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

Pull btrfs updates from Chris Mason:
 "This is against 3.11-rc7, but was pulled and tested against your tree
  as of yesterday.  We do have two small incrementals queued up, but I
  wanted to get this bunch out the door before I hop on an airplane.

  This is a fairly large batch of fixes, performance improvements, and
  cleanups from the usual Btrfs suspects.

  We've included Stefan Behren's work to index subvolume UUIDs, which is
  targeted at speeding up send/receive with many subvolumes or snapshots
  in place.  It closes a long standing performance issue that was built
  in to the disk format.

  Mark Fasheh's offline dedup work is also here.  In this case offline
  means the FS is mounted and active, but the dedup work is not done
  inline during file IO.  This is a building block where utilities are
  able to ask the FS to dedup a series of extents.  The kernel takes
  care of verifying the data involved really is the same.  Today this
  involves reading both extents, but we'll continue to evolve the
  patches"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (118 commits)
  Btrfs: optimize key searches in btrfs_search_slot
  Btrfs: don't use an async starter for most of our workers
  Btrfs: only update disk_i_size as we remove extents
  Btrfs: fix deadlock in uuid scan kthread
  Btrfs: stop refusing the relocation of chunk 0
  Btrfs: fix memory leak of uuid_root in free_fs_info
  btrfs: reuse kbasename helper
  btrfs: return btrfs error code for dev excl ops err
  Btrfs: allow partial ordered extent completion
  Btrfs: convert all bug_ons in free-space-cache.c
  Btrfs: add support for asserts
  Btrfs: adjust the fs_devices->missing count on unmount
  Btrf: cleanup: don't check for root_refs == 0 twice
  Btrfs: fix for patch "cleanup: don't check the same thing twice"
  Btrfs: get rid of one BUG() in write_all_supers()
  Btrfs: allocate prelim_ref with a slab allocater
  Btrfs: pass gfp_t to __add_prelim_ref() to avoid always using GFP_ATOMIC
  Btrfs: fix race conditions in BTRFS_IOC_FS_INFO ioctl
  Btrfs: fix race between removing a dev and writing sbs
  Btrfs: remove ourselves from the cluster list under lock
  ...

fs/btrfs/Kconfig

@@ -72,3 +72,12 @@ config BTRFS_DEBUG
 	  performance, or export extra information via sysfs.
 
 	  If unsure, say N.
+
+config BTRFS_ASSERT
+	bool "Btrfs assert support"
+	depends on BTRFS_FS
+	help
+	  Enable run-time assertion checking.  This will result in panics if
+	  any of the assertions trip.  This is meant for btrfs developers only.
+
+	  If unsure, say N.

fs/btrfs/Makefile

@@ -8,7 +8,10 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \
 	   extent_io.o volumes.o async-thread.o ioctl.o locking.o orphan.o \
 	   export.o tree-log.o free-space-cache.o zlib.o lzo.o \
 	   compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o \
-	   reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o
+	   reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o \
+	   uuid-tree.o
 
 btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o
 btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o
+
+btrfs-$(CONFIG_BTRFS_FS_RUN_SANITY_TESTS) += tests/free-space-tests.o

fs/btrfs/backref.c

@@ -119,6 +119,26 @@ struct __prelim_ref {
 	u64 wanted_disk_byte;
 };
 
+static struct kmem_cache *btrfs_prelim_ref_cache;
+
+int __init btrfs_prelim_ref_init(void)
+{
+	btrfs_prelim_ref_cache = kmem_cache_create("btrfs_prelim_ref",
+					sizeof(struct __prelim_ref),
+					0,
+					SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD,
+					NULL);
+	if (!btrfs_prelim_ref_cache)
+		return -ENOMEM;
+	return 0;
+}
+
+void btrfs_prelim_ref_exit(void)
+{
+	if (btrfs_prelim_ref_cache)
+		kmem_cache_destroy(btrfs_prelim_ref_cache);
+}
+
 /*
  * the rules for all callers of this function are:
  * - obtaining the parent is the goal
@@ -160,12 +180,12 @@ struct __prelim_ref {
 
 static int __add_prelim_ref(struct list_head *head, u64 root_id,
 			    struct btrfs_key *key, int level,
-			    u64 parent, u64 wanted_disk_byte, int count)
+			    u64 parent, u64 wanted_disk_byte, int count,
+			    gfp_t gfp_mask)
 {
 	struct __prelim_ref *ref;
 
