Btrfs: introduce a tree for UUID to subvol ID mapping

Mapping UUIDs to subvolume IDs is an operation with a high effort today. Today, the algorithm even has quadratic effort (based on the number of existing subvolumes), which means, that it takes minutes to send/receive a single subvolume if 10,000 subvolumes exist. But even linear effort would be too much since it is a waste. And these data structures to allow mapping UUIDs to subvolume IDs are created every time a btrfs send/receive instance is started.

It is much more efficient to maintain a searchable persistent data structure in the filesystem, one that is updated whenever a subvolume/snapshot is created and deleted, and when the received subvolume UUID is set by the btrfs-receive tool.

Therefore kernel code is added with this commit that is able to maintain data structures in the filesystem that allow to quickly search for a given UUID and to retrieve data that is assigned to this UUID, like which subvolume ID is related to this UUID.

This commit adds a new tree to hold UUID-to-data mapping items. The key of the items is the full UUID plus the key type BTRFS_UUID_KEY. Multiple data blocks can be stored for a given UUID, a type/length/ value scheme is used.

Now follows the lengthy justification, why a new tree was added instead of using the existing root tree:

The first approach was to not create another tree that holds UUID items. Instead, the items should just go into the top root tree. Unfortunately this confused the algorithm to assign the objectid of subvolumes and snapshots. The reason is that btrfs_find_free_objectid() calls btrfs_find_highest_objectid() for the first created subvol or snapshot after mounting a filesystem, and this function simply searches for the largest used objectid in the root tree keys to pick the next objectid to assign. Of course, the UUID keys have always been the ones with the highest offset value, and the next assigned subvol ID was wastefully huge.

To use any other existing tree did not look proper. To apply a workaround such as setting the objectid to zero in the UUID item key and to implement collision handling would either add limitations (in case of a btrfs_extend_item() approach to handle the collisions) or a lot of complexity and source code (in case a key would be looked up that is free of collisions). Adding new code that introduces limitations is not good, and adding code that is complex and lengthy for no good reason is also not good. That's the justification why a completely new tree was introduced.

Signed-off-by Stefan Behrens <[email protected]>

---
 fs/btrfs/Makefile    |   3 +-
 fs/btrfs/ctree.h     |  50 ++++++
 fs/btrfs/uuid-tree.c | 481 +++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 533 insertions(+), 1 deletion(-)

fs/btrfs/Makefile +2 -1

index 3932224..a550dfc 100644
@@ -8,7 +8,8 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \
 	   extent_io.o volumes.o async-thread.o ioctl.o locking.o orphan.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
+	   reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o \
+	   uuid-tree.o
 
 btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o
 btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o

fs/btrfs/ctree.h +3 -0

index f415377..ca9149a 100644
@@ -91,6 +91,9 @@ struct btrfs_ordered_sum;
 /* holds quota configuration and tracking */
 #define BTRFS_QUOTA_TREE_OBJECTID 8ULL
 
+/* for storing items that use the BTRFS_UUID_KEY */
+#define BTRFS_UUID_TREE_OBJECTID 9ULL
+
 /* for storing balance parameters in the root tree */
 #define BTRFS_BALANCE_OBJECTID -4ULL
 

@@ -953,6 +956,18 @@ struct btrfs_dev_replace_item {
 	__le64 num_uncorrectable_read_errors;
 } __attribute__ ((__packed__));

fs/btrfs/ctree.h +38 -0

+	__le64 subid[0];	/* sequence of subids */
+} __attribute__ ((__packed__));
+
 /* different types of block groups (and chunks) */
 #define BTRFS_BLOCK_GROUP_DATA		(1ULL << 0)
 #define BTRFS_BLOCK_GROUP_SYSTEM	(1ULL << 1)

@@ -1901,6 +1916,17 @@ struct btrfs_ioctl_defrag_range_args {
 #define BTRFS_DEV_REPLACE_KEY	250
 
 /*
+ * Stores items that allow to quickly map UUIDs to something else.
+ * These items are part of the filesystem UUID tree.
+ * The key is built like this:
+ * (UUID_upper_64_bits, BTRFS_UUID_KEY, UUID_lower_64_bits).
+ */
+#if BTRFS_UUID_SIZE != 16
+#error "UUID items require BTRFS_UUID_SIZE == 16!"
+#endif
+#define BTRFS_UUID_KEY	251
+
+/*
  * string items are for debugging.  They just store a short string of
  * data in the FS
  */

@@ -2956,6 +2982,12 @@ BTRFS_SETGET_FUNCS(dev_replace_cursor_left, struct btrfs_dev_replace_item,
 BTRFS_SETGET_FUNCS(dev_replace_cursor_right, struct btrfs_dev_replace_item,
 		   cursor_right, 64);
 
+/* btrfs_uuid_item */
+BTRFS_SETGET_FUNCS(uuid_type, struct btrfs_uuid_item, type, 64);
+BTRFS_SETGET_FUNCS(uuid_len, struct btrfs_uuid_item, len, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_uuid_type, struct btrfs_uuid_item, type, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_uuid_len, struct btrfs_uuid_item, len, 64);
+
 BTRFS_SETGET_STACK_FUNCS(stack_dev_replace_src_devid,
 			 struct btrfs_dev_replace_item, src_devid, 64);
 BTRFS_SETGET_STACK_FUNCS(stack_dev_replace_cont_reading_from_srcdev_mode,

@@ -3372,6 +3404,24 @@ void btrfs_check_and_init_root_item(struct btrfs_root_item *item);
 void btrfs_update_root_times(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root);
 
