slots[0]--;
+		src = src_path->nodes[0];
+		btrfs_item_key_to_cpu(src, &key, src_path->slots[0]);
+		if (key.objectid != btrfs_ino(inode) ||
+		    key.type != BTRFS_EXTENT_DATA_KEY)
+			goto fill_holes;
+		extent = btrfs_item_ptr(src, src_path->slots[0],
+					struct btrfs_file_extent_item);
+		if (btrfs_file_extent_type(src, extent) ==
+		    BTRFS_FILE_EXTENT_INLINE) {
+			len = btrfs_file_extent_inline_len(src,
+							   src_path->slots[0],
+							   extent);
+			*last_extent = ALIGN(key.offset + len,
+					     log->sectorsize);
+		} else {
+			len = btrfs_file_extent_num_bytes(src, extent);
+			*last_extent = key.offset + len;
+		}
+	}
+fill_holes:
+	/* So we did prev_leaf, now we need to move to the next leaf, but a few
+	 * things could have happened
+	 *
+	 * 1) A merge could have happened, so we could currently be on a leaf
+	 * that holds what we were copying in the first place.
+	 * 2) A split could have happened, and now not all of the items we want
+	 * are on the same leaf.
+	 *
+	 * So we need to adjust how we search for holes, we need to drop the
+	 * path and re-search for the first extent key we found, and then walk
+	 * forward until we hit the last one we copied.
+	 */
+	if (need_find_last_extent) {
+		/* btrfs_prev_leaf could return 1 without releasing the path */
+		btrfs_release_path(src_path);
+		ret = btrfs_search_slot(NULL, BTRFS_I(inode)->root, &first_key,
+					src_path, 0, 0);
+		if (ret < 0)
+			return ret;
+		ASSERT(ret == 0);
+		src = src_path->nodes[0];
+		i = src_path->slots[0];
+	} else {
+		i = start_slot;
+	}
+
+	/*
+	 * Ok so here we need to go through and fill in any holes we may have
+	 * to make sure that holes are punched for those areas in case they had
+	 * extents previously.
+	 */
+	while (!done) {
+		u64 offset, len;
+		u64 extent_end;
+
+		if (i >= btrfs_header_nritems(src_path->nodes[0])) {
+			ret = btrfs_next_leaf(BTRFS_I(inode)->root, src_path);
+			if (ret < 0)
+				return ret;
+			ASSERT(ret == 0);
+			src = src_path->nodes[0];
+			i = 0;
+		}
+
+		btrfs_item_key_to_cpu(src, &key, i);
+		if (!btrfs_comp_cpu_keys(&key, &last_key))
+			done = true;
+		if (key.objectid != btrfs_ino(inode) ||
+		    key.type != BTRFS_EXTENT_DATA_KEY) {
+			i++;
+			continue;
+		}
+		extent = btrfs_item_ptr(src, i, struct btrfs_file_extent_item);
+		if (btrfs_file_extent_type(src, extent) ==
+		    BTRFS_FILE_EXTENT_INLINE) {
+			len = btrfs_file_extent_inline_len(src, i, extent);
+			extent_end = ALIGN(key.offset + len, log->sectorsize);
+		} else {
+			len = btrfs_file_extent_num_bytes(src, extent);
+			extent_end = key.offset + len;
+		}
+		i++;
+
+		if (*last_extent == key.offset) {
+			*last_extent = extent_end;
+			continue;
+		}
+		offset = *last_extent;
+		len = key.offset - *last_extent;
+		ret = btrfs_insert_file_extent(trans, log, btrfs_ino(inode),
+					       offset, 0, 0, len, 0, len, 0,
+					       0, 0);
+		if (ret)
+			break;
+		*last_extent = offset + len;
+	}
+	/*
+	 * Need to let the callers know we dropped the path so they should
+	 * re-search.
+	 */
+	if (!ret && need_find_last_extent)
+		ret = 1;
 	return ret;
 }
 
@@ -3349,21 +3497,27 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 	int ret;
 	int index = log->log_transid % 2;
 	bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
-
-	ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
-				   em->start + em->len, NULL, 0);
-	if (ret)
-		return ret;
+	int extent_inserted = 0;
 
 	INIT_LIST_HEAD(&ordered_sums);
 	btrfs_init_map_token(&token);
-	key.objectid = btrfs_ino(inode);
-	key.type = BTRFS_EXTENT_DATA_KEY;
-	key.offset = em->start;
 
-	ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*fi));
+	ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
+				   em->start + em->len, NULL, 0, 1,
+				   sizeof(*fi), &extent_inserted);
 	if (ret)
 		return ret;
+
+	if (!extent_inserted) {
+		key.objectid = btrfs_ino(inode);
+		key.type = BTRFS_EXTENT_DATA_KEY;
+		key.offset = em->start;
+
+		ret = btrfs_insert_empty_item(trans, log, path, &key,
+					      sizeof(*fi));
+		if (ret)
+			return ret;
+	}
 	leaf = path->nodes[0];
 	fi = btrfs_item_ptr(leaf, path->slots[0],
 			    struct btrfs_file_extent_item);
@@ -3485,7 +3639,11 @@ again:
 		 * start over after this.
 		 */
 
