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

Pull btrfs update from Chris Mason:
 "These are mostly fixes.  The biggest exceptions are Josef's skinny
  extents and Jan Schmidt's code to rebuild our quota indexes if they
  get out of sync (or you enable quotas on an existing filesystem).

  The skinny extents are off by default because they are a new variation
  on the extent allocation tree format.  btrfstune -x enables them, and
  the new format makes the extent allocation tree about 30% smaller.

  I rebased this a few days ago to rework Dave Sterba's crc checks on
  the super block, but almost all of these go back to rc6, since I
  though 3.9 was due any minute.

  The biggest missing fix is the tracepoint bug that was hit late in
  3.9.  I ran into problems with that in overnight testing and I'm still
  tracking it down.  I'll definitely have that fixed for rc2."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (101 commits)
  Btrfs: allow superblock mismatch from older mkfs
  btrfs: enhance superblock checks
  btrfs: fix misleading variable name for flags
  btrfs: use unsigned long type for extent state bits
  Btrfs: improve the loop of scrub_stripe
  btrfs: read entire device info under lock
  btrfs: remove unused gfp mask parameter from release_extent_buffer callchain
  btrfs: handle errors returned from get_tree_block_key
  btrfs: make static code static & remove dead code
  Btrfs: deal with errors in write_dev_supers
  Btrfs: remove almost all of the BUG()'s from tree-log.c
  Btrfs: deal with free space cache errors while replaying log
  Btrfs: automatic rescan after "quota enable" command
  Btrfs: rescan for qgroups
  Btrfs: split btrfs_qgroup_account_ref into four functions
  Btrfs: allocate new chunks if the space is not enough for global rsv
  Btrfs: separate sequence numbers for delayed ref tracking and tree mod log
  btrfs: move leak debug code to functions
  Btrfs: return free space in cow error path
  Btrfs: set UUID in root_item for created trees
  ...

Documentation/filesystems/btrfs.txt

@@ -1,8 +1,8 @@
 
-	BTRFS
-	=====
+BTRFS
+=====
 
-Btrfs is a new copy on write filesystem for Linux aimed at
+Btrfs is a copy on write filesystem for Linux aimed at
 implementing advanced features while focusing on fault tolerance,
 repair and easy administration. Initially developed by Oracle, Btrfs
 is licensed under the GPL and open for contribution from anyone.
@@ -34,9 +34,175 @@ The main Btrfs features include:
     * Online filesystem defragmentation
 
 
+Mount Options
+=============
 
-	MAILING LIST
-	============
+When mounting a btrfs filesystem, the following option are accepted.
+Unless otherwise specified, all options default to off.
+
+  alloc_start=<bytes>
+	Debugging option to force all block allocations above a certain
+	byte threshold on each block device.  The value is specified in
+	bytes, optionally with a K, M, or G suffix, case insensitive.
+	Default is 1MB.
+
+  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.
+
+  check_int
+  check_int_data
+  check_int_print_mask=<value>
+	These debugging options control the behavior of the integrity checking
+	module (the BTRFS_FS_CHECK_INTEGRITY config option required).
+
+	check_int enables the integrity checker module, which examines all
+	block write requests to ensure on-disk consistency, at a large
+	memory and CPU cost.  
+
+	check_int_data includes extent data in the integrity checks, and
+	implies the check_int option.
+
+	check_int_print_mask takes a bitmask of BTRFSIC_PRINT_MASK_* values
+	as defined in fs/btrfs/check-integrity.c, to control the integrity
+	checker module behavior.
+
+	See comments at the top of fs/btrfs/check-integrity.c for more info.
+
+  compress
+  compress=<type>
+  compress-force
+  compress-force=<type>
+	Control BTRFS file data compression.  Type may be specified as "zlib"
+	"lzo" or "no" (for no compression, used for remounting).  If no type
+	is specified, zlib is used.  If compress-force is specified,
+	all files will be compressed, whether or not they compress well.
+	If compression is enabled, nodatacow and nodatasum are disabled.
+
+  degraded
+	Allow mounts to continue with missing devices.  A read-write mount may
+	fail with too many devices missing, for example if a stripe member
+	is completely missing.
+
+  device=<devicepath>
+	Specify a device during mount so that ioctls on the control device
+	can be avoided.  Especialy useful when trying to mount a multi-device
+	setup as root.  May be specified multiple times for multiple devices.
+
+  discard
+	Issue 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).
+
+  enospc_debug
+	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.
+
+  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
+	the committed state a fully consistent view of the file system from the
+	application's perspective (i.e., it includes all completed file system
+	operations).  This was previously the behavior only when a snapshot is
+	created.
+
+  inode_cache
+	Enable free inode number caching.   Defaults to off due to an overflow
+	problem when the free space crcs don't fit inside a single page.
+
+  max_inline=<bytes>
+	Specify the maximum amount of space, in bytes, that can be inlined in
+	a metadata B-tree leaf.  The value is specified in bytes, optionally 
+	with a K, M, or G suffix, case insensitive.  In practice, this value
+	is limited by the root sector size, with some space unavailable due
+	to leaf headers.  For a 4k sectorsize, max inline data is ~3900 bytes.
+
+  metadata_ratio=<value>
+	Specify that 1 metadata chunk should be allocated after every <value>
+	data chunks.  Off by default.
+
+  noacl
+	Disable support for Posix Access Control Lists (ACLs).  See the
+	acl(5) manual page for more information about ACLs.
+
+  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.
+
+  nodatacow
+	Disable data copy-on-write for newly created files.  Implies nodatasum,
+	and disables all compression.
+
+  nodatasum
+	Disable data checksumming for newly created files.
+
+  notreelog
+	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.
+	Currently this scans a list of several previous tree roots and tries to 
+	use the first readable.
+
+ skip_balance
+	Skip automatic resume of interrupted balance operation after mount.
+	May be resumed with "btrfs balance resume."
+
+  space_cache (*)
+	Enable the on-disk freespace cache.
+  nospace_cache
+	Disable freespace cache loading without clearing the cache.
+  clear_cache
+	Force clearing and rebuilding of the disk space cache if something
+	has gone wrong.
+
+  ssd
+  nossd
+  ssd_spread
+	Options to control ssd allocation schemes.  By default, BTRFS will
+	enable or disable ssd allocation heuristics depending on whether a
+	rotational or nonrotational disk is in use.  The ssd and nossd options
+	can override this autodetection.
+
+	The ssd_spread mount option attempts to allocate into big chunks
+	of unused space, and may perform better on low-end ssds.  ssd_spread
+	implies ssd, enabling all other ssd heuristics as well.
+
+  subvol=<path>
+	Mount subvolume at <path> rather than the root subvolume.  <path> is
+	relative to the top level subvolume.
+
+  subvolid=<ID>
+	Mount subvolume specified by an ID number rather than the root subvolume.
+	This allows mounting of subvolumes which are not in the root of the mounted
+	filesystem.
+	You can use "btrfs subvolume list" to see subvolume ID numbers.
+
+  subvolrootid=<objectid> (deprecated)
+	Mount subvolume specified by <objectid> rather than the root subvolume.
+	This allows mounting of subvolumes which are not in the root of the mounted
+	filesystem.
+	You can use "btrfs subvolume show " to see the object ID for a subvolume.
+	
+  thread_pool=<number>
+	The number of worker threads to allocate.  The default number is equal
+	to the number of CPUs + 2, or 8, whichever is smaller.
+
+  user_subvol_rm_allowed
+	Allow subvolumes to be deleted by a non-root user. Use with caution. 
+
+MAILING LIST
+============
 
 There is a Btrfs mailing list hosted on vger.kernel.org. You can
 find details on how to subscribe here:
@@ -49,8 +215,8 @@ http://dir.gmane.org/gmane.comp.file-systems.btrfs
 
 
 
-	IRC
-	===
+IRC
+===
 
 Discussion of Btrfs also occurs on the #btrfs channel of the Freenode
 IRC network.

fs/btrfs/Kconfig

@@ -1,5 +1,5 @@
 config BTRFS_FS
-	tristate "Btrfs filesystem Unstable disk format"
+	tristate "Btrfs filesystem support"
 	select LIBCRC32C
 	select ZLIB_INFLATE
 	select ZLIB_DEFLATE
@@ -52,3 +52,23 @@ config BTRFS_FS_CHECK_INTEGRITY
 	  In most cases, unless you are a btrfs developer who needs
 	  to verify the integrity of (super)-block write requests
 	  during the run of a regression test, say N
+
+config BTRFS_FS_RUN_SANITY_TESTS
+	bool "Btrfs will run sanity tests upon loading"
+	depends on BTRFS_FS
+	help
+	  This will run some basic sanity tests on the free space cache
+	  code to make sure it is acting as it should.  These are mostly
+	  regression tests and are only really interesting to btrfs devlopers.
+
+	  If unsure, say N.
+
+config BTRFS_DEBUG
+	bool "Btrfs debugging support"
+	depends on BTRFS_FS
+	help
+	  Enable run-time debugging support for the btrfs filesystem. This may
+	  enable additional and expensive checks with negative impact on
+	  performance, or export extra information via sysfs.
+
+	  If unsure, say N.

fs/btrfs/backref.c

@@ -352,6 +352,8 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 		err = __resolve_indirect_ref(fs_info, search_commit_root,
 					     time_seq, ref, parents,
 					     extent_item_pos);
+		if (err == -ENOMEM)
+			goto out;
 		if (err)
 			continue;
 
@@ -367,7 +369,7 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 			new_ref = kmalloc(sizeof(*new_ref), GFP_NOFS);
 			if (!new_ref) {
 				ret = -ENOMEM;
-				break;
+				goto out;
 			}
 			memcpy(new_ref, ref, sizeof(*ref));
 			new_ref->parent = node->val;
@@ -377,7 +379,7 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 		}
 		ulist_reinit(parents);
 	}
-
+out:
 	ulist_free(parents);
 	return ret;
 }
@@ -421,7 +423,10 @@ static int __add_missing_keys(struct btrfs_fs_info *fs_info,
 		BUG_ON(!ref->wanted_disk_byte);
 		eb = read_tree_block(fs_info->tree_root, ref->wanted_disk_byte,
 				     fs_info->tree_root->leafsize, 0);
-		BUG_ON(!eb);
+		if (!eb || !extent_buffer_uptodate(eb)) {
+			free_extent_buffer(eb);
+			return -EIO;
+		}
 		btrfs_tree_read_lock(eb);
 		if (btrfs_header_level(eb) == 0)
 			btrfs_item_key_to_cpu(eb, &ref->key_for_search, 0);
@@ -443,7 +448,7 @@ static int __add_missing_keys(struct btrfs_fs_info *fs_info,
  *           having a parent).
  * mode = 2: merge identical parents
  */
-static int __merge_refs(struct list_head *head, int mode)
+static void __merge_refs(struct list_head *head, int mode)
 {
 	struct list_head *pos1;
 
@@ -489,7 +494,6 @@ static int __merge_refs(struct list_head *head, int mode)
 		}
 
 	}
-	return 0;
 }
 
 /*
@@ -582,7 +586,8 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 		default:
 			WARN_ON(1);
 		}
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 	}
 
 	return 0;
@@ -680,7 +685,8 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 		default:
 			WARN_ON(1);
 		}
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 		ptr += btrfs_extent_inline_ref_size(type);
 	}
 
@@ -762,7 +768,9 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 		default:
 			WARN_ON(1);
 		}
-		BUG_ON(ret);
+		if (ret)
+			return ret;
+
 	}
 
 	return ret;
@@ -880,18 +888,14 @@ again:
 	if (ret)
 		goto out;
 
-	ret = __merge_refs(&prefs, 1);
-	if (ret)
-		goto out;
+	__merge_refs(&prefs, 1);
 
 	ret = __resolve_indirect_refs(fs_info, search_commit_root, time_seq,
 				      &prefs, extent_item_pos);
 	if (ret)
 		goto out;
 
-	ret = __merge_refs(&prefs, 2);
-	if (ret)
-		goto out;
+	__merge_refs(&prefs, 2);
 
 	while (!list_empty(&prefs)) {
 		ref = list_first_entry(&prefs, struct __prelim_ref, list);
@@ -900,7 +904,8 @@ again:
 		if (ref->count && ref->root_id && ref->parent == 0) {
 			/* no parent == root of tree */
 			ret = ulist_add(roots, ref->root_id, 0, GFP_NOFS);
-			BUG_ON(ret < 0);
+			if (ret < 0)
+				goto out;
 		}
 		if (ref->count && ref->parent) {
 			struct extent_inode_elem *eie = NULL;
@@ -911,7 +916,10 @@ again:
 							info_level);
 				eb = read_tree_block(fs_info->extent_root,
 							   ref->parent, bsz, 0);
-				BUG_ON(!eb);
+				if (!eb || !extent_buffer_uptodate(eb)) {
+					free_extent_buffer(eb);
+					return -EIO;
+				}
 				ret = find_extent_in_eb(eb, bytenr,
 							*extent_item_pos, &eie);
 				ref->inode_list = eie;
@@ -920,6 +928,8 @@ again:
 			ret = ulist_add_merge(refs, ref->parent,
 					      (uintptr_t)ref->inode_list,
 					      (u64 *)&eie, GFP_NOFS);
+			if (ret < 0)
+				goto out;
 			if (!ret && extent_item_pos) {
 				/*
 				 * we've recorded that parent, so we must extend
@@ -930,7 +940,6 @@ again:
 					eie = eie->next;
 				eie->next = ref->inode_list;
 			}
-			BUG_ON(ret < 0);
 		}
 		kfree(ref);
 	}
@@ -1180,6 +1189,20 @@ int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
 	return ret;
 }
 
+/*
+ * this iterates to turn a name (from iref/extref) into a full filesystem path.
+ * Elements of the path are separated by '/' and the path is guaranteed to be
+ * 0-terminated. the path is only given within the current file system.
+ * Therefore, it never starts with a '/'. the caller is responsible to provide
+ * "size" bytes in "dest". the dest buffer will be filled backwards. finally,
+ * the start point of the resulting string is returned. this pointer is within
+ * dest, normally.
+ * in case the path buffer would overflow, the pointer is decremented further
+ * as if output was written to the buffer, though no more output is actually
+ * generated. that way, the caller can determine how much space would be
+ * required for the path to fit into the buffer. in that case, the returned
+ * value will be smaller than dest. callers must check this!
+ */
 char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path,
 			u32 name_len, unsigned long name_off,
 			struct extent_buffer *eb_in, u64 parent,
@@ -1248,32 +1271,6 @@ char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path,
 	return dest + bytes_left;
 }
 
-/*
- * this iterates to turn a btrfs_inode_ref into a full filesystem path. elements
- * of the path are separated by '/' and the path is guaranteed to be
- * 0-terminated. the path is only given within the current file system.
- * Therefore, it never starts with a '/'. the caller is responsible to provide
- * "size" bytes in "dest". the dest buffer will be filled backwards. finally,
- * the start point of the resulting string is returned. this pointer is within
- * dest, normally.
- * in case the path buffer would overflow, the pointer is decremented further
- * as if output was written to the buffer, though no more output is actually
- * generated. that way, the caller can determine how much space would be
- * required for the path to fit into the buffer. in that case, the returned
- * value will be smaller than dest. callers must check this!
- */
-char *btrfs_iref_to_path(struct btrfs_root *fs_root,
-			 struct btrfs_path *path,
-			 struct btrfs_inode_ref *iref,
-			 struct extent_buffer *eb_in, u64 parent,
-			 char *dest, u32 size)
-{
-	return btrfs_ref_to_path(fs_root, path,
-				 btrfs_inode_ref_name_len(eb_in, iref),
-				 (unsigned long)(iref + 1),
-				 eb_in, parent, dest, size);
-}
-
 /*
  * this makes the path point to (logical EXTENT_ITEM *)
  * returns BTRFS_EXTENT_FLAG_DATA for data, BTRFS_EXTENT_FLAG_TREE_BLOCK for
@@ -1461,8 +1458,6 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
 				iterate_extent_inodes_t *iterate, void *ctx)
 {
 	int ret;
-	struct list_head data_refs = LIST_HEAD_INIT(data_refs);
-	struct list_head shared_refs = LIST_HEAD_INIT(shared_refs);
 	struct btrfs_trans_handle *trans;
 	struct ulist *refs = NULL;
 	struct ulist *roots = NULL;
@@ -1508,11 +1503,9 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
 						iterate, ctx);
 		}
 		ulist_free(roots);
-		roots = NULL;
 	}
 
 	free_leaf_list(refs);
-	ulist_free(roots);
 out:
 	if (!search_commit_root) {
 		btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem);

fs/btrfs/backref.h

@@ -59,9 +59,6 @@ int paths_from_inode(u64 inum, struct inode_fs_paths *ipath);
 int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
 				struct btrfs_fs_info *fs_info, u64 bytenr,
 				u64 time_seq, struct ulist **roots);
-char *btrfs_iref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path,
-			 struct btrfs_inode_ref *iref, struct extent_buffer *eb,
-			 u64 parent, char *dest, u32 size);
 char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path,
 			u32 name_len, unsigned long name_off,
 			struct extent_buffer *eb_in, u64 parent,

fs/btrfs/btrfs_inode.h

@@ -93,7 +93,7 @@ struct btrfs_inode {
 
 	unsigned long runtime_flags;
 
-	/* Keep track of who's O_SYNC/fsycing currently */
+	/* Keep track of who's O_SYNC/fsyncing currently */
 	atomic_t sync_writers;
 
 	/* full 64 bit generation number, struct vfs_inode doesn't have a big

fs/btrfs/compression.c

@@ -82,6 +82,10 @@ struct compressed_bio {
 	u32 sums;
 };
 
+static int btrfs_decompress_biovec(int type, struct page **pages_in,
+				   u64 disk_start, struct bio_vec *bvec,
+				   int vcnt, size_t srclen);
+
 static inline int compressed_bio_size(struct btrfs_root *root,
 				      unsigned long disk_size)
 {
@@ -106,7 +110,6 @@ static int check_compressed_csum(struct inode *inode,
 				 u64 disk_start)
 {
 	int ret;
-	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct page *page;
 	unsigned long i;
 	char *kaddr;
@@ -121,7 +124,7 @@ static int check_compressed_csum(struct inode *inode,
 		csum = ~(u32)0;
 
 		kaddr = kmap_atomic(page);
-		csum = btrfs_csum_data(root, kaddr, csum, PAGE_CACHE_SIZE);
+		csum = btrfs_csum_data(kaddr, csum, PAGE_CACHE_SIZE);
 		btrfs_csum_final(csum, (char *)&csum);
 		kunmap_atomic(kaddr);
 
@@ -739,7 +742,7 @@ static int comp_num_workspace[BTRFS_COMPRESS_TYPES];
 static atomic_t comp_alloc_workspace[BTRFS_COMPRESS_TYPES];
 static wait_queue_head_t comp_workspace_wait[BTRFS_COMPRESS_TYPES];
 
-struct btrfs_compress_op *btrfs_compress_op[] = {
+static struct btrfs_compress_op *btrfs_compress_op[] = {
 	&btrfs_zlib_compress,
 	&btrfs_lzo_compress,
 };
@@ -910,8 +913,9 @@ int btrfs_compress_pages(int type, struct address_space *mapping,
  * be contiguous.  They all correspond to the range of bytes covered by
  * the compressed extent.
  */
-int btrfs_decompress_biovec(int type, struct page **pages_in, u64 disk_start,
-			    struct bio_vec *bvec, int vcnt, size_t srclen)
+static int btrfs_decompress_biovec(int type, struct page **pages_in,
+				   u64 disk_start, struct bio_vec *bvec,
+				   int vcnt, size_t srclen)
 {
 	struct list_head *workspace;
 	int ret;

fs/btrfs/compression.h

@@ -30,8 +30,6 @@ int btrfs_compress_pages(int type, struct address_space *mapping,
 			 unsigned long *total_in,
 			 unsigned long *total_out,
 			 unsigned long max_out);
-int btrfs_decompress_biovec(int type, struct page **pages_in, u64 disk_start,
-			    struct bio_vec *bvec, int vcnt, size_t srclen);
 int btrfs_decompress(int type, unsigned char *data_in, struct page *dest_page,
 		     unsigned long start_byte, size_t srclen, size_t destlen);
 int btrfs_decompress_buf2page(char *buf, unsigned long buf_start,

fs/btrfs/ctree.c

@@ -37,16 +37,11 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root,
 			      struct extent_buffer *dst_buf,
 			      struct extent_buffer *src_buf);
-static void del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
-		    struct btrfs_path *path, int level, int slot);
+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,
 				 struct extent_buffer *eb);
-struct extent_buffer *read_old_tree_block(struct btrfs_root *root, u64 bytenr,
-					  u32 blocksize, u64 parent_transid,
-					  u64 time_seq);
-struct extent_buffer *btrfs_find_old_tree_block(struct btrfs_root *root,
-						u64 bytenr, u32 blocksize,
-						u64 time_seq);
+static int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path);
 
 struct btrfs_path *btrfs_alloc_path(void)
 {
@@ -208,7 +203,7 @@ struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root)
  * tree until you end up with a lock on the root.  A locked buffer
  * is returned, with a reference held.
  */
-struct extent_buffer *btrfs_read_lock_root_node(struct btrfs_root *root)
+static struct extent_buffer *btrfs_read_lock_root_node(struct btrfs_root *root)
 {
 	struct extent_buffer *eb;
 
@@ -360,6 +355,44 @@ static inline void tree_mod_log_write_unlock(struct btrfs_fs_info *fs_info)
 	write_unlock(&fs_info->tree_mod_log_lock);
 }
 
+/*
+ * Increment the upper half of tree_mod_seq, set lower half zero.
+ *
+ * Must be called with fs_info->tree_mod_seq_lock held.
+ */
+static inline u64 btrfs_inc_tree_mod_seq_major(struct btrfs_fs_info *fs_info)
+{
+	u64 seq = atomic64_read(&fs_info->tree_mod_seq);
+	seq &= 0xffffffff00000000ull;
+	seq += 1ull << 32;
+	atomic64_set(&fs_info->tree_mod_seq, seq);
+	return seq;
+}
+
+/*
+ * Increment the lower half of tree_mod_seq.
+ *
+ * Must be called with fs_info->tree_mod_seq_lock held. The way major numbers
+ * are generated should not technically require a spin lock here. (Rationale:
+ * incrementing the minor while incrementing the major seq number is between its
+ * atomic64_read and atomic64_set calls doesn't duplicate sequence numbers, it
+ * just returns a unique sequence number as usual.) We have decided to leave
+ * that requirement in here and rethink it once we notice it really imposes a
+ * problem on some workload.
+ */
+static inline u64 btrfs_inc_tree_mod_seq_minor(struct btrfs_fs_info *fs_info)
+{
+	return atomic64_inc_return(&fs_info->tree_mod_seq);
+}
+
+/*
+ * return the last minor in the previous major tree_mod_seq number
+ */
+u64 btrfs_tree_mod_seq_prev(u64 seq)
+{
+	return (seq & 0xffffffff00000000ull) - 1ull;
+}
+
 /*
  * This adds a new blocker to the tree mod log's blocker list if the @elem
  * passed does not already have a sequence number set. So when a caller expects
@@ -376,10 +409,10 @@ u64 btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info,
 	tree_mod_log_write_lock(fs_info);
 	spin_lock(&fs_info->tree_mod_seq_lock);
 	if (!elem->seq) {
-		elem->seq = btrfs_inc_tree_mod_seq(fs_info);
+		elem->seq = btrfs_inc_tree_mod_seq_major(fs_info);
 		list_add_tail(&elem->list, &fs_info->tree_mod_seq_list);
 	}
-	seq = btrfs_inc_tree_mod_seq(fs_info);
+	seq = btrfs_inc_tree_mod_seq_minor(fs_info);
 	spin_unlock(&fs_info->tree_mod_seq_lock);
 	tree_mod_log_write_unlock(fs_info);
 
@@ -524,7 +557,10 @@ static inline int tree_mod_alloc(struct btrfs_fs_info *fs_info, gfp_t flags,
 	if (!tm)
 		return -ENOMEM;
 
-	tm->seq = btrfs_inc_tree_mod_seq(fs_info);
+	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);
+
 	return tm->seq;
 }
 
@@ -643,7 +679,8 @@ __tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 static noinline int
 tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 			 struct extent_buffer *old_root,
-			 struct extent_buffer *new_root, gfp_t flags)
+			 struct extent_buffer *new_root, gfp_t flags,
+			 int log_removal)
 {
 	struct tree_mod_elem *tm;
 	int ret;
@@ -651,7 +688,8 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 	if (tree_mod_dont_log(fs_info, NULL))
 		return 0;
 
-	__tree_mod_log_free_eb(fs_info, old_root);
+	if (log_removal)
+		__tree_mod_log_free_eb(fs_info, old_root);
 
 	ret = tree_mod_alloc(fs_info, flags, &tm);
 	if (ret < 0)
@@ -738,7 +776,7 @@ tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq)
 static noinline void
 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 log_removal)
+		     unsigned long src_offset, int nr_items)
 {
 	int ret;
 	int i;
@@ -752,12 +790,10 @@ tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
 	}
 
 	for (i = 0; i < nr_items; i++) {
-		if (log_removal) {
-			ret = tree_mod_log_insert_key_locked(fs_info, src,
-							i + src_offset,
-							MOD_LOG_KEY_REMOVE);
-			BUG_ON(ret < 0);
-		}
+		ret = tree_mod_log_insert_key_locked(fs_info, src,
+						i + src_offset,
+						MOD_LOG_KEY_REMOVE);
+		BUG_ON(ret < 0);
 		ret = tree_mod_log_insert_key_locked(fs_info, dst,
 						     i + dst_offset,
 						     MOD_LOG_KEY_ADD);
@@ -802,11 +838,12 @@ tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 
 static noinline void
 tree_mod_log_set_root_pointer(struct btrfs_root *root,
-			      struct extent_buffer *new_root_node)
+			      struct extent_buffer *new_root_node,
+			      int log_removal)
 {
 	int ret;
 	ret = tree_mod_log_insert_root(root->fs_info, root->node,
-				       new_root_node, GFP_NOFS);
+				       new_root_node, GFP_NOFS, log_removal);
 	BUG_ON(ret < 0);
 }
 
@@ -867,7 +904,8 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 
 	if (btrfs_block_can_be_shared(root, buf)) {
 		ret = btrfs_lookup_extent_info(trans, root, buf->start,
-					       buf->len, &refs, &flags);
+					       btrfs_header_level(buf), 1,
+					       &refs, &flags);
 		if (ret)
 			return ret;
 		if (refs == 0) {
@@ -1028,7 +1066,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 			parent_start = 0;
 
 		extent_buffer_get(cow);
-		tree_mod_log_set_root_pointer(root, cow);
+		tree_mod_log_set_root_pointer(root, cow, 1);
 		rcu_assign_pointer(root->node, cow);
 
 		btrfs_free_tree_block(trans, root, buf, parent_start,
@@ -1067,11 +1105,11 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
  */
 static struct tree_mod_elem *
 __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
-			   struct btrfs_root *root, u64 time_seq)
+			   struct extent_buffer *eb_root, u64 time_seq)
 {
 	struct tree_mod_elem *tm;
 	struct tree_mod_elem *found = NULL;
-	u64 root_logical = root->node->start;
+	u64 root_logical = eb_root->start;
 	int looped = 0;
 
 	if (!time_seq)
@@ -1105,7 +1143,6 @@ __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
 
 		found = tm;
 		root_logical = tm->old_root.logical;
-		BUG_ON(root_logical == root->node->start);
 		looped = 1;
 	}
 
@@ -1190,6 +1227,13 @@ __tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq,
 	btrfs_set_header_nritems(eb, n);
 }
 
+/*
+ * Called with eb read locked. If the buffer cannot be rewinded, the same buffer
+ * is returned. If rewind operations happen, a fresh buffer is returned. The
+ * returned buffer is always read-locked. If the returned buffer is not the
+ * input buffer, the lock on the input buffer is released and the input buffer
+ * is freed (its refcount is decremented).
+ */
 static struct extent_buffer *
 tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
 		    u64 time_seq)
@@ -1223,8 +1267,11 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
 	}
 
 	extent_buffer_get(eb_rewin);
+	btrfs_tree_read_unlock(eb);
 	free_extent_buffer(eb);
 
+	extent_buffer_get(eb_rewin);
+	btrfs_tree_read_lock(eb_rewin);
 	__tree_mod_log_rewind(eb_rewin, time_seq, tm);
 	WARN_ON(btrfs_header_nritems(eb_rewin) >
 		BTRFS_NODEPTRS_PER_BLOCK(fs_info->tree_root));
@@ -1243,33 +1290,35 @@ static inline struct extent_buffer *
 get_old_root(struct btrfs_root *root, u64 time_seq)
 {
 	struct tree_mod_elem *tm;
-	struct extent_buffer *eb;
+	struct extent_buffer *eb = NULL;
+	struct extent_buffer *eb_root;
 	struct extent_buffer *old;
 	struct tree_mod_root *old_root = NULL;
 	u64 old_generation = 0;
 	u64 logical;
 	u32 blocksize;
 
-	eb = btrfs_read_lock_root_node(root);
-	tm = __tree_mod_log_oldest_root(root->fs_info, root, time_seq);
+	eb_root = btrfs_read_lock_root_node(root);
+	tm = __tree_mod_log_oldest_root(root->fs_info, eb_root, time_seq);
 	if (!tm)
-		return root->node;
+		return eb_root;
 
 	if (tm->op == MOD_LOG_ROOT_REPLACE) {
 		old_root = &tm->old_root;
 		old_generation = tm->generation;
 		logical = old_root->logical;
 	} else {
-		logical = root->node->start;
+		logical = eb_root->start;
 	}
 
 	tm = tree_mod_log_search(root->fs_info, logical, time_seq);
 	if (old_root && tm && tm->op != MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
-		btrfs_tree_read_unlock(root->node);
-		free_extent_buffer(root->node);
+		btrfs_tree_read_unlock(eb_root);
+		free_extent_buffer(eb_root);
 		blocksize = btrfs_level_size(root, old_root->level);
 		old = read_tree_block(root, logical, blocksize, 0);
-		if (!old) {
+		if (!old || !extent_buffer_uptodate(old)) {
+			free_extent_buffer(old);
 			pr_warn("btrfs: failed to read tree block %llu from get_old_root\n",
 				logical);
 			WARN_ON(1);
@@ -1278,13 +1327,13 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 			free_extent_buffer(old);
 		}
 	} else if (old_root) {
-		btrfs_tree_read_unlock(root->node);
-		free_extent_buffer(root->node);
+		btrfs_tree_read_unlock(eb_root);
+		free_extent_buffer(eb_root);
 		eb = alloc_dummy_extent_buffer(logical, root->nodesize);
 	} else {
-		eb = btrfs_clone_extent_buffer(root->node);
-		btrfs_tree_read_unlock(root->node);
-		free_extent_buffer(root->node);
+		eb = btrfs_clone_extent_buffer(eb_root);
+		btrfs_tree_read_unlock(eb_root);
+		free_extent_buffer(eb_root);
 	}
 
 	if (!eb)
@@ -1294,7 +1343,7 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 	if (old_root) {
 		btrfs_set_header_bytenr(eb, eb->start);
 		btrfs_set_header_backref_rev(eb, BTRFS_MIXED_BACKREF_REV);
-		btrfs_set_header_owner(eb, root->root_key.objectid);
+		btrfs_set_header_owner(eb, btrfs_header_owner(eb_root));
 		btrfs_set_header_level(eb, old_root->level);
 		btrfs_set_header_generation(eb, old_generation);
 	}
@@ -1311,15 +1360,15 @@ int btrfs_old_root_level(struct btrfs_root *root, u64 time_seq)
 {
 	struct tree_mod_elem *tm;
 	int level;
+	struct extent_buffer *eb_root = btrfs_root_node(root);
 
-	tm = __tree_mod_log_oldest_root(root->fs_info, root, time_seq);
+	tm = __tree_mod_log_oldest_root(root->fs_info, eb_root, time_seq);
 	if (tm && tm->op == MOD_LOG_ROOT_REPLACE) {
 		level = tm->old_root.level;
 	} else {
-		rcu_read_lock();
-		level = btrfs_header_level(root->node);
-		rcu_read_unlock();
+		level = btrfs_header_level(eb_root);
 	}
+	free_extent_buffer(eb_root);
 
 	return level;
 }
@@ -1514,8 +1563,10 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
 			if (!cur) {
 				cur = read_tree_block(root, blocknr,
 							 blocksize, gen);
-				if (!cur)
+				if (!cur || !extent_buffer_uptodate(cur)) {
+					free_extent_buffer(cur);
 					return -EIO;
+				}
 			} else if (!uptodate) {
 				err = btrfs_read_buffer(cur, gen);
 				if (err) {
@@ -1680,6 +1731,8 @@ static noinline struct extent_buffer *read_node_slot(struct btrfs_root *root,
 				   struct extent_buffer *parent, int slot)
 {
 	int level = btrfs_header_level(parent);
+	struct extent_buffer *eb;
+
 	if (slot < 0)
 		return NULL;
 	if (slot >= btrfs_header_nritems(parent))
@@ -1687,9 +1740,15 @@ static noinline struct extent_buffer *read_node_slot(struct btrfs_root *root,
 
 	BUG_ON(level == 0);
 
-	return read_tree_block(root, btrfs_node_blockptr(parent, slot),
-		       btrfs_level_size(root, level - 1),
-		       btrfs_node_ptr_generation(parent, slot));
+	eb = read_tree_block(root, btrfs_node_blockptr(parent, slot),
+			     btrfs_level_size(root, level - 1),
+			     btrfs_node_ptr_generation(parent, slot));
+	if (eb && !extent_buffer_uptodate(eb)) {
+		free_extent_buffer(eb);
+		eb = NULL;
+	}
+
+	return eb;
 }
 
 /*
@@ -1754,7 +1813,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 			goto enospc;
 		}
 
-		tree_mod_log_set_root_pointer(root, child);
+		tree_mod_log_set_root_pointer(root, child, 1);
 		rcu_assign_pointer(root->node, child);
 
 		add_root_to_dirty_list(root);
@@ -1818,7 +1877,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 		if (btrfs_header_nritems(right) == 0) {
 			clean_tree_block(trans, root, right);
 			btrfs_tree_unlock(right);
-			del_ptr(trans, root, path, level + 1, pslot + 1);
+			del_ptr(root, path, level + 1, pslot + 1);
 			root_sub_used(root, right->len);
 			btrfs_free_tree_block(trans, root, right, 0, 1);
 			free_extent_buffer_stale(right);
@@ -1862,7 +1921,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 	if (btrfs_header_nritems(mid) == 0) {
 		clean_tree_block(trans, root, mid);
 		btrfs_tree_unlock(mid);
-		del_ptr(trans, root, path, level + 1, pslot);
+		del_ptr(root, path, level + 1, pslot);
 		root_sub_used(root, mid->len);
 		btrfs_free_tree_block(trans, root, mid, 0, 1);
 		free_extent_buffer_stale(mid);
@@ -2210,9 +2269,6 @@ static noinline void unlock_up(struct btrfs_path *path, int level,
 	int no_skips = 0;
 	struct extent_buffer *t;
 
-	if (path->really_keep_locks)
-		return;
-
 	for (i = level; i < BTRFS_MAX_LEVEL; i++) {
 		if (!path->nodes[i])
 			break;
@@ -2260,7 +2316,7 @@ noinline void btrfs_unlock_up_safe(struct btrfs_path *path, int level)
 {
 	int i;
 
-	if (path->keep_locks || path->really_keep_locks)
+	if (path->keep_locks)
 		return;
 
 	for (i = level; i < BTRFS_MAX_LEVEL; i++) {
@@ -2493,7 +2549,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
 	if (!cow)
 		write_lock_level = -1;
 
-	if (cow && (p->really_keep_locks || p->keep_locks || p->lowest_level))
+	if (cow && (p->keep_locks || p->lowest_level))
 		write_lock_level = BTRFS_MAX_LEVEL;
 
 	min_write_lock_level = write_lock_level;
@@ -2795,15 +2851,9 @@ again:
 				btrfs_clear_path_blocking(p, b,
 							  BTRFS_READ_LOCK);
 			}
+			b = tree_mod_log_rewind(root->fs_info, b, time_seq);
 			p->locks[level] = BTRFS_READ_LOCK;
 			p->nodes[level] = b;
-			b = tree_mod_log_rewind(root->fs_info, b, time_seq);
-			if (b != p->nodes[level]) {
-				btrfs_tree_unlock_rw(p->nodes[level],
-						     p->locks[level]);
-				p->locks[level] = 0;
-				p->nodes[level] = b;
-			}
 		} else {
 			p->slots[level] = slot;
 			unlock_up(p, level, lowest_unlock, 0, NULL);
@@ -2902,8 +2952,7 @@ again:
  * higher levels
  *
  */
-static void fixup_low_keys(struct btrfs_trans_handle *trans,
-			   struct btrfs_root *root, struct btrfs_path *path,
+static void fixup_low_keys(struct btrfs_root *root, struct btrfs_path *path,
 			   struct btrfs_disk_key *key, int level)
 {
 	int i;
@@ -2928,8 +2977,7 @@ static void fixup_low_keys(struct btrfs_trans_handle *trans,
  * This function isn't completely safe. It's the caller's responsibility
  * that the new key won't break the order
  */
-void btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *root, struct btrfs_path *path,
+void btrfs_set_item_key_safe(struct btrfs_root *root, struct btrfs_path *path,
 			     struct btrfs_key *new_key)
 {
 	struct btrfs_disk_key disk_key;
@@ -2951,7 +2999,7 @@ void btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
 	btrfs_set_item_key(eb, &disk_key, slot);
 	btrfs_mark_buffer_dirty(eb);
 	if (slot == 0)
-		fixup_low_keys(trans, root, path, &disk_key, 1);
+		fixup_low_keys(root, path, &disk_key, 1);
 }
 
 /*
@@ -2998,7 +3046,7 @@ static int push_node_left(struct btrfs_trans_handle *trans,
 		push_items = min(src_nritems - 8, push_items);
 
 	tree_mod_log_eb_copy(root->fs_info, dst, src, dst_nritems, 0,
-			     push_items, 1);
+			     push_items);
 	copy_extent_buffer(dst, src,
 			   btrfs_node_key_ptr_offset(dst_nritems),
 			   btrfs_node_key_ptr_offset(0),
@@ -3069,7 +3117,7 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
 				      sizeof(struct btrfs_key_ptr));
 
 	tree_mod_log_eb_copy(root->fs_info, dst, src, 0,
-			     src_nritems - push_items, push_items, 1);
+			     src_nritems - push_items, push_items);
 	copy_extent_buffer(dst, src,
 			   btrfs_node_key_ptr_offset(0),
 			   btrfs_node_key_ptr_offset(src_nritems - push_items),
@@ -3093,7 +3141,7 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
  */
 static noinline int insert_new_root(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root,
-			   struct btrfs_path *path, int level)
+			   struct btrfs_path *path, int level, int log_removal)
 {
 	u64 lower_gen;
 	struct extent_buffer *lower;
@@ -3144,7 +3192,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
 	btrfs_mark_buffer_dirty(c);
 
 	old = root->node;
-	tree_mod_log_set_root_pointer(root, c);
+	tree_mod_log_set_root_pointer(root, c, log_removal);
 	rcu_assign_pointer(root->node, c);
 
 	/* the super has an extra ref to root->node */
@@ -3221,18 +3269,21 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 	int mid;
 	int ret;
 	u32 c_nritems;
-	int tree_mod_log_removal = 1;
 
 	c = path->nodes[level];
 	WARN_ON(btrfs_header_generation(c) != trans->transid);
 	if (c == root->node) {
-		/* trying to split the root, lets make a new one */
-		ret = insert_new_root(trans, root, path, level + 1);
 		/*
-		 * removal of root nodes has been logged by
-		 * tree_mod_log_set_root_pointer due to locking
+		 * trying to split the root, lets make a new one
+		 *
+		 * tree mod log: We pass 0 as log_removal parameter to
+		 * insert_new_root, because that root buffer will be kept as a
+		 * normal node. We are going to log removal of half of the
+		 * elements below with tree_mod_log_eb_copy. We're holding a
+		 * tree lock on the buffer, which is why we cannot race with
+		 * other tree_mod_log users.
 		 */
-		tree_mod_log_removal = 0;
+		ret = insert_new_root(trans, root, path, level + 1, 0);
 		if (ret)
 			return ret;
 	} else {
@@ -3270,8 +3321,7 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 			    (unsigned long)btrfs_header_chunk_tree_uuid(split),
 			    BTRFS_UUID_SIZE);
 
-	tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid,
-			     tree_mod_log_removal);
+	tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid);
 	copy_extent_buffer(split, c,
 			   btrfs_node_key_ptr_offset(0),
 			   btrfs_node_key_ptr_offset(mid),
@@ -3687,7 +3737,7 @@ static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
 		clean_tree_block(trans, root, right);
 
 	btrfs_item_key(right, &disk_key, 0);
-	fixup_low_keys(trans, root, path, &disk_key, 1);
+	fixup_low_keys(root, path, &disk_key, 1);
 