+/* uuid-tree.c */
+int btrfs_lookup_uuid_subvol_item(struct btrfs_root *uuid_root, u8 *uuid,
+				  u64 *subvol_id);
+int btrfs_insert_uuid_subvol_item(struct btrfs_trans_handle *trans,
+				  struct btrfs_root *uuid_root, u8 *uuid,
+				  u64 subvol_id);
+int btrfs_del_uuid_subvol_item(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *uuid_root, u8 *uuid,
+			       u64 subvol_id);
+int btrfs_lookup_uuid_received_subvol_item(struct btrfs_root *uuid_root,
+					   u8 *uuid, u64 *subvol_id);
+int btrfs_insert_uuid_received_subvol_item(struct btrfs_trans_handle *trans,
+					   struct btrfs_root *uuid_root,
+					   u8 *uuid, u64 subvol_id);
+int btrfs_del_uuid_received_subvol_item(struct btrfs_trans_handle *trans,
+					struct btrfs_root *uuid_root, u8 *uuid,
+					u64 subvol_id);
+
 /* dir-item.c */
 int btrfs_check_dir_item_collision(struct btrfs_root *root, u64 dir,
 			  const char *name, int name_len);

fs/btrfs/uuid-tree.c new file +481 -0

new file mode 100644
index 0000000..a44e13f
@@ -0,0 +1,481 @@
+/*
+ * Copyright (C) STRATO AG 2013.  All rights reserved.
+ *
+ * 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/uuid.h>
+#include <asm/unaligned.h>
+#include "ctree.h"
+#include "transaction.h"
+#include "disk-io.h"
+#include "print-tree.h"
+
+
+/*
+ * One key is used to store a sequence of btrfs_uuid_item items.
+ * Each item in the sequence contains a type information and a sequence of
+ * ids (together with the information about the size of the sequence of ids).
+ * {[btrfs_uuid_item type0 {id0, id1, ..., idN}],
+ *  ...,
+ *  [btrfs_uuid_item typeZ {id0, id1, ..., idN}]}
+ *
+ * It is forbidden to put multiple items with the same type under the same key.
+ * Instead the sequence of ids is extended and used to store any additional
+ * ids for the same item type.
+ */
+
+static void btrfs_uuid_to_key(u8 *uuid, struct btrfs_key *key)
+{
+	key->type = BTRFS_UUID_KEY;
+	key->objectid = get_unaligned_le64(uuid);
+	key->offset = get_unaligned_le64(uuid + sizeof(u64));
+}
+
+static struct btrfs_uuid_item *btrfs_match_uuid_item_type(
+		struct btrfs_path *path, u64 type)
+{
+	struct extent_buffer *eb;
+	int slot;
+	struct btrfs_uuid_item *ptr;
+	u32 item_size;
+
+	eb = path->nodes[0];
+	slot = path->slots[0];
+	ptr = btrfs_item_ptr(eb, slot, struct btrfs_uuid_item);
+	item_size = btrfs_item_size_nr(eb, slot);
+	do {
+		u64 sub_item_type;
+		u64 sub_item_len;
+
+		if (item_size < sizeof(*ptr)) {
+			pr_warn("btrfs: uuid item too short (%llu < %d)!\n",
+				(unsigned long long)item_size,
+				(int)sizeof(*ptr));
+			return NULL;
+		}
+		item_size -= sizeof(*ptr);
+		sub_item_type = btrfs_uuid_type(eb, ptr);
+		sub_item_len = btrfs_uuid_len(eb, ptr);
+		if (sub_item_len * sizeof(u64) > item_size) {
+			pr_warn("btrfs: uuid item too short (%llu > %llu)!\n",
+				(unsigned long long)(sub_item_len *
+						     sizeof(u64)),
+				(unsigned long long)item_size);
+			return NULL;
+		}
+		if (sub_item_type == type)
+			return ptr;
+		item_size -= sub_item_len * sizeof(u64);
+		ptr = 1 + (struct btrfs_uuid_item *)
+			(((char *)ptr) + (sub_item_len * sizeof(u64)));
+	} while (item_size);
+
+	return NULL;
+}
+
+static int btrfs_uuid_tree_lookup_prepare(struct btrfs_root *uuid_root,
+					  u8 *uuid, u64 type,
+					  struct btrfs_path *path,
+					  struct btrfs_uuid_item **ptr)
+{
+	int ret;
+	struct btrfs_key key;
+
+	if (!uuid_root) {
+		WARN_ON_ONCE(1);
+		ret = -ENOENT;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, &key);
+
+	ret = btrfs_search_slot(NULL, uuid_root, &key, path, 0, 0);
+	if (ret < 0)
+		goto out;
+	if (ret > 0) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	*ptr = btrfs_match_uuid_item_type(path, type);
+	if (!*ptr) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	ret = 0;
+
+out:
+	return ret;
+}
+
+/* return -ENOENT for !found, < 0 for errors, or 0 if an item was found */
+static int btrfs_uuid_tree_lookup(struct btrfs_root *uuid_root, u8 *uuid,
+				  u64 type, u64 subid)
+{
+	int ret;
+	struct btrfs_path *path;
+	struct extent_buffer *eb;
+	struct btrfs_uuid_item *ptr;
+	u64 sub_item_len;
+	unsigned long offset;
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = btrfs_uuid_tree_lookup_prepare(uuid_root, uuid, type, path, &ptr);
+	if (ret < 0)
+		goto out;
+
+	eb = path->nodes[0];
+	sub_item_len = btrfs_uuid_len(eb, ptr);
+	WARN_ON_ONCE(sub_item_len == 0);
+	ret = -ENOENT;
+	ptr++;
+	offset = (unsigned long)ptr;
+	while (sub_item_len > 0) {
+		u64 data;
+
+		read_extent_buffer(eb, &data, offset, sizeof(data));
+		data = le64_to_cpu(data);
+		if (data == subid) {
+			ret = 0;
+			break;
+		}
+		offset += sizeof(data);
+		sub_item_len--;
+	}
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+/* return -ENOENT for !found, < 0 for errors, or 0 if an item was found */
+static int btrfs_uuid_tree_lookup_any(struct btrfs_root *uuid_root, u8 *uuid,
+				      u64 type, u64 *subid)
+{
+	int ret;
+	struct btrfs_path *path;
+	struct extent_buffer *eb;
+	struct btrfs_uuid_item *ptr;
+	u64 sub_item_len;
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = btrfs_uuid_tree_lookup_prepare(uuid_root, uuid, type, path, &ptr);
+	if (ret < 0)
+		goto out;
+
+	eb = path->nodes[0];
+	sub_item_len = btrfs_uuid_len(eb, ptr);
+	WARN_ON_ONCE(sub_item_len == 0);
+	if (sub_item_len > 0) {
+		/* return first stored id */
+		read_extent_buffer(eb, subid, (unsigned long)(ptr + 1),
+				   sizeof(*subid));
+		*subid = le64_to_cpu(*subid);
+		ret = 0;
+	} else {
+		ret = -ENOENT;
+	}
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+/* it is not checked whether the entry to add already exists */
+static int btrfs_uuid_tree_add(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *uuid_root, u8 *uuid,
+			       u64 type, u64 subid)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct btrfs_key key;
+	struct extent_buffer *eb;
+	int slot;
+	struct btrfs_uuid_item *ptr;
+	u32 item_size;
+
+	if (!uuid_root) {
+		WARN_ON_ONCE(1);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, &key);
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = btrfs_insert_empty_item(trans, uuid_root, path, &key,
+				      sizeof(*ptr) + sizeof(subid));
+	if (ret == -EEXIST) {
+		ptr = btrfs_match_uuid_item_type(path, type);
+		if (ptr) {
+			/*
+			 * An item with that type already exists.