-		wait_event(ordered->wait, ordered->csum_bytes_left == 0);
+		if (ordered->csum_bytes_left) {
+			btrfs_start_ordered_extent(inode, ordered, 0);
+			wait_event(ordered->wait,
+				   ordered->csum_bytes_left == 0);
+		}
 
 		list_for_each_entry(sum, &ordered->list, list) {
 			ret = btrfs_csum_file_blocks(trans, log, sum);
@@ -3630,6 +3788,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 	struct btrfs_key max_key;
 	struct btrfs_root *log = root->log_root;
 	struct extent_buffer *src = NULL;
+	u64 last_extent = 0;
 	int err = 0;
 	int ret;
 	int nritems;
@@ -3745,11 +3904,15 @@ again:
 			goto next_slot;
 		}
 
-		ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
-				 ins_nr, inode_only);
-		if (ret) {
+		ret = copy_items(trans, inode, dst_path, path, &last_extent,
+				 ins_start_slot, ins_nr, inode_only);
+		if (ret < 0) {
 			err = ret;
 			goto out_unlock;
+		} if (ret) {
+			ins_nr = 0;
+			btrfs_release_path(path);
+			continue;
 		}
 		ins_nr = 1;
 		ins_start_slot = path->slots[0];
@@ -3763,13 +3926,14 @@ next_slot:
 			goto again;
 		}
 		if (ins_nr) {
-			ret = copy_items(trans, inode, dst_path, src,
-					 ins_start_slot,
+			ret = copy_items(trans, inode, dst_path, path,
+					 &last_extent, ins_start_slot,
 					 ins_nr, inode_only);
-			if (ret) {
+			if (ret < 0) {
 				err = ret;
 				goto out_unlock;
 			}
+			ret = 0;
 			ins_nr = 0;
 		}
 		btrfs_release_path(path);
@@ -3784,12 +3948,13 @@ next_slot:
 		}
 	}
 	if (ins_nr) {
-		ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
-				 ins_nr, inode_only);
-		if (ret) {
+		ret = copy_items(trans, inode, dst_path, path, &last_extent,
+				 ins_start_slot, ins_nr, inode_only);
+		if (ret < 0) {
 			err = ret;
 			goto out_unlock;
 		}
+		ret = 0;
 		ins_nr = 0;
 	}
 

fs/btrfs/ulist.c

@@ -5,8 +5,8 @@
  */
 
 #include <linux/slab.h>
-#include <linux/export.h>
 #include "ulist.h"
+#include "ctree.h"
 
 /*
  * ulist is a generic data structure to hold a collection of unique u64
@@ -14,10 +14,6 @@
  * enumerating it.
  * It is possible to store an auxiliary value along with the key.
  *
- * The implementation is preliminary and can probably be sped up
- * significantly. A first step would be to store the values in an rbtree
- * as soon as ULIST_SIZE is exceeded.
- *
  * A sample usage for ulists is the enumeration of directed graphs without
  * visiting a node twice. The pseudo-code could look like this:
  *
@@ -50,12 +46,10 @@
  */
 void ulist_init(struct ulist *ulist)
 {
-	ulist->nnodes = 0;
-	ulist->nodes = ulist->int_nodes;
-	ulist->nodes_alloced = ULIST_SIZE;
+	INIT_LIST_HEAD(&ulist->nodes);
 	ulist->root = RB_ROOT;
+	ulist->nnodes = 0;
 }
-EXPORT_SYMBOL(ulist_init);
 
 /**
  * ulist_fini - free up additionally allocated memory for the ulist
@@ -64,18 +58,17 @@ EXPORT_SYMBOL(ulist_init);
  * This is useful in cases where the base 'struct ulist' has been statically
  * allocated.
  */
-void ulist_fini(struct ulist *ulist)
+static void ulist_fini(struct ulist *ulist)
 {
-	/*
-	 * The first ULIST_SIZE elements are stored inline in struct ulist.
-	 * Only if more elements are alocated they need to be freed.
-	 */
-	if (ulist->nodes_alloced > ULIST_SIZE)
-		kfree(ulist->nodes);
-	ulist->nodes_alloced = 0;	/* in case ulist_fini is called twice */
+	struct ulist_node *node;
+	struct ulist_node *next;
+
+	list_for_each_entry_safe(node, next, &ulist->nodes, list) {
+		kfree(node);
+	}
 	ulist->root = RB_ROOT;
+	INIT_LIST_HEAD(&ulist->nodes);
 }
-EXPORT_SYMBOL(ulist_fini);
 
 /**
  * ulist_reinit - prepare a ulist for reuse
@@ -89,7 +82,6 @@ void ulist_reinit(struct ulist *ulist)
 	ulist_fini(ulist);
 	ulist_init(ulist);
 }
-EXPORT_SYMBOL(ulist_reinit);
 
 /**
  * ulist_alloc - dynamically allocate a ulist
@@ -108,7 +100,6 @@ struct ulist *ulist_alloc(gfp_t gfp_mask)
 
 	return ulist;
 }
-EXPORT_SYMBOL(ulist_alloc);
 
 /**
  * ulist_free - free dynamically allocated ulist
@@ -123,7 +114,6 @@ void ulist_free(struct ulist *ulist)
 	ulist_fini(ulist);
 	kfree(ulist);
 }
-EXPORT_SYMBOL(ulist_free);
 
 static struct ulist_node *ulist_rbtree_search(struct ulist *ulist, u64 val)
 {
@@ -192,63 +182,32 @@ int ulist_add(struct ulist *ulist, u64 val, u64 aux, gfp_t gfp_mask)
 int ulist_add_merge(struct ulist *ulist, u64 val, u64 aux,
 		    u64 *old_aux, gfp_t gfp_mask)
 {
-	int ret = 0;
-	struct ulist_node *node = NULL;
+	int ret;
+	struct ulist_node *node;
+
 	node = ulist_rbtree_search(ulist, val);
 	if (node) {
 		if (old_aux)
 			*old_aux = node->aux;
 		return 0;
 	}
+	node = kmalloc(sizeof(*node), gfp_mask);
+	if (!node)
+		return -ENOMEM;
 