 	/* then fixup the leaf pointer in the path */
 	if (path->slots[0] < push_items) {
@@ -3953,7 +4003,7 @@ static noinline int split_leaf(struct btrfs_trans_handle *trans,
 	}
 
 	if (!path->nodes[1]) {
-		ret = insert_new_root(trans, root, path, 1);
+		ret = insert_new_root(trans, root, path, 1, 1);
 		if (ret)
 			return ret;
 	}
@@ -4047,8 +4097,7 @@ again:
 			path->nodes[0] = right;
 			path->slots[0] = 0;
 			if (path->slots[1] == 0)
-				fixup_low_keys(trans, root, path,
-					       &disk_key, 1);
+				fixup_low_keys(root, path, &disk_key, 1);
 		}
 		btrfs_mark_buffer_dirty(right);
 		return ret;
@@ -4264,7 +4313,7 @@ int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
 		return ret;
 
 	path->slots[0]++;
-	setup_items_for_insert(trans, root, path, new_key, &item_size,
+	setup_items_for_insert(root, path, new_key, &item_size,
 			       item_size, item_size +
 			       sizeof(struct btrfs_item), 1);
 	leaf = path->nodes[0];
@@ -4281,9 +4330,7 @@ int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
  * off the end of the item or if we shift the item to chop bytes off
  * the front.
  */
-void btrfs_truncate_item(struct btrfs_trans_handle *trans,
-			 struct btrfs_root *root,
-			 struct btrfs_path *path,
+void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
 			 u32 new_size, int from_end)
 {
 	int slot;
@@ -4367,7 +4414,7 @@ void btrfs_truncate_item(struct btrfs_trans_handle *trans,
 		btrfs_set_disk_key_offset(&disk_key, offset + size_diff);
 		btrfs_set_item_key(leaf, &disk_key, slot);
 		if (slot == 0)
-			fixup_low_keys(trans, root, path, &disk_key, 1);
+			fixup_low_keys(root, path, &disk_key, 1);
 	}
 
 	item = btrfs_item_nr(leaf, slot);
@@ -4383,8 +4430,7 @@ void btrfs_truncate_item(struct btrfs_trans_handle *trans,
 /*
  * make the item pointed to by the path bigger, data_size is the new size.
  */
-void btrfs_extend_item(struct btrfs_trans_handle *trans,
-		       struct btrfs_root *root, struct btrfs_path *path,
+void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
 		       u32 data_size)
 {
 	int slot;
@@ -4454,8 +4500,7 @@ void btrfs_extend_item(struct btrfs_trans_handle *trans,
  * to save stack depth by doing the bulk of the work in a function
  * that doesn't call btrfs_search_slot
  */
-void setup_items_for_insert(struct btrfs_trans_handle *trans,
-			    struct btrfs_root *root, struct btrfs_path *path,
+void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 			    struct btrfs_key *cpu_key, u32 *data_size,
 			    u32 total_data, u32 total_size, int nr)
 {
@@ -4531,7 +4576,7 @@ void setup_items_for_insert(struct btrfs_trans_handle *trans,
 
 	if (slot == 0) {
 		btrfs_cpu_key_to_disk(&disk_key, cpu_key);
-		fixup_low_keys(trans, root, path, &disk_key, 1);
+		fixup_low_keys(root, path, &disk_key, 1);
 	}
 	btrfs_unlock_up_safe(path, 1);
 	btrfs_mark_buffer_dirty(leaf);
@@ -4571,7 +4616,7 @@ int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
 	slot = path->slots[0];
 	BUG_ON(slot < 0);
 
-	setup_items_for_insert(trans, root, path, cpu_key, data_size,
+	setup_items_for_insert(root, path, cpu_key, data_size,
 			       total_data, total_size, nr);
 	return 0;
 }
@@ -4609,8 +4654,8 @@ int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
  * the tree should have been previously balanced so the deletion does not
  * empty a node.
  */
-static void del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
-		    struct btrfs_path *path, int level, int slot)
+static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
+		    int level, int slot)
 {
 	struct extent_buffer *parent = path->nodes[level];
 	u32 nritems;
@@ -4642,7 +4687,7 @@ static void del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		struct btrfs_disk_key disk_key;
 
 		btrfs_node_key(parent, &disk_key, 0);
-		fixup_low_keys(trans, root, path, &disk_key, level + 1);
+		fixup_low_keys(root, path, &disk_key, level + 1);
 	}
 	btrfs_mark_buffer_dirty(parent);
 }
@@ -4663,7 +4708,7 @@ static noinline void btrfs_del_leaf(struct btrfs_trans_handle *trans,
 				    struct extent_buffer *leaf)
 {
 	WARN_ON(btrfs_header_generation(leaf) != trans->transid);
-	del_ptr(trans, root, path, 1, path->slots[1]);
+	del_ptr(root, path, 1, path->slots[1]);
 
 	/*
 	 * btrfs_free_extent is expensive, we want to make sure we
@@ -4744,7 +4789,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 			struct btrfs_disk_key disk_key;
 
 			btrfs_item_key(leaf, &disk_key, 0);
-			fixup_low_keys(trans, root, path, &disk_key, 1);
+			fixup_low_keys(root, path, &disk_key, 1);
 		}
 
 		/* delete the leaf if it is mostly empty */
@@ -5464,139 +5509,6 @@ int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path)
 	return btrfs_next_old_leaf(root, path, 0);
 }
 
-/* Release the path up to but not including the given level */
-static void btrfs_release_level(struct btrfs_path *path, int level)
-{
-	int i;
-
-	for (i = 0; i < level; i++) {
-		path->slots[i] = 0;
-		if (!path->nodes[i])
-			continue;
-		if (path->locks[i]) {
-			btrfs_tree_unlock_rw(path->nodes[i], path->locks[i]);
-			path->locks[i] = 0;
-		}
-		free_extent_buffer(path->nodes[i]);
-		path->nodes[i] = NULL;
-	}
-}
-
-/*
- * This function assumes 2 things
- *
- * 1) You are using path->keep_locks
- * 2) You are not inserting items.
- *
- * If either of these are not true do not use this function. If you need a next
- * leaf with either of these not being true then this function can be easily
- * adapted to do that, but at the moment these are the limitations.
- */
-int btrfs_next_leaf_write(struct btrfs_trans_handle *trans,
-			  struct btrfs_root *root, struct btrfs_path *path,
-			  int del)
-{
-	struct extent_buffer *b;
-	struct btrfs_key key;
-	u32 nritems;
-	int level = 1;
-	int slot;
-	int ret = 1;
-	int write_lock_level = BTRFS_MAX_LEVEL;
-	int ins_len = del ? -1 : 0;
-
-	WARN_ON(!(path->keep_locks || path->really_keep_locks));
-
-	nritems = btrfs_header_nritems(path->nodes[0]);
-	btrfs_item_key_to_cpu(path->nodes[0], &key, nritems - 1);
-
-	while (path->nodes[level]) {
-		nritems = btrfs_header_nritems(path->nodes[level]);
-		if (!(path->locks[level] & BTRFS_WRITE_LOCK)) {
-search:
-			btrfs_release_path(path);
-			ret = btrfs_search_slot(trans, root, &key, path,
-						ins_len, 1);
-			if (ret < 0)
-				goto out;
-			level = 1;
-			continue;
-		}
-
-		if (path->slots[level] >= nritems - 1) {
-			level++;
-			continue;
-		}
-
-		btrfs_release_level(path, level);
-		break;
-	}
-
-	if (!path->nodes[level]) {
-		ret = 1;
-		goto out;
-	}
-
-	path->slots[level]++;
-	b = path->nodes[level];
-
-	while (b) {
-		level = btrfs_header_level(b);
-
-		if (!should_cow_block(trans, root, b))
-			goto cow_done;
-
-		btrfs_set_path_blocking(path);
-		ret = btrfs_cow_block(trans, root, b,
-				      path->nodes[level + 1],
-				      path->slots[level + 1], &b);
-		if (ret)
-			goto out;
-cow_done:
-		path->nodes[level] = b;
-		btrfs_clear_path_blocking(path, NULL, 0);
-		if (level != 0) {
-			ret = setup_nodes_for_search(trans, root, path, b,
-						     level, ins_len,
-						     &write_lock_level);
-			if (ret == -EAGAIN)
-				goto search;
-			if (ret)
-				goto out;
-
-			b = path->nodes[level];
-			slot = path->slots[level];
-
-			ret = read_block_for_search(trans, root, path,
-						    &b, level, slot, &key, 0);
-			if (ret == -EAGAIN)
-				goto search;
-			if (ret)
-				goto out;
-			level = btrfs_header_level(b);
-			if (!btrfs_try_tree_write_lock(b)) {
-				btrfs_set_path_blocking(path);
-				btrfs_tree_lock(b);
-				btrfs_clear_path_blocking(path, b,
-							  BTRFS_WRITE_LOCK);
-			}
-			path->locks[level] = BTRFS_WRITE_LOCK;
-			path->nodes[level] = b;
-			path->slots[level] = 0;
-		} else {
-			path->slots[level] = 0;
-			ret = 0;
-			break;
-		}
-	}
-
-out:
-	if (ret)
-		btrfs_release_path(path);
-
-	return ret;
-}
-
 int btrfs_next_old_leaf(struct btrfs_root *root, struct btrfs_path *path,
 			u64 time_seq)
 {

fs/btrfs/ctree.h

@@ -340,6 +340,7 @@ static inline unsigned long btrfs_chunk_item_size(int num_stripes)
  */
 #define BTRFS_FS_STATE_ERROR		0
 #define BTRFS_FS_STATE_REMOUNTING	1
+#define BTRFS_FS_STATE_TRANS_ABORTED	2
 
 /* Super block flags */
 /* Errors detected */
@@ -508,6 +509,7 @@ 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_COMPAT_SUPP		0ULL
 #define BTRFS_FEATURE_COMPAT_RO_SUPP		0ULL
@@ -518,7 +520,8 @@ struct btrfs_super_block {
 	 BTRFS_FEATURE_INCOMPAT_BIG_METADATA |		\
 	 BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO |		\
 	 BTRFS_FEATURE_INCOMPAT_RAID56 |		\
-	 BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF)
+	 BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF |		\
+	 BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA)
 
 /*
  * A leaf is full of items. offset and size tell us where to find
@@ -583,7 +586,6 @@ struct btrfs_path {
 	unsigned int skip_locking:1;
 	unsigned int leave_spinning:1;
 	unsigned int search_commit_root:1;
-	unsigned int really_keep_locks:1;
 };
 
 /*
@@ -1019,9 +1021,9 @@ struct btrfs_block_group_item {
  */
 #define BTRFS_QGROUP_STATUS_FLAG_ON		(1ULL << 0)
 /*
- * SCANNING is set during the initialization phase
+ * RESCAN is set during the initialization phase
  */
-#define BTRFS_QGROUP_STATUS_FLAG_SCANNING	(1ULL << 1)
+#define BTRFS_QGROUP_STATUS_FLAG_RESCAN		(1ULL << 1)
 /*
  * Some qgroup entries are known to be out of date,
  * either because the configuration has changed in a way that
@@ -1050,7 +1052,7 @@ struct btrfs_qgroup_status_item {
 	 * only used during scanning to record the progress
 	 * of the scan. It contains a logical address
 	 */
-	__le64 scan;
+	__le64 rescan;
 } __attribute__ ((__packed__));
 
 struct btrfs_qgroup_info_item {
@@ -1360,6 +1362,17 @@ struct btrfs_fs_info {
 	wait_queue_head_t transaction_blocked_wait;
 	wait_queue_head_t async_submit_wait;
 
+	/*
+	 * Used to protect the incompat_flags, compat_flags, compat_ro_flags
+	 * when they are updated.
+	 *
+	 * Because we do not clear the flags for ever, so we needn't use
+	 * the lock on the read side.
+	 *
+	 * We also needn't use the lock when we mount the fs, because
+	 * there is no other task which will update the flag.
+	 */
+	spinlock_t super_lock;
 	struct btrfs_super_block *super_copy;
 	struct btrfs_super_block *super_for_commit;
 	struct block_device *__bdev;
@@ -1409,7 +1422,7 @@ struct btrfs_fs_info {
 
 	/* this protects tree_mod_seq_list */
 	spinlock_t tree_mod_seq_lock;
-	atomic_t tree_mod_seq;
+	atomic64_t tree_mod_seq;
 	struct list_head tree_mod_seq_list;
 	struct seq_list tree_mod_seq_elem;
 
@@ -1581,12 +1594,20 @@ struct btrfs_fs_info {
 	struct rb_root qgroup_tree;
 	spinlock_t qgroup_lock;
 
+	/* protect user change for quota operations */
+	struct mutex qgroup_ioctl_lock;
+
 	/* list of dirty qgroups to be written at next commit */
 	struct list_head dirty_qgroups;
 
 	/* used by btrfs_qgroup_record_ref for an efficient tree traversal */
 	u64 qgroup_seq;
 
+	/* qgroup rescan items */
+	struct mutex qgroup_rescan_lock; /* protects the progress item */
+	struct btrfs_key qgroup_rescan_progress;
+	struct btrfs_workers qgroup_rescan_workers;
+
 	/* filesystem state */
 	unsigned long fs_state;
 
@@ -1808,6 +1829,12 @@ struct btrfs_ioctl_defrag_range_args {
  */
 #define BTRFS_EXTENT_ITEM_KEY	168
 
+/*
+ * The same as the BTRFS_EXTENT_ITEM_KEY, except it's metadata we already know
+ * the length, so we save the level in key->offset instead of the length.
+ */
+#define BTRFS_METADATA_ITEM_KEY	169
+
 #define BTRFS_TREE_BLOCK_REF_KEY	176
 
 #define BTRFS_EXTENT_DATA_REF_KEY	178
@@ -2766,8 +2793,10 @@ BTRFS_SETGET_STACK_FUNCS(super_cache_generation, struct btrfs_super_block,
 
 static inline int btrfs_super_csum_size(struct btrfs_super_block *s)
 {
-	int t = btrfs_super_csum_type(s);
-	BUG_ON(t >= ARRAY_SIZE(btrfs_csum_sizes));
+	u16 t = btrfs_super_csum_type(s);
+	/*
+	 * csum type is validated at mount time
+	 */
 	return btrfs_csum_sizes[t];
 }
 
@@ -2864,8 +2893,8 @@ BTRFS_SETGET_FUNCS(qgroup_status_version, struct btrfs_qgroup_status_item,
 		   version, 64);
 BTRFS_SETGET_FUNCS(qgroup_status_flags, struct btrfs_qgroup_status_item,
 		   flags, 64);
-BTRFS_SETGET_FUNCS(qgroup_status_scan, struct btrfs_qgroup_status_item,
-		   scan, 64);
+BTRFS_SETGET_FUNCS(qgroup_status_rescan, struct btrfs_qgroup_status_item,
+		   rescan, 64);
 
 /* btrfs_qgroup_info_item */
 BTRFS_SETGET_FUNCS(qgroup_info_generation, struct btrfs_qgroup_info_item,
@@ -3005,7 +3034,7 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 int btrfs_lookup_extent(struct btrfs_root *root, u64 start, u64 len);
 int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root, u64 bytenr,
-			     u64 num_bytes, u64 *refs, u64 *flags);
+			     u64 offset, int metadata, u64 *refs, u64 *flags);
 int btrfs_pin_extent(struct btrfs_root *root,
 		     u64 bytenr, u64 num, int reserved);
 int btrfs_pin_extent_for_log_replay(struct btrfs_root *root,
@@ -3017,8 +3046,6 @@ 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);
-u64 btrfs_find_block_group(struct btrfs_root *root,
-			   u64 search_start, u64 search_hint, int owner);
 struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
 					struct btrfs_root *root, u32 blocksize,
 					u64 parent, u64 root_objectid,
@@ -3028,10 +3055,6 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root,
 			   struct extent_buffer *buf,
 			   u64 parent, int last_ref);
-struct extent_buffer *btrfs_init_new_buffer(struct btrfs_trans_handle *trans,
-					    struct btrfs_root *root,
-					    u64 bytenr, u32 blocksize,
-					    int level);
 int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root,
 				     u64 root_objectid, u64 owner,
@@ -3044,7 +3067,7 @@ int btrfs_reserve_extent(struct btrfs_trans_handle *trans,
 				  struct btrfs_root *root,
 				  u64 num_bytes, u64 min_alloc_size,
 				  u64 empty_size, u64 hint_byte,
-				  struct btrfs_key *ins, u64 data);
+				  struct btrfs_key *ins, int is_data);
 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		  struct extent_buffer *buf, int full_backref, int for_cow);
 int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
@@ -3084,7 +3107,6 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root, u64 group_start);
 void btrfs_create_pending_block_groups(struct btrfs_trans_handle *trans,
 				       struct btrfs_root *root);
-u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags);
 u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data);
 void btrfs_clear_space_info_full(struct btrfs_fs_info *info);
 
@@ -3161,8 +3183,7 @@ 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);
-void btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *root, struct btrfs_path *path,
+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);
 struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root);
@@ -3198,12 +3219,9 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans,
 		      struct extent_buffer **cow_ret, u64 new_root_objectid);
 int btrfs_block_can_be_shared(struct btrfs_root *root,
 			      struct extent_buffer *buf);
-void btrfs_extend_item(struct btrfs_trans_handle *trans,
-		       struct btrfs_root *root, struct btrfs_path *path,
+void btrfs_extend_item(struct btrfs_root *root, struct btrfs_path *path,
 		       u32 data_size);
-void btrfs_truncate_item(struct btrfs_trans_handle *trans,
-			 struct btrfs_root *root,
-			 struct btrfs_path *path,
+void btrfs_truncate_item(struct btrfs_root *root, struct btrfs_path *path,
 			 u32 new_size, int from_end);
 int btrfs_split_item(struct btrfs_trans_handle *trans,
 		     struct btrfs_root *root,
@@ -3243,8 +3261,7 @@ static inline int btrfs_del_item(struct btrfs_trans_handle *trans,
 	return btrfs_del_items(trans, root, path, path->slots[0], 1);
 }
 
-void setup_items_for_insert(struct btrfs_trans_handle *trans,
-			    struct btrfs_root *root, struct btrfs_path *path,
+void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 			    struct btrfs_key *cpu_key, u32 *data_size,
 			    u32 total_data, u32 total_size, int nr);
 int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
@@ -3264,9 +3281,6 @@ 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_next_leaf_write(struct btrfs_trans_handle *trans,
-			  struct btrfs_root *root, struct btrfs_path *path,
-			  int del);
 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,
@@ -3281,7 +3295,6 @@ static inline int btrfs_next_item(struct btrfs_root *root, struct btrfs_path *p)
 {
 	return btrfs_next_old_item(root, p, 0);
 }
-int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path);
 int btrfs_leaf_free_space(struct btrfs_root *root, struct extent_buffer *leaf);
 int __must_check btrfs_drop_snapshot(struct btrfs_root *root,
 				     struct btrfs_block_rsv *block_rsv,
@@ -3318,10 +3331,7 @@ u64 btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info,
 			   struct seq_list *elem);
 void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,
 			    struct seq_list *elem);
-static inline u64 btrfs_inc_tree_mod_seq(struct btrfs_fs_info *fs_info)
-{
-	return atomic_inc_return(&fs_info->tree_mod_seq);
-}
+u64 btrfs_tree_mod_seq_prev(u64 seq);
 int btrfs_old_root_level(struct btrfs_root *root, u64 time_seq);
 
 /* root-item.c */
@@ -3345,9 +3355,8 @@ int __must_check btrfs_update_root(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   struct btrfs_key *key,
 				   struct btrfs_root_item *item);
-void btrfs_read_root_item(struct btrfs_root *root,
-			 struct extent_buffer *eb, int slot,
-			 struct btrfs_root_item *item);
+void btrfs_read_root_item(struct extent_buffer *eb, int slot,
+			  struct btrfs_root_item *item);
 int btrfs_find_last_root(struct btrfs_root *root, u64 objectid, struct
 			 btrfs_root_item *item, struct btrfs_key *key);
 int btrfs_find_dead_roots(struct btrfs_root *root, u64 objectid);
@@ -3380,9 +3389,6 @@ struct btrfs_dir_item *
 btrfs_search_dir_index_item(struct btrfs_root *root,
 			    struct btrfs_path *path, u64 dirid,
 			    const char *name, int name_len);
-struct btrfs_dir_item *btrfs_match_dir_item_name(struct btrfs_root *root,
-			      struct btrfs_path *path,
-			      const char *name, int name_len);
 int btrfs_delete_one_dir_name(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root,
 			      struct btrfs_path *path,
@@ -3460,16 +3466,11 @@ int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root,
 			     struct btrfs_path *path, u64 objectid,
 			     u64 bytenr, int mod);
-u64 btrfs_file_extent_length(struct btrfs_path *path);
 int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans,
 			   struct btrfs_root *root,
 			   struct btrfs_ordered_sum *sums);
 int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
 		       struct bio *bio, u64 file_start, int contig);
-struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans,
-					  struct btrfs_root *root,
-					  struct btrfs_path *path,
-					  u64 bytenr, int cow);
 int btrfs_csum_truncate(struct btrfs_trans_handle *trans,
 			struct btrfs_root *root, struct btrfs_path *path,
 			u64 isize);
@@ -3531,8 +3532,6 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans,
 int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput);
 int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end,
 			      struct extent_state **cached_state);
-int btrfs_writepages(struct address_space *mapping,
-		     struct writeback_control *wbc);
 int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *new_root, u64 new_dirid);
 int btrfs_merge_bio_hook(int rw, struct page *page, unsigned long offset,
@@ -3542,7 +3541,6 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf);
 int btrfs_readpage(struct file *file, struct page *page);
 void btrfs_evict_inode(struct inode *inode);
 int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc);
-int btrfs_dirty_inode(struct inode *inode);
 struct inode *btrfs_alloc_inode(struct super_block *sb);
 void btrfs_destroy_inode(struct inode *inode);
 int btrfs_drop_inode(struct inode *inode);
@@ -3560,7 +3558,6 @@ int btrfs_update_inode(struct btrfs_trans_handle *trans,
 int btrfs_update_inode_fallback(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root, struct inode *inode);
 int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode);
-int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode);
 int btrfs_orphan_cleanup(struct btrfs_root *root);
 void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root);
@@ -3611,7 +3608,6 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 int btrfs_mark_extent_written(struct btrfs_trans_handle *trans,
 			      struct inode *inode, u64 start, u64 end);
 int btrfs_release_file(struct inode *inode, struct file *file);
-void btrfs_drop_pages(struct page **pages, size_t num_pages);
 int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode,
 		      struct page **pages, size_t num_pages,
 		      loff_t pos, size_t write_bytes,
@@ -3634,14 +3630,31 @@ int btrfs_sync_fs(struct super_block *sb, int wait);
 
 #ifdef CONFIG_PRINTK
 __printf(2, 3)
-void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...);
+void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...);
 #else
 static inline __printf(2, 3)
-void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...)
+void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...)
 {
 }
 #endif
 
+#define btrfs_emerg(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_EMERG fmt, ##args)
+#define btrfs_alert(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_ALERT fmt, ##args)
+#define btrfs_crit(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_CRIT fmt, ##args)
+#define btrfs_err(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_ERR fmt, ##args)
+#define btrfs_warn(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_WARNING fmt, ##args)
+#define btrfs_notice(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_NOTICE fmt, ##args)
+#define btrfs_info(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_INFO fmt, ##args)
+#define btrfs_debug(fs_info, fmt, args...) \
+	btrfs_printk(fs_info, KERN_DEBUG fmt, ##args)
+
 __printf(5, 6)
 void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		     unsigned int line, int errno, const char *fmt, ...);
@@ -3663,11 +3676,28 @@ static inline void __btrfs_set_fs_incompat(struct btrfs_fs_info *fs_info,
 	disk_super = fs_info->super_copy;
 	features = btrfs_super_incompat_flags(disk_super);
 	if (!(features & flag)) {
-		features |= flag;
-		btrfs_set_super_incompat_flags(disk_super, features);
+		spin_lock(&fs_info->super_lock);
+		features = btrfs_super_incompat_flags(disk_super);
+		if (!(features & flag)) {
+			features |= flag;
+			btrfs_set_super_incompat_flags(disk_super, features);
+			printk(KERN_INFO "btrfs: setting %llu feature flag\n",
+					 flag);
+		}
+		spin_unlock(&fs_info->super_lock);
 	}
 }
 
+#define btrfs_fs_incompat(fs_info, opt) \
+	__btrfs_fs_incompat((fs_info), BTRFS_FEATURE_INCOMPAT_##opt)
+
+static inline int __btrfs_fs_incompat(struct btrfs_fs_info *fs_info, u64 flag)
+{
+	struct btrfs_super_block *disk_super;
+	disk_super = fs_info->super_copy;
+	return !!(btrfs_super_incompat_flags(disk_super) & flag);
+}
+
 /*
  * Call btrfs_abort_transaction as early as possible when an error condition is
  * detected, that way the exact line number is reported.
@@ -3753,7 +3783,6 @@ void btrfs_scrub_continue_super(struct btrfs_root *root);
 int btrfs_scrub_cancel(struct btrfs_fs_info *info);
 int btrfs_scrub_cancel_dev(struct btrfs_fs_info *info,
 			   struct btrfs_device *dev);
-int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid);
 int btrfs_scrub_progress(struct btrfs_root *root, u64 devid,
 			 struct btrfs_scrub_progress *progress);
 
@@ -3784,7 +3813,7 @@ int btrfs_quota_enable(struct btrfs_trans_handle *trans,
 		       struct btrfs_fs_info *fs_info);
 int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 			struct btrfs_fs_info *fs_info);
-int btrfs_quota_rescan(struct btrfs_fs_info *fs_info);
+int btrfs_qgroup_rescan(struct btrfs_fs_info *fs_info);
 int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans,
 			      struct btrfs_fs_info *fs_info, u64 src, u64 dst);
 int btrfs_del_qgroup_relation(struct btrfs_trans_handle *trans,

fs/btrfs/delayed-inode.c

@@ -202,7 +202,7 @@ static void btrfs_dequeue_delayed_node(struct btrfs_delayed_root *root,
 	spin_unlock(&root->lock);
 }
 
-struct btrfs_delayed_node *btrfs_first_delayed_node(
+static struct btrfs_delayed_node *btrfs_first_delayed_node(
 			struct btrfs_delayed_root *delayed_root)
 {
 	struct list_head *p;
@@ -221,7 +221,7 @@ out:
 	return node;
 }
 
-struct btrfs_delayed_node *btrfs_next_delayed_node(
+static struct btrfs_delayed_node *btrfs_next_delayed_node(
 						struct btrfs_delayed_node *node)
 {
 	struct btrfs_delayed_root *delayed_root;
@@ -282,7 +282,7 @@ static inline void btrfs_release_delayed_node(struct btrfs_delayed_node *node)
 	__btrfs_release_delayed_node(node, 0);
 }
 
-struct btrfs_delayed_node *btrfs_first_prepared_delayed_node(
+static struct btrfs_delayed_node *btrfs_first_prepared_delayed_node(
 					struct btrfs_delayed_root *delayed_root)
 {
 	struct list_head *p;
@@ -308,7 +308,7 @@ static inline void btrfs_release_prepared_delayed_node(
 	__btrfs_release_delayed_node(node, 1);
 }
 
-struct btrfs_delayed_item *btrfs_alloc_delayed_item(u32 data_len)
+static struct btrfs_delayed_item *btrfs_alloc_delayed_item(u32 data_len)
 {
 	struct btrfs_delayed_item *item;
 	item = kmalloc(sizeof(*item) + data_len, GFP_NOFS);
@@ -383,7 +383,7 @@ static struct btrfs_delayed_item *__btrfs_lookup_delayed_item(
 	return NULL;
 }
 
-struct btrfs_delayed_item *__btrfs_lookup_delayed_insertion_item(
+static struct btrfs_delayed_item *__btrfs_lookup_delayed_insertion_item(
 					struct btrfs_delayed_node *delayed_node,
 					struct btrfs_key *key)
 {
@@ -394,45 +394,6 @@ struct btrfs_delayed_item *__btrfs_lookup_delayed_insertion_item(
 	return item;
 }
 
-struct btrfs_delayed_item *__btrfs_lookup_delayed_deletion_item(
-					struct btrfs_delayed_node *delayed_node,
-					struct btrfs_key *key)
-{
-	struct btrfs_delayed_item *item;
-
-	item = __btrfs_lookup_delayed_item(&delayed_node->del_root, key,
-					   NULL, NULL);
-	return item;
-}
-
-struct btrfs_delayed_item *__btrfs_search_delayed_insertion_item(
-					struct btrfs_delayed_node *delayed_node,
-					struct btrfs_key *key)
-{
-	struct btrfs_delayed_item *item, *next;
-
-	item = __btrfs_lookup_delayed_item(&delayed_node->ins_root, key,
-					   NULL, &next);
-	if (!item)
-		item = next;
-
-	return item;
-}
-
-struct btrfs_delayed_item *__btrfs_search_delayed_deletion_item(
-					struct btrfs_delayed_node *delayed_node,
-					struct btrfs_key *key)
-{
-	struct btrfs_delayed_item *item, *next;
-
-	item = __btrfs_lookup_delayed_item(&delayed_node->del_root, key,
-					   NULL, &next);
-	if (!item)
-		item = next;
-
-	return item;
-}
-
 static int __btrfs_add_delayed_item(struct btrfs_delayed_node *delayed_node,
 				    struct btrfs_delayed_item *ins,
 				    int action)
@@ -535,7 +496,7 @@ static void btrfs_release_delayed_item(struct btrfs_delayed_item *item)
 	}
 }
 
-struct btrfs_delayed_item *__btrfs_first_delayed_insertion_item(
+static struct btrfs_delayed_item *__btrfs_first_delayed_insertion_item(
 					struct btrfs_delayed_node *delayed_node)
 {
 	struct rb_node *p;
@@ -548,7 +509,7 @@ struct btrfs_delayed_item *__btrfs_first_delayed_insertion_item(
 	return item;
 }
 
-struct btrfs_delayed_item *__btrfs_first_delayed_deletion_item(
+static struct btrfs_delayed_item *__btrfs_first_delayed_deletion_item(
 					struct btrfs_delayed_node *delayed_node)
 {
 	struct rb_node *p;
@@ -561,7 +522,7 @@ struct btrfs_delayed_item *__btrfs_first_delayed_deletion_item(
 	return item;
 }
 
-struct btrfs_delayed_item *__btrfs_next_delayed_item(
+static struct btrfs_delayed_item *__btrfs_next_delayed_item(
 						struct btrfs_delayed_item *item)
 {
 	struct rb_node *p;
@@ -766,10 +727,9 @@ static void btrfs_delayed_inode_release_metadata(struct btrfs_root *root,
  * This helper will insert some continuous items into the same leaf according
  * to the free space of the leaf.
  */
-static int btrfs_batch_insert_items(struct btrfs_trans_handle *trans,
-				struct btrfs_root *root,
-				struct btrfs_path *path,
-				struct btrfs_delayed_item *item)
+static int btrfs_batch_insert_items(struct btrfs_root *root,
+				    struct btrfs_path *path,
+				    struct btrfs_delayed_item *item)
 {
 	struct btrfs_delayed_item *curr, *next;
 	int free_space;
@@ -848,7 +808,7 @@ static int btrfs_batch_insert_items(struct btrfs_trans_handle *trans,
 	btrfs_clear_path_blocking(path, NULL, 0);
 
 	/* insert the keys of the items */
-	setup_items_for_insert(trans, root, path, keys, data_size,
+	setup_items_for_insert(root, path, keys, data_size,
 			       total_data_size, total_size, nitems);
 
 	/* insert the dir index items */
@@ -932,7 +892,7 @@ do_again:
 	if (curr && btrfs_is_continuous_delayed_item(prev, curr)) {
 		/* insert the continuous items into the same leaf */
 		path->slots[0]++;
-		btrfs_batch_insert_items(trans, root, path, curr);
+		btrfs_batch_insert_items(root, path, curr);
 	}
 	btrfs_release_delayed_item(prev);
 	btrfs_mark_buffer_dirty(path->nodes[0]);

fs/btrfs/delayed-ref.c

@@ -40,16 +40,19 @@ struct kmem_cache *btrfs_delayed_extent_op_cachep;
  * compare two delayed tree backrefs with same bytenr and type
  */
 static int comp_tree_refs(struct btrfs_delayed_tree_ref *ref2,
-			  struct btrfs_delayed_tree_ref *ref1)
+			  struct btrfs_delayed_tree_ref *ref1, int type)
 {
-	if (ref1->root < ref2->root)
-		return -1;
-	if (ref1->root > ref2->root)
-		return 1;
-	if (ref1->parent < ref2->parent)
-		return -1;
-	if (ref1->parent > ref2->parent)
-		return 1;
+	if (type == BTRFS_TREE_BLOCK_REF_KEY) {
+		if (ref1->root < ref2->root)
+			return -1;
+		if (ref1->root > ref2->root)
+			return 1;
+	} else {
+		if (ref1->parent < ref2->parent)
+			return -1;
+		if (ref1->parent > ref2->parent)
+			return 1;
+	}
 	return 0;
 }
 
@@ -113,7 +116,8 @@ static int comp_entry(struct btrfs_delayed_ref_node *ref2,
 	if (ref1->type == BTRFS_TREE_BLOCK_REF_KEY ||
 	    ref1->type == BTRFS_SHARED_BLOCK_REF_KEY) {
 		return comp_tree_refs(btrfs_delayed_node_to_tree_ref(ref2),
-				      btrfs_delayed_node_to_tree_ref(ref1));
+				      btrfs_delayed_node_to_tree_ref(ref1),
+				      ref1->type);
 	} else if (ref1->type == BTRFS_EXTENT_DATA_REF_KEY ||
 		   ref1->type == BTRFS_SHARED_DATA_REF_KEY) {
 		return comp_data_refs(btrfs_delayed_node_to_data_ref(ref2),
@@ -357,8 +361,10 @@ int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,
 		elem = list_first_entry(&fs_info->tree_mod_seq_list,
 					struct seq_list, list);
 		if (seq >= elem->seq) {
-			pr_debug("holding back delayed_ref %llu, lowest is "
-				 "%llu (%p)\n", seq, elem->seq, delayed_refs);
+			pr_debug("holding back delayed_ref %#x.%x, lowest is %#x.%x (%p)\n",
+				 (u32)(seq >> 32), (u32)seq,
+				 (u32)(elem->seq >> 32), (u32)elem->seq,
+				 delayed_refs);
 			ret = 1;
 		}
 	}

fs/btrfs/dir-item.c

@@ -21,6 +21,10 @@
 #include "hash.h"
 #include "transaction.h"
 
+static struct btrfs_dir_item *btrfs_match_dir_item_name(struct btrfs_root *root,
+			      struct btrfs_path *path,
+			      const char *name, int name_len);
+
 /*
  * insert a name into a directory, doing overflow properly if there is a hash
  * collision.  data_size indicates how big the item inserted should be.  On
@@ -49,7 +53,7 @@ static struct btrfs_dir_item *insert_with_overflow(struct btrfs_trans_handle
 		di = btrfs_match_dir_item_name(root, path, name, name_len);
 		if (di)
 			return ERR_PTR(-EEXIST);
-		btrfs_extend_item(trans, root, path, data_size);
+		btrfs_extend_item(root, path, data_size);
 	} else if (ret < 0)
 		return ERR_PTR(ret);
 	WARN_ON(ret > 0);
@@ -379,7 +383,7 @@ struct btrfs_dir_item *btrfs_lookup_xattr(struct btrfs_trans_handle *trans,
  * this walks through all the entries in a dir item and finds one
  * for a specific name.
  */
-struct btrfs_dir_item *btrfs_match_dir_item_name(struct btrfs_root *root,
+static struct btrfs_dir_item *btrfs_match_dir_item_name(struct btrfs_root *root,
 			      struct btrfs_path *path,
 			      const char *name, int name_len)
 {
@@ -442,8 +446,7 @@ int btrfs_delete_one_dir_name(struct btrfs_trans_handle *trans,
 		start = btrfs_item_ptr_offset(leaf, path->slots[0]);
 		memmove_extent_buffer(leaf, ptr, ptr + sub_item_len,
 			item_len - (ptr + sub_item_len - start));
-		btrfs_truncate_item(trans, root, path,
-				    item_len - sub_item_len, 1);
+		btrfs_truncate_item(root, path, item_len - sub_item_len, 1);
 	}
 	return ret;
 }

fs/btrfs/disk-io.c

@@ -30,6 +30,7 @@
 #include <linux/slab.h>
 #include <linux/migrate.h>
 #include <linux/ratelimit.h>
+#include <linux/uuid.h>
 #include <asm/unaligned.h>
 #include "compat.h"
 #include "ctree.h"
@@ -69,6 +70,8 @@ static int btrfs_destroy_marked_extents(struct btrfs_root *root,
 					int mark);
 static int btrfs_destroy_pinned_extent(struct btrfs_root *root,
 				       struct extent_io_tree *pinned_extents);
+static int btrfs_cleanup_transaction(struct btrfs_root *root);
+static void btrfs_error_commit_super(struct btrfs_root *root);
 
 /*
  * end_io_wq structs are used to do processing in task context when an IO is
@@ -222,7 +225,7 @@ static struct extent_map *btree_get_extent(struct inode *inode,
 	em->bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
 
 	write_lock(&em_tree->lock);
-	ret = add_extent_mapping(em_tree, em);
+	ret = add_extent_mapping(em_tree, em, 0);
 	if (ret == -EEXIST) {
 		free_extent_map(em);
 		em = lookup_extent_mapping(em_tree, start, len);
@@ -238,7 +241,7 @@ out:
 	return em;
 }
 
-u32 btrfs_csum_data(struct btrfs_root *root, char *data, u32 seed, size_t len)
+u32 btrfs_csum_data(char *data, u32 seed, size_t len)
 {
 	return crc32c(seed, data, len);
 }
@@ -274,7 +277,7 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
 		if (err)
 			return 1;
 		cur_len = min(len, map_len - (offset - map_start));
-		crc = btrfs_csum_data(root, kaddr + offset - map_start,
+		crc = btrfs_csum_data(kaddr + offset - map_start,
 				      crc, cur_len);
 		len -= cur_len;
 		offset += cur_len;
@@ -353,6 +356,49 @@ out:
 	return ret;
 }
 
+/*
+ * Return 0 if the superblock checksum type matches the checksum value of that
+ * algorithm. Pass the raw disk superblock data.
+ */
+static int btrfs_check_super_csum(char *raw_disk_sb)
+{
+	struct btrfs_super_block *disk_sb =
+		(struct btrfs_super_block *)raw_disk_sb;
+	u16 csum_type = btrfs_super_csum_type(disk_sb);
+	int ret = 0;
+
+	if (csum_type == BTRFS_CSUM_TYPE_CRC32) {
+		u32 crc = ~(u32)0;
+		const int csum_size = sizeof(crc);
+		char result[csum_size];
+
+		/*
+		 * The super_block structure does not span the whole
+		 * BTRFS_SUPER_INFO_SIZE range, we expect that the unused space
+		 * is filled with zeros and is included in the checkum.
+		 */
+		crc = btrfs_csum_data(raw_disk_sb + BTRFS_CSUM_SIZE,
+				crc, BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE);
+		btrfs_csum_final(crc, result);
+
+		if (memcmp(raw_disk_sb, result, csum_size))
+			ret = 1;
+
+		if (ret && btrfs_super_generation(disk_sb) < 10) {
+			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",
+				csum_type);
+		ret = 1;
+	}
+
+	return ret;
+}
+
 /*
  * helper to read a given tree block, doing retries as required when
  * the checksums don't match and we have alternate mirrors to try.
@@ -530,41 +576,6 @@ static noinline int check_leaf(struct btrfs_root *root,
 	return 0;
 }
 
-struct extent_buffer *find_eb_for_page(struct extent_io_tree *tree,
-				       struct page *page, int max_walk)
-{
-	struct extent_buffer *eb;
-	u64 start = page_offset(page);
-	u64 target = start;
-	u64 min_start;
-
-	if (start < max_walk)
-		min_start = 0;
-	else
-		min_start = start - max_walk;
-
-	while (start >= min_start) {
-		eb = find_extent_buffer(tree, start, 0);
-		if (eb) {
-			/*
-			 * we found an extent buffer and it contains our page
-			 * horray!
-			 */
-			if (eb->start <= target &&
-			    eb->start + eb->len > target)
-				return eb;
-
-			/* we found an extent buffer that wasn't for us */
-			free_extent_buffer(eb);
-			return NULL;
-		}
-		if (start == 0)
-			break;
-		start -= PAGE_CACHE_SIZE;
-	}
-	return NULL;
-}
-
 static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 			       struct extent_state *state, int mirror)
 {
@@ -613,6 +624,12 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 		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",
+			   (int)btrfs_header_level(eb));
+		ret = -EIO;
+		goto err;
+	}
 
 	btrfs_set_buffer_lockdep_class(btrfs_header_owner(eb),
 				       eb, found_level);
@@ -636,10 +653,9 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 	if (!ret)
 		set_extent_buffer_uptodate(eb);
 err:
-	if (test_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags)) {
-		clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags);
+	if (reads_done &&
+	    test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags))
 		btree_readahead_hook(root, eb, eb->start, ret);
-	}
 
 	if (ret) {
 		/*
@@ -993,14 +1009,8 @@ static int btree_releasepage(struct page *page, gfp_t gfp_flags)
 {
 	if (PageWriteback(page) || PageDirty(page))
 		return 0;
-	/*
-	 * We need to mask out eg. __GFP_HIGHMEM and __GFP_DMA32 as we're doing
-	 * slab allocation from alloc_extent_state down the callchain where
-	 * it'd hit a BUG_ON as those flags are not allowed.
-	 */
-	gfp_flags &= ~GFP_SLAB_BUG_MASK;
 
-	return try_release_extent_buffer(page, gfp_flags);
+	return try_release_extent_buffer(page);
 }
 
 static void btree_invalidatepage(struct page *page, unsigned long offset)
@@ -1275,6 +1285,7 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 	struct btrfs_key key;
 	int ret = 0;
 	u64 bytenr;
+	uuid_le uuid;
 
 	root = btrfs_alloc_root(fs_info);
 	if (!root)
@@ -1324,6 +1335,8 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 	btrfs_set_root_used(&root->root_item, leaf->len);
 	btrfs_set_root_last_snapshot(&root->root_item, 0);
 	btrfs_set_root_dirid(&root->root_item, 0);
+	uuid_le_gen(&uuid);
+	memcpy(root->root_item.uuid, uuid.b, BTRFS_UUID_SIZE);
 	root->root_item.drop_level = 0;
 
 	key.objectid = objectid;
@@ -1476,7 +1489,7 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root,
 	if (ret == 0) {
 		l = path->nodes[0];
 		slot = path->slots[0];
-		btrfs_read_root_item(tree_root, l, slot, &root->root_item);
+		btrfs_read_root_item(l, slot, &root->root_item);
 		memcpy(&root->root_key, location, sizeof(*location));
 	}
 	btrfs_free_path(path);
@@ -1491,6 +1504,14 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root,
 	blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item));
 	root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item),
 				     blocksize, generation);
+	if (!root->node || !extent_buffer_uptodate(root->node)) {
+		ret = (!root->node) ? -ENOMEM : -EIO;
+
+		free_extent_buffer(root->node);
+		kfree(root);
+		return ERR_PTR(ret);
+	}
+
 	root->commit_root = btrfs_root_node(root);
 	BUG_ON(!root->node); /* -ENOMEM */
 out:
@@ -1658,15 +1679,20 @@ static int cleaner_kthread(void *arg)
 	struct btrfs_root *root = arg;
 
 	do {
+		int again = 0;
+
 		if (!(root->fs_info->sb->s_flags & MS_RDONLY) &&
-		    mutex_trylock(&root->fs_info->cleaner_mutex)) {
-			btrfs_run_delayed_iputs(root);
-			btrfs_clean_old_snapshots(root);
-			mutex_unlock(&root->fs_info->cleaner_mutex);
+		    down_read_trylock(&root->fs_info->sb->s_umount)) {
+			if (mutex_trylock(&root->fs_info->cleaner_mutex)) {
+				btrfs_run_delayed_iputs(root);
+				again = btrfs_clean_one_deleted_snapshot(root);
+				mutex_unlock(&root->fs_info->cleaner_mutex);
+			}
 			btrfs_run_defrag_inodes(root->fs_info);
+			up_read(&root->fs_info->sb->s_umount);
 		}
 