-	/* in case we're adding delayed refs, we're holding the refs spinlock */
-	ref = kmalloc(sizeof(*ref), GFP_ATOMIC);
+	ref = kmem_cache_alloc(btrfs_prelim_ref_cache, gfp_mask);
 	if (!ref)
 		return -ENOMEM;
 
@@ -295,10 +315,9 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
 	ret = btrfs_search_old_slot(root, &ref->key_for_search, path, time_seq);
 	pr_debug("search slot in root %llu (level %d, ref count %d) returned "
 		 "%d for key (%llu %u %llu)\n",
-		 (unsigned long long)ref->root_id, level, ref->count, ret,
-		 (unsigned long long)ref->key_for_search.objectid,
-		 ref->key_for_search.type,
-		 (unsigned long long)ref->key_for_search.offset);
+		 ref->root_id, level, ref->count, ret,
+		 ref->key_for_search.objectid, ref->key_for_search.type,
+		 ref->key_for_search.offset);
 	if (ret < 0)
 		goto out;
 
@@ -365,11 +384,12 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
 		node = ulist_next(parents, &uiter);
 		ref->parent = node ? node->val : 0;
 		ref->inode_list = node ?
-			(struct extent_inode_elem *)(uintptr_t)node->aux : 0;
+			(struct extent_inode_elem *)(uintptr_t)node->aux : NULL;
 
 		/* additional parents require new refs being added here */
 		while ((node = ulist_next(parents, &uiter))) {
-			new_ref = kmalloc(sizeof(*new_ref), GFP_NOFS);
+			new_ref = kmem_cache_alloc(btrfs_prelim_ref_cache,
+						   GFP_NOFS);
 			if (!new_ref) {
 				ret = -ENOMEM;
 				goto out;
@@ -493,7 +513,7 @@ static void __merge_refs(struct list_head *head, int mode)
 			ref1->count += ref2->count;
 
 			list_del(&ref2->list);
-			kfree(ref2);
+			kmem_cache_free(btrfs_prelim_ref_cache, ref2);
 		}
 
 	}
@@ -548,7 +568,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			ref = btrfs_delayed_node_to_tree_ref(node);
 			ret = __add_prelim_ref(prefs, ref->root, &op_key,
 					       ref->level + 1, 0, node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		case BTRFS_SHARED_BLOCK_REF_KEY: {
@@ -558,7 +578,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			ret = __add_prelim_ref(prefs, ref->root, NULL,
 					       ref->level + 1, ref->parent,
 					       node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		case BTRFS_EXTENT_DATA_REF_KEY: {
@@ -570,7 +590,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			key.offset = ref->offset;
 			ret = __add_prelim_ref(prefs, ref->root, &key, 0, 0,
 					       node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		case BTRFS_SHARED_DATA_REF_KEY: {
@@ -583,7 +603,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
 			key.offset = ref->offset;
 			ret = __add_prelim_ref(prefs, ref->root, &key, 0,
 					       ref->parent, node->bytenr,
-					       node->ref_mod * sgn);
+					       node->ref_mod * sgn, GFP_ATOMIC);
 			break;
 		}
 		default:
@@ -657,7 +677,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 		case BTRFS_SHARED_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, 0, NULL,
 						*info_level + 1, offset,
-						bytenr, 1);
+						bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_SHARED_DATA_REF_KEY: {
 			struct btrfs_shared_data_ref *sdref;
@@ -666,13 +686,13 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 			sdref = (struct btrfs_shared_data_ref *)(iref + 1);
 			count = btrfs_shared_data_ref_count(leaf, sdref);
 			ret = __add_prelim_ref(prefs, 0, NULL, 0, offset,
-					       bytenr, count);
+					       bytenr, count, GFP_NOFS);
 			break;
 		}
 		case BTRFS_TREE_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, offset, NULL,
 					       *info_level + 1, 0,
-					       bytenr, 1);
+					       bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_EXTENT_DATA_REF_KEY: {
 			struct btrfs_extent_data_ref *dref;
@@ -687,7 +707,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
 			root = btrfs_extent_data_ref_root(leaf, dref);
 			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
-					       bytenr, count);
+					       bytenr, count, GFP_NOFS);
 			break;
 		}
 		default:
@@ -738,7 +758,7 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 		case BTRFS_SHARED_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, 0, NULL,
 						info_level + 1, key.offset,
-						bytenr, 1);
+						bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_SHARED_DATA_REF_KEY: {
 			struct btrfs_shared_data_ref *sdref;
@@ -748,13 +768,13 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 					      struct btrfs_shared_data_ref);
 			count = btrfs_shared_data_ref_count(leaf, sdref);
 			ret = __add_prelim_ref(prefs, 0, NULL, 0, key.offset,
-						bytenr, count);
+						bytenr, count, GFP_NOFS);
 			break;
 		}
 		case BTRFS_TREE_BLOCK_REF_KEY:
 			ret = __add_prelim_ref(prefs, key.offset, NULL,
 					       info_level + 1, 0,
-					       bytenr, 1);
+					       bytenr, 1, GFP_NOFS);
 			break;
 		case BTRFS_EXTENT_DATA_REF_KEY: {
 			struct btrfs_extent_data_ref *dref;
@@ -770,7 +790,7 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
 			key.offset = btrfs_extent_data_ref_offset(leaf, dref);
 			root = btrfs_extent_data_ref_root(leaf, dref);
 			ret = __add_prelim_ref(prefs, root, &key, 0, 0,
-					       bytenr, count);
+					       bytenr, count, GFP_NOFS);
 			break;
 		}
 		default:
@@ -911,7 +931,6 @@ again:
 