-	if (ulist->nnodes >= ulist->nodes_alloced) {
-		u64 new_alloced = ulist->nodes_alloced + 128;
-		struct ulist_node *new_nodes;
-		void *old = NULL;
-		int i;
-
-		for (i = 0; i < ulist->nnodes; i++)
-			rb_erase(&ulist->nodes[i].rb_node, &ulist->root);
-
-		/*
-		 * if nodes_alloced == ULIST_SIZE no memory has been allocated
-		 * yet, so pass NULL to krealloc
-		 */
-		if (ulist->nodes_alloced > ULIST_SIZE)
-			old = ulist->nodes;
+	node->val = val;
+	node->aux = aux;
+#ifdef CONFIG_BTRFS_DEBUG
+	node->seqnum = ulist->nnodes;
+#endif
 
-		new_nodes = krealloc(old, sizeof(*new_nodes) * new_alloced,
-				     gfp_mask);
-		if (!new_nodes)
-			return -ENOMEM;
-
-		if (!old)
-			memcpy(new_nodes, ulist->int_nodes,
-			       sizeof(ulist->int_nodes));
-
-		ulist->nodes = new_nodes;
-		ulist->nodes_alloced = new_alloced;
-
-		/*
-		 * krealloc actually uses memcpy, which does not copy rb_node
-		 * pointers, so we have to do it ourselves.  Otherwise we may
-		 * be bitten by crashes.
-		 */
-		for (i = 0; i < ulist->nnodes; i++) {
-			ret = ulist_rbtree_insert(ulist, &ulist->nodes[i]);
-			if (ret < 0)
-				return ret;
-		}
-	}
-	ulist->nodes[ulist->nnodes].val = val;
-	ulist->nodes[ulist->nnodes].aux = aux;
-	ret = ulist_rbtree_insert(ulist, &ulist->nodes[ulist->nnodes]);
-	BUG_ON(ret);
-	++ulist->nnodes;
+	ret = ulist_rbtree_insert(ulist, node);
+	ASSERT(!ret);
+	list_add_tail(&node->list, &ulist->nodes);
+	ulist->nnodes++;
 
 	return 1;
 }
-EXPORT_SYMBOL(ulist_add);
 
 /**
  * ulist_next - iterate ulist
@@ -268,11 +227,25 @@ EXPORT_SYMBOL(ulist_add);
  */
 struct ulist_node *ulist_next(struct ulist *ulist, struct ulist_iterator *uiter)
 {
-	if (ulist->nnodes == 0)
+	struct ulist_node *node;
+
+	if (list_empty(&ulist->nodes))
 		return NULL;
-	if (uiter->i < 0 || uiter->i >= ulist->nnodes)
+	if (uiter->cur_list && uiter->cur_list->next == &ulist->nodes)
 		return NULL;
-
-	return &ulist->nodes[uiter->i++];
+	if (uiter->cur_list) {
+		uiter->cur_list = uiter->cur_list->next;
+	} else {
+		uiter->cur_list = ulist->nodes.next;
+#ifdef CONFIG_BTRFS_DEBUG
+		uiter->i = 0;
+#endif
+	}
+	node = list_entry(uiter->cur_list, struct ulist_node, list);
+#ifdef CONFIG_BTRFS_DEBUG
+	ASSERT(node->seqnum == uiter->i);
+	ASSERT(uiter->i >= 0 && uiter->i < ulist->nnodes);
+	uiter->i++;
+#endif
+	return node;
 }
-EXPORT_SYMBOL(ulist_next);

fs/btrfs/ulist.h

@@ -17,18 +17,12 @@
  * enumerating it.
  * It is possible to store an auxiliary value along with the key.
  *
- * The implementation is preliminary and can probably be sped up
- * significantly. A first step would be to store the values in an rbtree
- * as soon as ULIST_SIZE is exceeded.
  */
-
-/*
- * number of elements statically allocated inside struct ulist
- */
-#define ULIST_SIZE 16
-
 struct ulist_iterator {
+#ifdef CONFIG_BTRFS_DEBUG
 	int i;
+#endif
+	struct list_head *cur_list;  /* hint to start search */
 };
 
 /*
@@ -37,6 +31,12 @@ struct ulist_iterator {
 struct ulist_node {
 	u64 val;		/* value to store */
 	u64 aux;		/* auxiliary value saved along with the val */
+
+#ifdef CONFIG_BTRFS_DEBUG
+	int seqnum;		/* sequence number this node is added */
+#endif
+
+	struct list_head list;  /* used to link node */
 	struct rb_node rb_node;	/* used to speed up search */
 };
 
@@ -46,28 +46,11 @@ struct ulist {
 	 */
 	unsigned long nnodes;
 
-	/*
-	 * number of nodes we already have room for
-	 */
-	unsigned long nodes_alloced;
-
-	/*
-	 * pointer to the array storing the elements. The first ULIST_SIZE
-	 * elements are stored inline. In this case the it points to int_nodes.
-	 * After exceeding ULIST_SIZE, dynamic memory is allocated.
-	 */
-	struct ulist_node *nodes;
-
+	struct list_head nodes;
 	struct rb_root root;
-
-	/*
-	 * inline storage space for the first ULIST_SIZE entries
-	 */
-	struct ulist_node int_nodes[ULIST_SIZE];
 };
 
 void ulist_init(struct ulist *ulist);
-void ulist_fini(struct ulist *ulist);
 void ulist_reinit(struct ulist *ulist);
 struct ulist *ulist_alloc(gfp_t gfp_mask);
 void ulist_free(struct ulist *ulist);
@@ -77,6 +60,6 @@ int ulist_add_merge(struct ulist *ulist, u64 val, u64 aux,
 struct ulist_node *ulist_next(struct ulist *ulist,
 			      struct ulist_iterator *uiter);
 