-		if (!try_to_freeze()) {
+		if (!try_to_freeze() && !again) {
 			set_current_state(TASK_INTERRUPTIBLE);
 			if (!kthread_should_stop())
 				schedule();
@@ -1935,6 +1961,28 @@ static noinline int next_root_backup(struct btrfs_fs_info *info,
 	return 0;
 }
 
+/* helper to cleanup workers */
+static void btrfs_stop_all_workers(struct btrfs_fs_info *fs_info)
+{
+	btrfs_stop_workers(&fs_info->generic_worker);
+	btrfs_stop_workers(&fs_info->fixup_workers);
+	btrfs_stop_workers(&fs_info->delalloc_workers);
+	btrfs_stop_workers(&fs_info->workers);
+	btrfs_stop_workers(&fs_info->endio_workers);
+	btrfs_stop_workers(&fs_info->endio_meta_workers);
+	btrfs_stop_workers(&fs_info->endio_raid56_workers);
+	btrfs_stop_workers(&fs_info->rmw_workers);
+	btrfs_stop_workers(&fs_info->endio_meta_write_workers);
+	btrfs_stop_workers(&fs_info->endio_write_workers);
+	btrfs_stop_workers(&fs_info->endio_freespace_worker);
+	btrfs_stop_workers(&fs_info->submit_workers);
+	btrfs_stop_workers(&fs_info->delayed_workers);
+	btrfs_stop_workers(&fs_info->caching_workers);
+	btrfs_stop_workers(&fs_info->readahead_workers);
+	btrfs_stop_workers(&fs_info->flush_workers);
+	btrfs_stop_workers(&fs_info->qgroup_rescan_workers);
+}
+
 /* helper to cleanup tree roots */
 static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
 {
@@ -1972,6 +2020,36 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
 	}
 }
 
+static void del_fs_roots(struct btrfs_fs_info *fs_info)
+{
+	int ret;
+	struct btrfs_root *gang[8];
+	int i;
+
+	while (!list_empty(&fs_info->dead_roots)) {
+		gang[0] = list_entry(fs_info->dead_roots.next,
+				     struct btrfs_root, root_list);
+		list_del(&gang[0]->root_list);
+
+		if (gang[0]->in_radix) {
+			btrfs_free_fs_root(fs_info, gang[0]);
+		} else {
+			free_extent_buffer(gang[0]->node);
+			free_extent_buffer(gang[0]->commit_root);
+			kfree(gang[0]);
+		}
+	}
+
+	while (1) {
+		ret = radix_tree_gang_lookup(&fs_info->fs_roots_radix,
+					     (void **)gang, 0,
+					     ARRAY_SIZE(gang));
+		if (!ret)
+			break;
+		for (i = 0; i < ret; i++)
+			btrfs_free_fs_root(fs_info, gang[i]);
+	}
+}
 
 int open_ctree(struct super_block *sb,
 	       struct btrfs_fs_devices *fs_devices,
@@ -2060,6 +2138,7 @@ int open_ctree(struct super_block *sb,
 	spin_lock_init(&fs_info->defrag_inodes_lock);
 	spin_lock_init(&fs_info->free_chunk_lock);
 	spin_lock_init(&fs_info->tree_mod_seq_lock);
+	spin_lock_init(&fs_info->super_lock);
 	rwlock_init(&fs_info->tree_mod_log_lock);
 	mutex_init(&fs_info->reloc_mutex);
 	seqlock_init(&fs_info->profiles_lock);
@@ -2083,7 +2162,7 @@ int open_ctree(struct super_block *sb,
 	atomic_set(&fs_info->async_submit_draining, 0);
 	atomic_set(&fs_info->nr_async_bios, 0);
 	atomic_set(&fs_info->defrag_running, 0);
-	atomic_set(&fs_info->tree_mod_seq, 0);
+	atomic64_set(&fs_info->tree_mod_seq, 0);
 	fs_info->sb = sb;
 	fs_info->max_inline = 8192 * 1024;
 	fs_info->metadata_ratio = 0;
@@ -2187,11 +2266,13 @@ int open_ctree(struct super_block *sb,
 	mutex_init(&fs_info->dev_replace.lock);
 
 	spin_lock_init(&fs_info->qgroup_lock);
+	mutex_init(&fs_info->qgroup_ioctl_lock);
 	fs_info->qgroup_tree = RB_ROOT;
 	INIT_LIST_HEAD(&fs_info->dirty_qgroups);
 	fs_info->qgroup_seq = 1;
 	fs_info->quota_enabled = 0;
 	fs_info->pending_quota_state = 0;
+	mutex_init(&fs_info->qgroup_rescan_lock);
 
 	btrfs_init_free_cluster(&fs_info->meta_alloc_cluster);
 	btrfs_init_free_cluster(&fs_info->data_alloc_cluster);
@@ -2211,12 +2292,31 @@ int open_ctree(struct super_block *sb,
 		     fs_info, BTRFS_ROOT_TREE_OBJECTID);
 
 	invalidate_bdev(fs_devices->latest_bdev);
+
+	/*
+	 * Read super block and check the signature bytes only
+	 */
 	bh = btrfs_read_dev_super(fs_devices->latest_bdev);
 	if (!bh) {
 		err = -EINVAL;
 		goto fail_alloc;
 	}
 
+	/*
+	 * We want to check superblock checksum, the type is stored inside.
+	 * 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");
+		err = -EINVAL;
+		goto fail_alloc;
+	}
+
+	/*
+	 * super_copy is zeroed at allocation time and we never touch the
+	 * following bytes up to INFO_SIZE, the checksum is calculated from
+	 * the whole block of INFO_SIZE
+	 */
 	memcpy(fs_info->super_copy, bh->b_data, sizeof(*fs_info->super_copy));
 	memcpy(fs_info->super_for_commit, fs_info->super_copy,
 	       sizeof(*fs_info->super_for_commit));
@@ -2224,6 +2324,13 @@ int open_ctree(struct super_block *sb,
 
 	memcpy(fs_info->fsid, fs_info->super_copy->fsid, BTRFS_FSID_SIZE);
 
+	ret = btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY);
+	if (ret) {
+		printk(KERN_ERR "btrfs: superblock contains fatal errors\n");
+		err = -EINVAL;
+		goto fail_alloc;
+	}
+
 	disk_super = fs_info->super_copy;
 	if (!btrfs_super_root(disk_super))
 		goto fail_alloc;
@@ -2232,13 +2339,6 @@ int open_ctree(struct super_block *sb,
 	if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_ERROR)
 		set_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state);
 
-	ret = btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY);
-	if (ret) {
-		printk(KERN_ERR "btrfs: superblock contains fatal errors\n");
-		err = ret;
-		goto fail_alloc;
-	}
-
 	/*
 	 * run through our array of backup supers and setup
 	 * our ring pointer to the oldest one
@@ -2290,6 +2390,9 @@ int open_ctree(struct super_block *sb,
 	if (tree_root->fs_info->compress_type == BTRFS_COMPRESS_LZO)
 		features |= BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO;
 
+	if (features & BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA)
+		printk(KERN_ERR "btrfs: has skinny extents\n");
+
 	/*
 	 * flag our filesystem as having big metadata blocks if
 	 * they are bigger than the page size
@@ -2319,6 +2422,10 @@ int open_ctree(struct super_block *sb,
 		goto fail_alloc;
 	}
 
+	/*
+	 * Needn't use the lock because there is no other task which will
+	 * update the flag.
+	 */
 	btrfs_set_super_incompat_flags(disk_super, features);
 
 	features = btrfs_super_compat_ro_flags(disk_super) &
@@ -2394,6 +2501,8 @@ int open_ctree(struct super_block *sb,
 	btrfs_init_workers(&fs_info->readahead_workers, "readahead",
 			   fs_info->thread_pool_size,
 			   &fs_info->generic_worker);
+	btrfs_init_workers(&fs_info->qgroup_rescan_workers, "qgroup-rescan", 1,
+			   &fs_info->generic_worker);
 
 	/*
 	 * endios are largely parallel and should have a very
@@ -2428,6 +2537,7 @@ int open_ctree(struct super_block *sb,
 	ret |= btrfs_start_workers(&fs_info->caching_workers);
 	ret |= btrfs_start_workers(&fs_info->readahead_workers);
 	ret |= btrfs_start_workers(&fs_info->flush_workers);
+	ret |= btrfs_start_workers(&fs_info->qgroup_rescan_workers);
 	if (ret) {
 		err = -ENOMEM;
 		goto fail_sb_buffer;
@@ -2475,8 +2585,8 @@ int open_ctree(struct super_block *sb,
 	chunk_root->node = read_tree_block(chunk_root,
 					   btrfs_super_chunk_root(disk_super),
 					   blocksize, generation);
-	BUG_ON(!chunk_root->node); /* -ENOMEM */
-	if (!test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) {
+	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",
 		       sb->s_id);
 		goto fail_tree_roots;
@@ -2661,6 +2771,13 @@ retry_root_backup:
 		log_tree_root->node = read_tree_block(tree_root, bytenr,
 						      blocksize,
 						      generation + 1);
+		if (!log_tree_root->node ||
+		    !extent_buffer_uptodate(log_tree_root->node)) {
+			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;
+		}
 		/* returns with log_tree_root freed on success */
 		ret = btrfs_recover_log_trees(log_tree_root);
 		if (ret) {
@@ -2740,6 +2857,8 @@ fail_qgroup:
 	btrfs_free_qgroup_config(fs_info);
 fail_trans_kthread:
 	kthread_stop(fs_info->transaction_kthread);
+	del_fs_roots(fs_info);
+	btrfs_cleanup_transaction(fs_info->tree_root);
 fail_cleaner:
 	kthread_stop(fs_info->cleaner_kthread);
 
@@ -2750,6 +2869,7 @@ fail_cleaner:
 	filemap_write_and_wait(fs_info->btree_inode->i_mapping);
 
 fail_block_groups:
+	btrfs_put_block_group_cache(fs_info);
 	btrfs_free_block_groups(fs_info);
 
 fail_tree_roots:
@@ -2757,22 +2877,7 @@ fail_tree_roots:
 	invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
 
 fail_sb_buffer:
-	btrfs_stop_workers(&fs_info->generic_worker);
-	btrfs_stop_workers(&fs_info->readahead_workers);
-	btrfs_stop_workers(&fs_info->fixup_workers);
-	btrfs_stop_workers(&fs_info->delalloc_workers);
-	btrfs_stop_workers(&fs_info->workers);
-	btrfs_stop_workers(&fs_info->endio_workers);
-	btrfs_stop_workers(&fs_info->endio_meta_workers);
-	btrfs_stop_workers(&fs_info->endio_raid56_workers);
-	btrfs_stop_workers(&fs_info->rmw_workers);
-	btrfs_stop_workers(&fs_info->endio_meta_write_workers);
-	btrfs_stop_workers(&fs_info->endio_write_workers);
-	btrfs_stop_workers(&fs_info->endio_freespace_worker);
-	btrfs_stop_workers(&fs_info->submit_workers);
-	btrfs_stop_workers(&fs_info->delayed_workers);
-	btrfs_stop_workers(&fs_info->caching_workers);
-	btrfs_stop_workers(&fs_info->flush_workers);
+	btrfs_stop_all_workers(fs_info);
 fail_alloc:
 fail_iput:
 	btrfs_mapping_tree_free(&fs_info->mapping_tree);
@@ -2904,7 +3009,10 @@ static int write_dev_supers(struct btrfs_device *device,
 		if (wait) {
 			bh = __find_get_block(device->bdev, bytenr / 4096,
 					      BTRFS_SUPER_INFO_SIZE);
-			BUG_ON(!bh);
+			if (!bh) {
+				errors++;
+				continue;
+			}
 			wait_on_buffer(bh);
 			if (!buffer_uptodate(bh))
 				errors++;
@@ -2919,7 +3027,7 @@ static int write_dev_supers(struct btrfs_device *device,
 			btrfs_set_super_bytenr(sb, bytenr);
 
 			crc = ~(u32)0;
-			crc = btrfs_csum_data(NULL, (char *)sb +
+			crc = btrfs_csum_data((char *)sb +
 					      BTRFS_CSUM_SIZE, crc,
 					      BTRFS_SUPER_INFO_SIZE -
 					      BTRFS_CSUM_SIZE);
@@ -2931,6 +3039,13 @@ 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 "
+				       "buffer head for bytenr %Lu\n", bytenr);
+				errors++;
+				continue;
+			}
+
 			memcpy(bh->b_data, sb, BTRFS_SUPER_INFO_SIZE);
 
 			/* one reference for submit_bh */
@@ -3153,7 +3268,7 @@ int btrfs_calc_num_tolerated_disk_barrier_failures(
 	return num_tolerated_disk_barrier_failures;
 }
 
-int write_all_supers(struct btrfs_root *root, int max_mirrors)
+static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 {
 	struct list_head *head;
 	struct btrfs_device *dev;
@@ -3283,37 +3398,6 @@ static void free_fs_root(struct btrfs_root *root)
 	kfree(root);
 }
 
-static void del_fs_roots(struct btrfs_fs_info *fs_info)
-{
-	int ret;
-	struct btrfs_root *gang[8];
-	int i;
-
-	while (!list_empty(&fs_info->dead_roots)) {
-		gang[0] = list_entry(fs_info->dead_roots.next,
-				     struct btrfs_root, root_list);
-		list_del(&gang[0]->root_list);
-
-		if (gang[0]->in_radix) {
-			btrfs_free_fs_root(fs_info, gang[0]);
-		} else {
-			free_extent_buffer(gang[0]->node);
-			free_extent_buffer(gang[0]->commit_root);
-			kfree(gang[0]);
-		}
-	}
-
-	while (1) {
-		ret = radix_tree_gang_lookup(&fs_info->fs_roots_radix,
-					     (void **)gang, 0,
-					     ARRAY_SIZE(gang));
-		if (!ret)
-			break;
-		for (i = 0; i < ret; i++)
-			btrfs_free_fs_root(fs_info, gang[i]);
-	}
-}
-
 int btrfs_cleanup_fs_roots(struct btrfs_fs_info *fs_info)
 {
 	u64 root_objectid = 0;
@@ -3349,8 +3433,8 @@ int btrfs_commit_super(struct btrfs_root *root)
 
 	mutex_lock(&root->fs_info->cleaner_mutex);
 	btrfs_run_delayed_iputs(root);
-	btrfs_clean_old_snapshots(root);
 	mutex_unlock(&root->fs_info->cleaner_mutex);
+	wake_up_process(root->fs_info->cleaner_kthread);
 
 	/* wait until ongoing cleanup work done */
 	down_write(&root->fs_info->cleanup_work_sem);
@@ -3426,20 +3510,7 @@ int close_ctree(struct btrfs_root *root)
 		       percpu_counter_sum(&fs_info->delalloc_bytes));
 	}
 
-	free_extent_buffer(fs_info->extent_root->node);
-	free_extent_buffer(fs_info->extent_root->commit_root);
-	free_extent_buffer(fs_info->tree_root->node);
-	free_extent_buffer(fs_info->tree_root->commit_root);
-	free_extent_buffer(fs_info->chunk_root->node);
-	free_extent_buffer(fs_info->chunk_root->commit_root);
-	free_extent_buffer(fs_info->dev_root->node);
-	free_extent_buffer(fs_info->dev_root->commit_root);
-	free_extent_buffer(fs_info->csum_root->node);
-	free_extent_buffer(fs_info->csum_root->commit_root);
-	if (fs_info->quota_root) {
-		free_extent_buffer(fs_info->quota_root->node);
-		free_extent_buffer(fs_info->quota_root->commit_root);
-	}
+	free_root_pointers(fs_info, 1);
 
 	btrfs_free_block_groups(fs_info);
 
@@ -3447,22 +3518,7 @@ int close_ctree(struct btrfs_root *root)
 
 	iput(fs_info->btree_inode);
 
-	btrfs_stop_workers(&fs_info->generic_worker);
-	btrfs_stop_workers(&fs_info->fixup_workers);
-	btrfs_stop_workers(&fs_info->delalloc_workers);
-	btrfs_stop_workers(&fs_info->workers);
-	btrfs_stop_workers(&fs_info->endio_workers);
-	btrfs_stop_workers(&fs_info->endio_meta_workers);
-	btrfs_stop_workers(&fs_info->endio_raid56_workers);
-	btrfs_stop_workers(&fs_info->rmw_workers);
-	btrfs_stop_workers(&fs_info->endio_meta_write_workers);
-	btrfs_stop_workers(&fs_info->endio_write_workers);
-	btrfs_stop_workers(&fs_info->endio_freespace_worker);
-	btrfs_stop_workers(&fs_info->submit_workers);
-	btrfs_stop_workers(&fs_info->delayed_workers);
-	btrfs_stop_workers(&fs_info->caching_workers);
-	btrfs_stop_workers(&fs_info->readahead_workers);
-	btrfs_stop_workers(&fs_info->flush_workers);
+	btrfs_stop_all_workers(fs_info);
 
 #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
 	if (btrfs_test_opt(root, CHECK_INTEGRITY))
@@ -3567,18 +3623,13 @@ int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid)
 static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
 			      int read_only)
 {
-	if (btrfs_super_csum_type(fs_info->super_copy) >= ARRAY_SIZE(btrfs_csum_sizes)) {
-		printk(KERN_ERR "btrfs: unsupported checksum algorithm\n");
-		return -EINVAL;
-	}
-
-	if (read_only)
-		return 0;
-
+	/*
+	 * Placeholder for checks
+	 */
 	return 0;
 }
 
-void btrfs_error_commit_super(struct btrfs_root *root)
+static void btrfs_error_commit_super(struct btrfs_root *root)
 {
 	mutex_lock(&root->fs_info->cleaner_mutex);
 	btrfs_run_delayed_iputs(root);
@@ -3669,6 +3720,9 @@ int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 				continue;
 			}
 
+			if (head->must_insert_reserved)
+				btrfs_pin_extent(root, ref->bytenr,
+						 ref->num_bytes, 1);
 			btrfs_free_delayed_extent_op(head->extent_op);
 			delayed_refs->num_heads--;
 			if (list_empty(&head->cluster))
@@ -3740,13 +3794,9 @@ static int btrfs_destroy_marked_extents(struct btrfs_root *root,
 					int mark)
 {
 	int ret;
-	struct page *page;
-	struct inode *btree_inode = root->fs_info->btree_inode;
 	struct extent_buffer *eb;
 	u64 start = 0;
 	u64 end;
-	u64 offset;
-	unsigned long index;
 
 	while (1) {
 		ret = find_first_extent_bit(dirty_pages, start, &start, &end,
@@ -3756,36 +3806,17 @@ static int btrfs_destroy_marked_extents(struct btrfs_root *root,
 
 		clear_extent_bits(dirty_pages, start, end, mark, GFP_NOFS);
 		while (start <= end) {
-			index = start >> PAGE_CACHE_SHIFT;
-			start = (u64)(index + 1) << PAGE_CACHE_SHIFT;
-			page = find_get_page(btree_inode->i_mapping, index);
-			if (!page)
+			eb = btrfs_find_tree_block(root, start,
+						   root->leafsize);
+			start += eb->len;
+			if (!eb)
 				continue;
-			offset = page_offset(page);
-
-			spin_lock(&dirty_pages->buffer_lock);
-			eb = radix_tree_lookup(
-			     &(&BTRFS_I(page->mapping->host)->io_tree)->buffer,
-					       offset >> PAGE_CACHE_SHIFT);
-			spin_unlock(&dirty_pages->buffer_lock);
-			if (eb)
-				ret = test_and_clear_bit(EXTENT_BUFFER_DIRTY,
-							 &eb->bflags);
-			if (PageWriteback(page))
-				end_page_writeback(page);
-
-			lock_page(page);
-			if (PageDirty(page)) {
-				clear_page_dirty_for_io(page);
-				spin_lock_irq(&page->mapping->tree_lock);
-				radix_tree_tag_clear(&page->mapping->page_tree,
-							page_index(page),
-							PAGECACHE_TAG_DIRTY);
-				spin_unlock_irq(&page->mapping->tree_lock);
-			}
+			wait_on_extent_buffer_writeback(eb);
 
-			unlock_page(page);
-			page_cache_release(page);
+			if (test_and_clear_bit(EXTENT_BUFFER_DIRTY,
+					       &eb->bflags))
+				clear_extent_buffer_dirty(eb);
+			free_extent_buffer_stale(eb);
 		}
 	}
 
@@ -3866,7 +3897,7 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 	*/
 }
 
-int btrfs_cleanup_transaction(struct btrfs_root *root)
+static int btrfs_cleanup_transaction(struct btrfs_root *root)
 {
 	struct btrfs_transaction *t;
 	LIST_HEAD(list);
@@ -3887,10 +3918,6 @@ int btrfs_cleanup_transaction(struct btrfs_root *root)
 
 		btrfs_destroy_delayed_refs(t, root);
 
-		btrfs_block_rsv_release(root,
-					&root->fs_info->trans_block_rsv,
-					t->dirty_pages.dirty_bytes);
-
 		/* FIXME: cleanup wait for commit */
 		t->in_commit = 1;
 		t->blocked = 1;

fs/btrfs/disk-io.h

@@ -61,7 +61,6 @@ int write_ctree_super(struct btrfs_trans_handle *trans,
 		      struct btrfs_root *root, int max_mirrors);
 struct buffer_head *btrfs_read_dev_super(struct block_device *bdev);
 int btrfs_commit_super(struct btrfs_root *root);
-void btrfs_error_commit_super(struct btrfs_root *root);
 struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
 					    u64 bytenr, u32 blocksize);
 struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root,
@@ -77,7 +76,7 @@ int btrfs_buffer_uptodate(struct extent_buffer *buf, u64 parent_transid,
 			  int atomic);
 int btrfs_set_buffer_uptodate(struct extent_buffer *buf);
 int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid);
-u32 btrfs_csum_data(struct btrfs_root *root, char *data, u32 seed, size_t len);
+u32 btrfs_csum_data(char *data, u32 seed, size_t len);
 void btrfs_csum_final(u32 crc, char *result);
 int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
 			int metadata);
@@ -93,10 +92,8 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info);
 int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 		       struct btrfs_root *root);
-int btrfs_cleanup_transaction(struct btrfs_root *root);
 void btrfs_cleanup_one_transaction(struct btrfs_transaction *trans,
 				  struct btrfs_root *root);
-void btrfs_abort_devices(struct btrfs_root *root);
 struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 				     struct btrfs_fs_info *fs_info,
 				     u64 objectid);

fs/btrfs/extent_io.c

@@ -24,12 +24,62 @@
 static struct kmem_cache *extent_state_cache;
 static struct kmem_cache *extent_buffer_cache;
 
+#ifdef CONFIG_BTRFS_DEBUG
 static LIST_HEAD(buffers);
 static LIST_HEAD(states);
 
-#define LEAK_DEBUG 0
-#if LEAK_DEBUG
 static DEFINE_SPINLOCK(leak_lock);
+
+static inline
+void btrfs_leak_debug_add(struct list_head *new, struct list_head *head)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&leak_lock, flags);
+	list_add(new, head);
+	spin_unlock_irqrestore(&leak_lock, flags);
+}
+
+static inline
+void btrfs_leak_debug_del(struct list_head *entry)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&leak_lock, flags);
+	list_del(entry);
+	spin_unlock_irqrestore(&leak_lock, flags);
+}
+
+static inline
+void btrfs_leak_debug_check(void)
+{
+	struct extent_state *state;
+	struct extent_buffer *eb;
+
+	while (!list_empty(&states)) {
+		state = list_entry(states.next, struct extent_state, leak_list);
+		printk(KERN_ERR "btrfs state leak: start %llu end %llu "
+		       "state %lu in tree %p refs %d\n",
+		       (unsigned long long)state->start,
+		       (unsigned long long)state->end,
+		       state->state, state->tree, atomic_read(&state->refs));
+		list_del(&state->leak_list);
+		kmem_cache_free(extent_state_cache, state);
+	}
+
+	while (!list_empty(&buffers)) {
+		eb = list_entry(buffers.next, struct extent_buffer, leak_list);
+		printk(KERN_ERR "btrfs buffer leak start %llu len %lu "
+		       "refs %d\n", (unsigned long long)eb->start,
+		       eb->len, atomic_read(&eb->refs));
+		list_del(&eb->leak_list);
+		kmem_cache_free(extent_buffer_cache, eb);
+	}
+}
+#else
+#define btrfs_leak_debug_add(new, head)	do {} while (0)
+#define btrfs_leak_debug_del(entry)	do {} while (0)
+#define btrfs_leak_debug_check()	do {} while (0)
 #endif
 
 #define BUFFER_LRU_MAX 64
@@ -84,29 +134,7 @@ free_state_cache:
 
 void extent_io_exit(void)
 {
-	struct extent_state *state;
-	struct extent_buffer *eb;
-
-	while (!list_empty(&states)) {
-		state = list_entry(states.next, struct extent_state, leak_list);
-		printk(KERN_ERR "btrfs state leak: start %llu end %llu "
-		       "state %lu in tree %p refs %d\n",
-		       (unsigned long long)state->start,
-		       (unsigned long long)state->end,
-		       state->state, state->tree, atomic_read(&state->refs));
-		list_del(&state->leak_list);
-		kmem_cache_free(extent_state_cache, state);
-
-	}
-
-	while (!list_empty(&buffers)) {
-		eb = list_entry(buffers.next, struct extent_buffer, leak_list);
-		printk(KERN_ERR "btrfs buffer leak start %llu len %lu "
-		       "refs %d\n", (unsigned long long)eb->start,
-		       eb->len, atomic_read(&eb->refs));
-		list_del(&eb->leak_list);
-		kmem_cache_free(extent_buffer_cache, eb);
-	}
+	btrfs_leak_debug_check();
 
 	/*
 	 * Make sure all delayed rcu free are flushed before we
@@ -134,9 +162,6 @@ void extent_io_tree_init(struct extent_io_tree *tree,
 static struct extent_state *alloc_extent_state(gfp_t mask)
 {
 	struct extent_state *state;
-#if LEAK_DEBUG
-	unsigned long flags;
-#endif
 
 	state = kmem_cache_alloc(extent_state_cache, mask);
 	if (!state)
@@ -144,11 +169,7 @@ static struct extent_state *alloc_extent_state(gfp_t mask)
 	state->state = 0;
 	state->private = 0;
 	state->tree = NULL;
-#if LEAK_DEBUG
-	spin_lock_irqsave(&leak_lock, flags);
-	list_add(&state->leak_list, &states);
-	spin_unlock_irqrestore(&leak_lock, flags);
-#endif
+	btrfs_leak_debug_add(&state->leak_list, &states);
 	atomic_set(&state->refs, 1);
 	init_waitqueue_head(&state->wq);
 	trace_alloc_extent_state(state, mask, _RET_IP_);
@@ -160,15 +181,8 @@ void free_extent_state(struct extent_state *state)
 	if (!state)
 		return;
 	if (atomic_dec_and_test(&state->refs)) {
-#if LEAK_DEBUG
-		unsigned long flags;
-#endif
 		WARN_ON(state->tree);
-#if LEAK_DEBUG
-		spin_lock_irqsave(&leak_lock, flags);
-		list_del(&state->leak_list);
-		spin_unlock_irqrestore(&leak_lock, flags);
-#endif
+		btrfs_leak_debug_del(&state->leak_list);
 		trace_free_extent_state(state, _RET_IP_);
 		kmem_cache_free(extent_state_cache, state);
 	}
@@ -308,21 +322,21 @@ static void merge_state(struct extent_io_tree *tree,
 }
 
 static void set_state_cb(struct extent_io_tree *tree,
-			 struct extent_state *state, int *bits)
+			 struct extent_state *state, unsigned long *bits)
 {
 	if (tree->ops && tree->ops->set_bit_hook)
 		tree->ops->set_bit_hook(tree->mapping->host, state, bits);
 }
 
 static void clear_state_cb(struct extent_io_tree *tree,
-			   struct extent_state *state, int *bits)
+			   struct extent_state *state, unsigned long *bits)
 {
 	if (tree->ops && tree->ops->clear_bit_hook)
 		tree->ops->clear_bit_hook(tree->mapping->host, state, bits);
 }
 
 static void set_state_bits(struct extent_io_tree *tree,
-			   struct extent_state *state, int *bits);
+			   struct extent_state *state, unsigned long *bits);
 
 /*
  * insert an extent_state struct into the tree.  'bits' are set on the
@@ -336,7 +350,7 @@ 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,
-			int *bits)
+			unsigned long *bits)
 {
 	struct rb_node *node;
 
@@ -424,10 +438,10 @@ static struct extent_state *next_state(struct extent_state *state)
  */
 static struct extent_state *clear_state_bit(struct extent_io_tree *tree,
 					    struct extent_state *state,
-					    int *bits, int wake)
+					    unsigned long *bits, int wake)
 {
 	struct extent_state *next;
-	int bits_to_clear = *bits & ~EXTENT_CTLBITS;
+	unsigned long bits_to_clear = *bits & ~EXTENT_CTLBITS;
 
 	if ((bits_to_clear & EXTENT_DIRTY) && (state->state & EXTENT_DIRTY)) {
 		u64 range = state->end - state->start + 1;
@@ -463,7 +477,7 @@ alloc_extent_state_atomic(struct extent_state *prealloc)
 	return prealloc;
 }
 
-void extent_io_tree_panic(struct extent_io_tree *tree, int err)
+static void extent_io_tree_panic(struct extent_io_tree *tree, int err)
 {
 	btrfs_panic(tree_fs_info(tree), err, "Locking error: "
 		    "Extent tree was modified by another "
@@ -483,7 +497,7 @@ void extent_io_tree_panic(struct extent_io_tree *tree, int err)
  * This takes the tree lock, and returns 0 on success and < 0 on error.
  */
 int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		     int bits, int wake, int delete,
+		     unsigned long bits, int wake, int delete,
 		     struct extent_state **cached_state,
 		     gfp_t mask)
 {
@@ -644,7 +658,8 @@ static void wait_on_state(struct extent_io_tree *tree,
  * The range [start, end] is inclusive.
  * The tree lock is taken by this function
  */
-void wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits)
+static void wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
+			    unsigned long bits)
 {
 	struct extent_state *state;
 	struct rb_node *node;
@@ -685,9 +700,9 @@ out:
 
 static void set_state_bits(struct extent_io_tree *tree,
 			   struct extent_state *state,
-			   int *bits)
+			   unsigned long *bits)
 {
-	int bits_to_set = *bits & ~EXTENT_CTLBITS;
+	unsigned long bits_to_set = *bits & ~EXTENT_CTLBITS;
 
 	set_state_cb(tree, state, bits);
 	if ((bits_to_set & EXTENT_DIRTY) && !(state->state & EXTENT_DIRTY)) {
@@ -730,8 +745,9 @@ static void uncache_state(struct extent_state **cached_ptr)
 
 static int __must_check
 __set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		 int bits, int exclusive_bits, u64 *failed_start,
-		 struct extent_state **cached_state, gfp_t mask)
+		 unsigned long bits, unsigned long exclusive_bits,
+		 u64 *failed_start, struct extent_state **cached_state,
+		 gfp_t mask)
 {
 	struct extent_state *state;
 	struct extent_state *prealloc = NULL;
@@ -923,9 +939,9 @@ search_again:
 	goto again;
 }
 
-int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits,
-		   u64 *failed_start, struct extent_state **cached_state,
-		   gfp_t mask)
+int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
+		   unsigned long bits, u64 * failed_start,
+		   struct extent_state **cached_state, gfp_t mask)
 {
 	return __set_extent_bit(tree, start, end, bits, 0, failed_start,
 				cached_state, mask);
@@ -950,7 +966,7 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits,
  * boundary bits like LOCK.
  */
 int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		       int bits, int clear_bits,
+		       unsigned long bits, unsigned long clear_bits,
 		       struct extent_state **cached_state, gfp_t mask)
 {
 	struct extent_state *state;
@@ -1143,14 +1159,14 @@ int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
 }
 
 int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		    int bits, gfp_t mask)
+		    unsigned long bits, gfp_t mask)
 {
 	return set_extent_bit(tree, start, end, bits, NULL,
 			      NULL, mask);
 }
 
 int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		      int bits, gfp_t mask)
+		      unsigned long bits, gfp_t mask)
 {
 	return clear_extent_bit(tree, start, end, bits, 0, 0, NULL, mask);
 }
@@ -1189,7 +1205,7 @@ int set_extent_new(struct extent_io_tree *tree, u64 start, u64 end,
 int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
 			struct extent_state **cached_state, gfp_t mask)
 {
-	return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0,
+	return set_extent_bit(tree, start, end, EXTENT_UPTODATE, NULL,
 			      cached_state, mask);
 }
 
@@ -1205,7 +1221,7 @@ int clear_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
  * us if waiting is desired.
  */
 int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		     int bits, struct extent_state **cached_state)
+		     unsigned long bits, struct extent_state **cached_state)
 {
 	int err;
 	u64 failed_start;
@@ -1313,8 +1329,9 @@ static int set_range_writeback(struct extent_io_tree *tree, u64 start, u64 end)
  * return it.  tree->lock must be held.  NULL will returned if
  * nothing was found after 'start'
  */
-struct extent_state *find_first_extent_bit_state(struct extent_io_tree *tree,
-						 u64 start, int bits)
+static struct extent_state *
+find_first_extent_bit_state(struct extent_io_tree *tree,
+			    u64 start, unsigned long bits)
 {
 	struct rb_node *node;
 	struct extent_state *state;
@@ -1348,7 +1365,7 @@ out:
  * If nothing was found, 1 is returned. If found something, return 0.
  */
 int find_first_extent_bit(struct extent_io_tree *tree, u64 start,
-			  u64 *start_ret, u64 *end_ret, int bits,
+			  u64 *start_ret, u64 *end_ret, unsigned long bits,
 			  struct extent_state **cached_state)
 {
 	struct extent_state *state;
@@ -1638,7 +1655,7 @@ int extent_clear_unlock_delalloc(struct inode *inode,
 	unsigned long end_index = end >> PAGE_CACHE_SHIFT;
 	unsigned long nr_pages = end_index - index + 1;
 	int i;
-	int clear_bits = 0;
+	unsigned long clear_bits = 0;
 
 	if (op & EXTENT_CLEAR_UNLOCK)
 		clear_bits |= EXTENT_LOCKED;
@@ -1777,6 +1794,64 @@ out:
 	return ret;
 }
 
+void extent_cache_csums_dio(struct extent_io_tree *tree, u64 start, u32 csums[],
+			    int count)
+{
+	struct rb_node *node;
+	struct extent_state *state;
+
+	spin_lock(&tree->lock);
+	/*
+	 * this search will find all the extents that end after
+	 * our range starts.
+	 */
+	node = tree_search(tree, start);
+	BUG_ON(!node);
+
+	state = rb_entry(node, struct extent_state, rb_node);
+	BUG_ON(state->start != start);
+
+	while (count) {
+		state->private = *csums++;
+		count--;
+		state = next_state(state);
+	}
+	spin_unlock(&tree->lock);
+}
+
+static inline u64 __btrfs_get_bio_offset(struct bio *bio, int bio_index)
+{
+	struct bio_vec *bvec = bio->bi_io_vec + bio_index;
+
+	return page_offset(bvec->bv_page) + bvec->bv_offset;
+}
+
+void extent_cache_csums(struct extent_io_tree *tree, struct bio *bio, int bio_index,
+			u32 csums[], int count)
+{
+	struct rb_node *node;
+	struct extent_state *state = NULL;
+	u64 start;
+
+	spin_lock(&tree->lock);
+	do {
+		start = __btrfs_get_bio_offset(bio, bio_index);
+		if (state == NULL || state->start != start) {
+			node = tree_search(tree, start);
+			BUG_ON(!node);
+
+			state = rb_entry(node, struct extent_state, rb_node);
+			BUG_ON(state->start != start);
+		}
+		state->private = *csums++;
+		count--;
+		bio_index++;
+
+		state = next_state(state);
+	} while (count);
+	spin_unlock(&tree->lock);
+}
+
 int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private)
 {
 	struct rb_node *node;
@@ -1811,7 +1886,7 @@ out:
  * range is found set.
  */
 int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		   int bits, int filled, struct extent_state *cached)
+		   unsigned long bits, int filled, struct extent_state *cached)
 {
 	struct extent_state *state = NULL;
 	struct rb_node *node;
@@ -2595,7 +2670,8 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree,
 	return ret;
 }
 
-void attach_extent_buffer_page(struct extent_buffer *eb, struct page *page)
+static void attach_extent_buffer_page(struct extent_buffer *eb,
+				      struct page *page)
 {
 	if (!PagePrivate(page)) {
 		SetPagePrivate(page);
@@ -2625,7 +2701,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 				   struct page *page,
 				   get_extent_t *get_extent,
 				   struct bio **bio, int mirror_num,
-				   unsigned long *bio_flags)
+				   unsigned long *bio_flags, int rw)
 {
 	struct inode *inode = page->mapping->host;
 	u64 start = page_offset(page);
@@ -2771,7 +2847,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 		}
 
 		pnr -= page->index;
-		ret = submit_extent_page(READ, tree, page,
+		ret = submit_extent_page(rw, tree, page,
 					 sector, disk_io_size, pg_offset,
 					 bdev, bio, pnr,
 					 end_bio_extent_readpage, mirror_num,
@@ -2804,7 +2880,7 @@ int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
 	int ret;
 
 	ret = __extent_read_full_page(tree, page, get_extent, &bio, mirror_num,
-				      &bio_flags);
+				      &bio_flags, READ);
 	if (bio)
 		ret = submit_one_bio(READ, bio, mirror_num, bio_flags);
 	return ret;
@@ -3103,7 +3179,7 @@ static int eb_wait(void *word)
 	return 0;
 }
 
-static void wait_on_extent_buffer_writeback(struct extent_buffer *eb)
+void wait_on_extent_buffer_writeback(struct extent_buffer *eb)
 {
 	wait_on_bit(&eb->bflags, EXTENT_BUFFER_WRITEBACK, eb_wait,
 		    TASK_UNINTERRUPTIBLE);
@@ -3228,7 +3304,7 @@ static int write_one_eb(struct extent_buffer *eb,
 	u64 offset = eb->start;
 	unsigned long i, num_pages;
 	unsigned long bio_flags = 0;
-	int rw = (epd->sync_io ? WRITE_SYNC : WRITE);
+	int rw = (epd->sync_io ? WRITE_SYNC : WRITE) | REQ_META;
 	int ret = 0;
 
 	clear_bit(EXTENT_BUFFER_IOERR, &eb->bflags);
@@ -3665,14 +3741,14 @@ int extent_readpages(struct extent_io_tree *tree,
 			continue;
 		for (i = 0; i < nr; i++) {
 			__extent_read_full_page(tree, pagepool[i], get_extent,
-					&bio, 0, &bio_flags);
+					&bio, 0, &bio_flags, READ);
 			page_cache_release(pagepool[i]);
 		}
 		nr = 0;
 	}
 	for (i = 0; i < nr; i++) {
 		__extent_read_full_page(tree, pagepool[i], get_extent,
-					&bio, 0, &bio_flags);
+					&bio, 0, &bio_flags, READ);
 		page_cache_release(pagepool[i]);
 	}
 
@@ -3713,9 +3789,9 @@ int extent_invalidatepage(struct extent_io_tree *tree,
  * are locked or under IO and drops the related state bits if it is safe
  * to drop the page.
  */
-int try_release_extent_state(struct extent_map_tree *map,
-			     struct extent_io_tree *tree, struct page *page,
-			     gfp_t mask)
+static int try_release_extent_state(struct extent_map_tree *map,
+				    struct extent_io_tree *tree,
+				    struct page *page, gfp_t mask)
 {
 	u64 start = page_offset(page);
 	u64 end = start + PAGE_CACHE_SIZE - 1;
@@ -4006,12 +4082,7 @@ out:
 
 static void __free_extent_buffer(struct extent_buffer *eb)
 {
-#if LEAK_DEBUG
-	unsigned long flags;
-	spin_lock_irqsave(&leak_lock, flags);
-	list_del(&eb->leak_list);
-	spin_unlock_irqrestore(&leak_lock, flags);
-#endif
+	btrfs_leak_debug_del(&eb->leak_list);
 	kmem_cache_free(extent_buffer_cache, eb);
 }
 
@@ -4021,9 +4092,6 @@ static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree,
 						   gfp_t mask)
 {
 	struct extent_buffer *eb = NULL;
-#if LEAK_DEBUG
-	unsigned long flags;
-#endif
 
 	eb = kmem_cache_zalloc(extent_buffer_cache, mask);
 	if (eb == NULL)
@@ -4043,11 +4111,8 @@ static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree,
 	init_waitqueue_head(&eb->write_lock_wq);
 	init_waitqueue_head(&eb->read_lock_wq);
 
-#if LEAK_DEBUG
-	spin_lock_irqsave(&leak_lock, flags);
-	list_add(&eb->leak_list, &buffers);
-	spin_unlock_irqrestore(&leak_lock, flags);
-#endif
+	btrfs_leak_debug_add(&eb->leak_list, &buffers);
+
 	spin_lock_init(&eb->refs_lock);
 	atomic_set(&eb->refs, 1);
 	atomic_set(&eb->io_pages, 0);
@@ -4385,7 +4450,7 @@ static inline void btrfs_release_extent_buffer_rcu(struct rcu_head *head)
 }
 