+			 * Extend the item and store the subid at the
+			 * location of the first id of this type.
+			 */
+			unsigned long start;
+			unsigned long move_dst;
+			unsigned long move_src;
+			unsigned long move_len;
+
+			btrfs_extend_item(uuid_root, path, sizeof(subid));
+			ptr = (struct btrfs_uuid_item *)
+				(((char *)ptr) - sizeof(subid));
+			eb = path->nodes[0];
+			slot = path->slots[0];
+			item_size = btrfs_item_size_nr(eb, slot);
+			WARN_ON(btrfs_uuid_len(eb, ptr) == 0);
+			btrfs_set_uuid_len(eb, ptr,
+					   btrfs_uuid_len(eb, ptr) + 1);
+			ptr++;
+			start = btrfs_item_ptr_offset(eb, slot);
+			move_dst = ((unsigned long)ptr) + sizeof(subid);
+			move_src = (unsigned long)ptr;
+			move_len = item_size - (move_dst - start);
+			memmove_extent_buffer(eb, move_dst, move_src, move_len);
+
+			goto write_subid;
+		} else {
+			btrfs_extend_item(uuid_root, path,
+					  sizeof(*ptr) + sizeof(subid));
+		}
+	} else if (ret < 0) {
+		pr_warn("btrfs: insert uuid item failed %d (0x%016llx, 0x%016llx) type %llu!\n",
+			ret, (unsigned long long)key.objectid,
+			(unsigned long long)key.offset,
+			(unsigned long long)type);
+		goto out;
+	}
+
+	/*
+	 * Add an item for the type for the first time. Either add it behind
+	 * items with different types, or write the very first item.
+	 */
+	eb = path->nodes[0];
+	slot = path->slots[0];
+	ptr = btrfs_item_ptr(eb, slot, struct btrfs_uuid_item);
+	item_size = btrfs_item_size_nr(eb, slot);
+	ptr = (struct btrfs_uuid_item *)
+		(((char *)ptr) + item_size - (sizeof(*ptr) + sizeof(subid)));
+	btrfs_set_uuid_type(eb, ptr, type);
+	btrfs_set_uuid_len(eb, ptr, 1);
+	ptr++;
+
+write_subid:
+	ret = 0;
+	subid = cpu_to_le64(subid);
+	write_extent_buffer(eb, &subid, (unsigned long)ptr, sizeof(subid));
+	btrfs_mark_buffer_dirty(eb);
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+static int btrfs_uuid_tree_rem(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *uuid_root, u8 *uuid,
+			       u64 type, u64 subid)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct btrfs_key key;
+	struct extent_buffer *eb;
+	int slot;
+	struct btrfs_uuid_item *ptr_start;
+	unsigned long offset;
+	u32 item_size;
+	u64 id_index;
+	unsigned long start;
+	unsigned long move_dst;
+	unsigned long move_src;
+	unsigned long move_len;
+
+	if (!uuid_root) {
+		WARN_ON_ONCE(1);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, &key);
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	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",
+			ret);
+		goto out;
+	}
+	if (ret > 0) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	ptr_start = btrfs_match_uuid_item_type(path, type);
+	if (!ptr_start) {
+		/* case 1: no entry for that type is found */
+		ret = -ENOENT;
+		goto out;
+	}
+
+	eb = path->nodes[0];
+	slot = path->slots[0];
+	start = btrfs_item_ptr_offset(eb, slot);
+	item_size = btrfs_item_size_nr(eb, slot);
+	id_index = btrfs_uuid_len(eb, ptr_start);
+	offset = (unsigned long)(ptr_start + 1);
+	while (id_index > 0) {
+		u64 found_subid;
+
+		read_extent_buffer(eb, &found_subid, offset,
+				   sizeof(found_subid));
+		found_subid = le64_to_cpu(found_subid);
+		if (found_subid == subid)
+			break;
+		offset += sizeof(found_subid);
+		id_index--;
+	}
+
+	if (!id_index) {
+		/*
+		 * case 2: an entry for that type is found, but no match
+		 * for the id
+		 */
+		ret = -ENOENT;
+		goto out;
+	}
+
+	if (btrfs_uuid_len(eb, ptr_start) == 1) {
+		if (item_size == sizeof(*ptr_start) + sizeof(subid)) {
+			/*
+			 * case 3: no other types and ids are stored,
+			 * delete the full item
+			 */
+			ret = btrfs_del_item(trans, uuid_root, path);
+			goto out;
+		} else {
+			/*
+			 * case 4: No other ids for the particular type
+			 * are stored, but other types exist. Shrink the
+			 * item by the size of an id plus the size of the
+			 * item for that type.
+			 */
+			move_dst = (unsigned long)ptr_start;
+		}
+	} else {
+		/*
+		 * case 5: Other ids for that particular type are stored.
+		 * Shrink the item by the size of an id.
+		 */
+		move_dst = offset;
+		WARN_ON(btrfs_uuid_len(eb, ptr_start) == 0);
+		btrfs_set_uuid_len(eb, ptr_start,
+				   btrfs_uuid_len(eb, ptr_start) - 1);
+	}
+	move_src = offset + sizeof(subid);
+	move_len = item_size - (move_src - start);
+
+	memmove_extent_buffer(eb, move_dst, move_src, move_len);
+	btrfs_truncate_item(uuid_root, path, item_size - (move_src - move_dst),
+			    1);
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+int btrfs_lookup_uuid_subvol_item(struct btrfs_root *uuid_root, u8 *uuid,
+				  u64 *subvol_id)
+{
+	return btrfs_uuid_tree_lookup_any(uuid_root, uuid,
+					  BTRFS_UUID_ITEM_TYPE_SUBVOL,
+					  subvol_id);
+}
+
+int btrfs_insert_uuid_subvol_item(struct btrfs_trans_handle *trans,
+				  struct btrfs_root *uuid_root, u8 *uuid,
+				  u64 subvol_id)
+{
+	int ret;
+
+	ret = btrfs_uuid_tree_lookup(uuid_root, uuid,
+				     BTRFS_UUID_ITEM_TYPE_SUBVOL, subvol_id);
+	if (ret == -ENOENT)
+		ret = btrfs_uuid_tree_add(trans, uuid_root, uuid,
+					  BTRFS_UUID_ITEM_TYPE_SUBVOL,
+					  subvol_id);
+	return ret;
+}
+
+int btrfs_del_uuid_subvol_item(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *uuid_root, u8 *uuid,
+			       u64 subvol_id)
+{
+	return btrfs_uuid_tree_rem(trans, uuid_root, uuid,
+				   BTRFS_UUID_ITEM_TYPE_SUBVOL, subvol_id);
+}
+
+int btrfs_lookup_uuid_received_subvol_item(struct btrfs_root *uuid_root,
+					   u8 *uuid, u64 *subvol_id)
+{
+	return btrfs_uuid_tree_lookup_any(uuid_root, uuid,
+					  BTRFS_UUID_ITEM_TYPE_RECEIVED_SUBVOL,
+					  subvol_id);
+}
+
+int btrfs_insert_uuid_received_subvol_item(struct btrfs_trans_handle *trans,
+					   struct btrfs_root *uuid_root,
+					   u8 *uuid, u64 subvol_id)
+{
+	int ret;
+
+	ret = btrfs_uuid_tree_lookup(uuid_root, uuid,
+				     BTRFS_UUID_ITEM_TYPE_RECEIVED_SUBVOL,
+				     subvol_id);
+	if (ret == -ENOENT)
+		ret = btrfs_uuid_tree_add(trans, uuid_root, uuid,
+					  BTRFS_UUID_ITEM_TYPE_RECEIVED_SUBVOL,
+					  subvol_id);
+	return ret;
+}
+
+int btrfs_del_uuid_received_subvol_item(struct btrfs_trans_handle *trans,
+					struct btrfs_root *uuid_root, u8 *uuid,
+					u64 subvol_id)
+{
+	return btrfs_uuid_tree_rem(trans, uuid_root, uuid,
+				   BTRFS_UUID_ITEM_TYPE_RECEIVED_SUBVOL,
+				   subvol_id);
+}