-#define ULIST_ITER_INIT(uiter) ((uiter)->i = 0)
+#define ULIST_ITER_INIT(uiter) ((uiter)->cur_list = NULL)
 
 #endif

fs/btrfs/uuid-tree.c

@@ -69,7 +69,7 @@ static int btrfs_uuid_tree_lookup(struct btrfs_root *uuid_root, u8 *uuid,
 	ret = -ENOENT;
 
 	if (!IS_ALIGNED(item_size, sizeof(u64))) {
-		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+		btrfs_warn(uuid_root->fs_info, "uuid item with illegal size %lu!",
 			(unsigned long)item_size);
 		goto out;
 	}
@@ -137,7 +137,8 @@ int btrfs_uuid_tree_add(struct btrfs_trans_handle *trans,
 		offset = btrfs_item_ptr_offset(eb, slot);
 		offset += btrfs_item_size_nr(eb, slot) - sizeof(subid_le);
 	} else if (ret < 0) {
-		pr_warn("btrfs: insert uuid item failed %d (0x%016llx, 0x%016llx) type %u!\n",
+		btrfs_warn(uuid_root->fs_info, "insert uuid item failed %d "
+			"(0x%016llx, 0x%016llx) type %u!",
 			ret, (unsigned long long)key.objectid,
 			(unsigned long long)key.offset, type);
 		goto out;
@@ -183,7 +184,7 @@ int btrfs_uuid_tree_rem(struct btrfs_trans_handle *trans,
 
 	ret = btrfs_search_slot(trans, uuid_root, &key, path, -1, 1);
 	if (ret < 0) {
-		pr_warn("btrfs: error %d while searching for uuid item!\n",
+		btrfs_warn(uuid_root->fs_info, "error %d while searching for uuid item!",
 			ret);
 		goto out;
 	}
@@ -197,7 +198,7 @@ int btrfs_uuid_tree_rem(struct btrfs_trans_handle *trans,
 	offset = btrfs_item_ptr_offset(eb, slot);
 	item_size = btrfs_item_size_nr(eb, slot);
 	if (!IS_ALIGNED(item_size, sizeof(u64))) {
-		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+		btrfs_warn(uuid_root->fs_info, "uuid item with illegal size %lu!",
 			(unsigned long)item_size);
 		ret = -ENOENT;
 		goto out;
@@ -299,7 +300,7 @@ again_search_slot:
 		offset = btrfs_item_ptr_offset(leaf, slot);
 		item_size = btrfs_item_size_nr(leaf, slot);
 		if (!IS_ALIGNED(item_size, sizeof(u64))) {
-			pr_warn("btrfs: uuid item with illegal size %lu!\n",
+			btrfs_warn(fs_info, "uuid item with illegal size %lu!",
 				(unsigned long)item_size);
 			goto skip;
 		}
@@ -349,6 +350,6 @@ skip:
 out:
 	btrfs_free_path(path);
 	if (ret)
-		pr_warn("btrfs: btrfs_uuid_tree_iterate failed %d\n", ret);
+		btrfs_warn(fs_info, "btrfs_uuid_tree_iterate failed %d", ret);
 	return 0;
 }

fs/btrfs/volumes.c

@@ -125,7 +125,7 @@ static void btrfs_kobject_uevent(struct block_device *bdev,
 
 	ret = kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, action);
 	if (ret)
-		pr_warn("Sending event '%d' to kobject: '%s' (%p): failed\n",
+		pr_warn("BTRFS: Sending event '%d' to kobject: '%s' (%p): failed\n",
 			action,
 			kobject_name(&disk_to_dev(bdev->bd_disk)->kobj),
 			&disk_to_dev(bdev->bd_disk)->kobj);
@@ -200,7 +200,7 @@ btrfs_get_bdev_and_sb(const char *device_path, fmode_t flags, void *holder,
 
 	if (IS_ERR(*bdev)) {
 		ret = PTR_ERR(*bdev);
-		printk(KERN_INFO "btrfs: open %s failed\n", device_path);
+		printk(KERN_INFO "BTRFS: open %s failed\n", device_path);
 		goto error;
 	}
 
@@ -912,9 +912,9 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 	if (disk_super->label[0]) {
 		if (disk_super->label[BTRFS_LABEL_SIZE - 1])
 			disk_super->label[BTRFS_LABEL_SIZE - 1] = '\0';
-		printk(KERN_INFO "btrfs: device label %s ", disk_super->label);
+		printk(KERN_INFO "BTRFS: device label %s ", disk_super->label);
 	} else {
-		printk(KERN_INFO "btrfs: device fsid %pU ", disk_super->fsid);
+		printk(KERN_INFO "BTRFS: device fsid %pU ", disk_super->fsid);
 	}
 
 	printk(KERN_CONT "devid %llu transid %llu %s\n", devid, transid, path);
@@ -1813,7 +1813,7 @@ int btrfs_find_device_missing_or_by_path(struct btrfs_root *root,
 		}
 
 		if (!*device) {
-			pr_err("btrfs: no missing device found\n");
+			btrfs_err(root->fs_info, "no missing device found");
 			return -ENOENT;
 		}
 