 /* Expects to have eb->eb_lock already held */
-static int release_extent_buffer(struct extent_buffer *eb, gfp_t mask)
+static int release_extent_buffer(struct extent_buffer *eb)
 {
 	WARN_ON(atomic_read(&eb->refs) == 0);
 	if (atomic_dec_and_test(&eb->refs)) {
@@ -4443,7 +4508,7 @@ void free_extent_buffer(struct extent_buffer *eb)
 	 * I know this is terrible, but it's temporary until we stop tracking
 	 * the uptodate bits and such for the extent buffers.
 	 */
-	release_extent_buffer(eb, GFP_ATOMIC);
+	release_extent_buffer(eb);
 }
 
 void free_extent_buffer_stale(struct extent_buffer *eb)
@@ -4457,7 +4522,7 @@ void free_extent_buffer_stale(struct extent_buffer *eb)
 	if (atomic_read(&eb->refs) == 2 && !extent_buffer_under_io(eb) &&
 	    test_and_clear_bit(EXTENT_BUFFER_TREE_REF, &eb->bflags))
 		atomic_dec(&eb->refs);
-	release_extent_buffer(eb, GFP_NOFS);
+	release_extent_buffer(eb);
 }
 
 void clear_extent_buffer_dirty(struct extent_buffer *eb)
@@ -4509,17 +4574,6 @@ int set_extent_buffer_dirty(struct extent_buffer *eb)
 	return was_dirty;
 }
 
-static int range_straddles_pages(u64 start, u64 len)
-{
-	if (len < PAGE_CACHE_SIZE)
-		return 1;
-	if (start & (PAGE_CACHE_SIZE - 1))
-		return 1;
-	if ((start + len) & (PAGE_CACHE_SIZE - 1))
-		return 1;
-	return 0;
-}
-
 int clear_extent_buffer_uptodate(struct extent_buffer *eb)
 {
 	unsigned long i;
@@ -4551,37 +4605,6 @@ int set_extent_buffer_uptodate(struct extent_buffer *eb)
 	return 0;
 }
 
-int extent_range_uptodate(struct extent_io_tree *tree,
-			  u64 start, u64 end)
-{
-	struct page *page;
-	int ret;
-	int pg_uptodate = 1;
-	int uptodate;
-	unsigned long index;
-
-	if (range_straddles_pages(start, end - start + 1)) {
-		ret = test_range_bit(tree, start, end,
-				     EXTENT_UPTODATE, 1, NULL);
-		if (ret)
-			return 1;
-	}
-	while (start <= end) {
-		index = start >> PAGE_CACHE_SHIFT;
-		page = find_get_page(tree->mapping, index);
-		if (!page)
-			return 1;
-		uptodate = PageUptodate(page);
-		page_cache_release(page);
-		if (!uptodate) {
-			pg_uptodate = 0;
-			break;
-		}
-		start += PAGE_CACHE_SIZE;
-	}
-	return pg_uptodate;
-}
-
 int extent_buffer_uptodate(struct extent_buffer *eb)
 {
 	return test_bit(EXTENT_BUFFER_UPTODATE, &eb->bflags);
@@ -4644,7 +4667,8 @@ int read_extent_buffer_pages(struct extent_io_tree *tree,
 			ClearPageError(page);
 			err = __extent_read_full_page(tree, page,
 						      get_extent, &bio,
-						      mirror_num, &bio_flags);
+						      mirror_num, &bio_flags,
+						      READ | REQ_META);
 			if (err)
 				ret = err;
 		} else {
@@ -4653,7 +4677,8 @@ int read_extent_buffer_pages(struct extent_io_tree *tree,
 	}
 
 	if (bio) {
-		err = submit_one_bio(READ, bio, mirror_num, bio_flags);
+		err = submit_one_bio(READ | REQ_META, bio, mirror_num,
+				     bio_flags);
 		if (err)
 			return err;
 	}
@@ -5017,7 +5042,7 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 	}
 }
 
-int try_release_extent_buffer(struct page *page, gfp_t mask)
+int try_release_extent_buffer(struct page *page)
 {
 	struct extent_buffer *eb;
 
@@ -5047,9 +5072,6 @@ int try_release_extent_buffer(struct page *page, gfp_t mask)
 	}
 	spin_unlock(&page->mapping->private_lock);
 
-	if ((mask & GFP_NOFS) == GFP_NOFS)
-		mask = GFP_NOFS;
-
 	/*
 	 * If tree ref isn't set then we know the ref on this eb is a real ref,
 	 * so just return, this page will likely be freed soon anyway.
@@ -5059,5 +5081,5 @@ int try_release_extent_buffer(struct page *page, gfp_t mask)
 		return 0;
 	}
 
-	return release_extent_buffer(eb, mask);
+	return release_extent_buffer(eb);
 }

fs/btrfs/extent_io.h

@@ -81,9 +81,9 @@ struct extent_io_ops {
 	int (*writepage_end_io_hook)(struct page *page, u64 start, u64 end,
 				      struct extent_state *state, int uptodate);
 	void (*set_bit_hook)(struct inode *inode, struct extent_state *state,
-			     int *bits);
+			     unsigned long *bits);
 	void (*clear_bit_hook)(struct inode *inode, struct extent_state *state,
-			       int *bits);
+			       unsigned long *bits);
 	void (*merge_extent_hook)(struct inode *inode,
 				  struct extent_state *new,
 				  struct extent_state *other);
@@ -116,7 +116,9 @@ struct extent_state {
 	/* for use by the FS */
 	u64 private;
 
+#ifdef CONFIG_BTRFS_DEBUG
 	struct list_head leak_list;
+#endif
 };
 
 #define INLINE_EXTENT_BUFFER_PAGES 16
@@ -132,7 +134,6 @@ struct extent_buffer {
 	atomic_t refs;
 	atomic_t io_pages;
 	int read_mirror;
-	struct list_head leak_list;
 	struct rcu_head rcu_head;
 	pid_t lock_owner;
 
@@ -159,6 +160,9 @@ struct extent_buffer {
 	wait_queue_head_t read_lock_wq;
 	wait_queue_head_t lock_wq;
 	struct page *pages[INLINE_EXTENT_BUFFER_PAGES];
+#ifdef CONFIG_BTRFS_DEBUG
+	struct list_head leak_list;
+#endif
 };
 
 static inline void extent_set_compress_type(unsigned long *bio_flags,
@@ -185,13 +189,10 @@ void extent_io_tree_init(struct extent_io_tree *tree,
 int try_release_extent_mapping(struct extent_map_tree *map,
 			       struct extent_io_tree *tree, struct page *page,
 			       gfp_t mask);
-int try_release_extent_buffer(struct page *page, gfp_t mask);
-int try_release_extent_state(struct extent_map_tree *map,
-			     struct extent_io_tree *tree, struct page *page,
-			     gfp_t mask);
+int try_release_extent_buffer(struct page *page);
 int lock_extent(struct extent_io_tree *tree, u64 start, u64 end);
 int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		     int bits, struct extent_state **cached);
+		     unsigned long bits, struct extent_state **cached);
 int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end);
 int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end,
 			 struct extent_state **cached, gfp_t mask);
@@ -207,16 +208,17 @@ u64 count_range_bits(struct extent_io_tree *tree,
 
 void free_extent_state(struct extent_state *state);
 int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		   int bits, int filled, struct extent_state *cached_state);
+		   unsigned long bits, int filled,
+		   struct extent_state *cached_state);
 int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		      int bits, gfp_t mask);
+		      unsigned long bits, gfp_t mask);
 int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		     int bits, int wake, int delete, struct extent_state **cached,
-		     gfp_t mask);
+		     unsigned long bits, int wake, int delete,
+		     struct extent_state **cached, gfp_t mask);
 int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		    int bits, gfp_t mask);
+		    unsigned long bits, gfp_t mask);
 int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		   int bits, u64 *failed_start,
+		   unsigned long bits, u64 *failed_start,
 		   struct extent_state **cached_state, gfp_t mask);
 int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
 			struct extent_state **cached_state, gfp_t mask);
@@ -229,17 +231,15 @@ int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
 int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
 		       gfp_t mask);
 int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		       int bits, int clear_bits,
+		       unsigned long bits, unsigned long clear_bits,
 		       struct extent_state **cached_state, gfp_t mask);
 int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end,
 			struct extent_state **cached_state, gfp_t mask);
 int set_extent_defrag(struct extent_io_tree *tree, u64 start, u64 end,
 		      struct extent_state **cached_state, gfp_t mask);
 int find_first_extent_bit(struct extent_io_tree *tree, u64 start,
-			  u64 *start_ret, u64 *end_ret, int bits,
+			  u64 *start_ret, u64 *end_ret, unsigned long bits,
 			  struct extent_state **cached_state);
-struct extent_state *find_first_extent_bit_state(struct extent_io_tree *tree,
-						 u64 start, int bits);
 int extent_invalidatepage(struct extent_io_tree *tree,
 			  struct page *page, unsigned long offset);
 int extent_write_full_page(struct extent_io_tree *tree, struct page *page,
@@ -261,6 +261,10 @@ int extent_readpages(struct extent_io_tree *tree,
 int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len, get_extent_t *get_extent);
 int set_state_private(struct extent_io_tree *tree, u64 start, u64 private);
+void extent_cache_csums_dio(struct extent_io_tree *tree, u64 start, u32 csums[],
+			    int count);
+void extent_cache_csums(struct extent_io_tree *tree, struct bio *bio,
+			int bvec_index, u32 csums[], int count);
 int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private);
 void set_page_extent_mapped(struct page *page);
 
@@ -278,6 +282,7 @@ void free_extent_buffer_stale(struct extent_buffer *eb);
 int read_extent_buffer_pages(struct extent_io_tree *tree,
 			     struct extent_buffer *eb, u64 start, int wait,
 			     get_extent_t *get_extent, int mirror_num);
+void wait_on_extent_buffer_writeback(struct extent_buffer *eb);
 
 static inline unsigned long num_extent_pages(u64 start, u64 len)
 {
@@ -313,7 +318,6 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 			   unsigned long src_offset, unsigned long len);
 void memset_extent_buffer(struct extent_buffer *eb, char c,
 			  unsigned long start, unsigned long len);
-void wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits);
 void clear_extent_buffer_dirty(struct extent_buffer *eb);
 int set_extent_buffer_dirty(struct extent_buffer *eb);
 int set_extent_buffer_uptodate(struct extent_buffer *eb);
@@ -323,8 +327,6 @@ int map_private_extent_buffer(struct extent_buffer *eb, unsigned long offset,
 		      unsigned long min_len, char **map,
 		      unsigned long *map_start,
 		      unsigned long *map_len);
-int extent_range_uptodate(struct extent_io_tree *tree,
-			  u64 start, u64 end);
 int extent_range_clear_dirty_for_io(struct inode *inode, u64 start, u64 end);
 int extent_range_redirty_for_io(struct inode *inode, u64 start, u64 end);
 int extent_clear_unlock_delalloc(struct inode *inode,

fs/btrfs/extent_map.c

@@ -174,6 +174,14 @@ static int mergable_maps(struct extent_map *prev, struct extent_map *next)
 	    test_bit(EXTENT_FLAG_LOGGING, &next->flags))
 		return 0;
 
+	/*
+	 * We don't want to merge stuff that hasn't been written to the log yet
+	 * since it may not reflect exactly what is on disk, and that would be
+	 * bad.
+	 */
+	if (!list_empty(&prev->list) || !list_empty(&next->list))
+		return 0;
+
 	if (extent_map_end(prev) == next->start &&
 	    prev->flags == next->flags &&
 	    prev->bdev == next->bdev &&
@@ -209,9 +217,7 @@ static void try_merge_map(struct extent_map_tree *tree, struct extent_map *em)
 			em->mod_len = (em->mod_len + em->mod_start) - merge->mod_start;
 			em->mod_start = merge->mod_start;
 			em->generation = max(em->generation, merge->generation);
-			list_move(&em->list, &tree->modified_extents);
 
-			list_del_init(&merge->list);
 			rb_erase(&merge->rb_node, &tree->map);
 			free_extent_map(merge);
 		}
@@ -227,7 +233,6 @@ static void try_merge_map(struct extent_map_tree *tree, struct extent_map *em)
 		merge->in_tree = 0;
 		em->mod_len = (merge->mod_start + merge->mod_len) - em->mod_start;
 		em->generation = max(em->generation, merge->generation);
-		list_del_init(&merge->list);
 		free_extent_map(merge);
 	}
 }
@@ -302,7 +307,7 @@ void clear_em_logging(struct extent_map_tree *tree, struct extent_map *em)
  * reference dropped if the merge attempt was successful.
  */
 int add_extent_mapping(struct extent_map_tree *tree,
-		       struct extent_map *em)
+		       struct extent_map *em, int modified)
 {
 	int ret = 0;
 	struct rb_node *rb;
@@ -324,7 +329,10 @@ int add_extent_mapping(struct extent_map_tree *tree,
 	em->mod_start = em->start;
 	em->mod_len = em->len;
 
-	try_merge_map(tree, em);
+	if (modified)
+		list_move(&em->list, &tree->modified_extents);
+	else
+		try_merge_map(tree, em);
 out:
 	return ret;
 }
@@ -337,8 +345,9 @@ static u64 range_end(u64 start, u64 len)
 	return start + len;
 }
 
-struct extent_map *__lookup_extent_mapping(struct extent_map_tree *tree,
-					   u64 start, u64 len, int strict)
+static struct extent_map *
+__lookup_extent_mapping(struct extent_map_tree *tree,
+			u64 start, u64 len, int strict)
 {
 	struct extent_map *em;
 	struct rb_node *rb_node;

fs/btrfs/extent_map.h

@@ -26,6 +26,7 @@ struct extent_map {
 	u64 mod_len;
 	u64 orig_start;
 	u64 orig_block_len;
+	u64 ram_bytes;
 	u64 block_start;
 	u64 block_len;
 	u64 generation;
@@ -61,7 +62,7 @@ void extent_map_tree_init(struct extent_map_tree *tree);
 struct extent_map *lookup_extent_mapping(struct extent_map_tree *tree,
 					 u64 start, u64 len);
 int add_extent_mapping(struct extent_map_tree *tree,
-		       struct extent_map *em);
+		       struct extent_map *em, int modified);
 int remove_extent_mapping(struct extent_map_tree *tree, struct extent_map *em);
 
 struct extent_map *alloc_extent_map(void);

fs/btrfs/file-item.c

@@ -83,10 +83,11 @@ out:
 	return ret;
 }
 
-struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans,
-					  struct btrfs_root *root,
-					  struct btrfs_path *path,
-					  u64 bytenr, int cow)
+static struct btrfs_csum_item *
+btrfs_lookup_csum(struct btrfs_trans_handle *trans,
+		  struct btrfs_root *root,
+		  struct btrfs_path *path,
+		  u64 bytenr, int cow)
 {
 	int ret;
 	struct btrfs_key file_key;
@@ -152,32 +153,12 @@ int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
-u64 btrfs_file_extent_length(struct btrfs_path *path)
-{
-	int extent_type;
-	struct btrfs_file_extent_item *fi;
-	u64 len;
-
-	fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
-			    struct btrfs_file_extent_item);
-	extent_type = btrfs_file_extent_type(path->nodes[0], fi);
-
-	if (extent_type == BTRFS_FILE_EXTENT_REG ||
-	    extent_type == BTRFS_FILE_EXTENT_PREALLOC)
-		len = btrfs_file_extent_num_bytes(path->nodes[0], fi);
-	else if (extent_type == BTRFS_FILE_EXTENT_INLINE)
-		len = btrfs_file_extent_inline_len(path->nodes[0], fi);
-	else
-		BUG();
-
-	return len;
-}
-
 static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 				   struct inode *inode, struct bio *bio,
 				   u64 logical_offset, u32 *dst, int dio)
 {
-	u32 sum;
+	u32 sum[16];
+	int len;
 	struct bio_vec *bvec = bio->bi_io_vec;
 	int bio_index = 0;
 	u64 offset = 0;
@@ -186,7 +167,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 	u64 disk_bytenr;
 	u32 diff;
 	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
-	int ret;
+	int count;
 	struct btrfs_path *path;
 	struct btrfs_csum_item *item = NULL;
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
@@ -214,10 +195,12 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 	if (dio)
 		offset = logical_offset;
 	while (bio_index < bio->bi_vcnt) {
+		len = min_t(int, ARRAY_SIZE(sum), bio->bi_vcnt - bio_index);
 		if (!dio)
 			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
-		ret = btrfs_find_ordered_sum(inode, offset, disk_bytenr, &sum);
-		if (ret == 0)
+		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr, sum,
+					       len);
+		if (count)
 			goto found;
 
 		if (!item || disk_bytenr < item_start_offset ||
@@ -230,10 +213,8 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 			item = btrfs_lookup_csum(NULL, root->fs_info->csum_root,
 						 path, disk_bytenr, 0);
 			if (IS_ERR(item)) {
-				ret = PTR_ERR(item);
-				if (ret == -ENOENT || ret == -EFBIG)
-					ret = 0;
-				sum = 0;
+				count = 1;
+				sum[0] = 0;
 				if (BTRFS_I(inode)->root->root_key.objectid ==
 				    BTRFS_DATA_RELOC_TREE_OBJECTID) {
 					set_extent_bits(io_tree, offset,
@@ -269,19 +250,29 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 		diff = disk_bytenr - item_start_offset;
 		diff = diff / root->sectorsize;
 		diff = diff * csum_size;
-
-		read_extent_buffer(path->nodes[0], &sum,
+		count = min_t(int, len, (item_last_offset - disk_bytenr) >>
+					inode->i_sb->s_blocksize_bits);
+		read_extent_buffer(path->nodes[0], sum,
 				   ((unsigned long)item) + diff,
-				   csum_size);
+				   csum_size * count);
 found:
-		if (dst)
-			*dst++ = sum;
-		else
-			set_state_private(io_tree, offset, sum);
-		disk_bytenr += bvec->bv_len;
-		offset += bvec->bv_len;
-		bio_index++;
-		bvec++;
+		if (dst) {
+			memcpy(dst, sum, count * csum_size);
+			dst += count;
+		} else {
+			if (dio)
+				extent_cache_csums_dio(io_tree, offset, sum,
+						       count);
+			else
+				extent_cache_csums(io_tree, bio, bio_index, sum,
+					    count);
+		}
+		while (count--) {
+			disk_bytenr += bvec->bv_len;
+			offset += bvec->bv_len;
+			bio_index++;
+			bvec++;
+		}
 	}
 	btrfs_free_path(path);
 	return 0;
@@ -358,11 +349,8 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 
 		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
 		if (key.objectid != BTRFS_EXTENT_CSUM_OBJECTID ||
-		    key.type != BTRFS_EXTENT_CSUM_KEY)
-			break;
-
-		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
-		if (key.offset > end)
+		    key.type != BTRFS_EXTENT_CSUM_KEY ||
+		    key.offset > end)
 			break;
 
 		if (key.offset > start)
@@ -484,8 +472,7 @@ int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
 
 		data = kmap_atomic(bvec->bv_page);
 		sector_sum->sum = ~(u32)0;
-		sector_sum->sum = btrfs_csum_data(root,
-						  data + bvec->bv_offset,
+		sector_sum->sum = btrfs_csum_data(data + bvec->bv_offset,
 						  sector_sum->sum,
 						  bvec->bv_len);
 		kunmap_atomic(data);
@@ -518,8 +505,7 @@ int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
  * This calls btrfs_truncate_item with the correct args based on the
  * overlap, and fixes up the key as required.
  */
-static noinline void truncate_one_csum(struct btrfs_trans_handle *trans,
-				       struct btrfs_root *root,
+static noinline void truncate_one_csum(struct btrfs_root *root,
 				       struct btrfs_path *path,
 				       struct btrfs_key *key,
 				       u64 bytenr, u64 len)
@@ -544,7 +530,7 @@ static noinline void truncate_one_csum(struct btrfs_trans_handle *trans,
 		 */
 		u32 new_size = (bytenr - key->offset) >> blocksize_bits;
 		new_size *= csum_size;
-		btrfs_truncate_item(trans, root, path, new_size, 1);
+		btrfs_truncate_item(root, path, new_size, 1);
 	} else if (key->offset >= bytenr && csum_end > end_byte &&
 		   end_byte > key->offset) {
 		/*
@@ -556,10 +542,10 @@ static noinline void truncate_one_csum(struct btrfs_trans_handle *trans,
 		u32 new_size = (csum_end - end_byte) >> blocksize_bits;
 		new_size *= csum_size;
 
-		btrfs_truncate_item(trans, root, path, new_size, 0);
+		btrfs_truncate_item(root, path, new_size, 0);
 
 		key->offset = end_byte;
-		btrfs_set_item_key_safe(trans, root, path, key);
+		btrfs_set_item_key_safe(root, path, key);
 	} else {
 		BUG();
 	}
@@ -674,7 +660,7 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
 
 			key.offset = end_byte - 1;
 		} else {
-			truncate_one_csum(trans, root, path, &key, bytenr, len);
+			truncate_one_csum(root, path, &key, bytenr, len);
 			if (key.offset < bytenr)
 				break;
 		}
@@ -835,7 +821,7 @@ again:
 		diff /= csum_size;
 		diff *= csum_size;
 
-		btrfs_extend_item(trans, root, path, diff);
+		btrfs_extend_item(root, path, diff);
 		goto csum;
 	}
 

fs/btrfs/file.c

@@ -193,8 +193,8 @@ int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
  * the same inode in the tree, we will merge them together (by
  * __btrfs_add_inode_defrag()) and free the one that we want to requeue.
  */
-void btrfs_requeue_inode_defrag(struct inode *inode,
-				struct inode_defrag *defrag)
+static void btrfs_requeue_inode_defrag(struct inode *inode,
+				       struct inode_defrag *defrag)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	int ret;
@@ -474,7 +474,7 @@ static noinline int btrfs_copy_from_user(loff_t pos, int num_pages,
 /*
  * unlocks pages after btrfs_file_write is done with them
  */
-void btrfs_drop_pages(struct page **pages, size_t num_pages)
+static void btrfs_drop_pages(struct page **pages, size_t num_pages)
 {
 	size_t i;
 	for (i = 0; i < num_pages; i++) {
@@ -498,9 +498,9 @@ void btrfs_drop_pages(struct page **pages, size_t num_pages)
  * doing real data extents, marking pages dirty and delalloc as required.
  */
 int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode,
-		      struct page **pages, size_t num_pages,
-		      loff_t pos, size_t write_bytes,
-		      struct extent_state **cached)
+			     struct page **pages, size_t num_pages,
+			     loff_t pos, size_t write_bytes,
+			     struct extent_state **cached)
 {
 	int err = 0;
 	int i;
@@ -553,6 +553,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 	int testend = 1;
 	unsigned long flags;
 	int compressed = 0;
+	bool modified;
 
 	WARN_ON(end < start);
 	if (end == (u64)-1) {
@@ -562,6 +563,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 	while (1) {
 		int no_splits = 0;
 
+		modified = false;
 		if (!split)
 			split = alloc_extent_map();
 		if (!split2)
@@ -593,6 +595,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 		compressed = test_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
 		clear_bit(EXTENT_FLAG_PINNED, &em->flags);
 		clear_bit(EXTENT_FLAG_LOGGING, &flags);
+		modified = !list_empty(&em->list);
 		remove_extent_mapping(em_tree, em);
 		if (no_splits)
 			goto next;
@@ -608,15 +611,15 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 				split->block_len = em->block_len;
 			else
 				split->block_len = split->len;
+			split->ram_bytes = em->ram_bytes;
 			split->orig_block_len = max(split->block_len,
 						    em->orig_block_len);
 			split->generation = gen;
 			split->bdev = em->bdev;
 			split->flags = flags;
 			split->compress_type = em->compress_type;
-			ret = add_extent_mapping(em_tree, split);
+			ret = add_extent_mapping(em_tree, split, modified);
 			BUG_ON(ret); /* Logic error */
-			list_move(&split->list, &em_tree->modified_extents);
 			free_extent_map(split);
 			split = split2;
 			split2 = NULL;
@@ -633,6 +636,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 			split->generation = gen;
 			split->orig_block_len = max(em->block_len,
 						    em->orig_block_len);
+			split->ram_bytes = em->ram_bytes;
 
 			if (compressed) {
 				split->block_len = em->block_len;
@@ -644,9 +648,8 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 				split->orig_start = em->orig_start;
 			}
 
-			ret = add_extent_mapping(em_tree, split);
+			ret = add_extent_mapping(em_tree, split, modified);
 			BUG_ON(ret); /* Logic error */
-			list_move(&split->list, &em_tree->modified_extents);
 			free_extent_map(split);
 			split = NULL;
 		}
@@ -822,7 +825,7 @@ next_slot:
 
 			memcpy(&new_key, &key, sizeof(new_key));
 			new_key.offset = end;
-			btrfs_set_item_key_safe(trans, root, path, &new_key);
+			btrfs_set_item_key_safe(root, path, &new_key);
 
 			extent_offset += end - key.offset;
 			btrfs_set_file_extent_offset(leaf, fi, extent_offset);
@@ -1038,7 +1041,7 @@ again:
 				     ino, bytenr, orig_offset,
 				     &other_start, &other_end)) {
 			new_key.offset = end;
-			btrfs_set_item_key_safe(trans, root, path, &new_key);
+			btrfs_set_item_key_safe(root, path, &new_key);
 			fi = btrfs_item_ptr(leaf, path->slots[0],
 					    struct btrfs_file_extent_item);
 			btrfs_set_file_extent_generation(leaf, fi,
@@ -1072,7 +1075,7 @@ again:
 							 trans->transid);
 			path->slots[0]++;
 			new_key.offset = start;
-			btrfs_set_item_key_safe(trans, root, path, &new_key);
+			btrfs_set_item_key_safe(root, path, &new_key);
 
 			fi = btrfs_item_ptr(leaf, path->slots[0],
 					    struct btrfs_file_extent_item);
@@ -1883,7 +1886,7 @@ static int fill_holes(struct btrfs_trans_handle *trans, struct inode *inode,
 
 		path->slots[0]++;
 		key.offset = offset;
-		btrfs_set_item_key_safe(trans, root, path, &key);
+		btrfs_set_item_key_safe(root, path, &key);
 		fi = btrfs_item_ptr(leaf, path->slots[0],
 				    struct btrfs_file_extent_item);
 		num_bytes = btrfs_file_extent_num_bytes(leaf, fi) + end -
@@ -1913,6 +1916,7 @@ out:
 	} else {
 		hole_em->start = offset;
 		hole_em->len = end - offset;
+		hole_em->ram_bytes = hole_em->len;
 		hole_em->orig_start = offset;
 
 		hole_em->block_start = EXTENT_MAP_HOLE;
@@ -1925,10 +1929,7 @@ out:
 		do {
 			btrfs_drop_extent_cache(inode, offset, end - 1, 0);
 			write_lock(&em_tree->lock);
-			ret = add_extent_mapping(em_tree, hole_em);
-			if (!ret)
-				list_move(&hole_em->list,
-					  &em_tree->modified_extents);
+			ret = add_extent_mapping(em_tree, hole_em, 1);
 			write_unlock(&em_tree->lock);
 		} while (ret == -EEXIST);
 		free_extent_map(hole_em);

fs/btrfs/free-space-cache.c

@@ -104,7 +104,8 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root,
 
 	spin_lock(&block_group->lock);
 	if (!((BTRFS_I(inode)->flags & flags) == flags)) {
-		printk(KERN_INFO "Old style space inode found, converting.\n");
+		btrfs_info(root->fs_info,
+			"Old style space inode found, converting.");
 		BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM |
 			BTRFS_INODE_NODATACOW;
 		block_group->disk_cache_state = BTRFS_DC_CLEAR;
@@ -119,9 +120,10 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root,
 	return inode;
 }
 
-int __create_free_space_inode(struct btrfs_root *root,
-			      struct btrfs_trans_handle *trans,
-			      struct btrfs_path *path, u64 ino, u64 offset)
+static int __create_free_space_inode(struct btrfs_root *root,
+				     struct btrfs_trans_handle *trans,
+				     struct btrfs_path *path,
+				     u64 ino, u64 offset)
 {
 	struct btrfs_key key;
 	struct btrfs_disk_key disk_key;
@@ -431,7 +433,7 @@ static void io_ctl_set_crc(struct io_ctl *io_ctl, int index)
 	if (index == 0)
 		offset = sizeof(u32) * io_ctl->num_pages;
 
-	crc = btrfs_csum_data(io_ctl->root, io_ctl->orig + offset, crc,
+	crc = btrfs_csum_data(io_ctl->orig + offset, crc,
 			      PAGE_CACHE_SIZE - offset);
 	btrfs_csum_final(crc, (char *)&crc);
 	io_ctl_unmap_page(io_ctl);
@@ -461,7 +463,7 @@ static int io_ctl_check_crc(struct io_ctl *io_ctl, int index)
 	kunmap(io_ctl->pages[0]);
 
 	io_ctl_map_page(io_ctl, 0);
-	crc = btrfs_csum_data(io_ctl->root, io_ctl->orig + offset, crc,
+	crc = btrfs_csum_data(io_ctl->orig + offset, crc,
 			      PAGE_CACHE_SIZE - offset);
 	btrfs_csum_final(crc, (char *)&crc);
 	if (val != crc) {
@@ -624,9 +626,9 @@ next:
 	spin_unlock(&ctl->tree_lock);
 }
 
-int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
-			    struct btrfs_free_space_ctl *ctl,
-			    struct btrfs_path *path, u64 offset)
+static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
+				   struct btrfs_free_space_ctl *ctl,
+				   struct btrfs_path *path, u64 offset)
 {
 	struct btrfs_free_space_header *header;
 	struct extent_buffer *leaf;
@@ -669,10 +671,11 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 	btrfs_release_path(path);
 
 	if (BTRFS_I(inode)->generation != generation) {
-		printk(KERN_ERR "btrfs: free space inode generation (%llu) did"
-		       " not match free space cache generation (%llu)\n",
-		       (unsigned long long)BTRFS_I(inode)->generation,
-		       (unsigned long long)generation);
+		btrfs_err(root->fs_info,
+			"free space inode generation (%llu) "
+			"did not match free space cache generation (%llu)",
+			(unsigned long long)BTRFS_I(inode)->generation,
+			(unsigned long long)generation);
 		return 0;
 	}
 
@@ -721,8 +724,8 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 			ret = link_free_space(ctl, e);
 			spin_unlock(&ctl->tree_lock);
 			if (ret) {
-				printk(KERN_ERR "Duplicate entries in "
-				       "free space cache, dumping\n");
+				btrfs_err(root->fs_info,
+					"Duplicate entries in free space cache, dumping");
 				kmem_cache_free(btrfs_free_space_cachep, e);
 				goto free_cache;
 			}
@@ -741,8 +744,8 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 			ctl->op->recalc_thresholds(ctl);
 			spin_unlock(&ctl->tree_lock);
 			if (ret) {
-				printk(KERN_ERR "Duplicate entries in "
-				       "free space cache, dumping\n");
+				btrfs_err(root->fs_info,
+					"Duplicate entries in free space cache, dumping");
 				kmem_cache_free(btrfs_free_space_cachep, e);
 				goto free_cache;
 			}
@@ -833,8 +836,8 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info,
 
 	if (!matched) {
 		__btrfs_remove_free_space_cache(ctl);
-		printk(KERN_ERR "block group %llu has an wrong amount of free "
-		       "space\n", block_group->key.objectid);
+		btrfs_err(fs_info, "block group %llu has wrong amount of free space",
+			block_group->key.objectid);
 		ret = -1;
 	}
 out:
@@ -845,8 +848,8 @@ out:
 		spin_unlock(&block_group->lock);
 		ret = 0;
 
-		printk(KERN_ERR "btrfs: failed to load free space cache "
-		       "for block group %llu\n", block_group->key.objectid);
+		btrfs_err(fs_info, "failed to load free space cache for block group %llu",
+			block_group->key.objectid);
 	}
 
 	iput(inode);
@@ -866,11 +869,11 @@ out:
  * on mount.  This will return 0 if it was successfull in writing the cache out,
  * and -1 if it was not.
  */
-int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
-			    struct btrfs_free_space_ctl *ctl,
-			    struct btrfs_block_group_cache *block_group,
-			    struct btrfs_trans_handle *trans,
-			    struct btrfs_path *path, u64 offset)
+static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
+				   struct btrfs_free_space_ctl *ctl,
+				   struct btrfs_block_group_cache *block_group,
+				   struct btrfs_trans_handle *trans,
+				   struct btrfs_path *path, u64 offset)
 {
 	struct btrfs_free_space_header *header;
 	struct extent_buffer *leaf;
@@ -1104,8 +1107,9 @@ int btrfs_write_out_cache(struct btrfs_root *root,
 		spin_unlock(&block_group->lock);
 		ret = 0;
 #ifdef DEBUG
-		printk(KERN_ERR "btrfs: failed to write free space cache "
-		       "for block group %llu\n", block_group->key.objectid);
+		btrfs_err(root->fs_info,
+			"failed to write free space cache for block group %llu",
+			block_group->key.objectid);
 #endif
 	}
 
@@ -1564,7 +1568,8 @@ again:
 	search_bytes = ctl->unit;
 	search_bytes = min(search_bytes, end - search_start + 1);
 	ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes);
-	BUG_ON(ret < 0 || search_start != *offset);
+	if (ret < 0 || search_start != *offset)
+		return -EINVAL;
 
 	/* We may have found more bits than what we need */
 	search_bytes = min(search_bytes, *bytes);
@@ -1970,7 +1975,6 @@ again:
 		re_search = true;
 		goto again;
 	}
-	BUG_ON(ret); /* logic error */
 out_lock:
 	spin_unlock(&ctl->tree_lock);
 out:
@@ -2064,7 +2068,8 @@ out:
 	return 0;
 }
 
-void __btrfs_remove_free_space_cache_locked(struct btrfs_free_space_ctl *ctl)
+static void __btrfs_remove_free_space_cache_locked(
+				struct btrfs_free_space_ctl *ctl)
 {
 	struct btrfs_free_space *info;
 	struct rb_node *node;
@@ -2931,8 +2936,9 @@ int load_free_ino_cache(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
 	ret = __load_free_space_cache(root, inode, ctl, path, 0);
 
 	if (ret < 0)
-		printk(KERN_ERR "btrfs: failed to load free ino cache for "
-		       "root %llu\n", root->root_key.objectid);
+		btrfs_err(fs_info,
+			"failed to load free ino cache for root %llu",
+			root->root_key.objectid);
 out_put:
 	iput(inode);
 out:
@@ -2959,11 +2965,531 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root,
 	if (ret) {
 		btrfs_delalloc_release_metadata(inode, inode->i_size);
 #ifdef DEBUG
-		printk(KERN_ERR "btrfs: failed to write free ino cache "
-		       "for root %llu\n", root->root_key.objectid);
+		btrfs_err(root->fs_info,
+			"failed to write free ino cache for root %llu",
+			root->root_key.objectid);
 #endif
 	}
 
 	iput(inode);
 	return ret;
 }
+
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+static struct btrfs_block_group_cache *init_test_block_group(void)
+{
+	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 = 0;
+	cache->key.offset = 1024 * 1024 * 1024;
+	cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
+	cache->sectorsize = 4096;
+
+	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;
+}
+
+/*
+ * Checks to see if the given range is in the free space cache.  This is really
+ * just used to check the absence of space, so if there is free space in the
+ * range at all we will return 1.
+ */
+static int check_exists(struct btrfs_block_group_cache *cache, u64 offset,
+			u64 bytes)
+{
+	struct btrfs_free_space_ctl *ctl = cache->free_space_ctl;
+	struct btrfs_free_space *info;
+	int ret = 0;
+
+	spin_lock(&ctl->tree_lock);
+	info = tree_search_offset(ctl, offset, 0, 0);
+	if (!info) {
+		info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
+					  1, 0);
+		if (!info)
+			goto out;
+	}
+
+have_info:
+	if (info->bitmap) {
+		u64 bit_off, bit_bytes;
+		struct rb_node *n;
+		struct btrfs_free_space *tmp;
+
+		bit_off = offset;
+		bit_bytes = ctl->unit;
+		ret = search_bitmap(ctl, info, &bit_off, &bit_bytes);
+		if (!ret) {
+			if (bit_off == offset) {
+				ret = 1;
+				goto out;
+			} else if (bit_off > offset &&
+				   offset + bytes > bit_off) {
+				ret = 1;
+				goto out;
+			}
+		}
+
+		n = rb_prev(&info->offset_index);
+		while (n) {
+			tmp = rb_entry(n, struct btrfs_free_space,
+				       offset_index);
+			if (tmp->offset + tmp->bytes < offset)
+				break;
+			if (offset + bytes < tmp->offset) {
+				n = rb_prev(&info->offset_index);
+				continue;
+			}
+			info = tmp;
+			goto have_info;
+		}
+
+		n = rb_next(&info->offset_index);
+		while (n) {
+			tmp = rb_entry(n, struct btrfs_free_space,
+				       offset_index);
+			if (offset + bytes < tmp->offset)
+				break;
+			if (tmp->offset + tmp->bytes < offset) {
+				n = rb_next(&info->offset_index);
+				continue;
+			}
+			info = tmp;
+			goto have_info;
+		}
+
+		goto out;
+	}
+
+	if (info->offset == offset) {
+		ret = 1;
+		goto out;
+	}
+
+	if (offset > info->offset && offset < info->offset + info->bytes)
+		ret = 1;
+out:
+	spin_unlock(&ctl->tree_lock);
+	return ret;
+}
+
+/*
+ * Use this if you need to make a bitmap or extent entry specifically, it
+ * doesn't do any of the merging that add_free_space does, this acts a lot like
+ * how the free space cache loading stuff works, so you can get really weird
+ * configurations.
+ */
+static int add_free_space_entry(struct btrfs_block_group_cache *cache,
+				u64 offset, u64 bytes, bool bitmap)
+{
+	struct btrfs_free_space_ctl *ctl = cache->free_space_ctl;
+	struct btrfs_free_space *info = NULL, *bitmap_info;
+	void *map = NULL;
+	u64 bytes_added;
+	int ret;
+
+again:
+	if (!info) {
+		info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS);
+		if (!info)
+			return -ENOMEM;
+	}
+
+	if (!bitmap) {
+		spin_lock(&ctl->tree_lock);
+		info->offset = offset;
+		info->bytes = bytes;
+		ret = link_free_space(ctl, info);
+		spin_unlock(&ctl->tree_lock);
+		if (ret)
+			kmem_cache_free(btrfs_free_space_cachep, info);
+		return ret;
+	}
+
+	if (!map) {
+		map = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS);
+		if (!map) {
+			kmem_cache_free(btrfs_free_space_cachep, info);
+			return -ENOMEM;
+		}
+	}
+
+	spin_lock(&ctl->tree_lock);
+	bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
+					 1, 0);
+	if (!bitmap_info) {
+		info->bitmap = map;
+		map = NULL;
+		add_new_bitmap(ctl, info, offset);
+		bitmap_info = info;
+	}
+
+	bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes);
+	bytes -= bytes_added;
+	offset += bytes_added;
+	spin_unlock(&ctl->tree_lock);
+
+	if (bytes)
+		goto again;
+
+	if (map)
+		kfree(map);
+	return 0;
+}
+
+/*
+ * This test just does basic sanity checking, making sure we can add an exten
+ * entry and remove space from either end and the middle, and make sure we can
+ * remove space that covers adjacent extent entries.
+ */
+static int test_extents(struct btrfs_block_group_cache *cache)
+{
+	int ret = 0;
+
+	printk(KERN_ERR "Running extent only tests\n");
+
+	/* First just make sure we can remove an entire entry */
+	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Error adding initial extents %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Error removing extent %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, 0, 4 * 1024 * 1024)) {
+		printk(KERN_ERR "Full remove left some lingering space\n");
+		return -1;
+	}
+
+	/* Ok edge and middle cases now */
+	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Error adding half extent %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 1 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Error removing tail end %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Error removing front end %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 2 * 1024 * 1024, 4096);
+	if (ret) {
+		printk(KERN_ERR "Error removing middle peice %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, 0, 1 * 1024 * 1024)) {
+		printk(KERN_ERR "Still have space at the front\n");
+		return -1;
+	}
+
+	if (check_exists(cache, 2 * 1024 * 1024, 4096)) {
+		printk(KERN_ERR "Still have space in the middle\n");
+		return -1;
+	}
+
+	if (check_exists(cache, 3 * 1024 * 1024, 1 * 1024 * 1024)) {
+		printk(KERN_ERR "Still have space at the end\n");
+		return -1;
+	}
+
+	/* Cleanup */
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	return 0;
+}
+
+static int test_bitmaps(struct btrfs_block_group_cache *cache)
+{
+	u64 next_bitmap_offset;
+	int ret;
+
+	printk(KERN_ERR "Running bitmap only tests\n");
+
+	ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't create a bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Error removing bitmap full range %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, 0, 4 * 1024 * 1024)) {
+		printk(KERN_ERR "Left some space in bitmap\n");
+		return -1;
+	}
+
+	ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add to our bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 2 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Couldn't remove middle chunk %d\n", ret);
+		return ret;
+	}
+
+	/*
+	 * The first bitmap we have starts at offset 0 so the next one is just
+	 * at the end of the first bitmap.
+	 */
+	next_bitmap_offset = (u64)(BITS_PER_BITMAP * 4096);
+
+	/* Test a bit straddling two bitmaps */
+	ret = add_free_space_entry(cache, next_bitmap_offset -
+				   (2 * 1024 * 1024), 4 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add space that straddles two bitmaps"
+		       " %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, next_bitmap_offset -
+				      (1 * 1024 * 1024), 2 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Couldn't remove overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, next_bitmap_offset - (1 * 1024 * 1024),
+			 2 * 1024 * 1024)) {
+		printk(KERN_ERR "Left some space when removing overlapping\n");
+		return -1;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	return 0;
+}
+
+/* This is the high grade jackassery */
+static int test_bitmaps_and_extents(struct btrfs_block_group_cache *cache)
+{
+	u64 bitmap_offset = (u64)(BITS_PER_BITMAP * 4096);
+	int ret;
+
+	printk(KERN_ERR "Running bitmap and extent tests\n");
+
+	/*
+	 * First let's do something simple, an extent at the same offset as the
+	 * bitmap, but the free space completely in the extent and then
+	 * completely in the bitmap.
+	 */
+	ret = add_free_space_entry(cache, 4 * 1024 * 1024, 1 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't create bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Couldn't remove extent entry %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, 0, 1 * 1024 * 1024)) {
+		printk(KERN_ERR "Left remnants after our remove\n");
+		return -1;
+	}
+
+	/* Now to add back the extent entry and remove from the bitmap */
+	ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
+	if (ret) {
+		printk(KERN_ERR "Couldn't re-add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 4 * 1024 * 1024, 1 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Couldn't remove from bitmap %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, 4 * 1024 * 1024, 1 * 1024 * 1024)) {
+		printk(KERN_ERR "Left remnants in the bitmap\n");
+		return -1;
+	}
+
+	/*
+	 * Ok so a little more evil, extent entry and bitmap at the same offset,
+	 * removing an overlapping chunk.
+	 */
+	ret = add_free_space_entry(cache, 1 * 1024 * 1024, 4 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add to a bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 512 * 1024, 3 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Couldn't remove overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, 512 * 1024, 3 * 1024 * 1024)) {
+		printk(KERN_ERR "Left over peices after removing "
+		       "overlapping\n");
+		return -1;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	/* Now with the extent entry offset into the bitmap */
+	ret = add_free_space_entry(cache, 4 * 1024 * 1024, 4 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add space to the bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = add_free_space_entry(cache, 2 * 1024 * 1024, 2 * 1024 * 1024, 0);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add extent to the cache %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 4 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Problem removing overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, 3 * 1024 * 1024, 4 * 1024 * 1024)) {
+		printk(KERN_ERR "Left something behind when removing space");
+		return -1;
+	}
+
+	/*
+	 * This has blown up in the past, the extent entry starts before the
+	 * bitmap entry, but we're trying to remove an offset that falls
+	 * completely within the bitmap range and is in both the extent entry
+	 * and the bitmap entry, looks like this
+	 *
+	 *   [ extent ]
+	 *      [ bitmap ]
+	 *        [ del ]
+	 */
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	ret = add_free_space_entry(cache, bitmap_offset + 4 * 1024 * 1024,
+				   4 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = add_free_space_entry(cache, bitmap_offset - 1 * 1024 * 1024,
+				   5 * 1024 * 1024, 0);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, bitmap_offset + 1 * 1024 * 1024,
+				      5 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Failed to free our space %d\n", ret);
+		return ret;
+	}
+
+	if (check_exists(cache, bitmap_offset + 1 * 1024 * 1024,
+			 5 * 1024 * 1024)) {
+		printk(KERN_ERR "Left stuff over\n");
+		return -1;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+
+	/*
+	 * This blew up before, we have part of the free space in a bitmap and
+	 * then the entirety of the rest of the space in an extent.  This used
+	 * to return -EAGAIN back from btrfs_remove_extent, make sure this
+	 * doesn't happen.
+	 */
+	ret = add_free_space_entry(cache, 1 * 1024 * 1024, 2 * 1024 * 1024, 1);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = add_free_space_entry(cache, 3 * 1024 * 1024, 1 * 1024 * 1024, 0);
+	if (ret) {
+		printk(KERN_ERR "Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 3 * 1024 * 1024);
+	if (ret) {
+		printk(KERN_ERR "Error removing bitmap and extent "
+		       "overlapping %d\n", ret);
+		return ret;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	return 0;
+}
+
+void btrfs_test_free_space_cache(void)
+{
+	struct btrfs_block_group_cache *cache;
+
+	printk(KERN_ERR "Running btrfs free space cache tests\n");
+
+	cache = init_test_block_group();
+	if (!cache) {
+		printk(KERN_ERR "Couldn't run the tests\n");
+		return;
+	}
+
+	if (test_extents(cache))
+		goto out;
+	if (test_bitmaps(cache))
+		goto out;
+	if (test_bitmaps_and_extents(cache))
+		goto out;
+out:
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	kfree(cache->free_space_ctl);
+	kfree(cache);
+	printk(KERN_ERR "Free space cache tests finished\n");
+}
+#endif /* CONFIG_BTRFS_FS_RUN_SANITY_TESTS */

fs/btrfs/free-space-cache.h

@@ -110,4 +110,9 @@ int btrfs_return_cluster_to_free_space(
 			       struct btrfs_free_cluster *cluster);
 int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group,
 			   u64 *trimmed, u64 start, u64 end, u64 minlen);
+
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+void btrfs_test_free_space_cache(void);
+#endif
+
 #endif

fs/btrfs/inode-item.c

@@ -183,10 +183,11 @@ int btrfs_get_inode_ref_index(struct btrfs_trans_handle *trans,
 	return -ENOENT;
 }
 