 	while (!list_empty(&prefs)) {
 		ref = list_first_entry(&prefs, struct __prelim_ref, list);
-		list_del(&ref->list);
 		WARN_ON(ref->count < 0);
 		if (ref->count && ref->root_id && ref->parent == 0) {
 			/* no parent == root of tree */
@@ -935,8 +954,10 @@ again:
 				}
 				ret = find_extent_in_eb(eb, bytenr,
 							*extent_item_pos, &eie);
-				ref->inode_list = eie;
 				free_extent_buffer(eb);
+				if (ret < 0)
+					goto out;
+				ref->inode_list = eie;
 			}
 			ret = ulist_add_merge(refs, ref->parent,
 					      (uintptr_t)ref->inode_list,
@@ -954,7 +975,8 @@ again:
 				eie->next = ref->inode_list;
 			}
 		}
-		kfree(ref);
+		list_del(&ref->list);
+		kmem_cache_free(btrfs_prelim_ref_cache, ref);
 	}
 
 out:
@@ -962,13 +984,13 @@ out:
 	while (!list_empty(&prefs)) {
 		ref = list_first_entry(&prefs, struct __prelim_ref, list);
 		list_del(&ref->list);
-		kfree(ref);
+		kmem_cache_free(btrfs_prelim_ref_cache, ref);
 	}
 	while (!list_empty(&prefs_delayed)) {
 		ref = list_first_entry(&prefs_delayed, struct __prelim_ref,
 				       list);
 		list_del(&ref->list);
-		kfree(ref);
+		kmem_cache_free(btrfs_prelim_ref_cache, ref);
 	}
 
 	return ret;
@@ -1326,8 +1348,7 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
 	     found_key->type != BTRFS_METADATA_ITEM_KEY) ||
 	    found_key->objectid > logical ||
 	    found_key->objectid + size <= logical) {
-		pr_debug("logical %llu is not within any extent\n",
-			 (unsigned long long)logical);
+		pr_debug("logical %llu is not within any extent\n", logical);
 		return -ENOENT;
 	}
 
@@ -1340,11 +1361,8 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
 
 	pr_debug("logical %llu is at position %llu within the extent (%llu "
 		 "EXTENT_ITEM %llu) flags %#llx size %u\n",
-		 (unsigned long long)logical,
-		 (unsigned long long)(logical - found_key->objectid),
-		 (unsigned long long)found_key->objectid,
-		 (unsigned long long)found_key->offset,
-		 (unsigned long long)flags, item_size);
+		 logical, logical - found_key->objectid, found_key->objectid,
+		 found_key->offset, flags, item_size);
 