@@ -3052,7 +3052,7 @@ loop:
 error:
 	btrfs_free_path(path);
 	if (enospc_errors) {
-		printk(KERN_INFO "btrfs: %d enospc errors during balance\n",
+		btrfs_info(fs_info, "%d enospc errors during balance",
 		       enospc_errors);
 		if (!ret)
 			ret = -ENOSPC;
@@ -3138,8 +3138,8 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 		if (!(bctl->flags & BTRFS_BALANCE_DATA) ||
 		    !(bctl->flags & BTRFS_BALANCE_METADATA) ||
 		    memcmp(&bctl->data, &bctl->meta, sizeof(bctl->data))) {
-			printk(KERN_ERR "btrfs: with mixed groups data and "
-			       "metadata balance options must be the same\n");
+			btrfs_err(fs_info, "with mixed groups data and "
+				   "metadata balance options must be the same");
 			ret = -EINVAL;
 			goto out;
 		}
@@ -3165,8 +3165,8 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	if ((bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
 	    (!alloc_profile_is_valid(bctl->data.target, 1) ||
 	     (bctl->data.target & ~allowed))) {
-		printk(KERN_ERR "btrfs: unable to start balance with target "
-		       "data profile %llu\n",
+		btrfs_err(fs_info, "unable to start balance with target "
+			   "data profile %llu",
 		       bctl->data.target);
 		ret = -EINVAL;
 		goto out;
@@ -3174,8 +3174,8 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	if ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
 	    (!alloc_profile_is_valid(bctl->meta.target, 1) ||
 	     (bctl->meta.target & ~allowed))) {
-		printk(KERN_ERR "btrfs: unable to start balance with target "
-		       "metadata profile %llu\n",
+		btrfs_err(fs_info,
+			   "unable to start balance with target metadata profile %llu",
 		       bctl->meta.target);
 		ret = -EINVAL;
 		goto out;
@@ -3183,8 +3183,8 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	if ((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
 	    (!alloc_profile_is_valid(bctl->sys.target, 1) ||
 	     (bctl->sys.target & ~allowed))) {
-		printk(KERN_ERR "btrfs: unable to start balance with target "
-		       "system profile %llu\n",
+		btrfs_err(fs_info,
+			   "unable to start balance with target system profile %llu",
 		       bctl->sys.target);
 		ret = -EINVAL;
 		goto out;
@@ -3193,7 +3193,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	/* allow dup'ed data chunks only in mixed mode */
 	if (!mixed && (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
 	    (bctl->data.target & BTRFS_BLOCK_GROUP_DUP)) {
-		printk(KERN_ERR "btrfs: dup for data is not allowed\n");
+		btrfs_err(fs_info, "dup for data is not allowed");
 		ret = -EINVAL;
 		goto out;
 	}
@@ -3213,11 +3213,10 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 		     (fs_info->avail_metadata_alloc_bits & allowed) &&
 		     !(bctl->meta.target & allowed))) {
 			if (bctl->flags & BTRFS_BALANCE_FORCE) {
-				printk(KERN_INFO "btrfs: force reducing metadata "
-				       "integrity\n");
+				btrfs_info(fs_info, "force reducing metadata integrity");
 			} else {
-				printk(KERN_ERR "btrfs: balance will reduce metadata "
-				       "integrity, use force if you want this\n");
+				btrfs_err(fs_info, "balance will reduce metadata "
+					   "integrity, use force if you want this");
 				ret = -EINVAL;
 				goto out;
 			}
@@ -3303,7 +3302,7 @@ static int balance_kthread(void *data)
 	mutex_lock(&fs_info->balance_mutex);
 
 	if (fs_info->balance_ctl) {
-		printk(KERN_INFO "btrfs: continuing balance\n");
+		btrfs_info(fs_info, "continuing balance");
 		ret = btrfs_balance(fs_info->balance_ctl, NULL);
 	}
 
@@ -3325,7 +3324,7 @@ int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info)
 	spin_unlock(&fs_info->balance_lock);
 
 	if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) {
-		printk(KERN_INFO "btrfs: force skipping balance\n");
+		btrfs_info(fs_info, "force skipping balance");
 		return 0;
 	}
 