-int btrfs_del_inode_extref(struct btrfs_trans_handle *trans,
-			   struct btrfs_root *root,
-			   const char *name, int name_len,
-			   u64 inode_objectid, u64 ref_objectid, u64 *index)
+static int btrfs_del_inode_extref(struct btrfs_trans_handle *trans,
+				  struct btrfs_root *root,
+				  const char *name, int name_len,
+				  u64 inode_objectid, u64 ref_objectid,
+				  u64 *index)
 {
 	struct btrfs_path *path;
 	struct btrfs_key key;
@@ -246,7 +247,7 @@ int btrfs_del_inode_extref(struct btrfs_trans_handle *trans,
 	memmove_extent_buffer(leaf, ptr, ptr + del_len,
 			      item_size - (ptr + del_len - item_start));
 
-	btrfs_truncate_item(trans, root, path, item_size - del_len, 1);
+	btrfs_truncate_item(root, path, item_size - del_len, 1);
 
 out:
 	btrfs_free_path(path);
@@ -309,7 +310,7 @@ int btrfs_del_inode_ref(struct btrfs_trans_handle *trans,
 	item_start = btrfs_item_ptr_offset(leaf, path->slots[0]);
 	memmove_extent_buffer(leaf, ptr, ptr + sub_item_len,
 			      item_size - (ptr + sub_item_len - item_start));
-	btrfs_truncate_item(trans, root, path, item_size - sub_item_len, 1);
+	btrfs_truncate_item(root, path, item_size - sub_item_len, 1);
 out:
 	btrfs_free_path(path);
 
@@ -361,7 +362,7 @@ static int btrfs_insert_inode_extref(struct btrfs_trans_handle *trans,
 						   name, name_len, NULL))
 			goto out;
 
-		btrfs_extend_item(trans, root, path, ins_len);
+		btrfs_extend_item(root, path, ins_len);
 		ret = 0;
 	}
 	if (ret < 0)
@@ -417,7 +418,7 @@ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans,
 			goto out;
 
 		old_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]);
-		btrfs_extend_item(trans, root, path, ins_len);
+		btrfs_extend_item(root, path, ins_len);
 		ref = btrfs_item_ptr(path->nodes[0], path->slots[0],
 				     struct btrfs_inode_ref);
 		ref = (struct btrfs_inode_ref *)((unsigned long)ref + old_size);

fs/btrfs/inode.c

@@ -101,7 +101,10 @@ static noinline int cow_file_range(struct inode *inode,
 static struct extent_map *create_pinned_em(struct inode *inode, u64 start,
 					   u64 len, u64 orig_start,
 					   u64 block_start, u64 block_len,
-					   u64 orig_block_len, int type);
+					   u64 orig_block_len, u64 ram_bytes,
+					   int type);
+
+static int btrfs_dirty_inode(struct inode *inode);
 
 static int btrfs_init_inode_security(struct btrfs_trans_handle *trans,
 				     struct inode *inode,  struct inode *dir,
@@ -723,6 +726,7 @@ retry:
 		em->block_start = ins.objectid;
 		em->block_len = ins.offset;
 		em->orig_block_len = ins.offset;
+		em->ram_bytes = async_extent->ram_size;
 		em->bdev = root->fs_info->fs_devices->latest_bdev;
 		em->compress_type = async_extent->compress_type;
 		set_bit(EXTENT_FLAG_PINNED, &em->flags);
@@ -731,10 +735,7 @@ retry:
 
 		while (1) {
 			write_lock(&em_tree->lock);
-			ret = add_extent_mapping(em_tree, em);
-			if (!ret)
-				list_move(&em->list,
-					  &em_tree->modified_extents);
+			ret = add_extent_mapping(em_tree, em, 1);
 			write_unlock(&em_tree->lock);
 			if (ret != -EEXIST) {
 				free_extent_map(em);
@@ -922,7 +923,8 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 		}
 
 		em = alloc_extent_map();
-		BUG_ON(!em); /* -ENOMEM */
+		if (!em)
+			goto out_reserve;
 		em->start = start;
 		em->orig_start = em->start;
 		ram_size = ins.offset;
@@ -933,16 +935,14 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 		em->block_start = ins.objectid;
 		em->block_len = ins.offset;
 		em->orig_block_len = ins.offset;
+		em->ram_bytes = ram_size;
 		em->bdev = root->fs_info->fs_devices->latest_bdev;
 		set_bit(EXTENT_FLAG_PINNED, &em->flags);
 		em->generation = -1;
 
 		while (1) {
 			write_lock(&em_tree->lock);
-			ret = add_extent_mapping(em_tree, em);
-			if (!ret)
-				list_move(&em->list,
-					  &em_tree->modified_extents);
+			ret = add_extent_mapping(em_tree, em, 1);
 			write_unlock(&em_tree->lock);
 			if (ret != -EEXIST) {
 				free_extent_map(em);
@@ -951,11 +951,14 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 			btrfs_drop_extent_cache(inode, start,
 						start + ram_size - 1, 0);
 		}
+		if (ret)
+			goto out_reserve;
 
 		cur_alloc_size = ins.offset;
 		ret = btrfs_add_ordered_extent(inode, start, ins.objectid,
 					       ram_size, cur_alloc_size, 0);
-		BUG_ON(ret); /* -ENOMEM */
+		if (ret)
+			goto out_reserve;
 
 		if (root->root_key.objectid ==
 		    BTRFS_DATA_RELOC_TREE_OBJECTID) {
@@ -963,7 +966,7 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 						      cur_alloc_size);
 			if (ret) {
 				btrfs_abort_transaction(trans, root, ret);
-				goto out_unlock;
+				goto out_reserve;
 			}
 		}
 
@@ -992,6 +995,8 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans,
 out:
 	return ret;
 
+out_reserve:
+	btrfs_free_reserved_extent(root, ins.objectid, ins.offset);
 out_unlock:
 	extent_clear_unlock_delalloc(inode,
 		     &BTRFS_I(inode)->io_tree,
@@ -1195,6 +1200,7 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 	u64 disk_bytenr;
 	u64 num_bytes;
 	u64 disk_num_bytes;
+	u64 ram_bytes;
 	int extent_type;
 	int ret, err;
 	int type;
@@ -1291,6 +1297,7 @@ next_slot:
 				    struct btrfs_file_extent_item);
 		extent_type = btrfs_file_extent_type(leaf, fi);
 
+		ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
 		if (extent_type == BTRFS_FILE_EXTENT_REG ||
 		    extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
 			disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
@@ -1374,6 +1381,7 @@ out_check:
 			em->block_len = num_bytes;
 			em->block_start = disk_bytenr;
 			em->orig_block_len = disk_num_bytes;
+			em->ram_bytes = ram_bytes;
 			em->bdev = root->fs_info->fs_devices->latest_bdev;
 			em->mod_start = em->start;
 			em->mod_len = em->len;
@@ -1382,10 +1390,7 @@ out_check:
 			em->generation = -1;
 			while (1) {
 				write_lock(&em_tree->lock);
-				ret = add_extent_mapping(em_tree, em);
-				if (!ret)
-					list_move(&em->list,
-						  &em_tree->modified_extents);
+				ret = add_extent_mapping(em_tree, em, 1);
 				write_unlock(&em_tree->lock);
 				if (ret != -EEXIST) {
 					free_extent_map(em);
@@ -1526,7 +1531,7 @@ static void btrfs_merge_extent_hook(struct inode *inode,
  * have pending delalloc work to be done.
  */
 static void btrfs_set_bit_hook(struct inode *inode,
-			       struct extent_state *state, int *bits)
+			       struct extent_state *state, unsigned long *bits)
 {
 
 	/*
@@ -1570,7 +1575,8 @@ static void btrfs_set_bit_hook(struct inode *inode,
  * extent_io.c clear_bit_hook, see set_bit_hook for why
  */
 static void btrfs_clear_bit_hook(struct inode *inode,
-				 struct extent_state *state, int *bits)
+				 struct extent_state *state,
+				 unsigned long *bits)
 {
 	/*
 	 * set_bit and clear bit hooks normally require _irqsave/restore
@@ -2794,6 +2800,8 @@ static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 	int ret;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	u32 csum = ~(u32)0;
+	static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
+	                              DEFAULT_RATELIMIT_BURST);
 
 	if (PageChecked(page)) {
 		ClearPageChecked(page);
@@ -2820,7 +2828,7 @@ static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 	if (ret)
 		goto zeroit;
 
-	csum = btrfs_csum_data(root, kaddr + offset, csum,  end - start + 1);
+	csum = btrfs_csum_data(kaddr + offset, csum,  end - start + 1);
 	btrfs_csum_final(csum, (char *)&csum);
 	if (csum != private)
 		goto zeroit;
@@ -2830,11 +2838,11 @@ good:
 	return 0;
 
 zeroit:
-	printk_ratelimited(KERN_INFO "btrfs csum failed ino %llu off %llu csum %u "
-		       "private %llu\n",
-		       (unsigned long long)btrfs_ino(page->mapping->host),
-		       (unsigned long long)start, csum,
-		       (unsigned long long)private);
+	if (__ratelimit(&_rs))
+		btrfs_info(root->fs_info, "csum failed ino %llu off %llu csum %u private %llu",
+			(unsigned long long)btrfs_ino(page->mapping->host),
+			(unsigned long long)start, csum,
+			(unsigned long long)private);
 	memset(kaddr + offset, 1, end - start + 1);
 	flush_dcache_page(page);
 	kunmap_atomic(kaddr);
@@ -3020,7 +3028,8 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
  * We have done the truncate/delete so we can go ahead and remove the orphan
  * item for this particular inode.
  */
-int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode)
+static int btrfs_orphan_del(struct btrfs_trans_handle *trans,
+			    struct inode *inode)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	int delete_item = 0;
@@ -3115,8 +3124,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 		 */
 
 		if (found_key.offset == last_objectid) {
-			printk(KERN_ERR "btrfs: Error removing orphan entry, "
-			       "stopping orphan cleanup\n");
+			btrfs_err(root->fs_info,
+				"Error removing orphan entry, stopping orphan cleanup");
 			ret = -EINVAL;
 			goto out;
 		}
@@ -3173,8 +3182,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 				ret = PTR_ERR(trans);
 				goto out;
 			}
-			printk(KERN_ERR "auto deleting %Lu\n",
-			       found_key.objectid);
+			btrfs_debug(root->fs_info, "auto deleting %Lu",
+				found_key.objectid);
 			ret = btrfs_del_orphan_item(trans, root,
 						    found_key.objectid);
 			BUG_ON(ret); /* -ENOMEM or corruption (JDM: Recheck) */
@@ -3238,13 +3247,14 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 	}
 
 	if (nr_unlink)
-		printk(KERN_INFO "btrfs: unlinked %d orphans\n", nr_unlink);
+		btrfs_debug(root->fs_info, "unlinked %d orphans", nr_unlink);
 	if (nr_truncate)
-		printk(KERN_INFO "btrfs: truncated %d orphans\n", nr_truncate);
+		btrfs_debug(root->fs_info, "truncated %d orphans", nr_truncate);
 
 out:
 	if (ret)
-		printk(KERN_CRIT "btrfs: could not do orphan cleanup %d\n", ret);
+		btrfs_crit(root->fs_info,
+			"could not do orphan cleanup %d", ret);
 	btrfs_free_path(path);
 	return ret;
 }
@@ -3592,9 +3602,10 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans,
 	ret = btrfs_del_inode_ref(trans, root, name, name_len, ino,
 				  dir_ino, &index);
 	if (ret) {
-		printk(KERN_INFO "btrfs failed to delete reference to %.*s, "
-		       "inode %llu parent %llu\n", name_len, name,
-		       (unsigned long long)ino, (unsigned long long)dir_ino);
+		btrfs_info(root->fs_info,
+			"failed to delete reference to %.*s, inode %llu parent %llu",
+			name_len, name,
+			(unsigned long long)ino, (unsigned long long)dir_ino);
 		btrfs_abort_transaction(trans, root, ret);
 		goto err;
 	}
@@ -3616,6 +3627,8 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans,
 					   dir, index);
 	if (ret == -ENOENT)
 		ret = 0;
+	else if (ret)
+		btrfs_abort_transaction(trans, root, ret);
 err:
 	btrfs_free_path(path);
 	if (ret)
@@ -3661,7 +3674,7 @@ static int check_path_shared(struct btrfs_root *root,
 		eb = path->nodes[level];
 		if (!btrfs_block_can_be_shared(root, eb))
 			continue;
-		ret = btrfs_lookup_extent_info(NULL, root, eb->start, eb->len,
+		ret = btrfs_lookup_extent_info(NULL, root, eb->start, level, 1,
 					       &refs, NULL);
 		if (refs > 1)
 			return 1;
@@ -4176,8 +4189,7 @@ search_again:
 				}
 				size =
 				    btrfs_file_extent_calc_inline_size(size);
-				btrfs_truncate_item(trans, root, path,
-						    size, 1);
+				btrfs_truncate_item(root, path, size, 1);
 			} else if (root->ref_cows) {
 				inode_sub_bytes(inode, item_end + 1 -
 						found_key.offset);
@@ -4451,16 +4463,14 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 			hole_em->block_start = EXTENT_MAP_HOLE;
 			hole_em->block_len = 0;
 			hole_em->orig_block_len = 0;
+			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;
 
 			while (1) {
 				write_lock(&em_tree->lock);
-				err = add_extent_mapping(em_tree, hole_em);
-				if (!err)
-					list_move(&hole_em->list,
-						  &em_tree->modified_extents);
+				err = add_extent_mapping(em_tree, hole_em, 1);
 				write_unlock(&em_tree->lock);
 				if (err != -EEXIST)
 					break;
@@ -4671,8 +4681,9 @@ void btrfs_evict_inode(struct inode *inode)
 			ret = btrfs_block_rsv_migrate(global_rsv, rsv, min_size);
 
 		if (ret) {
-			printk(KERN_WARNING "Could not get space for a "
-			       "delete, will truncate on mount %d\n", ret);
+			btrfs_warn(root->fs_info,
+				"Could not get space for a delete, will truncate on mount %d",
+				ret);
 			btrfs_orphan_del(NULL, inode);
 			btrfs_free_block_rsv(root, rsv);
 			goto no_delete;
@@ -5336,7 +5347,7 @@ int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc)
  * FIXME, needs more benchmarking...there are no reasons other than performance
  * to keep or drop this code.
  */
-int btrfs_dirty_inode(struct inode *inode)
+static int btrfs_dirty_inode(struct inode *inode)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_trans_handle *trans;
@@ -5978,7 +5989,7 @@ static int merge_extent_mapping(struct extent_map_tree *em_tree,
 		em->block_start += start_diff;
 		em->block_len -= start_diff;
 	}
-	return add_extent_mapping(em_tree, em);
+	return add_extent_mapping(em_tree, em, 0);
 }
 
 static noinline int uncompress_inline(struct btrfs_path *path,
@@ -6152,6 +6163,7 @@ again:
 		goto not_found_em;
 	}
 
+	em->ram_bytes = btrfs_file_extent_ram_bytes(leaf, item);
 	if (found_type == BTRFS_FILE_EXTENT_REG ||
 	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
 		em->start = extent_start;
@@ -6260,18 +6272,18 @@ not_found_em:
 insert:
 	btrfs_release_path(path);
 	if (em->start > start || extent_map_end(em) <= start) {
-		printk(KERN_ERR "Btrfs: bad extent! em: [%llu %llu] passed "
-		       "[%llu %llu]\n", (unsigned long long)em->start,
-		       (unsigned long long)em->len,
-		       (unsigned long long)start,
-		       (unsigned long long)len);
+		btrfs_err(root->fs_info, "bad extent! em: [%llu %llu] passed [%llu %llu]",
+			(unsigned long long)em->start,
+			(unsigned long long)em->len,
+			(unsigned long long)start,
+			(unsigned long long)len);
 		err = -EIO;
 		goto out;
 	}
 
 	err = 0;
 	write_lock(&em_tree->lock);
-	ret = add_extent_mapping(em_tree, em);
+	ret = add_extent_mapping(em_tree, em, 0);
 	/* it is possible that someone inserted the extent into the tree
 	 * while we had the lock dropped.  It is also possible that
 	 * an overlapping map exists in the tree
@@ -6483,7 +6495,7 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode,
 	}
 
 	em = create_pinned_em(inode, start, ins.offset, start, ins.objectid,
-			      ins.offset, ins.offset, 0);
+			      ins.offset, ins.offset, ins.offset, 0);
 	if (IS_ERR(em))
 		goto out;
 
@@ -6503,7 +6515,9 @@ out:
  * block must be cow'd
  */
 static noinline int can_nocow_odirect(struct btrfs_trans_handle *trans,
-				      struct inode *inode, u64 offset, u64 len)
+				      struct inode *inode, u64 offset, u64 *len,
+				      u64 *orig_start, u64 *orig_block_len,
+				      u64 *ram_bytes)
 {
 	struct btrfs_path *path;
 	int ret;
@@ -6560,8 +6574,12 @@ static noinline int can_nocow_odirect(struct btrfs_trans_handle *trans,
 	disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
 	backref_offset = btrfs_file_extent_offset(leaf, fi);
 
+	*orig_start = key.offset - backref_offset;
+	*orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, fi);
+	*ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi);
+
 	extent_end = key.offset + btrfs_file_extent_num_bytes(leaf, fi);
-	if (extent_end < offset + len) {
+	if (extent_end < offset + *len) {
 		/* extent doesn't include our full range, must cow */
 		goto out;
 	}
@@ -6585,13 +6603,14 @@ static noinline int can_nocow_odirect(struct btrfs_trans_handle *trans,
 	 */
 	disk_bytenr += backref_offset;
 	disk_bytenr += offset - key.offset;
-	num_bytes = min(offset + len, extent_end) - offset;
+	num_bytes = min(offset + *len, extent_end) - offset;
 	if (csum_exist_in_range(root, disk_bytenr, num_bytes))
 				goto out;
 	/*
 	 * all of the above have passed, it is safe to overwrite this extent
 	 * without cow
 	 */
+	*len = num_bytes;
 	ret = 1;
 out:
 	btrfs_free_path(path);
@@ -6662,7 +6681,8 @@ static int lock_extent_direct(struct inode *inode, u64 lockstart, u64 lockend,
 static struct extent_map *create_pinned_em(struct inode *inode, u64 start,
 					   u64 len, u64 orig_start,
 					   u64 block_start, u64 block_len,
-					   u64 orig_block_len, int type)
+					   u64 orig_block_len, u64 ram_bytes,
+					   int type)
 {
 	struct extent_map_tree *em_tree;
 	struct extent_map *em;
@@ -6683,6 +6703,7 @@ static struct extent_map *create_pinned_em(struct inode *inode, u64 start,
 	em->block_start = block_start;
 	em->bdev = root->fs_info->fs_devices->latest_bdev;
 	em->orig_block_len = orig_block_len;
+	em->ram_bytes = ram_bytes;
 	em->generation = -1;
 	set_bit(EXTENT_FLAG_PINNED, &em->flags);
 	if (type == BTRFS_ORDERED_PREALLOC)
@@ -6692,10 +6713,7 @@ static struct extent_map *create_pinned_em(struct inode *inode, u64 start,
 		btrfs_drop_extent_cache(inode, em->start,
 				em->start + em->len - 1, 0);
 		write_lock(&em_tree->lock);
-		ret = add_extent_mapping(em_tree, em);
-		if (!ret)
-			list_move(&em->list,
-				  &em_tree->modified_extents);
+		ret = add_extent_mapping(em_tree, em, 1);
 		write_unlock(&em_tree->lock);
 	} while (ret == -EEXIST);
 
@@ -6790,7 +6808,7 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 	     em->block_start != EXTENT_MAP_HOLE)) {
 		int type;
 		int ret;
-		u64 block_start;
+		u64 block_start, orig_start, orig_block_len, ram_bytes;
 
 		if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags))
 			type = BTRFS_ORDERED_PREALLOC;
@@ -6808,16 +6826,15 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 		if (IS_ERR(trans))
 			goto must_cow;
 
-		if (can_nocow_odirect(trans, inode, start, len) == 1) {
-			u64 orig_start = em->orig_start;
-			u64 orig_block_len = em->orig_block_len;
-
+		if (can_nocow_odirect(trans, inode, start, &len, &orig_start,
+				      &orig_block_len, &ram_bytes) == 1) {
 			if (type == BTRFS_ORDERED_PREALLOC) {
 				free_extent_map(em);
 				em = create_pinned_em(inode, start, len,
 						       orig_start,
 						       block_start, len,
-						       orig_block_len, type);
+						       orig_block_len,
+						       ram_bytes, type);
 				if (IS_ERR(em)) {
 					btrfs_end_transaction(trans, root);
 					goto unlock_err;
@@ -6937,7 +6954,7 @@ static void btrfs_endio_direct_read(struct bio *bio, int err)
 				goto failed;
 			local_irq_save(flags);
 			kaddr = kmap_atomic(page);
-			csum = btrfs_csum_data(root, kaddr + bvec->bv_offset,
+			csum = btrfs_csum_data(kaddr + bvec->bv_offset,
 					       csum, bvec->bv_len);
 			btrfs_csum_final(csum, (char *)&csum);
 			kunmap_atomic(kaddr);
@@ -6946,11 +6963,10 @@ static void btrfs_endio_direct_read(struct bio *bio, int err)
 			flush_dcache_page(bvec->bv_page);
 			if (csum != private) {
 failed:
-				printk(KERN_ERR "btrfs csum failed ino %llu off"
-				      " %llu csum %u private %u\n",
-				      (unsigned long long)btrfs_ino(inode),
-				      (unsigned long long)start,
-				      csum, (unsigned)private);
+				btrfs_err(root->fs_info, "csum failed ino %llu off %llu csum %u private %u",
+					(unsigned long long)btrfs_ino(inode),
+					(unsigned long long)start,
+					csum, (unsigned)private);
 				err = -EIO;
 			}
 		}
@@ -7426,8 +7442,8 @@ static int btrfs_writepage(struct page *page, struct writeback_control *wbc)
 	return extent_write_full_page(tree, page, btrfs_get_extent, wbc);
 }
 
-int btrfs_writepages(struct address_space *mapping,
-		     struct writeback_control *wbc)
+static int btrfs_writepages(struct address_space *mapping,
+			    struct writeback_control *wbc)
 {
 	struct extent_io_tree *tree;
 
@@ -7942,8 +7958,8 @@ void btrfs_destroy_inode(struct inode *inode)
 
 	if (test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
 		     &BTRFS_I(inode)->runtime_flags)) {
-		printk(KERN_INFO "BTRFS: inode %llu still on the orphan list\n",
-		       (unsigned long long)btrfs_ino(inode));
+		btrfs_info(root->fs_info, "inode %llu still on the orphan list",
+			(unsigned long long)btrfs_ino(inode));
 		atomic_dec(&root->orphan_inodes);
 	}
 
@@ -7952,10 +7968,9 @@ void btrfs_destroy_inode(struct inode *inode)
 		if (!ordered)
 			break;
 		else {
-			printk(KERN_ERR "btrfs found ordered "
-			       "extent %llu %llu on inode cleanup\n",
-			       (unsigned long long)ordered->file_offset,
-			       (unsigned long long)ordered->len);
+			btrfs_err(root->fs_info, "found ordered extent %llu %llu on inode cleanup",
+				(unsigned long long)ordered->file_offset,
+				(unsigned long long)ordered->len);
 			btrfs_remove_ordered_extent(inode, ordered);
 			btrfs_put_ordered_extent(ordered);
 			btrfs_put_ordered_extent(ordered);
@@ -8572,16 +8587,14 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 		em->block_start = ins.objectid;
 		em->block_len = ins.offset;
 		em->orig_block_len = ins.offset;
+		em->ram_bytes = ins.offset;
 		em->bdev = root->fs_info->fs_devices->latest_bdev;
 		set_bit(EXTENT_FLAG_PREALLOC, &em->flags);
 		em->generation = trans->transid;
 
 		while (1) {
 			write_lock(&em_tree->lock);
-			ret = add_extent_mapping(em_tree, em);
-			if (!ret)
-				list_move(&em->list,
-					  &em_tree->modified_extents);
+			ret = add_extent_mapping(em_tree, em, 1);
 			write_unlock(&em_tree->lock);
 			if (ret != -EEXIST)
 				break;

fs/btrfs/ioctl.c

@@ -723,7 +723,9 @@ static noinline int btrfs_mksubvol(struct path *parent,
 	struct dentry *dentry;
 	int error;
 
-	mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
+	error = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
+	if (error == -EINTR)
+		return error;
 
 	dentry = lookup_one_len(name, parent->dentry, namelen);
 	error = PTR_ERR(dentry);
@@ -1152,8 +1154,11 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 	u64 new_align = ~((u64)128 * 1024 - 1);
 	struct page **pages = NULL;
 
-	if (extent_thresh == 0)
-		extent_thresh = 256 * 1024;
+	if (isize == 0)
+		return 0;
+
+	if (range->start >= isize)
+		return -EINVAL;
 
 	if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) {
 		if (range->compress_type > BTRFS_COMPRESS_TYPES)
@@ -1162,8 +1167,8 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			compress_type = range->compress_type;
 	}
 
-	if (isize == 0)
-		return 0;
+	if (extent_thresh == 0)
+		extent_thresh = 256 * 1024;
 
 	/*
 	 * if we were not given a file, allocate a readahead
@@ -2086,7 +2091,9 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file,
 	if (err)
 		goto out;
 
-	mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
+	err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
+	if (err == -EINTR)
+		goto out;
 	dentry = lookup_one_len(vol_args->name, parent, namelen);
 	if (IS_ERR(dentry)) {
 		err = PTR_ERR(dentry);
@@ -2425,7 +2432,6 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
 
 	mutex_lock(&fs_devices->device_list_mutex);
 	dev = btrfs_find_device(root->fs_info, di_args->devid, s_uuid, NULL);
-	mutex_unlock(&fs_devices->device_list_mutex);
 
 	if (!dev) {
 		ret = -ENODEV;
@@ -2449,6 +2455,7 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
 	}
 
 out:
+	mutex_unlock(&fs_devices->device_list_mutex);
 	if (ret == 0 && copy_to_user(arg, di_args, sizeof(*di_args)))
 		ret = -EFAULT;
 
@@ -3003,7 +3010,7 @@ void btrfs_get_block_group_info(struct list_head *groups_list,
 	}
 }
 
-long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
+static long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
 {
 	struct btrfs_ioctl_space_args space_args;
 	struct btrfs_ioctl_space_info space;
@@ -3693,12 +3700,11 @@ static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
 		goto drop_write;
 	}
 
-	if (sa->cmd != BTRFS_QUOTA_CTL_RESCAN) {
-		trans = btrfs_start_transaction(root, 2);
-		if (IS_ERR(trans)) {
-			ret = PTR_ERR(trans);
-			goto out;
-		}
+	down_write(&root->fs_info->subvol_sem);
+	trans = btrfs_start_transaction(root->fs_info->tree_root, 2);
+	if (IS_ERR(trans)) {
+		ret = PTR_ERR(trans);
+		goto out;
 	}
 
 	switch (sa->cmd) {
@@ -3708,9 +3714,6 @@ static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
 	case BTRFS_QUOTA_CTL_DISABLE:
 		ret = btrfs_quota_disable(trans, root->fs_info);
 		break;
-	case BTRFS_QUOTA_CTL_RESCAN:
-		ret = btrfs_quota_rescan(root->fs_info);
-		break;
 	default:
 		ret = -EINVAL;
 		break;
@@ -3719,13 +3722,12 @@ static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
 	if (copy_to_user(arg, sa, sizeof(*sa)))
 		ret = -EFAULT;
 
-	if (trans) {
-		err = btrfs_commit_transaction(trans, root);
-		if (err && !ret)
-			ret = err;
-	}
+	err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
+	if (err && !ret)
+		ret = err;
 out:
 	kfree(sa);
+	up_write(&root->fs_info->subvol_sem);
 drop_write:
 	mnt_drop_write_file(file);
 	return ret;
@@ -3877,6 +3879,64 @@ drop_write:
 	return ret;
 }
 
+static long btrfs_ioctl_quota_rescan(struct file *file, void __user *arg)
+{
+	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+	struct btrfs_ioctl_quota_rescan_args *qsa;
+	int ret;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	ret = mnt_want_write_file(file);
+	if (ret)
+		return ret;
+
+	qsa = memdup_user(arg, sizeof(*qsa));
+	if (IS_ERR(qsa)) {
+		ret = PTR_ERR(qsa);
+		goto drop_write;
+	}
+
+	if (qsa->flags) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	ret = btrfs_qgroup_rescan(root->fs_info);
+
+out:
+	kfree(qsa);
+drop_write:
+	mnt_drop_write_file(file);
+	return ret;
+}
+
+static long btrfs_ioctl_quota_rescan_status(struct file *file, void __user *arg)
+{
+	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+	struct btrfs_ioctl_quota_rescan_args *qsa;
+	int ret = 0;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	qsa = kzalloc(sizeof(*qsa), GFP_NOFS);
+	if (!qsa)
+		return -ENOMEM;
+
+	if (root->fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) {
+		qsa->flags = 1;
+		qsa->progress = root->fs_info->qgroup_rescan_progress.objectid;
+	}
+
+	if (copy_to_user(arg, qsa, sizeof(*qsa)))
+		ret = -EFAULT;
+
+	kfree(qsa);
+	return ret;
+}
+
 static long btrfs_ioctl_set_received_subvol(struct file *file,
 					    void __user *arg)
 {
@@ -4115,6 +4175,10 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_ioctl_qgroup_create(file, argp);
 	case BTRFS_IOC_QGROUP_LIMIT:
 		return btrfs_ioctl_qgroup_limit(file, argp);
+	case BTRFS_IOC_QUOTA_RESCAN:
+		return btrfs_ioctl_quota_rescan(file, argp);
+	case BTRFS_IOC_QUOTA_RESCAN_STATUS:
+		return btrfs_ioctl_quota_rescan_status(file, argp);
 	case BTRFS_IOC_DEV_REPLACE:
 		return btrfs_ioctl_dev_replace(root, argp);
 	case BTRFS_IOC_GET_FSLABEL:

fs/btrfs/locking.c

@@ -24,7 +24,7 @@
 #include "extent_io.h"
 #include "locking.h"
 
-void btrfs_assert_tree_read_locked(struct extent_buffer *eb);
+static void btrfs_assert_tree_read_locked(struct extent_buffer *eb);
 
 /*
  * if we currently have a spinning reader or writer lock
@@ -264,7 +264,7 @@ void btrfs_assert_tree_locked(struct extent_buffer *eb)
 	BUG_ON(!atomic_read(&eb->write_locks));
 }
 
-void btrfs_assert_tree_read_locked(struct extent_buffer *eb)
+static void btrfs_assert_tree_read_locked(struct extent_buffer *eb)
 {
 	BUG_ON(!atomic_read(&eb->read_locks));
 }

fs/btrfs/ordered-data.c

@@ -986,7 +986,7 @@ out:
  * be reclaimed before their checksum is actually put into the btree
  */
 int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
-			   u32 *sum)
+			   u32 *sum, int len)
 {
 	struct btrfs_ordered_sum *ordered_sum;
 	struct btrfs_sector_sum *sector_sums;
@@ -995,22 +995,28 @@ int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
 	unsigned long num_sectors;
 	unsigned long i;
 	u32 sectorsize = BTRFS_I(inode)->root->sectorsize;
-	int ret = 1;
+	int index = 0;
 
 	ordered = btrfs_lookup_ordered_extent(inode, offset);
 	if (!ordered)
-		return 1;
+		return 0;
 
 	spin_lock_irq(&tree->lock);
 	list_for_each_entry_reverse(ordered_sum, &ordered->list, list) {
-		if (disk_bytenr >= ordered_sum->bytenr) {
-			num_sectors = ordered_sum->len / sectorsize;
-			sector_sums = ordered_sum->sums;
-			for (i = 0; i < num_sectors; i++) {
+		if (disk_bytenr >= ordered_sum->bytenr &&
+		    disk_bytenr < ordered_sum->bytenr + ordered_sum->len) {
+			i = (disk_bytenr - ordered_sum->bytenr) >>
+			    inode->i_sb->s_blocksize_bits;
+			sector_sums = ordered_sum->sums + i;
+			num_sectors = ordered_sum->len >>
+				      inode->i_sb->s_blocksize_bits;
+			for (; i < num_sectors; i++) {
 				if (sector_sums[i].bytenr == disk_bytenr) {
-					*sum = sector_sums[i].sum;
-					ret = 0;
-					goto out;
+					sum[index] = sector_sums[i].sum;
+					index++;
+					if (index == len)
+						goto out;
+					disk_bytenr += sectorsize;
 				}
 			}
 		}
@@ -1018,7 +1024,7 @@ int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
 out:
 	spin_unlock_irq(&tree->lock);
 	btrfs_put_ordered_extent(ordered);
-	return ret;
+	return index;
 }
 
 

fs/btrfs/ordered-data.h

@@ -196,7 +196,8 @@ struct btrfs_ordered_extent *btrfs_lookup_ordered_range(struct inode *inode,
 							u64 len);
 int btrfs_ordered_update_i_size(struct inode *inode, u64 offset,
 				struct btrfs_ordered_extent *ordered);
-int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr, u32 *sum);
+int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
+			   u32 *sum, int len);
 int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans,
 				 struct btrfs_root *root, int wait);
 void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,

fs/btrfs/print-tree.c

@@ -176,7 +176,7 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 
 	nr = btrfs_header_nritems(l);
 
-	printk(KERN_INFO "leaf %llu total ptrs %d free space %d\n",
+	btrfs_info(root->fs_info, "leaf %llu total ptrs %d free space %d",
 		(unsigned long long)btrfs_header_bytenr(l), nr,
 		btrfs_leaf_free_space(root, l));
 	for (i = 0 ; i < nr ; i++) {
@@ -319,10 +319,9 @@ void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *c)
 		btrfs_print_leaf(root, c);
 		return;
 	}
-	printk(KERN_INFO "node %llu level %d total ptrs %d free spc %u\n",
-	       (unsigned long long)btrfs_header_bytenr(c),
-	      level, nr,
-	       (u32)BTRFS_NODEPTRS_PER_BLOCK(root) - nr);
+	btrfs_info(root->fs_info, "node %llu level %d total ptrs %d free spc %u",
+		(unsigned long long)btrfs_header_bytenr(c),
+		level, nr, (u32)BTRFS_NODEPTRS_PER_BLOCK(root) - nr);
 	for (i = 0; i < nr; i++) {
 		btrfs_node_key_to_cpu(c, &key, i);
 		printk(KERN_INFO "\tkey %d (%llu %u %llu) block %llu\n",

fs/btrfs/print-tree.h

@@ -19,5 +19,5 @@
 #ifndef __PRINT_TREE_
 #define __PRINT_TREE_
 void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l);
-void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *t);
+void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *c);
 #endif

fs/btrfs/qgroup.c

@@ -31,13 +31,13 @@
 #include "locking.h"
 #include "ulist.h"
 #include "backref.h"
+#include "extent_io.h"
 
 /* TODO XXX FIXME
  *  - subvol delete -> delete when ref goes to 0? delete limits also?
  *  - reorganize keys
  *  - compressed
  *  - sync
- *  - rescan
  *  - copy also limits on subvol creation
  *  - limit
  *  - caches fuer ulists
@@ -98,7 +98,15 @@ struct btrfs_qgroup_list {
 	struct btrfs_qgroup *member;
 };
 
-/* must be called with qgroup_lock held */
+struct qgroup_rescan {
+	struct btrfs_work	work;
+	struct btrfs_fs_info	*fs_info;
+};
+
+static void qgroup_rescan_start(struct btrfs_fs_info *fs_info,
+				struct qgroup_rescan *qscan);
+
+/* must be called with qgroup_ioctl_lock held */
 static struct btrfs_qgroup *find_qgroup_rb(struct btrfs_fs_info *fs_info,
 					   u64 qgroupid)
 {
@@ -298,7 +306,20 @@ int btrfs_read_qgroup_config(struct btrfs_fs_info *fs_info)
 			}
 			fs_info->qgroup_flags = btrfs_qgroup_status_flags(l,
 									  ptr);
-			/* FIXME read scan element */
+			fs_info->qgroup_rescan_progress.objectid =
+					btrfs_qgroup_status_rescan(l, ptr);
+			if (fs_info->qgroup_flags &
+			    BTRFS_QGROUP_STATUS_FLAG_RESCAN) {
+				struct qgroup_rescan *qscan =
+					kmalloc(sizeof(*qscan), GFP_NOFS);
+				if (!qscan) {
+					ret = -ENOMEM;
+					goto out;
+				}
+				fs_info->qgroup_rescan_progress.type = 0;
+				fs_info->qgroup_rescan_progress.offset = 0;
+				qgroup_rescan_start(fs_info, qscan);
+			}
 			goto next1;
 		}
 
@@ -420,8 +441,6 @@ void btrfs_free_qgroup_config(struct btrfs_fs_info *fs_info)
 		qgroup = rb_entry(n, struct btrfs_qgroup, node);
 		rb_erase(n, &fs_info->qgroup_tree);
 
-		WARN_ON(!list_empty(&qgroup->dirty));
-
 		while (!list_empty(&qgroup->groups)) {
 			list = list_first_entry(&qgroup->groups,
 						struct btrfs_qgroup_list,
@@ -721,7 +740,8 @@ static int update_qgroup_status_item(struct btrfs_trans_handle *trans,
 	ptr = btrfs_item_ptr(l, slot, struct btrfs_qgroup_status_item);
 	btrfs_set_qgroup_status_flags(l, ptr, fs_info->qgroup_flags);
 	btrfs_set_qgroup_status_generation(l, ptr, trans->transid);
-	/* XXX scan */
+	btrfs_set_qgroup_status_rescan(l, ptr,
+				fs_info->qgroup_rescan_progress.objectid);
 
 	btrfs_mark_buffer_dirty(l);
 