 	WARN_ON(!flags_ret);
 	if (flags_ret) {
@@ -1516,7 +1534,7 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
 		while (!ret && (root_node = ulist_next(roots, &root_uiter))) {
 			pr_debug("root %llu references leaf %llu, data list "
 				 "%#llx\n", root_node->val, ref_node->val,
-				 (long long)ref_node->aux);
+				 ref_node->aux);
 			ret = iterate_leaf_refs((struct extent_inode_elem *)
 						(uintptr_t)ref_node->aux,
 						root_node->val,
@@ -1608,9 +1626,8 @@ static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root,
 			name_len = btrfs_inode_ref_name_len(eb, iref);
 			/* path must be released before calling iterate()! */
 			pr_debug("following ref at offset %u for inode %llu in "
-				 "tree %llu\n", cur,
-				 (unsigned long long)found_key.objectid,
-				 (unsigned long long)fs_root->objectid);
+				 "tree %llu\n", cur, found_key.objectid,
+				 fs_root->objectid);
 			ret = iterate(parent, name_len,
 				      (unsigned long)(iref + 1), eb, ctx);
 			if (ret)

fs/btrfs/backref.h

@@ -72,4 +72,6 @@ int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid,
 			  struct btrfs_inode_extref **ret_extref,
 			  u64 *found_off);
 
+int __init btrfs_prelim_ref_init(void);
+void btrfs_prelim_ref_exit(void);
 #endif

fs/btrfs/btrfs_inode.h

@@ -218,6 +218,27 @@ static inline int btrfs_inode_in_log(struct inode *inode, u64 generation)
 	return 0;
 }
 
+struct btrfs_dio_private {
+	struct inode *inode;
+	u64 logical_offset;
+	u64 disk_bytenr;
+	u64 bytes;
+	void *private;
+
+	/* number of bios pending for this dio */
+	atomic_t pending_bios;
+
+	/* IO errors */
+	int errors;
+
+	/* orig_bio is our btrfs_io_bio */
+	struct bio *orig_bio;
+
+	/* dio_bio came from fs/direct-io.c */
+	struct bio *dio_bio;
+	u8 csum[0];
+};
+
 /*
  * Disable DIO read nolock optimization, so new dio readers will be forced
  * to grab i_mutex. It is used to avoid the endless truncate due to

fs/btrfs/compression.c

@@ -132,9 +132,8 @@ static int check_compressed_csum(struct inode *inode,
 			printk(KERN_INFO "btrfs csum failed ino %llu "
 			       "extent %llu csum %u "
 			       "wanted %u mirror %d\n",
-			       (unsigned long long)btrfs_ino(inode),
-			       (unsigned long long)disk_start,
-			       csum, *cb_sum, cb->mirror_num);
+			       btrfs_ino(inode), disk_start, csum, *cb_sum,
+			       cb->mirror_num);
 			ret = -EIO;
 			goto fail;
 		}
@@ -639,7 +638,11 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	faili = nr_pages - 1;
 	cb->nr_pages = nr_pages;
 
-	add_ra_bio_pages(inode, em_start + em_len, cb);
+	/* In the parent-locked case, we only locked the range we are
+	 * interested in.  In all other cases, we can opportunistically
+	 * cache decompressed data that goes beyond the requested range. */
+	if (!(bio_flags & EXTENT_BIO_PARENT_LOCKED))
+		add_ra_bio_pages(inode, em_start + em_len, cb);
 
 	/* include any pages we added in add_ra-bio_pages */
 	uncompressed_len = bio->bi_vcnt * PAGE_CACHE_SIZE;

fs/btrfs/ctree.c

@@ -274,8 +274,7 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans,
 	else
 		btrfs_set_header_owner(cow, new_root_objectid);
 
-	write_extent_buffer(cow, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(cow),
+	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(cow),
 			    BTRFS_FSID_SIZE);
 