Signature · 6 lines

1.8.2.2

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

This tree is not created by mkfs.btrfs. Therefore when a filesystem is mounted writable and the UUID tree does not exist, this tree is created if required. The tree is also added to the fs_info structure and initialized, but this commit does not yet read or write UUID tree elements.

Signed-off-by Stefan Behrens <[email protected]>

---
 fs/btrfs/ctree.h   |  1 +
 fs/btrfs/disk-io.c | 40 +++++++++++++++++++++++++++++++++++++++-
 fs/btrfs/volumes.c | 27 +++++++++++++++++++++++++++
 fs/btrfs/volumes.h |  1 +
 4 files changed, 68 insertions(+), 1 deletion(-)

fs/btrfs/ctree.h +1 -0

index ca9149a..6f9d760 100644
@@ -1305,6 +1305,7 @@ struct btrfs_fs_info {
 	struct btrfs_root *fs_root;
 	struct btrfs_root *csum_root;
 	struct btrfs_root *quota_root;
+	struct btrfs_root *uuid_root;
 
 	/* the log root tree is a directory of all the other log roots */
 	struct btrfs_root *log_root_tree;

fs/btrfs/disk-io.c +39 -1

index 4904d7f..3c2bf74 100644
@@ -1499,6 +1499,9 @@ struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info,
 	if (location->objectid == BTRFS_QUOTA_TREE_OBJECTID)
 		return fs_info->quota_root ? fs_info->quota_root :
 					     ERR_PTR(-ENOENT);
+	if (location->objectid == BTRFS_UUID_TREE_OBJECTID)
+		return fs_info->uuid_root ? fs_info->uuid_root :
+					    ERR_PTR(-ENOENT);
 again:
 	spin_lock(&fs_info->fs_roots_radix_lock);
 	root = radix_tree_lookup(&fs_info->fs_roots_radix,

@@ -1955,6 +1958,10 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
 		free_extent_buffer(info->quota_root->node);
 		free_extent_buffer(info->quota_root->commit_root);
 	}
+	if (info->uuid_root) {
+		free_extent_buffer(info->uuid_root->node);
+		free_extent_buffer(info->uuid_root->commit_root);
+	}
 
 	info->tree_root->node = NULL;
 	info->tree_root->commit_root = NULL;

@@ -1968,6 +1975,10 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
 		info->quota_root->node = NULL;
 		info->quota_root->commit_root = NULL;
 	}
+	if (info->uuid_root) {
+		info->uuid_root->node = NULL;
+		info->uuid_root->commit_root = NULL;
+	}
 
 	if (chunk_root) {
 		free_extent_buffer(info->chunk_root->node);

@@ -2029,11 +2040,13 @@ int open_ctree(struct super_block *sb,
 	struct btrfs_root *chunk_root;
 	struct btrfs_root *dev_root;
 	struct btrfs_root *quota_root;
+	struct btrfs_root *uuid_root;
 	struct btrfs_root *log_tree_root;
 	int ret;
 	int err = -EINVAL;
 	int num_backups_tried = 0;
 	int backup_index = 0;
+	bool create_uuid_tree = false;
 
 	tree_root = fs_info->tree_root = btrfs_alloc_root(fs_info);
 	extent_root = fs_info->extent_root = btrfs_alloc_root(fs_info);