@@ -3543,7 +3542,7 @@ update_tree:
 						  BTRFS_UUID_KEY_SUBVOL,
 						  key.objectid);
 			if (ret < 0) {
-				pr_warn("btrfs: uuid_tree_add failed %d\n",
+				btrfs_warn(fs_info, "uuid_tree_add failed %d",
 					ret);
 				break;
 			}
@@ -3555,7 +3554,7 @@ update_tree:
 						 BTRFS_UUID_KEY_RECEIVED_SUBVOL,
 						  key.objectid);
 			if (ret < 0) {
-				pr_warn("btrfs: uuid_tree_add failed %d\n",
+				btrfs_warn(fs_info, "uuid_tree_add failed %d",
 					ret);
 				break;
 			}
@@ -3590,7 +3589,7 @@ out:
 	if (trans && !IS_ERR(trans))
 		btrfs_end_transaction(trans, fs_info->uuid_root);
 	if (ret)
-		pr_warn("btrfs: btrfs_uuid_scan_kthread failed %d\n", ret);
+		btrfs_warn(fs_info, "btrfs_uuid_scan_kthread failed %d", ret);
 	else
 		fs_info->update_uuid_tree_gen = 1;
 	up(&fs_info->uuid_tree_rescan_sem);
@@ -3654,7 +3653,7 @@ static int btrfs_uuid_rescan_kthread(void *data)
 	 */
 	ret = btrfs_uuid_tree_iterate(fs_info, btrfs_check_uuid_tree_entry);
 	if (ret < 0) {
-		pr_warn("btrfs: iterating uuid_tree failed %d\n", ret);
+		btrfs_warn(fs_info, "iterating uuid_tree failed %d", ret);
 		up(&fs_info->uuid_tree_rescan_sem);
 		return ret;
 	}
@@ -3695,7 +3694,7 @@ int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info)
 	task = kthread_run(btrfs_uuid_scan_kthread, fs_info, "btrfs-uuid");
 	if (IS_ERR(task)) {
 		/* fs_info->update_uuid_tree_gen remains 0 in all error case */
-		pr_warn("btrfs: failed to start uuid_scan task\n");
+		btrfs_warn(fs_info, "failed to start uuid_scan task");
 		up(&fs_info->uuid_tree_rescan_sem);
 		return PTR_ERR(task);
 	}
@@ -3711,7 +3710,7 @@ int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info)
 	task = kthread_run(btrfs_uuid_rescan_kthread, fs_info, "btrfs-uuid");
 	if (IS_ERR(task)) {
 		/* fs_info->update_uuid_tree_gen remains 0 in all error case */
-		pr_warn("btrfs: failed to start uuid_rescan task\n");
+		btrfs_warn(fs_info, "failed to start uuid_rescan task");
 		up(&fs_info->uuid_tree_rescan_sem);
 		return PTR_ERR(task);
 	}
@@ -4033,7 +4032,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 		max_stripe_size = 32 * 1024 * 1024;
 		max_chunk_size = 2 * max_stripe_size;
 	} else {
-		printk(KERN_ERR "btrfs: invalid chunk type 0x%llx requested\n",
+		btrfs_err(info, "invalid chunk type 0x%llx requested\n",
 		       type);
 		BUG_ON(1);
 	}
@@ -4065,7 +4064,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 
 		if (!device->writeable) {
 			WARN(1, KERN_ERR
-			       "btrfs: read-only device in alloc_list\n");
+			       "BTRFS: read-only device in alloc_list\n");
 			continue;
 		}
 
@@ -5193,13 +5192,13 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
 	read_unlock(&em_tree->lock);
 
 	if (!em) {
-		printk(KERN_ERR "btrfs: couldn't find em for chunk %Lu\n",
+		printk(KERN_ERR "BTRFS: couldn't find em for chunk %Lu\n",
 		       chunk_start);
 		return -EIO;
 	}
 
 	if (em->start != chunk_start) {
-		printk(KERN_ERR "btrfs: bad chunk start, em=%Lu, wanted=%Lu\n",
+		printk(KERN_ERR "BTRFS: bad chunk start, em=%Lu, wanted=%Lu\n",
 		       em->start, chunk_start);
 		free_extent_map(em);
 		return -EIO;
@@ -6130,7 +6129,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 	BUG_ON(!path);
 	ret = btrfs_search_slot(trans, dev_root, &key, path, -1, 1);
 	if (ret < 0) {
-		printk_in_rcu(KERN_WARNING "btrfs: error %d while searching for dev_stats item for device %s!\n",
+		printk_in_rcu(KERN_WARNING "BTRFS: "
+			"error %d while searching for dev_stats item for device %s!\n",
 			      ret, rcu_str_deref(device->name));
 		goto out;
 	}
@@ -6140,7 +6140,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 		/* need to delete old one and insert a new one */
 		ret = btrfs_del_item(trans, dev_root, path);
 		if (ret != 0) {
-			printk_in_rcu(KERN_WARNING "btrfs: delete too small dev_stats item for device %s failed %d!\n",
+			printk_in_rcu(KERN_WARNING "BTRFS: "
+				"delete too small dev_stats item for device %s failed %d!\n",
 				      rcu_str_deref(device->name), ret);
 			goto out;
 		}
@@ -6153,7 +6154,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 		ret = btrfs_insert_empty_item(trans, dev_root, path,
 					      &key, sizeof(*ptr));
 		if (ret < 0) {
-			printk_in_rcu(KERN_WARNING "btrfs: insert dev_stats item for device %s failed %d!\n",
+			printk_in_rcu(KERN_WARNING "BTRFS: "
+					  "insert dev_stats item for device %s failed %d!\n",
 				      rcu_str_deref(device->name), ret);
 			goto out;
 		}
@@ -6206,16 +6208,14 @@ static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev)
 {
 	if (!dev->dev_stats_valid)
 		return;
-	printk_ratelimited_in_rcu(KERN_ERR
-			   "btrfs: bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
+	printk_ratelimited_in_rcu(KERN_ERR "BTRFS: "
+			   "bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
 			   rcu_str_deref(dev->name),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_FLUSH_ERRS),
-			   btrfs_dev_stat_read(dev,
-					       BTRFS_DEV_STAT_CORRUPTION_ERRS),
-			   btrfs_dev_stat_read(dev,
-					       BTRFS_DEV_STAT_GENERATION_ERRS));
+			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_CORRUPTION_ERRS),
+			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_GENERATION_ERRS));
 }
 
 static void btrfs_dev_stat_print_on_load(struct btrfs_device *dev)
@@ -6228,7 +6228,8 @@ static void btrfs_dev_stat_print_on_load(struct btrfs_device *dev)
 	if (i == BTRFS_DEV_STAT_VALUES_MAX)
 		return; /* all values == 0, suppress message */
 