 	WARN_ON(btrfs_header_generation(buf) > trans->transid);
@@ -484,8 +483,27 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
 	struct rb_node **new;
 	struct rb_node *parent = NULL;
 	struct tree_mod_elem *cur;
+	int ret = 0;
+
+	BUG_ON(!tm);
+
+	tree_mod_log_write_lock(fs_info);
+	if (list_empty(&fs_info->tree_mod_seq_list)) {
+		tree_mod_log_write_unlock(fs_info);
+		/*
+		 * Ok we no longer care about logging modifications, free up tm
+		 * and return 0.  Any callers shouldn't be using tm after
+		 * calling tree_mod_log_insert, but if they do we can just
+		 * change this to return a special error code to let the callers
+		 * do their own thing.
+		 */
+		kfree(tm);
+		return 0;
+	}
 
-	BUG_ON(!tm || !tm->seq);
+	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);
 
 	tm_root = &fs_info->tree_mod_log;
 	new = &tm_root->rb_node;
@@ -501,14 +519,17 @@ __tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
 		else if (cur->seq > tm->seq)
 			new = &((*new)->rb_right);
 		else {
+			ret = -EEXIST;
 			kfree(tm);
-			return -EEXIST;
+			goto out;
 		}
 	}
 
 	rb_link_node(&tm->node, parent, new);
 	rb_insert_color(&tm->node, tm_root);
-	return 0;
+out:
+	tree_mod_log_write_unlock(fs_info);
+	return ret;
 }
 
 /*
@@ -524,57 +545,19 @@ static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info,
 		return 1;
 	if (eb && btrfs_header_level(eb) == 0)
 		return 1;
-
-	tree_mod_log_write_lock(fs_info);
-	if (list_empty(&fs_info->tree_mod_seq_list)) {
-		/*
-		 * someone emptied the list while we were waiting for the lock.
-		 * we must not add to the list when no blocker exists.
-		 */
-		tree_mod_log_write_unlock(fs_info);
-		return 1;
-	}
-
 	return 0;
 }
 
-/*
- * This allocates memory and gets a tree modification sequence number.
- *
- * Returns <0 on error.
- * Returns >0 (the added sequence number) on success.
- */
-static inline int tree_mod_alloc(struct btrfs_fs_info *fs_info, gfp_t flags,
-				 struct tree_mod_elem **tm_ret)
-{
-	struct tree_mod_elem *tm;
-
-	/*
-	 * once we switch from spin locks to something different, we should
-	 * honor the flags parameter here.
-	 */
-	tm = *tm_ret = kzalloc(sizeof(*tm), GFP_ATOMIC);
-	if (!tm)
-		return -ENOMEM;
-
-	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;
-}
-
 static inline int
 __tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
 			  struct extent_buffer *eb, int slot,
 			  enum mod_log_op op, gfp_t flags)
 {
-	int ret;
 	struct tree_mod_elem *tm;
 
-	ret = tree_mod_alloc(fs_info, flags, &tm);
-	if (ret < 0)
-		return ret;
+	tm = kzalloc(sizeof(*tm), flags);
+	if (!tm)
+		return -ENOMEM;
 
 	tm->index = eb->start >> PAGE_CACHE_SHIFT;
 	if (op != MOD_LOG_KEY_ADD) {
@@ -589,34 +572,14 @@ __tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
 }
 
 static noinline int
-tree_mod_log_insert_key_mask(struct btrfs_fs_info *fs_info,
-			     struct extent_buffer *eb, int slot,
-			     enum mod_log_op op, gfp_t flags)
+tree_mod_log_insert_key(struct btrfs_fs_info *fs_info,
+			struct extent_buffer *eb, int slot,
+			enum mod_log_op op, gfp_t flags)
 {
-	int ret;
-
 	if (tree_mod_dont_log(fs_info, eb))
 		return 0;
 