@@ -2041,9 +2054,10 @@ int open_ctree(struct super_block *sb,
 	chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info);
 	dev_root = fs_info->dev_root = btrfs_alloc_root(fs_info);
 	quota_root = fs_info->quota_root = btrfs_alloc_root(fs_info);
+	uuid_root = fs_info->uuid_root = btrfs_alloc_root(fs_info);
 
 	if (!tree_root || !extent_root || !csum_root ||
-	    !chunk_root || !dev_root || !quota_root) {
+	    !chunk_root || !dev_root || !quota_root || !uuid_root) {
 		err = -ENOMEM;
 		goto fail;
 	}

@@ -2604,6 +2618,19 @@ retry_root_backup:
 		fs_info->pending_quota_state = 1;
 	}
 
+	ret = find_and_setup_root(tree_root, fs_info,
+				  BTRFS_UUID_TREE_OBJECTID, uuid_root);
+	if (ret) {
+		if (ret != -ENOENT)
+			goto recovery_tree_root;
+
+		kfree(uuid_root);
+		uuid_root = fs_info->uuid_root = NULL;
+		create_uuid_tree = true;
+	} else {
+		uuid_root->track_dirty = 1;
+	}
+
 	fs_info->generation = generation;
 	fs_info->last_trans_committed = generation;
 

@@ -2788,6 +2815,17 @@ retry_root_backup:
 		return ret;
 	}
 
+	if (create_uuid_tree) {
+		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",
+				ret);
+			close_ctree(tree_root);
+			return ret;
+		}
+	}
+
 	return 0;
 
 fail_qgroup:

fs/btrfs/volumes.c +14 -0

index a191bac..e6f5b88 100644
@@ -3427,6 +3427,33 @@ int btrfs_cancel_balance(struct btrfs_fs_info *fs_info)
 	return 0;
 }
 
+int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_trans_handle *trans;
+	struct btrfs_root *tree_root = fs_info->tree_root;
+	struct btrfs_root *uuid_root;
+
+	/*
+	 * 1 - root node
+	 * 1 - root item
+	 */
+	trans = btrfs_start_transaction(tree_root, 2);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+

fs/btrfs/volumes.c +3 -0

+
+	return btrfs_commit_transaction(trans, tree_root);
+}
 /*
  * shrinking a device means finding all of the device extents past
  * the new size, and then following the back refs to the chunks.

fs/btrfs/volumes.h +1 -0

index 845ccbb..3479532 100644
@@ -295,6 +295,7 @@ int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info);
 int btrfs_recover_balance(struct btrfs_fs_info *fs_info);
 int btrfs_pause_balance(struct btrfs_fs_info *fs_info);
 int btrfs_cancel_balance(struct btrfs_fs_info *fs_info);
+int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info);
 int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset);
 int find_free_dev_extent(struct btrfs_device *device, u64 num_bytes,
 			 u64 *start, u64 *max_avail);

Signature · 6 lines

1.8.2.2

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

When a new subvolume or snapshot is created, a new UUID item is added to the UUID tree. Such items are removed when the subvolume is deleted. The ioctl to set the received subvolume UUID is also touched and will now also add this received UUID into the UUID tree together with the ID of the subvolume. The latter is also done when read-only snapshots are created which inherit all the send/receive information from the parent subvolume.

User mode programs use the BTRFS_IOC_TREE_SEARCH ioctl to search and read in the UUID tree.

Signed-off-by Stefan Behrens <[email protected]>

---
 fs/btrfs/ctree.h       |  1 +
 fs/btrfs/ioctl.c       | 74 +++++++++++++++++++++++++++++++++++++++++++-------
 fs/btrfs/transaction.c | 19 ++++++++++++-
 3 files changed, 83 insertions(+), 11 deletions(-)

fs/btrfs/ctree.h +1 -0

index 6f9d760..73d797c 100644
@@ -3634,6 +3634,7 @@ extern const struct dentry_operations btrfs_dentry_operations;
 long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
 void btrfs_update_iflags(struct inode *inode);
 void btrfs_inherit_iflags(struct inode *inode, struct inode *dir);
+int btrfs_is_empty_uuid(u8 *uuid);
 int btrfs_defrag_file(struct inode *inode, struct file *file,
 		      struct btrfs_ioctl_defrag_range_args *range,
 		      u64 newer_than, unsigned long max_pages);

fs/btrfs/ioctl.c +8 -1

index b2537c7..47bef5a 100644
@@ -363,6 +363,13 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
 	return 0;
 }
 
+int btrfs_is_empty_uuid(u8 *uuid)
+{
+	static char empty_uuid[BTRFS_UUID_SIZE] = {0};
+
+	return !memcmp(uuid, empty_uuid, BTRFS_UUID_SIZE);
+}
+
 static noinline int create_subvol(struct inode *dir,
 				  struct dentry *dentry,
 				  char *name, int namelen,

@@ -396,7 +403,7 @@ static noinline int create_subvol(struct inode *dir,
 	 * of create_snapshot().
 	 */
 	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
-					       7, &qgroup_reserved);
+					       8, &qgroup_reserved);
 	if (ret)
 		return ret;
 

fs/btrfs/ioctl.c +56 -8

+	ret = btrfs_insert_uuid_subvol_item(trans, root->fs_info->uuid_root,
+					    root_item.uuid, objectid);
+	if (ret)
+		btrfs_abort_transaction(trans, root, ret);
+
 fail:
 	trans->block_rsv = NULL;
 	trans->bytes_reserved = 0;

@@ -567,9 +578,10 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	 * 1 - root item
 	 * 2 - root ref/backref
 	 * 1 - root of snapshot
+	 * 1 - UUID item
 	 */
 	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
-					&pending_snapshot->block_rsv, 7,
+					&pending_snapshot->block_rsv, 8,
 					&pending_snapshot->qgroup_reserved);
 	if (ret)
 		goto out;