-	printk_in_rcu(KERN_INFO "btrfs: bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
+	printk_in_rcu(KERN_INFO "BTRFS: "
+		   "bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
 	       rcu_str_deref(dev->name),
 	       btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
 	       btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
@@ -6249,12 +6250,10 @@ int btrfs_get_dev_stats(struct btrfs_root *root,
 	mutex_unlock(&fs_devices->device_list_mutex);
 
 	if (!dev) {
-		printk(KERN_WARNING
-		       "btrfs: get dev_stats failed, device not found\n");
+		btrfs_warn(root->fs_info, "get dev_stats failed, device not found");
 		return -ENODEV;
 	} else if (!dev->dev_stats_valid) {
-		printk(KERN_WARNING
-		       "btrfs: get dev_stats failed, not yet valid\n");
+		btrfs_warn(root->fs_info, "get dev_stats failed, not yet valid");
 		return -ENODEV;
 	} else if (stats->flags & BTRFS_DEV_STATS_RESET) {
 		for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++) {

fs/btrfs/xattr.c

@@ -28,6 +28,7 @@
 #include "transaction.h"
 #include "xattr.h"
 #include "disk-io.h"
+#include "props.h"
 
 
 ssize_t __btrfs_getxattr(struct inode *inode, const char *name,
@@ -332,7 +333,8 @@ static bool btrfs_is_valid_xattr(const char *name)
 			XATTR_SECURITY_PREFIX_LEN) ||
 	       !strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN) ||
 	       !strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN) ||
-	       !strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN);
+	       !strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN) ||
+		!strncmp(name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN);
 }
 
 ssize_t btrfs_getxattr(struct dentry *dentry, const char *name,
@@ -374,6 +376,10 @@ int btrfs_setxattr(struct dentry *dentry, const char *name, const void *value,
 	if (!btrfs_is_valid_xattr(name))
 		return -EOPNOTSUPP;
 
+	if (!strncmp(name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN))
+		return btrfs_set_prop(dentry->d_inode, name,
+				      value, size, flags);
+
 	if (size == 0)
 		value = "";  /* empty EA, do not remove */
 
@@ -403,6 +409,10 @@ int btrfs_removexattr(struct dentry *dentry, const char *name)
 	if (!btrfs_is_valid_xattr(name))
 		return -EOPNOTSUPP;
 
+	if (!strncmp(name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN))
+		return btrfs_set_prop(dentry->d_inode, name,
+				      NULL, 0, XATTR_REPLACE);
+
 	return __btrfs_setxattr(NULL, dentry->d_inode, name, NULL, 0,
 				XATTR_REPLACE);
 }

fs/btrfs/zlib.c

@@ -97,7 +97,7 @@ static int zlib_compress_pages(struct list_head *ws,
 	*total_in = 0;
 
 	if (Z_OK != zlib_deflateInit(&workspace->def_strm, 3)) {
-		printk(KERN_WARNING "btrfs: deflateInit failed\n");
+		printk(KERN_WARNING "BTRFS: deflateInit failed\n");
 		ret = -1;
 		goto out;
 	}
@@ -125,7 +125,7 @@ static int zlib_compress_pages(struct list_head *ws,
 	while (workspace->def_strm.total_in < len) {
 		ret = zlib_deflate(&workspace->def_strm, Z_SYNC_FLUSH);
 		if (ret != Z_OK) {
-			printk(KERN_DEBUG "btrfs: deflate in loop returned %d\n",
+			printk(KERN_DEBUG "BTRFS: deflate in loop returned %d\n",
 			       ret);
 			zlib_deflateEnd(&workspace->def_strm);
 			ret = -1;
@@ -252,7 +252,7 @@ static int zlib_decompress_biovec(struct list_head *ws, struct page **pages_in,
 	}
 
 	if (Z_OK != zlib_inflateInit2(&workspace->inf_strm, wbits)) {
-		printk(KERN_WARNING "btrfs: inflateInit failed\n");
+		printk(KERN_WARNING "BTRFS: inflateInit failed\n");
 		return -1;
 	}
 	while (workspace->inf_strm.total_in < srclen) {
@@ -336,7 +336,7 @@ static int zlib_decompress(struct list_head *ws, unsigned char *data_in,
 	}
 
 	if (Z_OK != zlib_inflateInit2(&workspace->inf_strm, wbits)) {
-		printk(KERN_WARNING "btrfs: inflateInit failed\n");
+		printk(KERN_WARNING "BTRFS: inflateInit failed\n");
 		return -1;
 	}
 

include/linux/rwsem.h

@@ -74,6 +74,17 @@ do {								\
 	__init_rwsem((sem), #sem, &__key);			\
 } while (0)
 
+/*
+ * This is the same regardless of which rwsem implementation that is being used.
+ * It is just a heuristic meant to be called by somebody alreadying holding the
+ * rwsem to see if somebody from an incompatible type is wanting access to the
+ * lock.
+ */
+static inline int rwsem_is_contended(struct rw_semaphore *sem)
+{
+	return !list_empty(&sem->wait_list);
+}
+
 /*
  * lock for reading
  */

include/trace/events/btrfs.h

@@ -208,17 +208,18 @@ TRACE_EVENT_CONDITION(btrfs_get_extent,
 		  __entry->refs, __entry->compress_type)
 );
 