-	ret = __tree_mod_log_insert_key(fs_info, eb, slot, op, flags);
-
-	tree_mod_log_write_unlock(fs_info);
-	return ret;
-}
-
-static noinline int
-tree_mod_log_insert_key(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
-			int slot, enum mod_log_op op)
-{
-	return tree_mod_log_insert_key_mask(fs_info, eb, slot, op, GFP_NOFS);
-}
-
-static noinline int
-tree_mod_log_insert_key_locked(struct btrfs_fs_info *fs_info,
-			     struct extent_buffer *eb, int slot,
-			     enum mod_log_op op)
-{
-	return __tree_mod_log_insert_key(fs_info, eb, slot, op, GFP_NOFS);
+	return __tree_mod_log_insert_key(fs_info, eb, slot, op, flags);
 }
 
 static noinline int
@@ -637,14 +600,14 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
 	 * buffer, i.e. dst_slot < src_slot.
 	 */
 	for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
-		ret = tree_mod_log_insert_key_locked(fs_info, eb, i + dst_slot,
-					      MOD_LOG_KEY_REMOVE_WHILE_MOVING);
+		ret = __tree_mod_log_insert_key(fs_info, eb, i + dst_slot,
+				MOD_LOG_KEY_REMOVE_WHILE_MOVING, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 
-	ret = tree_mod_alloc(fs_info, flags, &tm);
-	if (ret < 0)
-		goto out;
+	tm = kzalloc(sizeof(*tm), flags);
+	if (!tm)
+		return -ENOMEM;
 
 	tm->index = eb->start >> PAGE_CACHE_SHIFT;
 	tm->slot = src_slot;
@@ -652,10 +615,7 @@ tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
 	tm->move.nr_items = nr_items;
 	tm->op = MOD_LOG_MOVE_KEYS;
 
-	ret = __tree_mod_log_insert(fs_info, tm);
-out:
-	tree_mod_log_write_unlock(fs_info);
-	return ret;
+	return __tree_mod_log_insert(fs_info, tm);
 }
 
 static inline void
@@ -670,8 +630,8 @@ __tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 
 	nritems = btrfs_header_nritems(eb);
 	for (i = nritems - 1; i >= 0; i--) {
-		ret = tree_mod_log_insert_key_locked(fs_info, eb, i,
-					      MOD_LOG_KEY_REMOVE_WHILE_FREEING);
+		ret = __tree_mod_log_insert_key(fs_info, eb, i,
+				MOD_LOG_KEY_REMOVE_WHILE_FREEING, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 }
@@ -683,7 +643,6 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 			 int log_removal)
 {
 	struct tree_mod_elem *tm;
-	int ret;
 
 	if (tree_mod_dont_log(fs_info, NULL))
 		return 0;
@@ -691,9 +650,9 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 	if (log_removal)
 		__tree_mod_log_free_eb(fs_info, old_root);
 
-	ret = tree_mod_alloc(fs_info, flags, &tm);
-	if (ret < 0)
-		goto out;
+	tm = kzalloc(sizeof(*tm), flags);
+	if (!tm)
+		return -ENOMEM;
 
 	tm->index = new_root->start >> PAGE_CACHE_SHIFT;
 	tm->old_root.logical = old_root->start;
@@ -701,10 +660,7 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
 	tm->generation = btrfs_header_generation(old_root);
 	tm->op = MOD_LOG_ROOT_REPLACE;
 
-	ret = __tree_mod_log_insert(fs_info, tm);
-out:
-	tree_mod_log_write_unlock(fs_info);
-	return ret;
+	return __tree_mod_log_insert(fs_info, tm);
 }
 
 static struct tree_mod_elem *
@@ -784,23 +740,20 @@ tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
 	if (tree_mod_dont_log(fs_info, NULL))
 		return;
 
-	if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0) {
-		tree_mod_log_write_unlock(fs_info);
+	if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0)
 		return;
-	}
 
 	for (i = 0; i < nr_items; i++) {
-		ret = tree_mod_log_insert_key_locked(fs_info, src,
+		ret = __tree_mod_log_insert_key(fs_info, src,
 						i + src_offset,
-						MOD_LOG_KEY_REMOVE);
+						MOD_LOG_KEY_REMOVE, GFP_NOFS);
 		BUG_ON(ret < 0);
-		ret = tree_mod_log_insert_key_locked(fs_info, dst,
+		ret = __tree_mod_log_insert_key(fs_info, dst,
 						     i + dst_offset,
-						     MOD_LOG_KEY_ADD);
+						     MOD_LOG_KEY_ADD,
+						     GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
-
-	tree_mod_log_write_unlock(fs_info);
 }
 
 static inline void
@@ -819,9 +772,9 @@ tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info,
 {
 	int ret;
 