@@ -2207,6 +2219,27 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 			goto out_end_trans;
 		}
 	}
+
+	ret = btrfs_del_uuid_subvol_item(trans, root->fs_info->uuid_root,
+					 dest->root_item.uuid,
+					 dest->root_key.objectid);
+	if (ret && ret != -ENOENT) {
+		btrfs_abort_transaction(trans, root, ret);
+		err = ret;
+		goto out_end_trans;
+	}
+	if (!btrfs_is_empty_uuid(dest->root_item.received_uuid)) {
+		ret = btrfs_del_uuid_received_subvol_item(trans,
+				root->fs_info->uuid_root,
+				dest->root_item.received_uuid,
+				dest->root_key.objectid);
+		if (ret && ret != -ENOENT) {
+			btrfs_abort_transaction(trans, root, ret);
+			err = ret;
+			goto out_end_trans;
+		}
+	}
+
 out_end_trans:
 	trans->block_rsv = NULL;
 	trans->bytes_reserved = 0;

@@ -2418,7 +2451,6 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 	int ret = 0;
 	char *s_uuid = NULL;
-	char empty_uuid[BTRFS_UUID_SIZE] = {0};
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;

@@ -2427,7 +2459,7 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
 	if (IS_ERR(di_args))
 		return PTR_ERR(di_args);
 
-	if (memcmp(empty_uuid, di_args->uuid, BTRFS_UUID_SIZE) != 0)
+	if (!btrfs_is_empty_uuid(di_args->uuid))
 		s_uuid = di_args->uuid;
 
 	mutex_lock(&fs_devices->device_list_mutex);

@@ -3952,6 +3984,7 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 	struct btrfs_trans_handle *trans;
 	struct timespec ct = CURRENT_TIME;
 	int ret = 0;
+	int received_uuid_changed;
 
 	ret = mnt_want_write_file(file);
 	if (ret < 0)

@@ -3981,7 +4014,11 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 		goto out;
 	}
 
-	trans = btrfs_start_transaction(root, 1);
+	/*
+	 * 1 - root item
+	 * 2 - uuid items (received uuid + subvol uuid)
+	 */
+	trans = btrfs_start_transaction(root, 3);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
 		trans = NULL;

@@ -3992,6 +4029,14 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 	sa->rtime.sec = ct.tv_sec;
 	sa->rtime.nsec = ct.tv_nsec;
 
+	received_uuid_changed = memcmp(root_item->received_uuid, sa->uuid,
+				       BTRFS_UUID_SIZE);
+	if (received_uuid_changed &&
+	    !btrfs_is_empty_uuid(root_item->received_uuid))
+		btrfs_del_uuid_received_subvol_item(trans,
+						    root->fs_info->uuid_root,
+						    root_item->received_uuid,
+						    root->root_key.objectid);
 	memcpy(root_item->received_uuid, sa->uuid, BTRFS_UUID_SIZE);
 	btrfs_set_root_stransid(root_item, sa->stransid);
 	btrfs_set_root_rtransid(root_item, sa->rtransid);

@@ -4004,12 +4049,21 @@ static long btrfs_ioctl_set_received_subvol(struct file *file,
 				&root->root_key, &root->root_item);
 	if (ret < 0) {
 		btrfs_end_transaction(trans, root);
-		trans = NULL;
 		goto out;
-	} else {
-		ret = btrfs_commit_transaction(trans, root);
-		if (ret < 0)
+	}
+	if (received_uuid_changed && !btrfs_is_empty_uuid(sa->uuid)) {
+		ret = btrfs_insert_uuid_received_subvol_item(
+			trans, root->fs_info->uuid_root, sa->uuid,
+			root->root_key.objectid);
+		if (ret < 0 && ret != -EEXIST) {
+			btrfs_abort_transaction(trans, root, ret);
 			goto out;
+		}
+	}
+	ret = btrfs_commit_transaction(trans, root);
+	if (ret < 0) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto out;
 	}
 
 	ret = copy_to_user(arg, sa, sizeof(*sa));

fs/btrfs/transaction.c +18 -1

index 0544587..a3d92b3 100644
@@ -1263,8 +1263,25 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 					 dentry->d_name.len * 2);
 	parent_inode->i_mtime = parent_inode->i_ctime = CURRENT_TIME;
 	ret = btrfs_update_inode_fallback(trans, parent_root, parent_inode);
-	if (ret)
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto fail;
+	}
+	ret = btrfs_insert_uuid_subvol_item(trans, fs_info->uuid_root,
+					    new_uuid.b, objectid);
+	if (ret) {
 		btrfs_abort_transaction(trans, root, ret);
+		goto fail;
+	}
+	if (!btrfs_is_empty_uuid(new_root_item->received_uuid)) {
+		ret = btrfs_insert_uuid_received_subvol_item(
+			trans, fs_info->uuid_root, new_root_item->received_uuid,
+			objectid);
+		if (ret && ret != -EEXIST) {
+			btrfs_abort_transaction(trans, root, ret);
+			goto fail;
+		}
+	}
 fail:
 	pending->error = ret;
 dir_item_existed:

Signature · 6 lines

1.8.2.2

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Review static noinline int create_subvol(struct inode *dir, · line 408

Quoted 8 lines · Stefan Behrens

(quoted diff) +1 -1

@@ -396,7 +403,7 @@ static noinline int create_subvol(struct inode *dir,
 	 * of create_snapshot().
 	 */
 	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
-					       7, &qgroup_reserved);
+					       8, &qgroup_reserved);
 	if (ret)
 		return ret;

This uuid_root will not use trans->block_rsv but empty_rsv since it does not set ref_cow, so you don't need to add one more to block_rsv, and same for the below cases.