@@ -783,19 +803,21 @@ int btrfs_quota_enable(struct btrfs_trans_handle *trans,
 		       struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_root *quota_root;
+	struct btrfs_root *tree_root = fs_info->tree_root;
 	struct btrfs_path *path = NULL;
 	struct btrfs_qgroup_status_item *ptr;
 	struct extent_buffer *leaf;
 	struct btrfs_key key;
+	struct btrfs_key found_key;
+	struct btrfs_qgroup *qgroup = NULL;
 	int ret = 0;
+	int slot;
 
-	spin_lock(&fs_info->qgroup_lock);
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	if (fs_info->quota_root) {
 		fs_info->pending_quota_state = 1;
-		spin_unlock(&fs_info->qgroup_lock);
 		goto out;
 	}
-	spin_unlock(&fs_info->qgroup_lock);
 
 	/*
 	 * initially create the quota tree
@@ -830,10 +852,57 @@ int btrfs_quota_enable(struct btrfs_trans_handle *trans,
 	fs_info->qgroup_flags = BTRFS_QGROUP_STATUS_FLAG_ON |
 				BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
 	btrfs_set_qgroup_status_flags(leaf, ptr, fs_info->qgroup_flags);
-	btrfs_set_qgroup_status_scan(leaf, ptr, 0);
+	btrfs_set_qgroup_status_rescan(leaf, ptr, 0);
 
 	btrfs_mark_buffer_dirty(leaf);
 
+	key.objectid = 0;
+	key.type = BTRFS_ROOT_REF_KEY;
+	key.offset = 0;
+
+	btrfs_release_path(path);
+	ret = btrfs_search_slot_for_read(tree_root, &key, path, 1, 0);
+	if (ret > 0)
+		goto out_add_root;
+	if (ret < 0)
+		goto out_free_path;
+
+
+	while (1) {
+		slot = path->slots[0];
+		leaf = path->nodes[0];
+		btrfs_item_key_to_cpu(leaf, &found_key, slot);
+
+		if (found_key.type == BTRFS_ROOT_REF_KEY) {
+			ret = add_qgroup_item(trans, quota_root,
+					      found_key.offset);
+			if (ret)
+				goto out_free_path;
+
+			qgroup = add_qgroup_rb(fs_info, found_key.offset);
+			if (IS_ERR(qgroup)) {
+				ret = PTR_ERR(qgroup);
+				goto out_free_path;
+			}
+		}
+		ret = btrfs_next_item(tree_root, path);
+		if (ret < 0)
+			goto out_free_path;
+		if (ret)
+			break;
+	}
+
+out_add_root:
+	btrfs_release_path(path);
+	ret = add_qgroup_item(trans, quota_root, BTRFS_FS_TREE_OBJECTID);
+	if (ret)
+		goto out_free_path;
+
+	qgroup = add_qgroup_rb(fs_info, BTRFS_FS_TREE_OBJECTID);
+	if (IS_ERR(qgroup)) {
+		ret = PTR_ERR(qgroup);
+		goto out_free_path;
+	}
 	spin_lock(&fs_info->qgroup_lock);
 	fs_info->quota_root = quota_root;
 	fs_info->pending_quota_state = 1;
@@ -847,6 +916,7 @@ out_free_root:
 		kfree(quota_root);
 	}
 out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
@@ -857,11 +927,10 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 	struct btrfs_root *quota_root;
 	int ret = 0;
 
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
+	if (!fs_info->quota_root)
+		goto out;
 	spin_lock(&fs_info->qgroup_lock);
-	if (!fs_info->quota_root) {
-		spin_unlock(&fs_info->qgroup_lock);
-		return 0;
-	}
 	fs_info->quota_enabled = 0;
 	fs_info->pending_quota_state = 0;
 	quota_root = fs_info->quota_root;
@@ -869,8 +938,10 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 	btrfs_free_qgroup_config(fs_info);
 	spin_unlock(&fs_info->qgroup_lock);
 
-	if (!quota_root)
-		return -EINVAL;
+	if (!quota_root) {
+		ret = -EINVAL;
+		goto out;
+	}
 
 	ret = btrfs_clean_quota_tree(trans, quota_root);
 	if (ret)
@@ -891,39 +962,62 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 	free_extent_buffer(quota_root->commit_root);
 	kfree(quota_root);
 out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
-int btrfs_quota_rescan(struct btrfs_fs_info *fs_info)
+static void qgroup_dirty(struct btrfs_fs_info *fs_info,
+			 struct btrfs_qgroup *qgroup)
 {
-	/* FIXME */
-	return 0;
+	if (list_empty(&qgroup->dirty))
+		list_add(&qgroup->dirty, &fs_info->dirty_qgroups);
 }
 
 int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans,
 			      struct btrfs_fs_info *fs_info, u64 src, u64 dst)
 {
 	struct btrfs_root *quota_root;
+	struct btrfs_qgroup *parent;
+	struct btrfs_qgroup *member;
+	struct btrfs_qgroup_list *list;
 	int ret = 0;
 
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	quota_root = fs_info->quota_root;
-	if (!quota_root)
-		return -EINVAL;
+	if (!quota_root) {
+		ret = -EINVAL;
+		goto out;
+	}
+	member = find_qgroup_rb(fs_info, src);
+	parent = find_qgroup_rb(fs_info, dst);
+	if (!member || !parent) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* check if such qgroup relation exist firstly */
+	list_for_each_entry(list, &member->groups, next_group) {
+		if (list->group == parent) {
+			ret = -EEXIST;
+			goto out;
+		}
+	}
 
 	ret = add_qgroup_relation_item(trans, quota_root, src, dst);
 	if (ret)
-		return ret;
+		goto out;
 
 	ret = add_qgroup_relation_item(trans, quota_root, dst, src);
 	if (ret) {
 		del_qgroup_relation_item(trans, quota_root, src, dst);
-		return ret;
+		goto out;
 	}
 
 	spin_lock(&fs_info->qgroup_lock);
 	ret = add_relation_rb(quota_root->fs_info, src, dst);
 	spin_unlock(&fs_info->qgroup_lock);
-
+out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
@@ -931,13 +1025,34 @@ int btrfs_del_qgroup_relation(struct btrfs_trans_handle *trans,
 			      struct btrfs_fs_info *fs_info, u64 src, u64 dst)
 {
 	struct btrfs_root *quota_root;
+	struct btrfs_qgroup *parent;
+	struct btrfs_qgroup *member;
+	struct btrfs_qgroup_list *list;
 	int ret = 0;
 	int err;
 
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	quota_root = fs_info->quota_root;
-	if (!quota_root)
-		return -EINVAL;
+	if (!quota_root) {
+		ret = -EINVAL;
+		goto out;
+	}
 
+	member = find_qgroup_rb(fs_info, src);
+	parent = find_qgroup_rb(fs_info, dst);
+	if (!member || !parent) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* check if such qgroup relation exist firstly */
+	list_for_each_entry(list, &member->groups, next_group) {
+		if (list->group == parent)
+			goto exist;
+	}
+	ret = -ENOENT;
+	goto out;
+exist:
 	ret = del_qgroup_relation_item(trans, quota_root, src, dst);
 	err = del_qgroup_relation_item(trans, quota_root, dst, src);
 	if (err && !ret)
@@ -945,9 +1060,9 @@ int btrfs_del_qgroup_relation(struct btrfs_trans_handle *trans,
 
 	spin_lock(&fs_info->qgroup_lock);
 	del_relation_rb(fs_info, src, dst);
-
 	spin_unlock(&fs_info->qgroup_lock);
-
+out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
@@ -958,11 +1073,21 @@ int btrfs_create_qgroup(struct btrfs_trans_handle *trans,
 	struct btrfs_qgroup *qgroup;
 	int ret = 0;
 
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	quota_root = fs_info->quota_root;
-	if (!quota_root)
-		return -EINVAL;
+	if (!quota_root) {
+		ret = -EINVAL;
+		goto out;
+	}
+	qgroup = find_qgroup_rb(fs_info, qgroupid);
+	if (qgroup) {
+		ret = -EEXIST;
+		goto out;
+	}
 
 	ret = add_qgroup_item(trans, quota_root, qgroupid);
+	if (ret)
+		goto out;
 
 	spin_lock(&fs_info->qgroup_lock);
 	qgroup = add_qgroup_rb(fs_info, qgroupid);
@@ -970,7 +1095,8 @@ int btrfs_create_qgroup(struct btrfs_trans_handle *trans,
 
 	if (IS_ERR(qgroup))
 		ret = PTR_ERR(qgroup);
-
+out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
@@ -981,27 +1107,32 @@ int btrfs_remove_qgroup(struct btrfs_trans_handle *trans,
 	struct btrfs_qgroup *qgroup;
 	int ret = 0;
 
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	quota_root = fs_info->quota_root;
-	if (!quota_root)
-		return -EINVAL;
+	if (!quota_root) {
+		ret = -EINVAL;
+		goto out;
+	}
 
-	/* check if there are no relations to this qgroup */
-	spin_lock(&fs_info->qgroup_lock);
 	qgroup = find_qgroup_rb(fs_info, qgroupid);
-	if (qgroup) {
-		if (!list_empty(&qgroup->groups) || !list_empty(&qgroup->members)) {
-			spin_unlock(&fs_info->qgroup_lock);
-			return -EBUSY;
+	if (!qgroup) {
+		ret = -ENOENT;
+		goto out;
+	} else {
+		/* check if there are no relations to this qgroup */
+		if (!list_empty(&qgroup->groups) ||
+		    !list_empty(&qgroup->members)) {
+			ret = -EBUSY;
+			goto out;
 		}
 	}
-	spin_unlock(&fs_info->qgroup_lock);
-
 	ret = del_qgroup_item(trans, quota_root, qgroupid);
 
 	spin_lock(&fs_info->qgroup_lock);
 	del_qgroup_rb(quota_root->fs_info, qgroupid);
 	spin_unlock(&fs_info->qgroup_lock);
-
+out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
@@ -1009,13 +1140,22 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans,
 		       struct btrfs_fs_info *fs_info, u64 qgroupid,
 		       struct btrfs_qgroup_limit *limit)
 {
-	struct btrfs_root *quota_root = fs_info->quota_root;
+	struct btrfs_root *quota_root;
 	struct btrfs_qgroup *qgroup;
 	int ret = 0;
 
-	if (!quota_root)
-		return -EINVAL;
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
+	quota_root = fs_info->quota_root;
+	if (!quota_root) {
+		ret = -EINVAL;
+		goto out;
+	}
 
+	qgroup = find_qgroup_rb(fs_info, qgroupid);
+	if (!qgroup) {
+		ret = -ENOENT;
+		goto out;
+	}
 	ret = update_qgroup_limit_item(trans, quota_root, qgroupid,
 				       limit->flags, limit->max_rfer,
 				       limit->max_excl, limit->rsv_rfer,
@@ -1027,31 +1167,17 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans,
 	}
 
 	spin_lock(&fs_info->qgroup_lock);
-
-	qgroup = find_qgroup_rb(fs_info, qgroupid);
-	if (!qgroup) {
-		ret = -ENOENT;
-		goto unlock;
-	}
 	qgroup->lim_flags = limit->flags;
 	qgroup->max_rfer = limit->max_rfer;
 	qgroup->max_excl = limit->max_excl;
 	qgroup->rsv_rfer = limit->rsv_rfer;
 	qgroup->rsv_excl = limit->rsv_excl;
-
-unlock:
 	spin_unlock(&fs_info->qgroup_lock);
-
+out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
-static void qgroup_dirty(struct btrfs_fs_info *fs_info,
-			 struct btrfs_qgroup *qgroup)
-{
-	if (list_empty(&qgroup->dirty))
-		list_add(&qgroup->dirty, &fs_info->dirty_qgroups);
-}
-
 /*
  * btrfs_qgroup_record_ref is called when the ref is added or deleted. it puts
  * the modification into a list that's later used by btrfs_end_transaction to
@@ -1075,6 +1201,144 @@ int btrfs_qgroup_record_ref(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
+static int qgroup_account_ref_step1(struct btrfs_fs_info *fs_info,
+				    struct ulist *roots, struct ulist *tmp,
+				    u64 seq)
+{
+	struct ulist_node *unode;
+	struct ulist_iterator uiter;
+	struct ulist_node *tmp_unode;
+	struct ulist_iterator tmp_uiter;
+	struct btrfs_qgroup *qg;
+	int ret;
+
+	ULIST_ITER_INIT(&uiter);
+	while ((unode = ulist_next(roots, &uiter))) {
+		qg = find_qgroup_rb(fs_info, unode->val);
+		if (!qg)
+			continue;
+
+		ulist_reinit(tmp);
+						/* XXX id not needed */
+		ret = ulist_add(tmp, qg->qgroupid,
+				(u64)(uintptr_t)qg, GFP_ATOMIC);
+		if (ret < 0)
+			return ret;
+		ULIST_ITER_INIT(&tmp_uiter);
+		while ((tmp_unode = ulist_next(tmp, &tmp_uiter))) {
+			struct btrfs_qgroup_list *glist;
+
+			qg = (struct btrfs_qgroup *)(uintptr_t)tmp_unode->aux;
+			if (qg->refcnt < seq)
+				qg->refcnt = seq + 1;
+			else
+				++qg->refcnt;
+
+			list_for_each_entry(glist, &qg->groups, next_group) {
+				ret = ulist_add(tmp, glist->group->qgroupid,
+						(u64)(uintptr_t)glist->group,
+						GFP_ATOMIC);
+				if (ret < 0)
+					return ret;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int qgroup_account_ref_step2(struct btrfs_fs_info *fs_info,
+				    struct ulist *roots, struct ulist *tmp,
+				    u64 seq, int sgn, u64 num_bytes,
+				    struct btrfs_qgroup *qgroup)
+{
+	struct ulist_node *unode;
+	struct ulist_iterator uiter;
+	struct btrfs_qgroup *qg;
+	struct btrfs_qgroup_list *glist;
+	int ret;
+
+	ulist_reinit(tmp);
+	ret = ulist_add(tmp, qgroup->qgroupid, (uintptr_t)qgroup, GFP_ATOMIC);
+	if (ret < 0)
+		return ret;
+
+	ULIST_ITER_INIT(&uiter);
+	while ((unode = ulist_next(tmp, &uiter))) {
+		qg = (struct btrfs_qgroup *)(uintptr_t)unode->aux;
+		if (qg->refcnt < seq) {
+			/* not visited by step 1 */
+			qg->rfer += sgn * num_bytes;
+			qg->rfer_cmpr += sgn * num_bytes;
+			if (roots->nnodes == 0) {
+				qg->excl += sgn * num_bytes;
+				qg->excl_cmpr += sgn * num_bytes;
+			}
+			qgroup_dirty(fs_info, qg);
+		}
+		WARN_ON(qg->tag >= seq);
+		qg->tag = seq;
+
+		list_for_each_entry(glist, &qg->groups, next_group) {
+			ret = ulist_add(tmp, glist->group->qgroupid,
+					(uintptr_t)glist->group, GFP_ATOMIC);
+			if (ret < 0)
+				return ret;
+		}
+	}
+
+	return 0;
+}
+
+static int qgroup_account_ref_step3(struct btrfs_fs_info *fs_info,
+				    struct ulist *roots, struct ulist *tmp,
+				    u64 seq, int sgn, u64 num_bytes)
+{
+	struct ulist_node *unode;
+	struct ulist_iterator uiter;
+	struct btrfs_qgroup *qg;
+	struct ulist_node *tmp_unode;
+	struct ulist_iterator tmp_uiter;
+	int ret;
+
+	ULIST_ITER_INIT(&uiter);
+	while ((unode = ulist_next(roots, &uiter))) {
+		qg = find_qgroup_rb(fs_info, unode->val);
+		if (!qg)
+			continue;
+
+		ulist_reinit(tmp);
+		ret = ulist_add(tmp, qg->qgroupid, (uintptr_t)qg, GFP_ATOMIC);
+		if (ret < 0)
+			return ret;
+
+		ULIST_ITER_INIT(&tmp_uiter);
+		while ((tmp_unode = ulist_next(tmp, &tmp_uiter))) {
+			struct btrfs_qgroup_list *glist;
+
+			qg = (struct btrfs_qgroup *)(uintptr_t)tmp_unode->aux;
+			if (qg->tag == seq)
+				continue;
+
+			if (qg->refcnt - seq == roots->nnodes) {
+				qg->excl -= sgn * num_bytes;
+				qg->excl_cmpr -= sgn * num_bytes;
+				qgroup_dirty(fs_info, qg);
+			}
+
+			list_for_each_entry(glist, &qg->groups, next_group) {
+				ret = ulist_add(tmp, glist->group->qgroupid,
+						(uintptr_t)glist->group,
+						GFP_ATOMIC);
+				if (ret < 0)
+					return ret;
+			}
+		}
+	}
+
+	return 0;
+}
+
 /*
  * btrfs_qgroup_account_ref is called for every ref that is added to or deleted
  * from the fs. First, all roots referencing the extent are searched, and
@@ -1090,10 +1354,8 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans,
 	struct btrfs_root *quota_root;
 	u64 ref_root;
 	struct btrfs_qgroup *qgroup;
-	struct ulist_node *unode;
 	struct ulist *roots = NULL;
 	struct ulist *tmp = NULL;
-	struct ulist_iterator uiter;
 	u64 seq;
 	int ret = 0;
 	int sgn;
@@ -1132,9 +1394,11 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans,
 	case BTRFS_ADD_DELAYED_REF:
 	case BTRFS_ADD_DELAYED_EXTENT:
 		sgn = 1;
+		seq = btrfs_tree_mod_seq_prev(node->seq);
 		break;
 	case BTRFS_DROP_DELAYED_REF:
 		sgn = -1;
+		seq = node->seq;
 		break;
 	case BTRFS_UPDATE_DELAYED_HEAD:
 		return 0;
@@ -1142,20 +1406,37 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans,
 		BUG();
 	}
 
+	mutex_lock(&fs_info->qgroup_rescan_lock);
+	if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) {
+		if (fs_info->qgroup_rescan_progress.objectid <= node->bytenr) {
+			mutex_unlock(&fs_info->qgroup_rescan_lock);
+			return 0;
+		}
+	}
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
+
 	/*
 	 * the delayed ref sequence number we pass depends on the direction of
-	 * the operation. for add operations, we pass (node->seq - 1) to skip
+	 * the operation. for add operations, we pass
+	 * tree_mod_log_prev_seq(node->seq) to skip
 	 * the delayed ref's current sequence number, because we need the state
 	 * of the tree before the add operation. for delete operations, we pass
 	 * (node->seq) to include the delayed ref's current sequence number,
 	 * because we need the state of the tree after the delete operation.
 	 */
-	ret = btrfs_find_all_roots(trans, fs_info, node->bytenr,
-				   sgn > 0 ? node->seq - 1 : node->seq, &roots);
+	ret = btrfs_find_all_roots(trans, fs_info, node->bytenr, seq, &roots);
 	if (ret < 0)
 		return ret;
 
+	mutex_lock(&fs_info->qgroup_rescan_lock);
 	spin_lock(&fs_info->qgroup_lock);
+	if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) {
+		if (fs_info->qgroup_rescan_progress.objectid <= node->bytenr) {
+			ret = 0;
+			goto unlock;
+		}
+	}
+
 	quota_root = fs_info->quota_root;
 	if (!quota_root)
 		goto unlock;
@@ -1175,106 +1456,29 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans,
 	seq = fs_info->qgroup_seq;
 	fs_info->qgroup_seq += roots->nnodes + 1; /* max refcnt */
 
-	ULIST_ITER_INIT(&uiter);
-	while ((unode = ulist_next(roots, &uiter))) {
-		struct ulist_node *tmp_unode;
-		struct ulist_iterator tmp_uiter;
-		struct btrfs_qgroup *qg;
-
-		qg = find_qgroup_rb(fs_info, unode->val);
-		if (!qg)
-			continue;
-
-		ulist_reinit(tmp);
-						/* XXX id not needed */
-		ulist_add(tmp, qg->qgroupid, (u64)(uintptr_t)qg, GFP_ATOMIC);
-		ULIST_ITER_INIT(&tmp_uiter);
-		while ((tmp_unode = ulist_next(tmp, &tmp_uiter))) {
-			struct btrfs_qgroup_list *glist;
-
-			qg = (struct btrfs_qgroup *)(uintptr_t)tmp_unode->aux;
-			if (qg->refcnt < seq)
-				qg->refcnt = seq + 1;
-			else
-				++qg->refcnt;
-
-			list_for_each_entry(glist, &qg->groups, next_group) {
-				ulist_add(tmp, glist->group->qgroupid,
-					  (u64)(uintptr_t)glist->group,
-					  GFP_ATOMIC);
-			}
-		}
-	}
+	ret = qgroup_account_ref_step1(fs_info, roots, tmp, seq);
+	if (ret)
+		goto unlock;
 
 	/*
 	 * step 2: walk from the new root
 	 */
-	ulist_reinit(tmp);
-	ulist_add(tmp, qgroup->qgroupid, (uintptr_t)qgroup, GFP_ATOMIC);
-	ULIST_ITER_INIT(&uiter);
-	while ((unode = ulist_next(tmp, &uiter))) {
-		struct btrfs_qgroup *qg;
-		struct btrfs_qgroup_list *glist;
-
-		qg = (struct btrfs_qgroup *)(uintptr_t)unode->aux;
-		if (qg->refcnt < seq) {
-			/* not visited by step 1 */
-			qg->rfer += sgn * node->num_bytes;
-			qg->rfer_cmpr += sgn * node->num_bytes;
-			if (roots->nnodes == 0) {
-				qg->excl += sgn * node->num_bytes;
-				qg->excl_cmpr += sgn * node->num_bytes;
-			}
-			qgroup_dirty(fs_info, qg);
-		}
-		WARN_ON(qg->tag >= seq);
-		qg->tag = seq;
-
-		list_for_each_entry(glist, &qg->groups, next_group) {
-			ulist_add(tmp, glist->group->qgroupid,
-				  (uintptr_t)glist->group, GFP_ATOMIC);
-		}
-	}
+	ret = qgroup_account_ref_step2(fs_info, roots, tmp, seq, sgn,
+				       node->num_bytes, qgroup);
+	if (ret)
+		goto unlock;
 
 	/*
 	 * step 3: walk again from old refs
 	 */
-	ULIST_ITER_INIT(&uiter);
-	while ((unode = ulist_next(roots, &uiter))) {
-		struct btrfs_qgroup *qg;
-		struct ulist_node *tmp_unode;
-		struct ulist_iterator tmp_uiter;
-
-		qg = find_qgroup_rb(fs_info, unode->val);
-		if (!qg)
-			continue;
-
-		ulist_reinit(tmp);
-		ulist_add(tmp, qg->qgroupid, (uintptr_t)qg, GFP_ATOMIC);
-		ULIST_ITER_INIT(&tmp_uiter);
-		while ((tmp_unode = ulist_next(tmp, &tmp_uiter))) {
-			struct btrfs_qgroup_list *glist;
-
-			qg = (struct btrfs_qgroup *)(uintptr_t)tmp_unode->aux;
-			if (qg->tag == seq)
-				continue;
-
-			if (qg->refcnt - seq == roots->nnodes) {
-				qg->excl -= sgn * node->num_bytes;
-				qg->excl_cmpr -= sgn * node->num_bytes;
-				qgroup_dirty(fs_info, qg);
-			}
+	ret = qgroup_account_ref_step3(fs_info, roots, tmp, seq, sgn,
+				       node->num_bytes);
+	if (ret)
+		goto unlock;
 
-			list_for_each_entry(glist, &qg->groups, next_group) {
-				ulist_add(tmp, glist->group->qgroupid,
-					  (uintptr_t)glist->group,
-					  GFP_ATOMIC);
-			}
-		}
-	}
-	ret = 0;
 unlock:
 	spin_unlock(&fs_info->qgroup_lock);
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
 	ulist_free(roots);
 	ulist_free(tmp);
 
@@ -1289,10 +1493,14 @@ int btrfs_run_qgroups(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_root *quota_root = fs_info->quota_root;
 	int ret = 0;
+	int start_rescan_worker = 0;
 
 	if (!quota_root)
 		goto out;
 
+	if (!fs_info->quota_enabled && fs_info->pending_quota_state)
+		start_rescan_worker = 1;
+
 	fs_info->quota_enabled = fs_info->pending_quota_state;
 
 	spin_lock(&fs_info->qgroup_lock);
@@ -1318,6 +1526,13 @@ int btrfs_run_qgroups(struct btrfs_trans_handle *trans,
 	if (ret)
 		fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
 
+	if (!ret && start_rescan_worker) {
+		ret = btrfs_qgroup_rescan(fs_info);
+		if (ret)
+			pr_err("btrfs: start rescan quota failed: %d\n", ret);
+		ret = 0;
+	}
+
 out:
 
 	return ret;
@@ -1338,12 +1553,30 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans,
 	struct btrfs_qgroup *srcgroup;
 	struct btrfs_qgroup *dstgroup;
 	u32 level_size = 0;
+	u64 nums;
 
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	if (!fs_info->quota_enabled)
-		return 0;
+		goto out;
 
-	if (!quota_root)
-		return -EINVAL;
+	if (!quota_root) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (inherit) {
+		i_qgroups = (u64 *)(inherit + 1);
+		nums = inherit->num_qgroups + 2 * inherit->num_ref_copies +
+		       2 * inherit->num_excl_copies;
+		for (i = 0; i < nums; ++i) {
+			srcgroup = find_qgroup_rb(fs_info, *i_qgroups);
+			if (!srcgroup) {
+				ret = -EINVAL;
+				goto out;
+			}
+			++i_qgroups;
+		}
+	}
 
 	/*
 	 * create a tracking group for the subvol itself
@@ -1470,6 +1703,7 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans,
 unlock:
 	spin_unlock(&fs_info->qgroup_lock);
 out:
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
@@ -1514,7 +1748,10 @@ int btrfs_qgroup_reserve(struct btrfs_root *root, u64 num_bytes)
 		ret = -ENOMEM;
 		goto out;
 	}
-	ulist_add(ulist, qgroup->qgroupid, (uintptr_t)qgroup, GFP_ATOMIC);
+	ret = ulist_add(ulist, qgroup->qgroupid,
+			(uintptr_t)qgroup, GFP_ATOMIC);
+	if (ret < 0)
+		goto out;
 	ULIST_ITER_INIT(&uiter);
 	while ((unode = ulist_next(ulist, &uiter))) {
 		struct btrfs_qgroup *qg;
@@ -1523,25 +1760,27 @@ int btrfs_qgroup_reserve(struct btrfs_root *root, u64 num_bytes)
 		qg = (struct btrfs_qgroup *)(uintptr_t)unode->aux;
 
 		if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_RFER) &&
-		    qg->reserved + qg->rfer + num_bytes >
+		    qg->reserved + (s64)qg->rfer + num_bytes >
 		    qg->max_rfer) {
 			ret = -EDQUOT;
 			goto out;
 		}
 
 		if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_EXCL) &&
-		    qg->reserved + qg->excl + num_bytes >
+		    qg->reserved + (s64)qg->excl + num_bytes >
 		    qg->max_excl) {
 			ret = -EDQUOT;
 			goto out;
 		}
 
 		list_for_each_entry(glist, &qg->groups, next_group) {
-			ulist_add(ulist, glist->group->qgroupid,
-				  (uintptr_t)glist->group, GFP_ATOMIC);
+			ret = ulist_add(ulist, glist->group->qgroupid,
+					(uintptr_t)glist->group, GFP_ATOMIC);
+			if (ret < 0)
+				goto out;
 		}
 	}
-
+	ret = 0;
 	/*
 	 * no limits exceeded, now record the reservation into all qgroups
 	 */
@@ -1570,6 +1809,7 @@ void btrfs_qgroup_free(struct btrfs_root *root, u64 num_bytes)
 	struct ulist_node *unode;
 	struct ulist_iterator uiter;
 	u64 ref_root = root->root_key.objectid;
+	int ret = 0;
 
 	if (!is_fstree(ref_root))
 		return;
@@ -1592,7 +1832,10 @@ void btrfs_qgroup_free(struct btrfs_root *root, u64 num_bytes)
 		btrfs_std_error(fs_info, -ENOMEM);
 		goto out;
 	}
-	ulist_add(ulist, qgroup->qgroupid, (uintptr_t)qgroup, GFP_ATOMIC);
+	ret = ulist_add(ulist, qgroup->qgroupid,
+			(uintptr_t)qgroup, GFP_ATOMIC);
+	if (ret < 0)
+		goto out;
 	ULIST_ITER_INIT(&uiter);
 	while ((unode = ulist_next(ulist, &uiter))) {
 		struct btrfs_qgroup *qg;
@@ -1603,8 +1846,10 @@ void btrfs_qgroup_free(struct btrfs_root *root, u64 num_bytes)
 		qg->reserved -= num_bytes;
 
 		list_for_each_entry(glist, &qg->groups, next_group) {
-			ulist_add(ulist, glist->group->qgroupid,
-				  (uintptr_t)glist->group, GFP_ATOMIC);
+			ret = ulist_add(ulist, glist->group->qgroupid,
+					(uintptr_t)glist->group, GFP_ATOMIC);
+			if (ret < 0)
+				goto out;
 		}
 	}
 
@@ -1617,8 +1862,265 @@ void assert_qgroups_uptodate(struct btrfs_trans_handle *trans)
 {
 	if (list_empty(&trans->qgroup_ref_list) && !trans->delayed_ref_elem.seq)
 		return;
-	printk(KERN_ERR "btrfs: qgroups not uptodate in trans handle %p: list is%s empty, seq is %llu\n",
+	pr_err("btrfs: qgroups not uptodate in trans handle %p: list is%s empty, seq is %#x.%x\n",
 		trans, list_empty(&trans->qgroup_ref_list) ? "" : " not",
-		trans->delayed_ref_elem.seq);
+		(u32)(trans->delayed_ref_elem.seq >> 32),
+		(u32)trans->delayed_ref_elem.seq);
 	BUG();
 }
+
+/*
+ * returns < 0 on error, 0 when more leafs are to be scanned.
+ * returns 1 when done, 2 when done and FLAG_INCONSISTENT was cleared.
+ */
+static int
+qgroup_rescan_leaf(struct qgroup_rescan *qscan, struct btrfs_path *path,
+		   struct btrfs_trans_handle *trans, struct ulist *tmp,
+		   struct extent_buffer *scratch_leaf)
+{
+	struct btrfs_key found;
+	struct btrfs_fs_info *fs_info = qscan->fs_info;
+	struct ulist *roots = NULL;
+	struct ulist_node *unode;
+	struct ulist_iterator uiter;
+	struct seq_list tree_mod_seq_elem = {};
+	u64 seq;
+	int slot;
+	int ret;
+
+	path->leave_spinning = 1;
+	mutex_lock(&fs_info->qgroup_rescan_lock);
+	ret = btrfs_search_slot_for_read(fs_info->extent_root,
+					 &fs_info->qgroup_rescan_progress,
+					 path, 1, 0);
+
+	pr_debug("current progress key (%llu %u %llu), search_slot ret %d\n",
+		 (unsigned long long)fs_info->qgroup_rescan_progress.objectid,
+		 fs_info->qgroup_rescan_progress.type,
+		 (unsigned long long)fs_info->qgroup_rescan_progress.offset,
+		 ret);
+
+	if (ret) {
+		/*
+		 * The rescan is about to end, we will not be scanning any
+		 * further blocks. We cannot unset the RESCAN flag here, because
+		 * we want to commit the transaction if everything went well.
+		 * To make the live accounting work in this phase, we set our
+		 * scan progress pointer such that every real extent objectid
+		 * will be smaller.
+		 */
+		fs_info->qgroup_rescan_progress.objectid = (u64)-1;
+		btrfs_release_path(path);
+		mutex_unlock(&fs_info->qgroup_rescan_lock);
+		return ret;
+	}
+
+	btrfs_item_key_to_cpu(path->nodes[0], &found,
+			      btrfs_header_nritems(path->nodes[0]) - 1);
+	fs_info->qgroup_rescan_progress.objectid = found.objectid + 1;
+
+	btrfs_get_tree_mod_seq(fs_info, &tree_mod_seq_elem);
+	memcpy(scratch_leaf, path->nodes[0], sizeof(*scratch_leaf));
+	slot = path->slots[0];
+	btrfs_release_path(path);
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
+
+	for (; slot < btrfs_header_nritems(scratch_leaf); ++slot) {
+		btrfs_item_key_to_cpu(scratch_leaf, &found, slot);
+		if (found.type != BTRFS_EXTENT_ITEM_KEY)
+			continue;
+		ret = btrfs_find_all_roots(trans, fs_info, found.objectid,
+					   tree_mod_seq_elem.seq, &roots);
+		if (ret < 0)
+			goto out;
+		spin_lock(&fs_info->qgroup_lock);
+		seq = fs_info->qgroup_seq;
+		fs_info->qgroup_seq += roots->nnodes + 1; /* max refcnt */
+
+		ret = qgroup_account_ref_step1(fs_info, roots, tmp, seq);
+		if (ret) {
+			spin_unlock(&fs_info->qgroup_lock);
+			ulist_free(roots);
+			goto out;
+		}
+
+		/*
+		 * step2 of btrfs_qgroup_account_ref works from a single root,
+		 * we're doing all at once here.
+		 */
+		ulist_reinit(tmp);
+		ULIST_ITER_INIT(&uiter);
+		while ((unode = ulist_next(roots, &uiter))) {
+			struct btrfs_qgroup *qg;
+
+			qg = find_qgroup_rb(fs_info, unode->val);
+			if (!qg)
+				continue;
+
+			ret = ulist_add(tmp, qg->qgroupid, (uintptr_t)qg,
+					GFP_ATOMIC);
+			if (ret < 0) {
+				spin_unlock(&fs_info->qgroup_lock);
+				ulist_free(roots);
+				goto out;
+			}
+		}
+
+		/* this loop is similar to step 2 of btrfs_qgroup_account_ref */
+		ULIST_ITER_INIT(&uiter);
+		while ((unode = ulist_next(tmp, &uiter))) {
+			struct btrfs_qgroup *qg;
+			struct btrfs_qgroup_list *glist;
+
+			qg = (struct btrfs_qgroup *)(uintptr_t) unode->aux;
+			qg->rfer += found.offset;
+			qg->rfer_cmpr += found.offset;
+			WARN_ON(qg->tag >= seq);
+			if (qg->refcnt - seq == roots->nnodes) {
+				qg->excl += found.offset;
+				qg->excl_cmpr += found.offset;
+			}
+			qgroup_dirty(fs_info, qg);
+
+			list_for_each_entry(glist, &qg->groups, next_group) {
+				ret = ulist_add(tmp, glist->group->qgroupid,
+						(uintptr_t)glist->group,
+						GFP_ATOMIC);
+				if (ret < 0) {
+					spin_unlock(&fs_info->qgroup_lock);
+					ulist_free(roots);
+					goto out;
+				}
+			}
+		}
+
+		spin_unlock(&fs_info->qgroup_lock);
+		ulist_free(roots);
+		ret = 0;
+	}
+
+out:
+	btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem);
+
+	return ret;
+}
+
+static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
+{
+	struct qgroup_rescan *qscan = container_of(work, struct qgroup_rescan,
+						   work);
+	struct btrfs_path *path;
+	struct btrfs_trans_handle *trans = NULL;
+	struct btrfs_fs_info *fs_info = qscan->fs_info;
+	struct ulist *tmp = NULL;
+	struct extent_buffer *scratch_leaf = NULL;
+	int err = -ENOMEM;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		goto out;
+	tmp = ulist_alloc(GFP_NOFS);
+	if (!tmp)
+		goto out;
+	scratch_leaf = kmalloc(sizeof(*scratch_leaf), GFP_NOFS);
+	if (!scratch_leaf)
+		goto out;
+
+	err = 0;
+	while (!err) {
+		trans = btrfs_start_transaction(fs_info->fs_root, 0);
+		if (IS_ERR(trans)) {
+			err = PTR_ERR(trans);
+			break;
+		}
+		if (!fs_info->quota_enabled) {
+			err = -EINTR;
+		} else {
+			err = qgroup_rescan_leaf(qscan, path, trans,
+						 tmp, scratch_leaf);
+		}
+		if (err > 0)
+			btrfs_commit_transaction(trans, fs_info->fs_root);
+		else
+			btrfs_end_transaction(trans, fs_info->fs_root);
+	}
+
+out:
+	kfree(scratch_leaf);
+	ulist_free(tmp);
+	btrfs_free_path(path);
+	kfree(qscan);
+
+	mutex_lock(&fs_info->qgroup_rescan_lock);
+	fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN;
+
+	if (err == 2 &&
+	    fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT) {
+		fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
+	} else if (err < 0) {
+		fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
+	}
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
+
+	if (err >= 0) {
+		pr_info("btrfs: qgroup scan completed%s\n",
+			err == 2 ? " (inconsistency flag cleared)" : "");
+	} else {
+		pr_err("btrfs: qgroup scan failed with %d\n", err);
+	}
+}
+
+static void
+qgroup_rescan_start(struct btrfs_fs_info *fs_info, struct qgroup_rescan *qscan)
+{
+	memset(&qscan->work, 0, sizeof(qscan->work));
+	qscan->work.func = btrfs_qgroup_rescan_worker;
+	qscan->fs_info = fs_info;
+
+	pr_info("btrfs: qgroup scan started\n");
+	btrfs_queue_worker(&fs_info->qgroup_rescan_workers, &qscan->work);
+}
+
+int
+btrfs_qgroup_rescan(struct btrfs_fs_info *fs_info)
+{
+	int ret = 0;
+	struct rb_node *n;
+	struct btrfs_qgroup *qgroup;
+	struct qgroup_rescan *qscan = kmalloc(sizeof(*qscan), GFP_NOFS);
+
+	if (!qscan)
+		return -ENOMEM;
+
+	mutex_lock(&fs_info->qgroup_rescan_lock);
+	spin_lock(&fs_info->qgroup_lock);
+	if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN)
+		ret = -EINPROGRESS;
+	else if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_ON))
+		ret = -EINVAL;
+	if (ret) {
+		spin_unlock(&fs_info->qgroup_lock);
+		mutex_unlock(&fs_info->qgroup_rescan_lock);
+		kfree(qscan);
+		return ret;
+	}
+
+	fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_RESCAN;
+	memset(&fs_info->qgroup_rescan_progress, 0,
+		sizeof(fs_info->qgroup_rescan_progress));
+
+	/* clear all current qgroup tracking information */
+	for (n = rb_first(&fs_info->qgroup_tree); n; n = rb_next(n)) {
+		qgroup = rb_entry(n, struct btrfs_qgroup, node);
+		qgroup->rfer = 0;
+		qgroup->rfer_cmpr = 0;
+		qgroup->excl = 0;
+		qgroup->excl_cmpr = 0;
+	}
+	spin_unlock(&fs_info->qgroup_lock);
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
+
+	qgroup_rescan_start(fs_info, qscan);
+
+	return 0;
+}

fs/btrfs/raid56.c

@@ -410,7 +410,7 @@ static void remove_rbio_from_cache(struct btrfs_raid_bio *rbio)
 /*
  * remove everything in the cache
  */
-void btrfs_clear_rbio_cache(struct btrfs_fs_info *info)
+static void btrfs_clear_rbio_cache(struct btrfs_fs_info *info)
 {
 	struct btrfs_stripe_hash_table *table;
 	unsigned long flags;
@@ -1010,12 +1010,12 @@ static int alloc_rbio_parity_pages(struct btrfs_raid_bio *rbio)
  * this will try to merge into existing bios if possible, and returns
  * zero if all went well.
  */
-int rbio_add_io_page(struct btrfs_raid_bio *rbio,
-		     struct bio_list *bio_list,
-		     struct page *page,
-		     int stripe_nr,
-		     unsigned long page_index,
-		     unsigned long bio_max_len)
+static int rbio_add_io_page(struct btrfs_raid_bio *rbio,
+			    struct bio_list *bio_list,
+			    struct page *page,
+			    int stripe_nr,
+			    unsigned long page_index,
+			    unsigned long bio_max_len)
 {
 	struct bio *last = bio_list->tail;
 	u64 last_end = 0;

fs/btrfs/reada.c

@@ -955,10 +955,11 @@ int btrfs_reada_wait(void *handle)
 	while (atomic_read(&rc->elems)) {
 		wait_event_timeout(rc->wait, atomic_read(&rc->elems) == 0,
 				   5 * HZ);
-		dump_devs(rc->root->fs_info, rc->elems < 10 ? 1 : 0);
+		dump_devs(rc->root->fs_info,
+			  atomic_read(&rc->elems) < 10 ? 1 : 0);
 	}
 
-	dump_devs(rc->root->fs_info, rc->elems < 10 ? 1 : 0);
+	dump_devs(rc->root->fs_info, atomic_read(&rc->elems) < 10 ? 1 : 0);
 
 	kref_put(&rc->refcnt, reada_control_release);
 

fs/btrfs/relocation.c

@@ -326,8 +326,7 @@ static struct rb_node *tree_search(struct rb_root *root, u64 bytenr)
 	return NULL;
 }
 
-void backref_tree_panic(struct rb_node *rb_node, int errno,
-					  u64 bytenr)
+static void backref_tree_panic(struct rb_node *rb_node, int errno, u64 bytenr)
 {
 
 	struct btrfs_fs_info *fs_info = NULL;
@@ -619,10 +618,13 @@ static noinline_for_stack
 int find_inline_backref(struct extent_buffer *leaf, int slot,
 			unsigned long *ptr, unsigned long *end)
 {
+	struct btrfs_key key;
 	struct btrfs_extent_item *ei;
 	struct btrfs_tree_block_info *bi;
 	u32 item_size;
 
+	btrfs_item_key_to_cpu(leaf, &key, slot);
+
 	item_size = btrfs_item_size_nr(leaf, slot);
 #ifdef BTRFS_COMPAT_EXTENT_TREE_V0
 	if (item_size < sizeof(*ei)) {
@@ -634,13 +636,18 @@ int find_inline_backref(struct extent_buffer *leaf, int slot,
 	WARN_ON(!(btrfs_extent_flags(leaf, ei) &
 		  BTRFS_EXTENT_FLAG_TREE_BLOCK));
 