-	ret = tree_mod_log_insert_key_mask(fs_info, eb, slot,
-					   MOD_LOG_KEY_REPLACE,
-					   atomic ? GFP_ATOMIC : GFP_NOFS);
+	ret = __tree_mod_log_insert_key(fs_info, eb, slot,
+					MOD_LOG_KEY_REPLACE,
+					atomic ? GFP_ATOMIC : GFP_NOFS);
 	BUG_ON(ret < 0);
 }
 
@@ -830,10 +783,7 @@ tree_mod_log_free_eb(struct btrfs_fs_info *fs_info, struct extent_buffer *eb)
 {
 	if (tree_mod_dont_log(fs_info, eb))
 		return;
-
 	__tree_mod_log_free_eb(fs_info, eb);
-
-	tree_mod_log_write_unlock(fs_info);
 }
 
 static noinline void
@@ -1046,8 +996,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 	else
 		btrfs_set_header_owner(cow, root->root_key.objectid);
 
-	write_extent_buffer(cow, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(cow),
+	write_extent_buffer(cow, root->fs_info->fsid, btrfs_header_fsid(cow),
 			    BTRFS_FSID_SIZE);
 
 	ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
@@ -1083,7 +1032,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 
 		WARN_ON(trans->transid != btrfs_header_generation(parent));
 		tree_mod_log_insert_key(root->fs_info, parent, parent_slot,
-					MOD_LOG_KEY_REPLACE);
+					MOD_LOG_KEY_REPLACE, GFP_NOFS);
 		btrfs_set_node_blockptr(parent, parent_slot,
 					cow->start);
 		btrfs_set_node_ptr_generation(parent, parent_slot,
@@ -1116,7 +1065,7 @@ __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
 	int looped = 0;
 
 	if (!time_seq)
-		return 0;
+		return NULL;
 
 	/*
 	 * the very last operation that's logged for a root is the replacement
@@ -1127,7 +1076,7 @@ __tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
 		tm = tree_mod_log_search_oldest(fs_info, root_logical,
 						time_seq);
 		if (!looped && !tm)
-			return 0;
+			return NULL;
 		/*
 		 * if there are no tree operation for the oldest root, we simply
 		 * return it. this should only happen if that (old) root is at
@@ -1240,8 +1189,8 @@ __tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
  * 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)
+tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
+		    struct extent_buffer *eb, u64 time_seq)
 {
 	struct extent_buffer *eb_rewin;
 	struct tree_mod_elem *tm;
@@ -1256,11 +1205,18 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
 	if (!tm)
 		return eb;
 
+	btrfs_set_path_blocking(path);
+	btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
+
 	if (tm->op == MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
 		BUG_ON(tm->slot != 0);
 		eb_rewin = alloc_dummy_extent_buffer(eb->start,
 						fs_info->tree_root->nodesize);
-		BUG_ON(!eb_rewin);
+		if (!eb_rewin) {
+			btrfs_tree_read_unlock_blocking(eb);
+			free_extent_buffer(eb);
+			return NULL;
+		}
 		btrfs_set_header_bytenr(eb_rewin, eb->start);
 		btrfs_set_header_backref_rev(eb_rewin,
 					     btrfs_header_backref_rev(eb));
@@ -1268,10 +1224,15 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
 		btrfs_set_header_level(eb_rewin, btrfs_header_level(eb));
 	} else {
 		eb_rewin = btrfs_clone_extent_buffer(eb);
-		BUG_ON(!eb_rewin);
+		if (!eb_rewin) {
+			btrfs_tree_read_unlock_blocking(eb);
+			free_extent_buffer(eb);
+			return NULL;
+		}
 	}
 