-#define show_ordered_flags(flags)					\
-	__print_symbolic(flags,						\
-		{ BTRFS_ORDERED_IO_DONE, 	"IO_DONE" 	},	\
-		{ BTRFS_ORDERED_COMPLETE, 	"COMPLETE" 	},	\
-		{ BTRFS_ORDERED_NOCOW, 		"NOCOW" 	},	\
-		{ BTRFS_ORDERED_COMPRESSED, 	"COMPRESSED" 	},	\
-		{ BTRFS_ORDERED_PREALLOC, 	"PREALLOC" 	},	\
-		{ BTRFS_ORDERED_DIRECT, 	"DIRECT" 	},	\
-		{ BTRFS_ORDERED_IOERR, 		"IOERR" 	},	\
-		{ BTRFS_ORDERED_UPDATED_ISIZE, 	"UPDATED_ISIZE"	},	\
-		{ BTRFS_ORDERED_LOGGED_CSUM, 	"LOGGED_CSUM"	})
+#define show_ordered_flags(flags)					   \
+	__print_flags(flags, "|",					   \
+		{ (1 << BTRFS_ORDERED_IO_DONE), 	"IO_DONE" 	}, \
+		{ (1 << BTRFS_ORDERED_COMPLETE), 	"COMPLETE" 	}, \
+		{ (1 << BTRFS_ORDERED_NOCOW), 		"NOCOW" 	}, \
+		{ (1 << BTRFS_ORDERED_COMPRESSED), 	"COMPRESSED" 	}, \
+		{ (1 << BTRFS_ORDERED_PREALLOC), 	"PREALLOC" 	}, \
+		{ (1 << BTRFS_ORDERED_DIRECT),	 	"DIRECT" 	}, \
+		{ (1 << BTRFS_ORDERED_IOERR), 		"IOERR" 	}, \
+		{ (1 << BTRFS_ORDERED_UPDATED_ISIZE), 	"UPDATED_ISIZE"	}, \
+		{ (1 << BTRFS_ORDERED_LOGGED_CSUM), 	"LOGGED_CSUM"	}, \
+		{ (1 << BTRFS_ORDERED_TRUNCATED), 	"TRUNCATED"	})
 
 
 DECLARE_EVENT_CLASS(btrfs__ordered_extent,

include/uapi/linux/btrfs.h

@@ -184,6 +184,12 @@ struct btrfs_ioctl_fs_info_args {
 	__u64 reserved[124];			/* pad to 1k */
 };
 
+struct btrfs_ioctl_feature_flags {
+	__u64 compat_flags;
+	__u64 compat_ro_flags;
+	__u64 incompat_flags;
+};
+
 /* balance control ioctl modes */
 #define BTRFS_BALANCE_CTL_PAUSE		1
 #define BTRFS_BALANCE_CTL_CANCEL	2
@@ -552,6 +558,7 @@ static inline char *btrfs_err_str(enum btrfs_err_code err_code)
 #define BTRFS_IOC_DEFAULT_SUBVOL _IOW(BTRFS_IOCTL_MAGIC, 19, __u64)
 #define BTRFS_IOC_SPACE_INFO _IOWR(BTRFS_IOCTL_MAGIC, 20, \
 				    struct btrfs_ioctl_space_args)
+#define BTRFS_IOC_GLOBAL_RSV _IOR(BTRFS_IOCTL_MAGIC, 20, __u64)
 #define BTRFS_IOC_START_SYNC _IOR(BTRFS_IOCTL_MAGIC, 24, __u64)
 #define BTRFS_IOC_WAIT_SYNC  _IOW(BTRFS_IOCTL_MAGIC, 22, __u64)
 #define BTRFS_IOC_SNAP_CREATE_V2 _IOW(BTRFS_IOCTL_MAGIC, 23, \
@@ -606,5 +613,11 @@ static inline char *btrfs_err_str(enum btrfs_err_code err_code)
 				    struct btrfs_ioctl_dev_replace_args)
 #define BTRFS_IOC_FILE_EXTENT_SAME _IOWR(BTRFS_IOCTL_MAGIC, 54, \
 					 struct btrfs_ioctl_same_args)
+#define BTRFS_IOC_GET_FEATURES _IOR(BTRFS_IOCTL_MAGIC, 57, \
+				   struct btrfs_ioctl_feature_flags)
+#define BTRFS_IOC_SET_FEATURES _IOW(BTRFS_IOCTL_MAGIC, 57, \
+				   struct btrfs_ioctl_feature_flags[2])
+#define BTRFS_IOC_GET_SUPPORTED_FEATURES _IOR(BTRFS_IOCTL_MAGIC, 57, \
+				   struct btrfs_ioctl_feature_flags[3])
 
 #endif /* _UAPI_LINUX_BTRFS_H */

include/uapi/linux/xattr.h

@@ -20,6 +20,9 @@
 #define XATTR_MAC_OSX_PREFIX "osx."
 #define XATTR_MAC_OSX_PREFIX_LEN (sizeof(XATTR_MAC_OSX_PREFIX) - 1)
 
+#define XATTR_BTRFS_PREFIX "btrfs."
+#define XATTR_BTRFS_PREFIX_LEN (sizeof(XATTR_BTRFS_PREFIX) - 1)
+
 #define XATTR_SECURITY_PREFIX	"security."
 #define XATTR_SECURITY_PREFIX_LEN (sizeof(XATTR_SECURITY_PREFIX) - 1)
 

lib/kobject.c

@@ -779,6 +779,7 @@ const struct sysfs_ops kobj_sysfs_ops = {
 	.show	= kobj_attr_show,
 	.store	= kobj_attr_store,
 };
+EXPORT_SYMBOL_GPL(kobj_sysfs_ops);
 
 /**
  * kset_register - initialize and add a kset.