-	if (item_size <= sizeof(*ei) + sizeof(*bi)) {
+	if (key.type == BTRFS_EXTENT_ITEM_KEY &&
+	    item_size <= sizeof(*ei) + sizeof(*bi)) {
 		WARN_ON(item_size < sizeof(*ei) + sizeof(*bi));
 		return 1;
 	}
 
-	bi = (struct btrfs_tree_block_info *)(ei + 1);
-	*ptr = (unsigned long)(bi + 1);
+	if (key.type == BTRFS_EXTENT_ITEM_KEY) {
+		bi = (struct btrfs_tree_block_info *)(ei + 1);
+		*ptr = (unsigned long)(bi + 1);
+	} else {
+		*ptr = (unsigned long)(ei + 1);
+	}
 	*end = (unsigned long)ei + item_size;
 	return 0;
 }
@@ -708,7 +715,7 @@ again:
 	end = 0;
 	ptr = 0;
 	key.objectid = cur->bytenr;
-	key.type = BTRFS_EXTENT_ITEM_KEY;
+	key.type = BTRFS_METADATA_ITEM_KEY;
 	key.offset = (u64)-1;
 
 	path1->search_commit_root = 1;
@@ -766,7 +773,8 @@ again:
 				break;
 			}
 
-			if (key.type == BTRFS_EXTENT_ITEM_KEY) {
+			if (key.type == BTRFS_EXTENT_ITEM_KEY ||
+			    key.type == BTRFS_METADATA_ITEM_KEY) {
 				ret = find_inline_backref(eb, path1->slots[0],
 							  &ptr, &end);
 				if (ret)
@@ -1762,7 +1770,11 @@ again:
 
 			eb = read_tree_block(dest, old_bytenr, blocksize,
 					     old_ptr_gen);
-			BUG_ON(!eb);
+			if (!eb || !extent_buffer_uptodate(eb)) {
+				ret = (!eb) ? -ENOMEM : -EIO;
+				free_extent_buffer(eb);
+				return ret;
+			}
 			btrfs_tree_lock(eb);
 			if (cow) {
 				ret = btrfs_cow_block(trans, dest, eb, parent,
@@ -1915,6 +1927,10 @@ int walk_down_reloc_tree(struct btrfs_root *root, struct btrfs_path *path,
 		bytenr = btrfs_node_blockptr(eb, path->slots[i]);
 		blocksize = btrfs_level_size(root, i - 1);
 		eb = read_tree_block(root, bytenr, blocksize, ptr_gen);
+		if (!eb || !extent_buffer_uptodate(eb)) {
+			free_extent_buffer(eb);
+			return -EIO;
+		}
 		BUG_ON(btrfs_header_level(eb) != i - 1);
 		path->nodes[i - 1] = eb;
 		path->slots[i - 1] = 0;
@@ -2592,7 +2608,8 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 		blocksize = btrfs_level_size(root, node->level);
 		generation = btrfs_node_ptr_generation(upper->eb, slot);
 		eb = read_tree_block(root, bytenr, blocksize, generation);
-		if (!eb) {
+		if (!eb || !extent_buffer_uptodate(eb)) {
+			free_extent_buffer(eb);
 			err = -EIO;
 			goto next;
 		}
@@ -2753,7 +2770,10 @@ static int get_tree_block_key(struct reloc_control *rc,
 	BUG_ON(block->key_ready);
 	eb = read_tree_block(rc->extent_root, block->bytenr,
 			     block->key.objectid, block->key.offset);
-	BUG_ON(!eb);
+	if (!eb || !extent_buffer_uptodate(eb)) {
+		free_extent_buffer(eb);
+		return -EIO;
+	}
 	WARN_ON(btrfs_header_level(eb) != block->level);
 	if (block->level == 0)
 		btrfs_item_key_to_cpu(eb, &block->key, 0);
@@ -2768,8 +2788,13 @@ static int reada_tree_block(struct reloc_control *rc,
 			    struct tree_block *block)
 {
 	BUG_ON(block->key_ready);
-	readahead_tree_block(rc->extent_root, block->bytenr,
-			     block->key.objectid, block->key.offset);
+	if (block->key.type == BTRFS_METADATA_ITEM_KEY)
+		readahead_tree_block(rc->extent_root, block->bytenr,
+				     block->key.objectid,
+				     rc->extent_root->leafsize);
+	else
+		readahead_tree_block(rc->extent_root, block->bytenr,
+				     block->key.objectid, block->key.offset);
 	return 0;
 }
 
@@ -2850,7 +2875,7 @@ int relocate_tree_blocks(struct btrfs_trans_handle *trans,
 	path = btrfs_alloc_path();
 	if (!path) {
 		err = -ENOMEM;
-		goto out_path;
+		goto out_free_blocks;
 	}
 
 	rb_node = rb_first(blocks);
@@ -2864,8 +2889,11 @@ int relocate_tree_blocks(struct btrfs_trans_handle *trans,
 	rb_node = rb_first(blocks);
 	while (rb_node) {
 		block = rb_entry(rb_node, struct tree_block, rb_node);
-		if (!block->key_ready)
-			get_tree_block_key(rc, block);
+		if (!block->key_ready) {
+			err = get_tree_block_key(rc, block);
+			if (err)
+				goto out_free_path;
+		}
 		rb_node = rb_next(rb_node);
 	}
 
@@ -2892,8 +2920,9 @@ int relocate_tree_blocks(struct btrfs_trans_handle *trans,
 out:
 	err = finish_pending_nodes(trans, rc, path, err);
 
+out_free_path:
 	btrfs_free_path(path);
-out_path:
+out_free_blocks:
 	free_block_list(blocks);
 	return err;
 }
@@ -2965,7 +2994,7 @@ int setup_extent_mapping(struct inode *inode, u64 start, u64 end,
 	lock_extent(&BTRFS_I(inode)->io_tree, start, end);
 	while (1) {
 		write_lock(&em_tree->lock);
-		ret = add_extent_mapping(em_tree, em);
+		ret = add_extent_mapping(em_tree, em, 0);
 		write_unlock(&em_tree->lock);
 		if (ret != -EEXIST) {
 			free_extent_map(em);
@@ -3176,12 +3205,17 @@ static int add_tree_block(struct reloc_control *rc,
 	eb =  path->nodes[0];
 	item_size = btrfs_item_size_nr(eb, path->slots[0]);
 
-	if (item_size >= sizeof(*ei) + sizeof(*bi)) {
+	if (extent_key->type == BTRFS_METADATA_ITEM_KEY ||
+	    item_size >= sizeof(*ei) + sizeof(*bi)) {
 		ei = btrfs_item_ptr(eb, path->slots[0],
 				struct btrfs_extent_item);
-		bi = (struct btrfs_tree_block_info *)(ei + 1);
+		if (extent_key->type == BTRFS_EXTENT_ITEM_KEY) {
+			bi = (struct btrfs_tree_block_info *)(ei + 1);
+			level = btrfs_tree_block_level(eb, bi);
+		} else {
+			level = (int)extent_key->offset;
+		}
 		generation = btrfs_extent_generation(eb, ei);
-		level = btrfs_tree_block_level(eb, bi);
 	} else {
 #ifdef BTRFS_COMPAT_EXTENT_TREE_V0
 		u64 ref_owner;
@@ -3210,7 +3244,7 @@ static int add_tree_block(struct reloc_control *rc,
 		return -ENOMEM;
 
 	block->bytenr = extent_key->objectid;
-	block->key.objectid = extent_key->offset;
+	block->key.objectid = rc->extent_root->leafsize;
 	block->key.offset = generation;
 	block->level = level;
 	block->key_ready = 0;
@@ -3252,9 +3286,15 @@ static int __add_tree_block(struct reloc_control *rc,
 	ret = btrfs_search_slot(NULL, rc->extent_root, &key, path, 0, 0);
 	if (ret < 0)
 		goto out;
-	BUG_ON(ret);
 
 	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
+	if (ret > 0) {
+		if (key.objectid == bytenr &&
+		    key.type == BTRFS_METADATA_ITEM_KEY)
+			ret = 0;
+	}
+	BUG_ON(ret);
+
 	ret = add_tree_block(rc, &key, path, blocks);
 out:
 	btrfs_free_path(path);
@@ -3275,7 +3315,8 @@ static int block_use_full_backref(struct reloc_control *rc,
 		return 1;
 
 	ret = btrfs_lookup_extent_info(NULL, rc->extent_root,
-				       eb->start, eb->len, NULL, &flags);
+				       eb->start, btrfs_header_level(eb), 1,
+				       NULL, &flags);
 	BUG_ON(ret);
 
 	if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)
@@ -3644,12 +3685,25 @@ next:
 			break;
 		}
 
-		if (key.type != BTRFS_EXTENT_ITEM_KEY ||
+		if (key.type != BTRFS_EXTENT_ITEM_KEY &&
+		    key.type != BTRFS_METADATA_ITEM_KEY) {
+			path->slots[0]++;
+			goto next;
+		}
+
+		if (key.type == BTRFS_EXTENT_ITEM_KEY &&
 		    key.objectid + key.offset <= rc->search_start) {
 			path->slots[0]++;
 			goto next;
 		}
 
+		if (key.type == BTRFS_METADATA_ITEM_KEY &&
+		    key.objectid + rc->extent_root->leafsize <=
+		    rc->search_start) {
+			path->slots[0]++;
+			goto next;
+		}
+
 		ret = find_first_extent_bit(&rc->processed_blocks,
 					    key.objectid, &start, &end,
 					    EXTENT_DIRTY, NULL);
@@ -3658,7 +3712,11 @@ next:
 			btrfs_release_path(path);
 			rc->search_start = end + 1;
 		} else {
-			rc->search_start = key.objectid + key.offset;
+			if (key.type == BTRFS_EXTENT_ITEM_KEY)
+				rc->search_start = key.objectid + key.offset;
+			else
+				rc->search_start = key.objectid +
+					rc->extent_root->leafsize;
 			memcpy(extent_key, &key, sizeof(key));
 			return 0;
 		}
@@ -4105,10 +4163,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 
 	while (1) {
 		mutex_lock(&fs_info->cleaner_mutex);
-
-		btrfs_clean_old_snapshots(fs_info->tree_root);
 		ret = relocate_block_group(rc);
-
 		mutex_unlock(&fs_info->cleaner_mutex);
 		if (ret < 0) {
 			err = ret;

fs/btrfs/root-tree.c

@@ -29,9 +29,8 @@
  * generation numbers as then we know the root was once mounted with an older
  * kernel that was not aware of the root item structure change.
  */
-void btrfs_read_root_item(struct btrfs_root *root,
-			 struct extent_buffer *eb, int slot,
-			 struct btrfs_root_item *item)
+void btrfs_read_root_item(struct extent_buffer *eb, int slot,
+			  struct btrfs_root_item *item)
 {
 	uuid_le uuid;
 	int len;
@@ -104,7 +103,7 @@ int btrfs_find_last_root(struct btrfs_root *root, u64 objectid,
 		goto out;
 	}
 	if (item)
-		btrfs_read_root_item(root, l, slot, item);
+		btrfs_read_root_item(l, slot, item);
 	if (key)
 		memcpy(key, &found_key, sizeof(found_key));
 

fs/btrfs/scrub.c

@@ -1336,7 +1336,6 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
 	int page_num;
 	u8 calculated_csum[BTRFS_CSUM_SIZE];
 	u32 crc = ~(u32)0;
-	struct btrfs_root *root = fs_info->extent_root;
 	void *mapped_buffer;
 
 	WARN_ON(!sblock->pagev[0]->page);
@@ -1365,12 +1364,11 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
 
 	for (page_num = 0;;) {
 		if (page_num == 0 && is_metadata)
-			crc = btrfs_csum_data(root,
+			crc = btrfs_csum_data(
 				((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE,
 				crc, PAGE_SIZE - BTRFS_CSUM_SIZE);
 		else
-			crc = btrfs_csum_data(root, mapped_buffer, crc,
-					      PAGE_SIZE);
+			crc = btrfs_csum_data(mapped_buffer, crc, PAGE_SIZE);
 
 		kunmap_atomic(mapped_buffer);
 		page_num++;
@@ -1657,7 +1655,6 @@ static int scrub_checksum_data(struct scrub_block *sblock)
 	void *buffer;
 	u32 crc = ~(u32)0;
 	int fail = 0;
-	struct btrfs_root *root = sctx->dev_root;
 	u64 len;
 	int index;
 
@@ -1674,7 +1671,7 @@ static int scrub_checksum_data(struct scrub_block *sblock)
 	for (;;) {
 		u64 l = min_t(u64, len, PAGE_SIZE);
 
-		crc = btrfs_csum_data(root, buffer, crc, l);
+		crc = btrfs_csum_data(buffer, crc, l);
 		kunmap_atomic(buffer);
 		len -= l;
 		if (len == 0)
@@ -1744,7 +1741,7 @@ static int scrub_checksum_tree_block(struct scrub_block *sblock)
 	for (;;) {
 		u64 l = min_t(u64, len, mapped_size);
 
-		crc = btrfs_csum_data(root, p, crc, l);
+		crc = btrfs_csum_data(p, crc, l);
 		kunmap_atomic(mapped_buffer);
 		len -= l;
 		if (len == 0)
@@ -1805,7 +1802,7 @@ static int scrub_checksum_super(struct scrub_block *sblock)
 	for (;;) {
 		u64 l = min_t(u64, len, mapped_size);
 
-		crc = btrfs_csum_data(root, p, crc, l);
+		crc = btrfs_csum_data(p, crc, l);
 		kunmap_atomic(mapped_buffer);
 		len -= l;
 		if (len == 0)
@@ -2236,12 +2233,12 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 	u64 flags;
 	int ret;
 	int slot;
-	int i;
 	u64 nstripes;
 	struct extent_buffer *l;
 	struct btrfs_key key;
 	u64 physical;
 	u64 logical;
+	u64 logic_end;
 	u64 generation;
 	int mirror_num;
 	struct reada_control *reada1;
@@ -2255,6 +2252,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 	u64 extent_len;
 	struct btrfs_device *extent_dev;
 	int extent_mirror_num;
+	int stop_loop;
 
 	if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
 			 BTRFS_BLOCK_GROUP_RAID6)) {
@@ -2315,8 +2313,8 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 	key_start.type = BTRFS_EXTENT_ITEM_KEY;
 	key_start.offset = (u64)0;
 	key_end.objectid = base + offset + nstripes * increment;
-	key_end.type = BTRFS_EXTENT_ITEM_KEY;
-	key_end.offset = (u64)0;
+	key_end.type = BTRFS_METADATA_ITEM_KEY;
+	key_end.offset = (u64)-1;
 	reada1 = btrfs_reada_add(root, &key_start, &key_end);
 
 	key_start.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
@@ -2354,8 +2352,9 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 	 */
 	logical = base + offset;
 	physical = map->stripes[num].physical;
+	logic_end = logical + increment * nstripes;
 	ret = 0;
-	for (i = 0; i < nstripes; ++i) {
+	while (logical < logic_end) {
 		/*
 		 * canceled?
 		 */
@@ -2391,19 +2390,14 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 			wake_up(&fs_info->scrub_pause_wait);
 		}
 
-		ret = btrfs_lookup_csums_range(csum_root, logical,
-					       logical + map->stripe_len - 1,
-					       &sctx->csum_list, 1);
-		if (ret)
-			goto out;
-
 		key.objectid = logical;
 		key.type = BTRFS_EXTENT_ITEM_KEY;
-		key.offset = (u64)0;
+		key.offset = (u64)-1;
 
 		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 		if (ret < 0)
 			goto out;
+
 		if (ret > 0) {
 			ret = btrfs_previous_item(root, path, 0,
 						  BTRFS_EXTENT_ITEM_KEY);
@@ -2420,7 +2414,10 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 			}
 		}
 
+		stop_loop = 0;
 		while (1) {
+			u64 bytes;
+
 			l = path->nodes[0];
 			slot = path->slots[0];
 			if (slot >= btrfs_header_nritems(l)) {
@@ -2430,19 +2427,30 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 				if (ret < 0)
 					goto out;
 
+				stop_loop = 1;
 				break;
 			}
 			btrfs_item_key_to_cpu(l, &key, slot);
 
-			if (key.objectid + key.offset <= logical)
-				goto next;
+			if (key.type == BTRFS_METADATA_ITEM_KEY)
+				bytes = root->leafsize;
+			else
+				bytes = key.offset;
 
-			if (key.objectid >= logical + map->stripe_len)
-				break;
+			if (key.objectid + bytes <= logical)
+				goto next;
 
-			if (btrfs_key_type(&key) != BTRFS_EXTENT_ITEM_KEY)
+			if (key.type != BTRFS_EXTENT_ITEM_KEY &&
+			    key.type != BTRFS_METADATA_ITEM_KEY)
 				goto next;
 
+			if (key.objectid >= logical + map->stripe_len) {
+				/* out of this device extent */
+				if (key.objectid >= logic_end)
+					stop_loop = 1;
+				break;
+			}
+
 			extent = btrfs_item_ptr(l, slot,
 						struct btrfs_extent_item);
 			flags = btrfs_extent_flags(l, extent);
@@ -2458,22 +2466,24 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 				goto next;
 			}
 
+again:
+			extent_logical = key.objectid;
+			extent_len = bytes;
+
 			/*
 			 * trim extent to this stripe
 			 */
-			if (key.objectid < logical) {
-				key.offset -= logical - key.objectid;
-				key.objectid = logical;
+			if (extent_logical < logical) {
+				extent_len -= logical - extent_logical;
+				extent_logical = logical;
 			}
-			if (key.objectid + key.offset >
+			if (extent_logical + extent_len >
 			    logical + map->stripe_len) {
-				key.offset = logical + map->stripe_len -
-					     key.objectid;
+				extent_len = logical + map->stripe_len -
+					     extent_logical;
 			}
 
-			extent_logical = key.objectid;
-			extent_physical = key.objectid - logical + physical;
-			extent_len = key.offset;
+			extent_physical = extent_logical - logical + physical;
 			extent_dev = scrub_dev;
 			extent_mirror_num = mirror_num;
 			if (is_dev_replace)
@@ -2481,13 +2491,35 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 						   extent_len, &extent_physical,
 						   &extent_dev,
 						   &extent_mirror_num);
+
+			ret = btrfs_lookup_csums_range(csum_root, logical,
+						logical + map->stripe_len - 1,
+						&sctx->csum_list, 1);
+			if (ret)
+				goto out;
+
 			ret = scrub_extent(sctx, extent_logical, extent_len,
 					   extent_physical, extent_dev, flags,
 					   generation, extent_mirror_num,
-					   key.objectid - logical + physical);
+					   extent_physical);
 			if (ret)
 				goto out;
 
+			if (extent_logical + extent_len <
+			    key.objectid + bytes) {
+				logical += increment;
+				physical += map->stripe_len;
+
+				if (logical < key.objectid + bytes) {
+					cond_resched();
+					goto again;
+				}
+
+				if (logical >= logic_end) {
+					stop_loop = 1;
+					break;
+				}
+			}
 next:
 			path->slots[0]++;
 		}
@@ -2495,8 +2527,14 @@ next:
 		logical += increment;
 		physical += map->stripe_len;
 		spin_lock(&sctx->stat_lock);
-		sctx->stat.last_physical = physical;
+		if (stop_loop)
+			sctx->stat.last_physical = map->stripes[num].physical +
+						   length;
+		else
+			sctx->stat.last_physical = physical;
 		spin_unlock(&sctx->stat_lock);
+		if (stop_loop)
+			break;
 	}
 out:
 	/* push queued extents */
@@ -3005,28 +3043,6 @@ int btrfs_scrub_cancel_dev(struct btrfs_fs_info *fs_info,
 	return 0;
 }
 
-int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid)
-{
-	struct btrfs_fs_info *fs_info = root->fs_info;
-	struct btrfs_device *dev;
-	int ret;
-
-	/*
-	 * we have to hold the device_list_mutex here so the device
-	 * does not go away in cancel_dev. FIXME: find a better solution
-	 */
-	mutex_lock(&fs_info->fs_devices->device_list_mutex);
-	dev = btrfs_find_device(fs_info, devid, NULL, NULL);
-	if (!dev) {
-		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
-		return -ENODEV;
-	}
-	ret = btrfs_scrub_cancel_dev(fs_info, dev);
-	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
-
-	return ret;
-}
-
 int btrfs_scrub_progress(struct btrfs_root *root, u64 devid,
 			 struct btrfs_scrub_progress *progress)
 {

fs/btrfs/send.c

@@ -387,7 +387,7 @@ static struct btrfs_path *alloc_path_for_send(void)
 	return path;
 }
 
-int write_buf(struct file *filp, const void *buf, u32 len, loff_t *off)
+static int write_buf(struct file *filp, const void *buf, u32 len, loff_t *off)
 {
 	int ret;
 	mm_segment_t old_fs;
@@ -3479,7 +3479,6 @@ static int __process_changed_new_xattr(int num, struct btrfs_key *di_key,
 	struct send_ctx *sctx = ctx;
 	char *found_data = NULL;
 	int found_data_len  = 0;
-	struct fs_path *p = NULL;
 
 	ret = find_xattr(sctx, sctx->parent_root, sctx->right_path,
 			sctx->cmp_key, name, name_len, &found_data,
@@ -3498,7 +3497,6 @@ static int __process_changed_new_xattr(int num, struct btrfs_key *di_key,
 	}
 
 	kfree(found_data);
-	fs_path_free(sctx, p);
 	return ret;
 }
 
@@ -4529,9 +4527,11 @@ static int send_subvol(struct send_ctx *sctx)
 {
 	int ret;
 
-	ret = send_header(sctx);
-	if (ret < 0)
-		goto out;
+	if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_STREAM_HEADER)) {
+		ret = send_header(sctx);
+		if (ret < 0)
+			goto out;
+	}
 
 	ret = send_subvol_begin(sctx);
 	if (ret < 0)
@@ -4593,7 +4593,7 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 		goto out;
 	}
 
-	if (arg->flags & ~BTRFS_SEND_FLAG_NO_FILE_DATA) {
+	if (arg->flags & ~BTRFS_SEND_FLAG_MASK) {
 		ret = -EINVAL;
 		goto out;
 	}
@@ -4612,8 +4612,8 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 	sctx->flags = arg->flags;
 
 	sctx->send_filp = fget(arg->send_fd);
-	if (IS_ERR(sctx->send_filp)) {
-		ret = PTR_ERR(sctx->send_filp);
+	if (!sctx->send_filp) {
+		ret = -EBADF;
 		goto out;
 	}
 
@@ -4704,12 +4704,14 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 	if (ret < 0)
 		goto out;
 
-	ret = begin_cmd(sctx, BTRFS_SEND_C_END);
-	if (ret < 0)
-		goto out;
-	ret = send_cmd(sctx);
-	if (ret < 0)
-		goto out;
+	if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_END_CMD)) {
+		ret = begin_cmd(sctx, BTRFS_SEND_C_END);
+		if (ret < 0)
+			goto out;
+		ret = send_cmd(sctx);
+		if (ret < 0)
+			goto out;
+	}
 
 out:
 	kfree(arg);

fs/btrfs/send.h

@@ -131,5 +131,4 @@ enum {
 
 #ifdef __KERNEL__
 long btrfs_ioctl_send(struct file *mnt_file, void __user *arg);
-int write_buf(struct file *filp, const void *buf, u32 len, loff_t *off);
 #endif

fs/btrfs/super.c

@@ -56,6 +56,7 @@
 #include "compression.h"
 #include "rcu-string.h"
 #include "dev-replace.h"
+#include "free-space-cache.h"
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/btrfs.h>
@@ -63,9 +64,9 @@
 static const struct super_operations btrfs_super_ops;
 static struct file_system_type btrfs_fs_type;
 
-static const char *btrfs_decode_error(int errno, char nbuf[16])
+static const char *btrfs_decode_error(int errno)
 {
-	char *errstr = NULL;
+	char *errstr = "unknown";
 
 	switch (errno) {
 	case -EIO:
@@ -80,18 +81,18 @@ static const char *btrfs_decode_error(int errno, char nbuf[16])
 	case -EEXIST:
 		errstr = "Object already exists";
 		break;
-	default:
-		if (nbuf) {
-			if (snprintf(nbuf, 16, "error %d", -errno) >= 0)
-				errstr = nbuf;
-		}
+	case -ENOSPC:
+		errstr = "No space left";
+		break;
+	case -ENOENT:
+		errstr = "No such entry";
 		break;
 	}
 
 	return errstr;
 }
 
-static void __save_error_info(struct btrfs_fs_info *fs_info)
+static void save_error_info(struct btrfs_fs_info *fs_info)
 {
 	/*
 	 * today we only save the error info into ram.  Long term we'll
@@ -100,11 +101,6 @@ static void __save_error_info(struct btrfs_fs_info *fs_info)
 	set_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state);
 }
 
-static void save_error_info(struct btrfs_fs_info *fs_info)
-{
-	__save_error_info(fs_info);
-}
-
 /* btrfs handle error by forcing the filesystem readonly */
 static void btrfs_handle_error(struct btrfs_fs_info *fs_info)
 {
@@ -115,7 +111,7 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info)
 
 	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) {
 		sb->s_flags |= MS_RDONLY;
-		printk(KERN_INFO "btrfs is forced readonly\n");
+		btrfs_info(fs_info, "forced readonly");
 		/*
 		 * Note that a running device replace operation is not
 		 * canceled here although there is no way to update
@@ -126,7 +122,6 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info)
 		 * mounted writeable again, the device replace
 		 * operation continues.
 		 */
-//		WARN_ON(1);
 	}
 }
 
@@ -139,7 +134,6 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		       unsigned int line, int errno, const char *fmt, ...)
 {
 	struct super_block *sb = fs_info->sb;
-	char nbuf[16];
 	const char *errstr;
 
 	/*
@@ -149,7 +143,7 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 	if (errno == -EROFS && (sb->s_flags & MS_RDONLY))
   		return;
 
-  	errstr = btrfs_decode_error(errno, nbuf);
+	errstr = btrfs_decode_error(errno);
 	if (fmt) {
 		struct va_format vaf;
 		va_list args;
@@ -158,19 +152,18 @@ 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: %s (%pV)\n",
-			sb->s_id, function, line, errstr, &vaf);
+		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: %s\n",
-			sb->s_id, function, line, errstr);
+		printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: errno=%d %s\n",
+			sb->s_id, function, line, errno, errstr);
 	}
 
 	/* Don't go through full error handling during mount */
-	if (sb->s_flags & MS_BORN) {
-		save_error_info(fs_info);
+	save_error_info(fs_info);
+	if (sb->s_flags & MS_BORN)
 		btrfs_handle_error(fs_info);
-	}
 }
 
 static const char * const logtypes[] = {
@@ -184,7 +177,7 @@ static const char * const logtypes[] = {
 	"debug",
 };
 
-void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...)
+void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...)
 {
 	struct super_block *sb = fs_info->sb;
 	char lvl[4];
@@ -208,7 +201,7 @@ void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...)
 	vaf.fmt = fmt;
 	vaf.va = &args;
 
-	printk("%sBTRFS %s (device %s): %pV", lvl, type, sb->s_id, &vaf);
+	printk("%sBTRFS %s (device %s): %pV\n", lvl, type, sb->s_id, &vaf);
 
 	va_end(args);
 }
@@ -252,18 +245,24 @@ void __btrfs_abort_transaction(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root, const char *function,
 			       unsigned int line, int errno)
 {
-	WARN_ONCE(1, KERN_DEBUG "btrfs: Transaction aborted\n");
+	/*
+	 * Report first abort since mount
+	 */
+	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",
+				errno);
+	}
 	trans->aborted = errno;
 	/* Nothing used. The other threads that have joined this
 	 * transaction may be able to continue. */
 	if (!trans->blocks_used) {
-		char nbuf[16];
 		const char *errstr;
 
-		errstr = btrfs_decode_error(errno, nbuf);
-		btrfs_printk(root->fs_info,
-			     "%s:%d: Aborting unused transaction(%s).\n",
-			     function, line, errstr);
+		errstr = btrfs_decode_error(errno);
+		btrfs_warn(root->fs_info,
+		           "%s:%d: Aborting unused transaction(%s).",
+		           function, line, errstr);
 		return;
 	}
 	ACCESS_ONCE(trans->transaction->aborted) = errno;
@@ -276,7 +275,6 @@ void __btrfs_abort_transaction(struct btrfs_trans_handle *trans,
 void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function,
 		   unsigned int line, int errno, const char *fmt, ...)
 {
-	char nbuf[16];
 	char *s_id = "<unknown>";
 	const char *errstr;
 	struct va_format vaf = { .fmt = fmt };
@@ -288,13 +286,13 @@ void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function,
 	va_start(args, fmt);
 	vaf.va = &args;
 
-	errstr = btrfs_decode_error(errno, nbuf);
+	errstr = btrfs_decode_error(errno);
 	if (fs_info && (fs_info->mount_opt & BTRFS_MOUNT_PANIC_ON_FATAL_ERROR))
-		panic(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (%s)\n",
-			s_id, function, line, &vaf, errstr);
+		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 (%s)\n",
-	       s_id, function, line, &vaf, errstr);
+	printk(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (errno=%d %s)\n",
+	       s_id, function, line, &vaf, errno, errstr);
 	va_end(args);
 	/* Caller calls BUG() */
 }
@@ -650,7 +648,7 @@ out:
  */
 static int btrfs_parse_early_options(const char *options, fmode_t flags,
 		void *holder, char **subvol_name, u64 *subvol_objectid,
-		u64 *subvol_rootid, struct btrfs_fs_devices **fs_devices)
+		struct btrfs_fs_devices **fs_devices)
 {
 	substring_t args[MAX_OPT_ARGS];
 	char *device_name, *opts, *orig, *p;
@@ -693,16 +691,8 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags,
 			}
 			break;
 		case Opt_subvolrootid:
-			intarg = 0;
-			error = match_int(&args[0], &intarg);
-			if (!error) {
-				/* we want the original fs_tree */
-				if (!intarg)
-					*subvol_rootid =
-						BTRFS_FS_TREE_OBJECTID;
-				else
-					*subvol_rootid = intarg;
-			}
+			printk(KERN_WARNING
+				"btrfs: 'subvolrootid' mount option is deprecated and has no effect\n");
 			break;
 		case Opt_device:
 			device_name = match_strdup(&args[0]);
@@ -876,7 +866,7 @@ int btrfs_sync_fs(struct super_block *sb, int wait)
 		return 0;
 	}
 
-	btrfs_wait_ordered_extents(root, 0);
+	btrfs_wait_ordered_extents(root, 1);
 
 	trans = btrfs_attach_transaction_barrier(root);
 	if (IS_ERR(trans)) {
@@ -1080,7 +1070,6 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
 	fmode_t mode = FMODE_READ;
 	char *subvol_name = NULL;
 	u64 subvol_objectid = 0;
-	u64 subvol_rootid = 0;
 	int error = 0;
 
 	if (!(flags & MS_RDONLY))
@@ -1088,7 +1077,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
 
 	error = btrfs_parse_early_options(data, mode, fs_type,
 					  &subvol_name, &subvol_objectid,
-					  &subvol_rootid, &fs_devices);
+					  &fs_devices);
 	if (error) {
 		kfree(subvol_name);
 		return ERR_PTR(error);
@@ -1202,11 +1191,14 @@ static void btrfs_resize_thread_pool(struct btrfs_fs_info *fs_info,
 			      new_pool_size);
 }
 
-static inline void btrfs_remount_prepare(struct btrfs_fs_info *fs_info,
-					 unsigned long old_opts, int flags)
+static inline void btrfs_remount_prepare(struct btrfs_fs_info *fs_info)
 {
 	set_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state);
+}
 
+static inline void btrfs_remount_begin(struct btrfs_fs_info *fs_info,
+				       unsigned long old_opts, int flags)
+{
 	if (btrfs_raw_test_opt(old_opts, AUTO_DEFRAG) &&
 	    (!btrfs_raw_test_opt(fs_info->mount_opt, AUTO_DEFRAG) ||
 	     (flags & MS_RDONLY))) {
@@ -1247,7 +1239,7 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 	unsigned int old_metadata_ratio = fs_info->metadata_ratio;
 	int ret;
 
-	btrfs_remount_prepare(fs_info, old_opts, *flags);
+	btrfs_remount_prepare(fs_info);
 
 	ret = btrfs_parse_options(root, data);
 	if (ret) {
@@ -1255,6 +1247,7 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 		goto restore;
 	}
 
+	btrfs_remount_begin(fs_info, old_opts, *flags);
 	btrfs_resize_thread_pool(fs_info,
 		fs_info->thread_pool_size, old_thread_pool_size);
 
@@ -1739,6 +1732,10 @@ static int __init init_btrfs_fs(void)
 
 	btrfs_init_lockdep();
 
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+	btrfs_test_free_space_cache();
+#endif
+
 	printk(KERN_INFO "%s loaded\n", BTRFS_BUILD_VERSION);
 	return 0;
 

fs/btrfs/transaction.c

@@ -34,7 +34,7 @@
 
 #define BTRFS_ROOT_TRANS_TAG 0
 
-void put_transaction(struct btrfs_transaction *transaction)
+static void put_transaction(struct btrfs_transaction *transaction)
 {
 	WARN_ON(atomic_read(&transaction->use_count) == 0);
 	if (atomic_dec_and_test(&transaction->use_count)) {
@@ -162,7 +162,7 @@ loop:
 	if (!RB_EMPTY_ROOT(&fs_info->tree_mod_log))
 		WARN(1, KERN_ERR "btrfs: tree_mod_log rb tree not empty when "
 			"creating a fresh transaction\n");
-	atomic_set(&fs_info->tree_mod_seq, 0);
+	atomic64_set(&fs_info->tree_mod_seq, 0);
 
 	spin_lock_init(&cur_trans->commit_lock);
 	spin_lock_init(&cur_trans->delayed_refs.lock);
@@ -707,23 +707,13 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 int btrfs_end_transaction(struct btrfs_trans_handle *trans,
 			  struct btrfs_root *root)
 {
-	int ret;
-
-	ret = __btrfs_end_transaction(trans, root, 0);
-	if (ret)
-		return ret;
-	return 0;
+	return __btrfs_end_transaction(trans, root, 0);
 }
 
 int btrfs_end_transaction_throttle(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root)
 {
-	int ret;
-
-	ret = __btrfs_end_transaction(trans, root, 1);
-	if (ret)
-		return ret;
-	return 0;
+	return __btrfs_end_transaction(trans, root, 1);
 }
 
 int btrfs_end_transaction_dmeta(struct btrfs_trans_handle *trans,
@@ -948,7 +938,7 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans,
 int btrfs_add_dead_root(struct btrfs_root *root)
 {
 	spin_lock(&root->fs_info->trans_lock);
-	list_add(&root->root_list, &root->fs_info->dead_roots);
+	list_add_tail(&root->root_list, &root->fs_info->dead_roots);
 	spin_unlock(&root->fs_info->trans_lock);
 	return 0;
 }
@@ -1179,13 +1169,17 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	memcpy(new_root_item->uuid, new_uuid.b, BTRFS_UUID_SIZE);
 	memcpy(new_root_item->parent_uuid, root->root_item.uuid,
 			BTRFS_UUID_SIZE);
+	if (!(root_flags & BTRFS_ROOT_SUBVOL_RDONLY)) {
+		memset(new_root_item->received_uuid, 0,
+		       sizeof(new_root_item->received_uuid));
+		memset(&new_root_item->stime, 0, sizeof(new_root_item->stime));
+		memset(&new_root_item->rtime, 0, sizeof(new_root_item->rtime));
+		btrfs_set_root_stransid(new_root_item, 0);
+		btrfs_set_root_rtransid(new_root_item, 0);
+	}
 	new_root_item->otime.sec = cpu_to_le64(cur_time.tv_sec);
 	new_root_item->otime.nsec = cpu_to_le32(cur_time.tv_nsec);
 	btrfs_set_root_otransid(new_root_item, trans->transid);
-	memset(&new_root_item->stime, 0, sizeof(new_root_item->stime));
-	memset(&new_root_item->rtime, 0, sizeof(new_root_item->rtime));
-	btrfs_set_root_stransid(new_root_item, 0);
-	btrfs_set_root_rtransid(new_root_item, 0);
 
 	old = btrfs_lock_root_node(root);
 	ret = btrfs_cow_block(trans, root, old, NULL, 0, &old);
@@ -1487,6 +1481,10 @@ static void cleanup_transaction(struct btrfs_trans_handle *trans,
 		current->journal_info = NULL;
 
 	kmem_cache_free(btrfs_trans_handle_cachep, trans);
+
+	spin_lock(&root->fs_info->trans_lock);
+	root->fs_info->trans_no_join = 0;
+	spin_unlock(&root->fs_info->trans_lock);
 }
 
 static int btrfs_flush_all_pending_stuffs(struct btrfs_trans_handle *trans,
@@ -1808,7 +1806,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	ret = btrfs_write_and_wait_transaction(trans, root);
 	if (ret) {
 		btrfs_error(root->fs_info, ret,
-			    "Error while writing out transaction.");
+			    "Error while writing out transaction");
 		mutex_unlock(&root->fs_info->tree_log_mutex);
 		goto cleanup_transaction;
 	}
@@ -1864,8 +1862,7 @@ cleanup_transaction:
 		btrfs_qgroup_free(root, trans->qgroup_reserved);
 		trans->qgroup_reserved = 0;
 	}
-	btrfs_printk(root->fs_info, "Skipping commit of aborted transaction.\n");
-//	WARN_ON(1);
+	btrfs_warn(root->fs_info, "Skipping commit of aborted transaction.");
 	if (current->journal_info == trans)
 		current->journal_info = NULL;
 	cleanup_transaction(trans, root, ret);
@@ -1874,31 +1871,49 @@ cleanup_transaction:
 }
 
 /*
- * interface function to delete all the snapshots we have scheduled for deletion
+ * return < 0 if error
+ * 0 if there are no more dead_roots at the time of call
+ * 1 there are more to be processed, call me again
+ *
+ * The return value indicates there are certainly more snapshots to delete, but
+ * if there comes a new one during processing, it may return 0. We don't mind,
+ * because btrfs_commit_super will poke cleaner thread and it will process it a
+ * few seconds later.
  */
-int btrfs_clean_old_snapshots(struct btrfs_root *root)
+int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 {
-	LIST_HEAD(list);
+	int ret;
 	struct btrfs_fs_info *fs_info = root->fs_info;
 
+	if (fs_info->sb->s_flags & MS_RDONLY) {
+		pr_debug("btrfs: cleaner called for RO fs!\n");
+		return 0;
+	}
+
 	spin_lock(&fs_info->trans_lock);
-	list_splice_init(&fs_info->dead_roots, &list);
+	if (list_empty(&fs_info->dead_roots)) {
+		spin_unlock(&fs_info->trans_lock);
+		return 0;
+	}
+	root = list_first_entry(&fs_info->dead_roots,
+			struct btrfs_root, root_list);
+	list_del(&root->root_list);
 	spin_unlock(&fs_info->trans_lock);
 
-	while (!list_empty(&list)) {
-		int ret;
-
-		root = list_entry(list.next, struct btrfs_root, root_list);
-		list_del(&root->root_list);
+	pr_debug("btrfs: cleaner removing %llu\n",
+			(unsigned long long)root->objectid);
 
-		btrfs_kill_all_delayed_nodes(root);
+	btrfs_kill_all_delayed_nodes(root);
 
-		if (btrfs_header_backref_rev(root->node) <
-		    BTRFS_MIXED_BACKREF_REV)
-			ret = btrfs_drop_snapshot(root, NULL, 0, 0);
-		else
-			ret =btrfs_drop_snapshot(root, NULL, 1, 0);
-		BUG_ON(ret < 0);
-	}
-	return 0;
+	if (btrfs_header_backref_rev(root->node) <
+			BTRFS_MIXED_BACKREF_REV)
+		ret = btrfs_drop_snapshot(root, NULL, 0, 0);
+	else
+		ret = btrfs_drop_snapshot(root, NULL, 1, 0);
+	/*
+	 * If we encounter a transaction abort during snapshot cleaning, we
+	 * don't want to crash here
+	 */
+	BUG_ON(ret < 0 && ret != -EAGAIN && ret != -EROFS);
+	return 1;
 }

fs/btrfs/transaction.h

@@ -123,7 +123,7 @@ int btrfs_write_and_wait_transaction(struct btrfs_trans_handle *trans,
 
 int btrfs_add_dead_root(struct btrfs_root *root);
 int btrfs_defrag_root(struct btrfs_root *root);
-int btrfs_clean_old_snapshots(struct btrfs_root *root);
+int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root);
 int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root);
 int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans,
@@ -146,5 +146,4 @@ int btrfs_wait_marked_extents(struct btrfs_root *root,
 				struct extent_io_tree *dirty_pages, int mark);
 int btrfs_transaction_blocked(struct btrfs_fs_info *info);
 int btrfs_transaction_in_commit(struct btrfs_fs_info *info);
-void put_transaction(struct btrfs_transaction *transaction);
 #endif

fs/btrfs/tree-log.c

@@ -277,17 +277,19 @@ static int process_one_buffer(struct btrfs_root *log,
 			      struct extent_buffer *eb,
 			      struct walk_control *wc, u64 gen)
 {
+	int ret = 0;
+
 	if (wc->pin)
-		btrfs_pin_extent_for_log_replay(log->fs_info->extent_root,
-						eb->start, eb->len);
+		ret = btrfs_pin_extent_for_log_replay(log->fs_info->extent_root,
+						      eb->start, eb->len);
 
-	if (btrfs_buffer_uptodate(eb, gen, 0)) {
+	if (!ret && btrfs_buffer_uptodate(eb, gen, 0)) {
 		if (wc->write)
 			btrfs_write_tree_block(eb);
 		if (wc->wait)
 			btrfs_wait_tree_block_writeback(eb);
 	}
-	return 0;
+	return ret;
 }
 