Quoted 12 lines

(quoted diff) +2 -1

@@ -567,9 +578,10 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	 * 1 - root item
 	 * 2 - root ref/backref
 	 * 1 - root of snapshot
+	 * 1 - UUID item
 	 */
 	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
-					&pending_snapshot->block_rsv, 7,
+					&pending_snapshot->block_rsv, 8,
 					&pending_snapshot->qgroup_reserved);
 	if (ret)
 		goto out;

Quoted 13 lines · Liu Bo

Review static noinline int create_subvol(struct inode *dir, · line 408

Quoted 8 lines · Stefan Behrens

(quoted diff) +1 -1

@@ -396,7 +403,7 @@ static noinline int create_subvol(struct inode *dir,
 	 * of create_snapshot().
 	 */
 	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
-					       7, &qgroup_reserved);
+					       8, &qgroup_reserved);
 	if (ret)
 		return ret;

This uuid_root will not use trans->block_rsv but empty_rsv since it does not set ref_cow, so you don't need to add one more to block_rsv, and same for the below cases.

Hi Liu Bo,

Thanks for your review comments!

You are right, this won't work because the empty_block_rsv is used for the UUID tree as it is now.

I need to avoid ENOSPC in the middle of the transaction. Can you please acknowledge or comment the following addition to get_block_rsv()? And the "plus 1" that I have added to the transaction reservations for subvolume/snapshot creation/destruction will work correctly afterwards.

fs/btrfs/extent-tree.c +3 -0

@@ -4259,6 +4259,9 @@ static struct btrfs_block_rsv *get_block_rsv(
        if (root == root->fs_info->csum_root && trans->adding_csums)
                block_rsv = trans->block_rsv;
 
+       if (root == root->fs_info->uuid_root)
+               block_rsv = trans->block_rsv;
+
        if (!block_rsv)
                block_rsv = root->block_rsv;
 

Quoted 14 lines

Quoted 14 lines

(quoted diff) +2 -1

@@ -567,9 +578,10 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	 * 1 - root item
 	 * 2 - root ref/backref
 	 * 1 - root of snapshot
+	 * 1 - UUID item
 	 */
 	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
-					&pending_snapshot->block_rsv, 7,
+					&pending_snapshot->block_rsv, 8,
 					&pending_snapshot->qgroup_reserved);
 	if (ret)
 		goto out;

Mapping UUIDs to subvolume IDs is an operation with a high effort today. Today, the algorithm even has quadratic effort (based on the number of existing subvolumes), which means, that it takes minutes to send/receive a single subvolume if 10,000 subvolumes exist. But even linear effort would be too much since it is a waste. And these data structures to allow mapping UUIDs to subvolume IDs are created every time a btrfs send/receive instance is started.

So the issue to address is that Btrfs send / receive does not work as it is today when a high number of subvolumes exist.

It is much more efficient to maintain a searchable persistent data structure in the filesystem, one that is updated whenever a subvolume/snapshot is created and deleted, and when the received subvolume UUID is set by the btrfs-receive tool.

Therefore kernel code is added that is able to maintain data structures in the filesystem that allow to quickly search for a given UUID and to retrieve the subvol ID.

Now follows the lengthy justification, why a new tree was added instead of using the existing root tree:

The first approach was to not create another tree that holds UUID items. Instead, the items should just go into the top root tree. Unfortunately this confused the algorithm to assign the objectid of subvolumes and snapshots. The reason is that btrfs_find_free_objectid() calls btrfs_find_highest_objectid() for the first created subvol or snapshot after mounting a filesystem, and this function simply searches for the largest used objectid in the root tree keys to pick the next objectid to assign. Of course, the UUID keys have always been the ones with the highest offset value, and the next assigned subvol ID was wastefully huge.

To use any other existing tree did not look proper. To apply a workaround such as setting the objectid to zero in the UUID item key and to implement collision handling would either add limitations (in case of a btrfs_extend_item() approach to handle the collisions) or a lot of complexity and source code (in case a key would be looked up that is free of collisions). Adding new code that introduces limitations is not good, and adding code that is complex and lengthy for no good reason is also not good. That's the justification why a completely new tree was introduced.

v1 -> v2:
- All review comments from David Sterba, Josef Bacik and Jan Schmidt are addressed.
The hugest change was to add a mechanism that handles the case that the filesystem is mounted with an older kernel. Now that case is detected when the filesystem is mounted with a newer kernel again, and the UUID tree is updated in the background.

Stefan Behrens (8):
Btrfs: introduce a tree for items that map UUIDs to something
Btrfs: support printing UUID tree elements
Btrfs: create UUID tree if required
Btrfs: maintain subvolume items in the UUID tree
Btrfs: fill UUID tree initially
Btrfs: introduce uuid-tree-gen field
Btrfs: check UUID tree during mount if required
Btrfs: add mount option to force UUID tree checking

 fs/btrfs/Makefile      |   3 +-
 fs/btrfs/ctree.h       |  65 ++++-
 fs/btrfs/disk-io.c     |  62 ++++-
 fs/btrfs/ioctl.c       |  74 +++++-
 fs/btrfs/print-tree.c  |  82 +++++++
 fs/btrfs/super.c       |   8 +-
 fs/btrfs/transaction.c |  21 +-
 fs/btrfs/uuid-tree.c   | 637 +++++++++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/volumes.c     | 258 ++++++++++++++++++++
 fs/btrfs/volumes.h     |   2 +
 10 files changed, 1197 insertions(+), 15 deletions(-)
 create mode 100644 fs/btrfs/uuid-tree.c

Signature · 6 lines

1.8.2.2

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Quoted 34 lines · Stefan Behrens

Mapping UUIDs to subvolume IDs is an operation with a high effort today. Today, the algorithm even has quadratic effort (based on the number of existing subvolumes), which means, that it takes minutes to send/receive a single subvolume if 10,000 subvolumes exist. But even linear effort would be too much since it is a waste. And these data structures to allow mapping UUIDs to subvolume IDs are created every time a btrfs send/receive instance is started.

So the issue to address is that Btrfs send / receive does not work as it is today when a high number of subvolumes exist.

It is much more efficient to maintain a searchable persistent data structure in the filesystem, one that is updated whenever a subvolume/snapshot is created and deleted, and when the received subvolume UUID is set by the btrfs-receive tool.

Therefore kernel code is added that is able to maintain data structures in the filesystem that allow to quickly search for a given UUID and to retrieve the subvol ID.