 /*
@@ -408,9 +410,9 @@ insert:
 		found_size = btrfs_item_size_nr(path->nodes[0],
 						path->slots[0]);
 		if (found_size > item_size)
-			btrfs_truncate_item(trans, root, path, item_size, 1);
+			btrfs_truncate_item(root, path, item_size, 1);
 		else if (found_size < item_size)
-			btrfs_extend_item(trans, root, path,
+			btrfs_extend_item(root, path,
 					  item_size - found_size);
 	} else if (ret) {
 		return ret;
@@ -587,7 +589,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 
 	/* drop any overlapping extents */
 	ret = btrfs_drop_extents(trans, root, inode, start, extent_end, 1);
-	BUG_ON(ret);
+	if (ret)
+		goto out;
 
 	if (found_type == BTRFS_FILE_EXTENT_REG ||
 	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
@@ -597,7 +600,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 
 		ret = btrfs_insert_empty_item(trans, root, path, key,
 					      sizeof(*item));
-		BUG_ON(ret);
+		if (ret)
+			goto out;
 		dest_offset = btrfs_item_ptr_offset(path->nodes[0],
 						    path->slots[0]);
 		copy_extent_buffer(path->nodes[0], eb, dest_offset,
@@ -623,7 +627,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 						ins.objectid, ins.offset,
 						0, root->root_key.objectid,
 						key->objectid, offset, 0);
-				BUG_ON(ret);
+				if (ret)
+					goto out;
 			} else {
 				/*
 				 * insert the extent pointer in the extent
@@ -632,7 +637,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 				ret = btrfs_alloc_logged_file_extent(trans,
 						root, root->root_key.objectid,
 						key->objectid, offset, &ins);
-				BUG_ON(ret);
+				if (ret)
+					goto out;
 			}
 			btrfs_release_path(path);
 
@@ -649,26 +655,30 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 			ret = btrfs_lookup_csums_range(root->log_root,
 						csum_start, csum_end - 1,
 						&ordered_sums, 0);
-			BUG_ON(ret);
+			if (ret)
+				goto out;
 			while (!list_empty(&ordered_sums)) {
 				struct btrfs_ordered_sum *sums;
 				sums = list_entry(ordered_sums.next,
 						struct btrfs_ordered_sum,
 						list);
-				ret = btrfs_csum_file_blocks(trans,
+				if (!ret)
+					ret = btrfs_csum_file_blocks(trans,
 						root->fs_info->csum_root,
 						sums);
-				BUG_ON(ret);
 				list_del(&sums->list);
 				kfree(sums);
 			}
+			if (ret)
+				goto out;
 		} else {
 			btrfs_release_path(path);
 		}
 	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
 		/* inline extents are easy, we just overwrite them */
 		ret = overwrite_item(trans, root, path, eb, slot, key);
-		BUG_ON(ret);
+		if (ret)
+			goto out;
 	}
 
 	inode_add_bytes(inode, nbytes);
@@ -713,20 +723,21 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
 
 	inode = read_one_inode(root, location.objectid);
 	if (!inode) {
-		kfree(name);
-		return -EIO;
+		ret = -EIO;
+		goto out;
 	}
 
 	ret = link_to_fixup_dir(trans, root, path, location.objectid);
-	BUG_ON(ret);
+	if (ret)
+		goto out;
 
 	ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
-	BUG_ON(ret);
+	if (ret)
+		goto out;
+	btrfs_run_delayed_items(trans, root);
+out:
 	kfree(name);
-
 	iput(inode);
-
-	btrfs_run_delayed_items(trans, root);
 	return ret;
 }
 
@@ -879,7 +890,8 @@ again:
 			victim_name_len = btrfs_inode_ref_name_len(leaf,
 								   victim_ref);
 			victim_name = kmalloc(victim_name_len, GFP_NOFS);
-			BUG_ON(!victim_name);
+			if (!victim_name)
+				return -ENOMEM;
 
 			read_extent_buffer(leaf, victim_name,
 					   (unsigned long)(victim_ref + 1),
@@ -895,9 +907,10 @@ again:
 				ret = btrfs_unlink_inode(trans, root, dir,
 							 inode, victim_name,
 							 victim_name_len);
-				BUG_ON(ret);
-				btrfs_run_delayed_items(trans, root);
 				kfree(victim_name);
+				if (ret)
+					return ret;
+				btrfs_run_delayed_items(trans, root);
 				*search_done = 1;
 				goto again;
 			}
@@ -905,7 +918,6 @@ again:
 
 			ptr = (unsigned long)(victim_ref + 1) + victim_name_len;
 		}
-		BUG_ON(ret);
 
 		/*
 		 * NOTE: we have searched root tree and checked the
@@ -939,6 +951,8 @@ again:
 				goto next;
 
 			victim_name = kmalloc(victim_name_len, GFP_NOFS);
+			if (!victim_name)
+				return -ENOMEM;
 			read_extent_buffer(leaf, victim_name, (unsigned long)&extref->name,
 					   victim_name_len);
 
@@ -965,14 +979,16 @@ again:
 								 victim_name_len);
 					btrfs_run_delayed_items(trans, root);
 				}
-				BUG_ON(ret);
 				iput(victim_parent);
 				kfree(victim_name);
+				if (ret)
+					return ret;
 				*search_done = 1;
 				goto again;
 			}
 			kfree(victim_name);
-			BUG_ON(ret);
+			if (ret)
+				return ret;
 next:
 			cur_offset += victim_name_len + sizeof(*extref);
 		}
@@ -985,7 +1001,8 @@ next:
 					 ref_index, name, namelen, 0);
 	if (di && !IS_ERR(di)) {
 		ret = drop_one_dir_item(trans, root, path, dir, di);
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 	}
 	btrfs_release_path(path);
 
@@ -994,7 +1011,8 @@ next:
 				   name, namelen, 0);
 	if (di && !IS_ERR(di)) {
 		ret = drop_one_dir_item(trans, root, path, dir, di);
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 	}
 	btrfs_release_path(path);
 
@@ -1139,15 +1157,19 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 						      parent_objectid,
 						      ref_index, name, namelen,
 						      &search_done);
-				if (ret == 1)
+				if (ret == 1) {
+					ret = 0;
+					goto out;
+				}
+				if (ret)
 					goto out;
-				BUG_ON(ret);
 			}
 
 			/* insert our name */
 			ret = btrfs_add_link(trans, dir, inode, name, namelen,
 					     0, ref_index);
-			BUG_ON(ret);
+			if (ret)
+				goto out;
 
 			btrfs_update_inode(trans, root, inode);
 		}
@@ -1162,13 +1184,11 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 
 	/* finally write the back reference in the inode */
 	ret = overwrite_item(trans, root, path, eb, slot, key);
-	BUG_ON(ret);
-
 out:
 	btrfs_release_path(path);
 	iput(dir);
 	iput(inode);
-	return 0;
+	return ret;
 }
 
 static int insert_orphan_item(struct btrfs_trans_handle *trans,
@@ -1326,10 +1346,10 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans,
 		if (S_ISDIR(inode->i_mode)) {
 			ret = replay_dir_deletes(trans, root, NULL, path,
 						 ino, 1);
-			BUG_ON(ret);
+			if (ret)
+				goto out;
 		}
 		ret = insert_orphan_item(trans, root, ino);
-		BUG_ON(ret);
 	}
 
 out:
@@ -1374,9 +1394,9 @@ static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans,
 			return -EIO;
 
 		ret = fixup_inode_link_count(trans, root, inode);
-		BUG_ON(ret);
-
 		iput(inode);
+		if (ret)
+			goto out;
 
 		/*
 		 * fixup on a directory may create new entries,
@@ -1426,7 +1446,7 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
 	} else if (ret == -EEXIST) {
 		ret = 0;
 	} else {
-		BUG();
+		BUG(); /* Logic Error */
 	}
 	iput(inode);
 
@@ -1495,7 +1515,7 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 	struct inode *dir;
 	u8 log_type;
 	int exists;
-	int ret;
+	int ret = 0;
 
 	dir = read_one_inode(root, key->objectid);
 	if (!dir)
@@ -1527,7 +1547,9 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 						     key->offset, name,
 						     name_len, 1);
 	} else {
-		BUG();
+		/* Corruption */
+		ret = -EINVAL;
+		goto out;
 	}
 	if (IS_ERR_OR_NULL(dst_di)) {
 		/* we need a sequence number to insert, so we only
@@ -1555,7 +1577,8 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 		goto out;
 
 	ret = drop_one_dir_item(trans, root, path, dir, dst_di);
-	BUG_ON(ret);
+	if (ret)
+		goto out;
 
 	if (key->type == BTRFS_DIR_INDEX_KEY)
 		goto insert;
@@ -1563,14 +1586,15 @@ out:
 	btrfs_release_path(path);
 	kfree(name);
 	iput(dir);
-	return 0;
+	return ret;
 
 insert:
 	btrfs_release_path(path);
 	ret = insert_one_name(trans, root, path, key->objectid, key->offset,
 			      name, name_len, log_type, &log_key);
-
-	BUG_ON(ret && ret != -ENOENT);
+	if (ret && ret != -ENOENT)
+		goto out;
+	ret = 0;
 	goto out;
 }
 
@@ -1601,7 +1625,8 @@ static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans,
 			return -EIO;
 		name_len = btrfs_dir_name_len(eb, di);
 		ret = replay_one_name(trans, root, path, eb, di, key);
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 		ptr = (unsigned long)(di + 1);
 		ptr += name_len;
 	}
@@ -1762,16 +1787,21 @@ again:
 
 			ret = link_to_fixup_dir(trans, root,
 						path, location.objectid);
-			BUG_ON(ret);
+			if (ret) {
+				kfree(name);
+				iput(inode);
+				goto out;
+			}
+
 			btrfs_inc_nlink(inode);
 			ret = btrfs_unlink_inode(trans, root, dir, inode,
 						 name, name_len);
-			BUG_ON(ret);
-
-			btrfs_run_delayed_items(trans, root);
-
+			if (!ret)
+				btrfs_run_delayed_items(trans, root);
 			kfree(name);
 			iput(inode);
+			if (ret)
+				goto out;
 
 			/* there might still be more names under this key
 			 * check and repeat if required
@@ -1875,7 +1905,8 @@ again:
 			ret = check_item_in_log(trans, root, log, path,
 						log_path, dir,
 						&found_key);
-			BUG_ON(ret);
+			if (ret)
+				goto out;
 			if (found_key.offset == (u64)-1)
 				break;
 			dir_key.offset = found_key.offset + 1;
@@ -1952,11 +1983,13 @@ static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb,
 			if (S_ISDIR(mode)) {
 				ret = replay_dir_deletes(wc->trans,
 					 root, log, path, key.objectid, 0);
-				BUG_ON(ret);
+				if (ret)
+					break;
 			}
 			ret = overwrite_item(wc->trans, root, path,
 					     eb, i, &key);
-			BUG_ON(ret);
+			if (ret)
+				break;
 
 			/* for regular files, make sure corresponding
 			 * orhpan item exist. extents past the new EOF
@@ -1965,12 +1998,14 @@ static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb,
 			if (S_ISREG(mode)) {
 				ret = insert_orphan_item(wc->trans, root,
 							 key.objectid);
-				BUG_ON(ret);
+				if (ret)
+					break;
 			}
 
 			ret = link_to_fixup_dir(wc->trans, root,
 						path, key.objectid);
-			BUG_ON(ret);
+			if (ret)
+				break;
 		}
 		if (wc->stage < LOG_WALK_REPLAY_ALL)
 			continue;
@@ -1979,28 +2014,35 @@ static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb,
 		if (key.type == BTRFS_XATTR_ITEM_KEY) {
 			ret = overwrite_item(wc->trans, root, path,
 					     eb, i, &key);
-			BUG_ON(ret);
+			if (ret)
+				break;
 		} else if (key.type == BTRFS_INODE_REF_KEY) {
 			ret = add_inode_ref(wc->trans, root, log, path,
 					    eb, i, &key);
-			BUG_ON(ret && ret != -ENOENT);
+			if (ret && ret != -ENOENT)
+				break;
+			ret = 0;
 		} else if (key.type == BTRFS_INODE_EXTREF_KEY) {
 			ret = add_inode_ref(wc->trans, root, log, path,
 					    eb, i, &key);
-			BUG_ON(ret && ret != -ENOENT);
+			if (ret && ret != -ENOENT)
+				break;
+			ret = 0;
 		} else if (key.type == BTRFS_EXTENT_DATA_KEY) {
 			ret = replay_one_extent(wc->trans, root, path,
 						eb, i, &key);
-			BUG_ON(ret);
+			if (ret)
+				break;
 		} else if (key.type == BTRFS_DIR_ITEM_KEY ||
 			   key.type == BTRFS_DIR_INDEX_KEY) {
 			ret = replay_one_dir_item(wc->trans, root, path,
 						  eb, i, &key);
-			BUG_ON(ret);
+			if (ret)
+				break;
 		}
 	}
 	btrfs_free_path(path);
-	return 0;
+	return ret;
 }
 
 static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
@@ -2045,8 +2087,10 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 
 		if (*level == 1) {
 			ret = wc->process_func(root, next, wc, ptr_gen);
-			if (ret)
+			if (ret) {
+				free_extent_buffer(next);
 				return ret;
+			}
 
 			path->slots[*level]++;
 			if (wc->free) {
@@ -2066,7 +2110,10 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 					BTRFS_TREE_LOG_OBJECTID);
 				ret = btrfs_free_and_pin_reserved_extent(root,
 							 bytenr, blocksize);
-				BUG_ON(ret); /* -ENOMEM or logic errors */
+				if (ret) {
+					free_extent_buffer(next);
+					return ret;
+				}
 			}
 			free_extent_buffer(next);
 			continue;
@@ -2139,7 +2186,8 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
 				ret = btrfs_free_and_pin_reserved_extent(root,
 						path->nodes[*level]->start,
 						path->nodes[*level]->len);
-				BUG_ON(ret);
+				if (ret)
+					return ret;
 			}
 			free_extent_buffer(path->nodes[*level]);
 			path->nodes[*level] = NULL;
@@ -2161,7 +2209,6 @@ static int walk_log_tree(struct btrfs_trans_handle *trans,
 	int wret;
 	int level;
 	struct btrfs_path *path;
-	int i;
 	int orig_level;
 
 	path = btrfs_alloc_path();
@@ -2213,17 +2260,12 @@ static int walk_log_tree(struct btrfs_trans_handle *trans,
 				BTRFS_TREE_LOG_OBJECTID);
 			ret = btrfs_free_and_pin_reserved_extent(log, next->start,
 							 next->len);
-			BUG_ON(ret); /* -ENOMEM or logic errors */
+			if (ret)
+				goto out;
 		}
 	}
 
 out:
-	for (i = 0; i <= orig_level; i++) {
-		if (path->nodes[i]) {
-			free_extent_buffer(path->nodes[i]);
-			path->nodes[i] = NULL;
-		}
-	}
 	btrfs_free_path(path);
 	return ret;
 }
@@ -2507,7 +2549,10 @@ static void free_log_tree(struct btrfs_trans_handle *trans,
 
 	if (trans) {
 		ret = walk_log_tree(trans, log, &wc);
-		BUG_ON(ret);
+
+		/* I don't think this can happen but just in case */
+		if (ret)
+			btrfs_abort_transaction(trans, log, ret);
 	}
 
 	while (1) {
@@ -2615,7 +2660,10 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans,
 	if (di) {
 		ret = btrfs_delete_one_dir_name(trans, log, path, di);
 		bytes_del += name_len;
-		BUG_ON(ret);
+		if (ret) {
+			err = ret;
+			goto fail;
+		}
 	}
 	btrfs_release_path(path);
 	di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
@@ -2627,7 +2675,10 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans,
 	if (di) {
 		ret = btrfs_delete_one_dir_name(trans, log, path, di);
 		bytes_del += name_len;
-		BUG_ON(ret);
+		if (ret) {
+			err = ret;
+			goto fail;
+		}
 	}
 
 	/* update the directory size in the log to reflect the names
@@ -2966,7 +3017,7 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans,
 
 	while (1) {
 		ret = btrfs_search_slot(trans, log, &key, path, -1, 1);
-		BUG_ON(ret == 0);
+		BUG_ON(ret == 0); /* Logic error */
 		if (ret < 0)
 			break;
 
@@ -3169,7 +3220,11 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 						log->fs_info->csum_root,
 						ds + cs, ds + cs + cl - 1,
 						&ordered_sums, 0);
-				BUG_ON(ret);
+				if (ret) {
+					btrfs_release_path(dst_path);
+					kfree(ins_data);
+					return ret;
+				}
 			}
 		}
 	}
@@ -3209,115 +3264,6 @@ static int extent_cmp(void *priv, struct list_head *a, struct list_head *b)
 	return 0;
 }
 
-static int drop_adjacent_extents(struct btrfs_trans_handle *trans,
-				 struct btrfs_root *root, struct inode *inode,
-				 struct extent_map *em,
-				 struct btrfs_path *path)
-{
-	struct btrfs_file_extent_item *fi;
-	struct extent_buffer *leaf;
-	struct btrfs_key key, new_key;
-	struct btrfs_map_token token;
-	u64 extent_end;
-	u64 extent_offset = 0;
-	int extent_type;
-	int del_slot = 0;
-	int del_nr = 0;
-	int ret = 0;
-
-	while (1) {
-		btrfs_init_map_token(&token);
-		leaf = path->nodes[0];
-		path->slots[0]++;
-		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
-			if (del_nr) {
-				ret = btrfs_del_items(trans, root, path,
-						      del_slot, del_nr);
-				if (ret)
-					return ret;
-				del_nr = 0;
-			}
-
-			ret = btrfs_next_leaf_write(trans, root, path, 1);
-			if (ret < 0)
-				return ret;
-			if (ret > 0)
-				return 0;
-			leaf = path->nodes[0];
-		}
-
-		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
-		if (key.objectid != btrfs_ino(inode) ||
-		    key.type != BTRFS_EXTENT_DATA_KEY ||
-		    key.offset >= em->start + em->len)
-			break;
-
-		fi = btrfs_item_ptr(leaf, path->slots[0],
-				    struct btrfs_file_extent_item);
-		extent_type = btrfs_token_file_extent_type(leaf, fi, &token);
-		if (extent_type == BTRFS_FILE_EXTENT_REG ||
-		    extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
-			extent_offset = btrfs_token_file_extent_offset(leaf,
-								fi, &token);
-			extent_end = key.offset +
-				btrfs_token_file_extent_num_bytes(leaf, fi,
-								  &token);
-		} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
-			extent_end = key.offset +
-				btrfs_file_extent_inline_len(leaf, fi);
-		} else {
-			BUG();
-		}
-
-		if (extent_end <= em->len + em->start) {
-			if (!del_nr) {
-				del_slot = path->slots[0];
-			}
-			del_nr++;
-			continue;
-		}
-
-		/*
-		 * Ok so we'll ignore previous items if we log a new extent,
-		 * which can lead to overlapping extents, so if we have an
-		 * existing extent we want to adjust we _have_ to check the next
-		 * guy to make sure we even need this extent anymore, this keeps
-		 * us from panicing in set_item_key_safe.
-		 */
-		if (path->slots[0] < btrfs_header_nritems(leaf) - 1) {
-			struct btrfs_key tmp_key;
-
-			btrfs_item_key_to_cpu(leaf, &tmp_key,
-					      path->slots[0] + 1);
-			if (tmp_key.objectid == btrfs_ino(inode) &&
-			    tmp_key.type == BTRFS_EXTENT_DATA_KEY &&
-			    tmp_key.offset <= em->start + em->len) {
-				if (!del_nr)
-					del_slot = path->slots[0];
-				del_nr++;
-				continue;
-			}
-		}
-
-		BUG_ON(extent_type == BTRFS_FILE_EXTENT_INLINE);
-		memcpy(&new_key, &key, sizeof(new_key));
-		new_key.offset = em->start + em->len;
-		btrfs_set_item_key_safe(trans, root, path, &new_key);
-		extent_offset += em->start + em->len - key.offset;
-		btrfs_set_token_file_extent_offset(leaf, fi, extent_offset,
-						   &token);
-		btrfs_set_token_file_extent_num_bytes(leaf, fi, extent_end -
-						      (em->start + em->len),
-						      &token);
-		btrfs_mark_buffer_dirty(leaf);
-	}
-
-	if (del_nr)
-		ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
-
-	return ret;
-}
-
 static int log_one_extent(struct btrfs_trans_handle *trans,
 			  struct inode *inode, struct btrfs_root *root,
 			  struct extent_map *em, struct btrfs_path *path)
@@ -3339,39 +3285,24 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
 	int index = log->log_transid % 2;
 	bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
 
-insert:
+	ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
+				   em->start + em->len, NULL, 0);
+	if (ret)
+		return ret;
+
 	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;
-	path->really_keep_locks = 1;
 
 	ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*fi));
-	if (ret && ret != -EEXIST) {
-		path->really_keep_locks = 0;
+	if (ret)
 		return ret;
-	}
 	leaf = path->nodes[0];
 	fi = btrfs_item_ptr(leaf, path->slots[0],
 			    struct btrfs_file_extent_item);
 
-	/*
-	 * If we are overwriting an inline extent with a real one then we need
-	 * to just delete the inline extent as it may not be large enough to
-	 * have the entire file_extent_item.
-	 */
-	if (ret && btrfs_token_file_extent_type(leaf, fi, &token) ==
-	    BTRFS_FILE_EXTENT_INLINE) {
-		ret = btrfs_del_item(trans, log, path);
-		btrfs_release_path(path);
-		if (ret) {
-			path->really_keep_locks = 0;
-			return ret;
-		}
-		goto insert;
-	}
-
 	btrfs_set_token_file_extent_generation(leaf, fi, em->generation,
 					       &token);
 	if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
@@ -3410,22 +3341,14 @@ insert:
 					   em->start - em->orig_start,
 					   &token);
 	btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token);
-	btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->len, &token);
+	btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->ram_bytes, &token);
 	btrfs_set_token_file_extent_compression(leaf, fi, em->compress_type,
 						&token);
 	btrfs_set_token_file_extent_encryption(leaf, fi, 0, &token);
 	btrfs_set_token_file_extent_other_encoding(leaf, fi, 0, &token);
 	btrfs_mark_buffer_dirty(leaf);
 
-	/*
-	 * Have to check the extent to the right of us to make sure it doesn't
-	 * fall in our current range.  We're ok if the previous extent is in our
-	 * range since the recovery stuff will run us in key order and thus just
-	 * drop the part we overwrote.
-	 */
-	ret = drop_adjacent_extents(trans, log, inode, em, path);
 	btrfs_release_path(path);
-	path->really_keep_locks = 0;
 	if (ret) {
 		return ret;
 	}
@@ -3650,8 +3573,6 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 	bool fast_search = false;
 	u64 ino = btrfs_ino(inode);
 
-	log = root->log_root;
-
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -3918,9 +3839,9 @@ out:
  * only logging is done of any parent directories that are older than
  * the last committed transaction
  */
-int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
-		    struct btrfs_root *root, struct inode *inode,
-		    struct dentry *parent, int exists_only)
+static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
+			    	  struct btrfs_root *root, struct inode *inode,
+			    	  struct dentry *parent, int exists_only)
 {
 	int inode_only = exists_only ? LOG_INODE_EXISTS : LOG_INODE_ALL;
 	struct super_block *sb;
@@ -4111,6 +4032,9 @@ again:
 		wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key);
 		if (IS_ERR(wc.replay_dest)) {
 			ret = PTR_ERR(wc.replay_dest);
+			free_extent_buffer(log->node);
+			free_extent_buffer(log->commit_root);
+			kfree(log);
 			btrfs_error(fs_info, ret, "Couldn't read target root "
 				    "for tree log recovery.");
 			goto error;
@@ -4119,12 +4043,10 @@ again:
 		wc.replay_dest->log_root = log;
 		btrfs_record_root_in_trans(trans, wc.replay_dest);
 		ret = walk_log_tree(trans, log, &wc);
-		BUG_ON(ret);
 
-		if (wc.stage == LOG_WALK_REPLAY_ALL) {
+		if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) {
 			ret = fixup_inode_link_counts(trans, wc.replay_dest,
 						      path);
-			BUG_ON(ret);
 		}
 
 		key.offset = found_key.offset - 1;
@@ -4133,6 +4055,9 @@ again:
 		free_extent_buffer(log->commit_root);
 		kfree(log);
 
+		if (ret)
+			goto error;
+
 		if (found_key.offset == 0)
 			break;
 	}
@@ -4153,17 +4078,20 @@ again:
 
 	btrfs_free_path(path);
 
+	/* step 4: commit the transaction, which also unpins the blocks */
+	ret = btrfs_commit_transaction(trans, fs_info->tree_root);
+	if (ret)
+		return ret;
+
 	free_extent_buffer(log_root_tree->node);
 	log_root_tree->log_root = NULL;
 	fs_info->log_root_recovering = 0;
-
-	/* step 4: commit the transaction, which also unpins the blocks */
-	btrfs_commit_transaction(trans, fs_info->tree_root);
-
 	kfree(log_root_tree);
-	return 0;
 
+	return 0;
 error:
+	if (wc.trans)
+		btrfs_end_transaction(wc.trans, fs_info->tree_root);
 	btrfs_free_path(path);
 	return ret;
 }

fs/btrfs/tree-log.h

@@ -40,9 +40,6 @@ int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans,
 			       struct inode *inode, u64 dirid);
 void btrfs_end_log_trans(struct btrfs_root *root);
 int btrfs_pin_log_trans(struct btrfs_root *root);
-int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
-		    struct btrfs_root *root, struct inode *inode,
-		    struct dentry *parent, int exists_only);
 void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans,
 			     struct inode *dir, struct inode *inode,
 			     int for_rename);

fs/btrfs/ulist.c

@@ -53,6 +53,7 @@ void ulist_init(struct ulist *ulist)
 	ulist->nnodes = 0;
 	ulist->nodes = ulist->int_nodes;
 	ulist->nodes_alloced = ULIST_SIZE;
+	ulist->root = RB_ROOT;
 }
 EXPORT_SYMBOL(ulist_init);
 
@@ -72,6 +73,7 @@ void ulist_fini(struct ulist *ulist)
 	if (ulist->nodes_alloced > ULIST_SIZE)
 		kfree(ulist->nodes);
 	ulist->nodes_alloced = 0;	/* in case ulist_fini is called twice */
+	ulist->root = RB_ROOT;
 }
 EXPORT_SYMBOL(ulist_fini);
 
@@ -123,6 +125,45 @@ void ulist_free(struct ulist *ulist)
 }
 EXPORT_SYMBOL(ulist_free);
 
+static struct ulist_node *ulist_rbtree_search(struct ulist *ulist, u64 val)
+{
+	struct rb_node *n = ulist->root.rb_node;
+	struct ulist_node *u = NULL;
+
+	while (n) {
+		u = rb_entry(n, struct ulist_node, rb_node);
+		if (u->val < val)
+			n = n->rb_right;
+		else if (u->val > val)
+			n = n->rb_left;
+		else
+			return u;
+	}
+	return NULL;
+}
+
+static int ulist_rbtree_insert(struct ulist *ulist, struct ulist_node *ins)
+{
+	struct rb_node **p = &ulist->root.rb_node;
+	struct rb_node *parent = NULL;
+	struct ulist_node *cur = NULL;
+
+	while (*p) {
+		parent = *p;
+		cur = rb_entry(parent, struct ulist_node, rb_node);
+
+		if (cur->val < ins->val)
+			p = &(*p)->rb_right;
+		else if (cur->val > ins->val)
+			p = &(*p)->rb_left;
+		else
+			return -EEXIST;
+	}
+	rb_link_node(&ins->rb_node, parent, p);
+	rb_insert_color(&ins->rb_node, &ulist->root);
+	return 0;
+}
+
 /**
  * ulist_add - add an element to the ulist
  * @ulist:	ulist to add the element to
@@ -151,14 +192,13 @@ 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 i;
-
-	for (i = 0; i < ulist->nnodes; ++i) {
-		if (ulist->nodes[i].val == val) {
-			if (old_aux)
-				*old_aux = ulist->nodes[i].aux;
-			return 0;
-		}
+	int ret = 0;
+	struct ulist_node *node = NULL;
+	node = ulist_rbtree_search(ulist, val);
+	if (node) {
+		if (old_aux)
+			*old_aux = node->aux;
+		return 0;
 	}
 
 	if (ulist->nnodes >= ulist->nodes_alloced) {
@@ -187,6 +227,8 @@ int ulist_add_merge(struct ulist *ulist, u64 val, u64 aux,
 	}
 	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;
 
 	return 1;

fs/btrfs/ulist.h

@@ -8,6 +8,9 @@
 #ifndef __ULIST__
 #define __ULIST__
 
+#include <linux/list.h>
+#include <linux/rbtree.h>
+
 /*
  * ulist is a generic data structure to hold a collection of unique u64
  * values. The only operations it supports is adding to the list and
@@ -34,6 +37,7 @@ struct ulist_iterator {
 struct ulist_node {
 	u64 val;		/* value to store */
 	u64 aux;		/* auxiliary value saved along with the val */
+	struct rb_node rb_node;	/* used to speed up search */
 };
 
 struct ulist {
@@ -54,6 +58,8 @@ struct ulist {
 	 */
 	struct ulist_node *nodes;
 
+	struct rb_root root;
+
 	/*
 	 * inline storage space for the first ULIST_SIZE entries
 	 */

fs/btrfs/volumes.c

@@ -46,6 +46,7 @@ static int init_first_rw_device(struct btrfs_trans_handle *trans,
 				struct btrfs_device *device);
 static int btrfs_relocate_sys_chunks(struct btrfs_root *root);
 static void __btrfs_reset_dev_stats(struct btrfs_device *dev);
+static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev);
 static void btrfs_dev_stat_print_on_load(struct btrfs_device *device);
 
 static DEFINE_MUTEX(uuid_mutex);
@@ -717,9 +718,9 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 		if (!device->name)
 			continue;
 
-		ret = btrfs_get_bdev_and_sb(device->name->str, flags, holder, 1,
-					    &bdev, &bh);
-		if (ret)
+		/* Just open everything we can; ignore failures here */
+		if (btrfs_get_bdev_and_sb(device->name->str, flags, holder, 1,
+					    &bdev, &bh))
 			continue;
 
 		disk_super = (struct btrfs_super_block *)bh->b_data;
@@ -1199,10 +1200,10 @@ out:
 	return ret;
 }
 
-int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
-			   struct btrfs_device *device,
-			   u64 chunk_tree, u64 chunk_objectid,
-			   u64 chunk_offset, u64 start, u64 num_bytes)
+static int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
+				  struct btrfs_device *device,
+				  u64 chunk_tree, u64 chunk_objectid,
+				  u64 chunk_offset, u64 start, u64 num_bytes)
 {
 	int ret;
 	struct btrfs_path *path;
@@ -1329,9 +1330,9 @@ error:
  * the device information is stored in the chunk root
  * the btrfs_device struct should be fully filled in
  */
-int btrfs_add_device(struct btrfs_trans_handle *trans,
-		     struct btrfs_root *root,
-		     struct btrfs_device *device)
+static int btrfs_add_device(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root,
+			    struct btrfs_device *device)
 {
 	int ret;
 	struct btrfs_path *path;
@@ -1710,8 +1711,8 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 }
 
-int btrfs_find_device_by_path(struct btrfs_root *root, char *device_path,
-			      struct btrfs_device **device)
+static int btrfs_find_device_by_path(struct btrfs_root *root, char *device_path,
+				     struct btrfs_device **device)
 {
 	int ret = 0;
 	struct btrfs_super_block *disk_super;
@@ -3607,7 +3608,7 @@ static int btrfs_cmp_device_info(const void *a, const void *b)
 	return 0;
 }
 
-struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
+static struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
 	[BTRFS_RAID_RAID10] = {
 		.sub_stripes	= 2,
 		.dev_stripes	= 1,
@@ -3674,18 +3675,10 @@ static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target)
 
 static void check_raid56_incompat_flag(struct btrfs_fs_info *info, u64 type)
 {
-	u64 features;
-
 	if (!(type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)))
 		return;
 
-	features = btrfs_super_incompat_flags(info->super_copy);
-	if (features & BTRFS_FEATURE_INCOMPAT_RAID56)
-		return;
-
-	features |= BTRFS_FEATURE_INCOMPAT_RAID56;
-	btrfs_set_super_incompat_flags(info->super_copy, features);
-	printk(KERN_INFO "btrfs: setting RAID5/6 feature flag\n");
+	btrfs_set_fs_incompat(info, RAID56);
 }
 
 static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
@@ -3932,7 +3925,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 
 	em_tree = &extent_root->fs_info->mapping_tree.map_tree;
 	write_lock(&em_tree->lock);
-	ret = add_extent_mapping(em_tree, em);
+	ret = add_extent_mapping(em_tree, em, 0);
 	write_unlock(&em_tree->lock);
 	if (ret) {
 		free_extent_map(em);
@@ -4240,9 +4233,25 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len)
 	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, logical, len);
 	read_unlock(&em_tree->lock);
-	BUG_ON(!em);
 
-	BUG_ON(em->start > logical || em->start + em->len < logical);
+	/*
+	 * We could return errors for these cases, but that could get ugly and
+	 * we'd probably do the same thing which is just not do anything else
+	 * and exit, so return 1 so the callers don't try to use other copies.
+	 */
+	if (!em) {
+		btrfs_emerg(fs_info, "No mapping for %Lu-%Lu\n", logical,
+			    logical+len);
+		return 1;
+	}
+
+	if (em->start > logical || em->start + em->len < logical) {
+		btrfs_emerg(fs_info, "Invalid mapping for %Lu-%Lu, got "
+			    "%Lu-%Lu\n", logical, logical+len, em->start,
+			    em->start + em->len);
+		return 1;
+	}
+
 	map = (struct map_lookup *)em->bdev;
 	if (map->type & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1))
 		ret = map->num_stripes;
@@ -4411,13 +4420,19 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 	read_unlock(&em_tree->lock);
 
 	if (!em) {
-		printk(KERN_CRIT "btrfs: unable to find logical %llu len %llu\n",
-		       (unsigned long long)logical,
-		       (unsigned long long)*length);
-		BUG();
+		btrfs_crit(fs_info, "unable to find logical %llu len %llu",
+			(unsigned long long)logical,
+			(unsigned long long)*length);
+		return -EINVAL;
+	}
+
+	if (em->start > logical || em->start + em->len < logical) {
+		btrfs_crit(fs_info, "found a bad mapping, wanted %Lu, "
+			   "found %Lu-%Lu\n", logical, em->start,
+			   em->start + em->len);
+		return -EINVAL;
 	}
 
-	BUG_ON(em->start > logical || em->start + em->len < logical);
 	map = (struct map_lookup *)em->bdev;
 	offset = logical - em->start;
 
@@ -5106,9 +5121,9 @@ struct async_sched {
  * This will add one bio to the pending list for a device and make sure
  * the work struct is scheduled.
  */
-noinline void btrfs_schedule_bio(struct btrfs_root *root,
-				 struct btrfs_device *device,
-				 int rw, struct bio *bio)
+static noinline void btrfs_schedule_bio(struct btrfs_root *root,
+					struct btrfs_device *device,
+					int rw, struct bio *bio)
 {
 	int should_queue = 1;
 	struct btrfs_pending_bios *pending_bios;
@@ -5308,10 +5323,10 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
 	}
 
 	if (map_length < length) {
-		printk(KERN_CRIT "btrfs: mapping failed logical %llu bio len %llu "
-		       "len %llu\n", (unsigned long long)logical,
-		       (unsigned long long)length,
-		       (unsigned long long)map_length);
+		btrfs_crit(root->fs_info, "mapping failed logical %llu bio len %llu len %llu",
+			(unsigned long long)logical,
+			(unsigned long long)length,
+			(unsigned long long)map_length);
 		BUG();
 	}
 
@@ -5476,7 +5491,7 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
 	}
 
 	write_lock(&map_tree->map_tree.lock);
-	ret = add_extent_mapping(&map_tree->map_tree, em);
+	ret = add_extent_mapping(&map_tree->map_tree, em, 0);
 	write_unlock(&map_tree->map_tree.lock);
 	BUG_ON(ret); /* Tree corruption */
 	free_extent_map(em);
@@ -5583,8 +5598,8 @@ static int read_one_dev(struct btrfs_root *root,
 			return -EIO;
 
 		if (!device) {
-			printk(KERN_WARNING "warning devid %llu missing\n",
-			       (unsigned long long)devid);
+			btrfs_warn(root->fs_info, "devid %llu missing",
+				(unsigned long long)devid);
 			device = add_missing_dev(root, devid, dev_uuid);
 			if (!device)
 				return -ENOMEM;
@@ -5926,7 +5941,7 @@ void btrfs_dev_stat_inc_and_print(struct btrfs_device *dev, int index)
 	btrfs_dev_stat_print_on_error(dev);
 }
 
-void btrfs_dev_stat_print_on_error(struct btrfs_device *dev)
+static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev)
 {
 	if (!dev->dev_stats_valid)
 		return;

fs/btrfs/volumes.h

@@ -254,10 +254,6 @@ int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start,
 #define btrfs_bio_size(n) (sizeof(struct btrfs_bio) + \
 			    (sizeof(struct btrfs_bio_stripe) * (n)))
 
-int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
-			   struct btrfs_device *device,
-			   u64 chunk_tree, u64 chunk_objectid,
-			   u64 chunk_offset, u64 start, u64 num_bytes);
 int btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 		    u64 logical, u64 *length,
 		    struct btrfs_bio **bbio_ret, int mirror_num);
@@ -282,11 +278,6 @@ void btrfs_close_extra_devices(struct btrfs_fs_info *fs_info,
 int btrfs_find_device_missing_or_by_path(struct btrfs_root *root,
 					 char *device_path,
 					 struct btrfs_device **device);
-int btrfs_find_device_by_path(struct btrfs_root *root, char *device_path,
-			      struct btrfs_device **device);
-int btrfs_add_device(struct btrfs_trans_handle *trans,
-		     struct btrfs_root *root,
-		     struct btrfs_device *device);
 int btrfs_rm_device(struct btrfs_root *root, char *device_path);
 void btrfs_cleanup_fs_uuids(void);
 int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len);
@@ -307,7 +298,6 @@ int btrfs_cancel_balance(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);
-void btrfs_dev_stat_print_on_error(struct btrfs_device *device);
 void btrfs_dev_stat_inc_and_print(struct btrfs_device *dev, int index);
 int btrfs_get_dev_stats(struct btrfs_root *root,
 			struct btrfs_ioctl_get_dev_stats *stats);
@@ -321,9 +311,6 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 void btrfs_init_dev_replace_tgtdev_for_resume(struct btrfs_fs_info *fs_info,
 					      struct btrfs_device *tgtdev);
 int btrfs_scratch_superblock(struct btrfs_device *device);
-void btrfs_schedule_bio(struct btrfs_root *root,
-			struct btrfs_device *device,
-			int rw, struct bio *bio);
 int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree,
 			   u64 logical, u64 len, int mirror_num);
 unsigned long btrfs_full_stripe_len(struct btrfs_root *root,

fs/btrfs/xattr.c

@@ -406,8 +406,8 @@ int btrfs_removexattr(struct dentry *dentry, const char *name)
 				XATTR_REPLACE);
 }
 
-int btrfs_initxattrs(struct inode *inode, const struct xattr *xattr_array,
-		     void *fs_info)
+static int btrfs_initxattrs(struct inode *inode,
+			    const struct xattr *xattr_array, void *fs_info)
 {
 	const struct xattr *xattr;
 	struct btrfs_trans_handle *trans = fs_info;

include/uapi/linux/btrfs.h

@@ -376,12 +376,18 @@ struct btrfs_ioctl_get_dev_stats {
 
 #define BTRFS_QUOTA_CTL_ENABLE	1
 #define BTRFS_QUOTA_CTL_DISABLE	2
-#define BTRFS_QUOTA_CTL_RESCAN	3
+#define BTRFS_QUOTA_CTL_RESCAN__NOTUSED	3
 struct btrfs_ioctl_quota_ctl_args {
 	__u64 cmd;
 	__u64 status;
 };
 
+struct btrfs_ioctl_quota_rescan_args {
+	__u64	flags;
+	__u64   progress;
+	__u64   reserved[6];
+};
+
 struct btrfs_ioctl_qgroup_assign_args {
 	__u64 assign;
 	__u64 src;
@@ -412,7 +418,25 @@ struct btrfs_ioctl_received_subvol_args {
  * search of clone sources doesn't find an extent. UPDATE_EXTENT
  * commands will be sent instead of WRITE commands.
  */
-#define BTRFS_SEND_FLAG_NO_FILE_DATA     0x1
+#define BTRFS_SEND_FLAG_NO_FILE_DATA		0x1
+
+/*
+ * Do not add the leading stream header. Used when multiple snapshots
+ * are sent back to back.
+ */
+#define BTRFS_SEND_FLAG_OMIT_STREAM_HEADER	0x2
+
+/*
+ * Omit the command at the end of the stream that indicated the end
+ * of the stream. This option is used when multiple snapshots are
+ * sent back to back.
+ */
+#define BTRFS_SEND_FLAG_OMIT_END_CMD		0x4
+
+#define BTRFS_SEND_FLAG_MASK \
+	(BTRFS_SEND_FLAG_NO_FILE_DATA | \
+	 BTRFS_SEND_FLAG_OMIT_STREAM_HEADER | \
+	 BTRFS_SEND_FLAG_OMIT_END_CMD)
 
 struct btrfs_ioctl_send_args {
 	__s64 send_fd;			/* in */
@@ -502,6 +526,10 @@ struct btrfs_ioctl_send_args {
 			       struct btrfs_ioctl_qgroup_create_args)
 #define BTRFS_IOC_QGROUP_LIMIT _IOR(BTRFS_IOCTL_MAGIC, 43, \
 			       struct btrfs_ioctl_qgroup_limit_args)
+#define BTRFS_IOC_QUOTA_RESCAN _IOW(BTRFS_IOCTL_MAGIC, 44, \
+			       struct btrfs_ioctl_quota_rescan_args)
+#define BTRFS_IOC_QUOTA_RESCAN_STATUS _IOR(BTRFS_IOCTL_MAGIC, 45, \
+			       struct btrfs_ioctl_quota_rescan_args)
 #define BTRFS_IOC_GET_FSLABEL _IOR(BTRFS_IOCTL_MAGIC, 49, \
 				   char[BTRFS_LABEL_SIZE])
 #define BTRFS_IOC_SET_FSLABEL _IOW(BTRFS_IOCTL_MAGIC, 50, \