Now follows the lengthy justification, why a new tree was added instead of using the existing root tree:

The first approach was to not create another tree that holds UUID items. Instead, the items should just go into the top root tree. Unfortunately this confused the algorithm to assign the objectid of subvolumes and snapshots. The reason is that btrfs_find_free_objectid() calls btrfs_find_highest_objectid() for the first created subvol or snapshot after mounting a filesystem, and this function simply searches for the largest used objectid in the root tree keys to pick the next objectid to assign. Of course, the UUID keys have always been the ones with the highest offset value, and the next assigned subvol ID was wastefully huge.

Quoted 9 lines

To use any other existing tree did not look proper. To apply a workaround such as setting the objectid to zero in the UUID item key and to implement collision handling would either add limitations (in case of a btrfs_extend_item() approach to handle the collisions) or a lot of complexity and source code (in case a key would be looked up that is free of collisions). Adding new code that introduces limitations is not good, and adding code that is complex and lengthy for no good reason is also not good. That's the justification why a completely new tree was introduced.

I'd appreciate if some performance number appear here since it's a speedup.

Quoted 45 lines · Liu Bo

Quoted 34 lines · Stefan Behrens

Mapping UUIDs to subvolume IDs is an operation with a high effort today. Today, the algorithm even has quadratic effort (based on the number of existing subvolumes), which means, that it takes minutes to send/receive a single subvolume if 10,000 subvolumes exist. But even linear effort would be too much since it is a waste. And these data structures to allow mapping UUIDs to subvolume IDs are created every time a btrfs send/receive instance is started.

So the issue to address is that Btrfs send / receive does not work as it is today when a high number of subvolumes exist.

It is much more efficient to maintain a searchable persistent data structure in the filesystem, one that is updated whenever a subvolume/snapshot is created and deleted, and when the received subvolume UUID is set by the btrfs-receive tool.

Therefore kernel code is added that is able to maintain data structures in the filesystem that allow to quickly search for a given UUID and to retrieve the subvol ID.

Now follows the lengthy justification, why a new tree was added instead of using the existing root tree:

The first approach was to not create another tree that holds UUID items. Instead, the items should just go into the top root tree. Unfortunately this confused the algorithm to assign the objectid of subvolumes and snapshots. The reason is that btrfs_find_free_objectid() calls btrfs_find_highest_objectid() for the first created subvol or snapshot after mounting a filesystem, and this function simply searches for the largest used objectid in the root tree keys to pick the next objectid to assign. Of course, the UUID keys have always been the ones with the highest offset value, and the next assigned subvol ID was wastefully huge.

Quoted 9 lines

To use any other existing tree did not look proper. To apply a workaround such as setting the objectid to zero in the UUID item key and to implement collision handling would either add limitations (in case of a btrfs_extend_item() approach to handle the collisions) or a lot of complexity and source code (in case a key would be looked up that is free of collisions). Adding new code that introduces limitations is not good, and adding code that is complex and lengthy for no good reason is also not good. That's the justification why a completely new tree was introduced.

Quoted 1 line

I'd appreciate if some performance number appear here since it's a speedup.

That's a good idea. The numbers are below in the table and there's also a link to a chart.

I stopped the measurement with the old version after 10000 subvolumes because it already took almost 13 minutes to send a single, empty subvolume. All the time is spent building a database, and this is done each time the btrfs send or receive tool is started to send or receive a single subvolume.

The table shows the time it takes to send a single, empty subvolume depending on the number of subvolume that exist in the filesystem.

# of subvols  | without    | with
in filesystem | UUID tree  | UUID tree
--------------+------------+----------
            2 |  0m00.004s | 0m00.003s
         1000 |  0m07.010s | 0m00.004s
         2000 |  0m28.210s | 0m00.004s
         3000 |  1m04.872s | 0m00.004s
         4000 |  1m56.059s | 0m00.004s
         5000 |  3m00.489s | 0m00.004s
         6000 |  4m27.376s | 0m00.004s
         7000 |  6m08.938s | 0m00.004s
         8000 |  7m54.020s | 0m00.004s
         9000 | 10m05.108s | 0m00.004s
        10000 | 12m47.406s | 0m00.004s

Or as a chart:
http://btrfs.giantdisaster.de/Btrfs-send-recv-perf.pdf

The Hardware:
Intel(R) Xeon(R) CPU X3450 @ 2.67GHz
8 GB RAM
6 high performance SSDs

The script:
#!/bin/bash
set -e
MOUNT=/mnt2
umount $MOUNT || true
mkfs.btrfs -f -m raid0 -d raid0 -n 32768 /dev/sdc /dev/sdj /dev/sds /dev/sdt /dev/sdu /dev/sdv mount /dev/sdc $MOUNT
btrfs subv create $MOUNT/0
btrfs subv snapshot -r $MOUNT/0 $MOUNT/0ro > /dev/null
echo '2 subvols'
time btrfs send $MOUNT/0ro > /dev/null
umount $MOUNT
mkfs.btrfs -f -m raid0 -d raid0 -n 32768 /dev/sdc /dev/sdj /dev/sds /dev/sdt /dev/sdu /dev/sdv mount /dev/sdc $MOUNT
for i in seq 1000; do
btrfs subv create $MOUNT/$i
for j in seq 500; do
btrfs subv create $MOUNT/$i/$j > /dev/null
btrfs subv snapshot -r $MOUNT/$i/$j $MOUNT/$i/${j}ro > /dev/null
done
echo $i 'k subvols'
time btrfs send $MOUNT/$i/1ro > /dev/null
done

# of subvols  | without    | with
in filesystem | UUID tree  | UUID tree
--------------+------------+----------
            2 |  0m00.004s | 0m00.003s
         1000 |  0m07.010s | 0m00.004s
         2000 |  0m28.210s | 0m00.004s

         3000 |  1m04.872s | 0m00.004s
         4000 |  1m56.059s | 0m00.004s
         5000 |  3m00.489s | 0m00.004s
         6000 |  4m27.376s | 0m00.004s
         7000 |  6m08.938s | 0m00.004s
         8000 |  7m54.020s | 0m00.004s
         9000 | 10m05.108s | 0m00.004s
        10000 | 12m47.406s | 0m00.004s