-	btrfs_tree_read_unlock(eb);
+	btrfs_clear_path_blocking(path, NULL, BTRFS_READ_LOCK);
+	btrfs_tree_read_unlock_blocking(eb);
 	free_extent_buffer(eb);
 
 	extent_buffer_get(eb_rewin);
@@ -1335,8 +1296,9 @@ get_old_root(struct btrfs_root *root, u64 time_seq)
 		free_extent_buffer(eb_root);
 		eb = alloc_dummy_extent_buffer(logical, root->nodesize);
 	} else {
+		btrfs_set_lock_blocking_rw(eb_root, BTRFS_READ_LOCK);
 		eb = btrfs_clone_extent_buffer(eb_root);
-		btrfs_tree_read_unlock(eb_root);
+		btrfs_tree_read_unlock_blocking(eb_root);
 		free_extent_buffer(eb_root);
 	}
 
@@ -1419,14 +1381,12 @@ noinline int btrfs_cow_block(struct btrfs_trans_handle *trans,
 
 	if (trans->transaction != root->fs_info->running_transaction)
 		WARN(1, KERN_CRIT "trans %llu running %llu\n",
-		       (unsigned long long)trans->transid,
-		       (unsigned long long)
+		       trans->transid,
 		       root->fs_info->running_transaction->transid);
 
 	if (trans->transid != root->fs_info->generation)
 		WARN(1, KERN_CRIT "trans %llu running %llu\n",
-		       (unsigned long long)trans->transid,
-		       (unsigned long long)root->fs_info->generation);
+		       trans->transid, root->fs_info->generation);
 
 	if (!should_cow_block(trans, root, buf)) {
 		*cow_ret = buf;
@@ -2466,6 +2426,40 @@ done:
 	return ret;
 }
 
+static void key_search_validate(struct extent_buffer *b,
+				struct btrfs_key *key,
+				int level)
+{
+#ifdef CONFIG_BTRFS_ASSERT
+	struct btrfs_disk_key disk_key;
+
+	btrfs_cpu_key_to_disk(&disk_key, key);
+
+	if (level == 0)
+		ASSERT(!memcmp_extent_buffer(b, &disk_key,
+		    offsetof(struct btrfs_leaf, items[0].key),
+		    sizeof(disk_key)));
+	else
+		ASSERT(!memcmp_extent_buffer(b, &disk_key,
+		    offsetof(struct btrfs_node, ptrs[0].key),
+		    sizeof(disk_key)));
+#endif
+}
+
+static int key_search(struct extent_buffer *b, struct btrfs_key *key,
+		      int level, int *prev_cmp, int *slot)
+{
+	if (*prev_cmp != 0) {
+		*prev_cmp = bin_search(b, key, level, slot);
+		return *prev_cmp;
+	}
+
+	key_search_validate(b, key, level);
+	*slot = 0;
+
+	return 0;
+}
+
 /*
  * look for key in the tree.  path is filled in with nodes along the way
  * if key is found, we return zero and you can find the item in the leaf
@@ -2494,6 +2488,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
 	int write_lock_level = 0;
 	u8 lowest_level = 0;
 	int min_write_lock_level;
+	int prev_cmp;
 
 	lowest_level = p->lowest_level;
 	WARN_ON(lowest_level && ins_len > 0);
@@ -2524,6 +2519,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
 	min_write_lock_level = write_lock_level;
 
 again:
+	prev_cmp = -1;
 	/*
 	 * we try very hard to do read locks on the root
 	 */
@@ -2624,7 +2620,7 @@ cow_done:
 		if (!cow)
 			btrfs_unlock_up_safe(p, level + 1);
 
-		ret = bin_search(b, key, level, &slot);
+		ret = key_search(b, key, level, &prev_cmp, &slot);
 
 		if (level != 0) {
 			int dec = 0;
@@ -2759,6 +2755,7 @@ int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
 	int level;
 	int lowest_unlock = 1;
 	u8 lowest_level = 0;
+	int prev_cmp;
 
 	lowest_level = p->lowest_level;
 	WARN_ON(p->nodes[0] != NULL);
@@ -2769,6 +2766,7 @@ int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
 	}
 
 again:
+	prev_cmp = -1;
 	b = get_old_root(root, time_seq);
 	level = btrfs_header_level(b);
 	p->locks[level] = BTRFS_READ_LOCK;
@@ -2786,7 +2784,7 @@ again:
 		 */
 		btrfs_unlock_up_safe(p, level + 1);
 
-		ret = bin_search(b, key, level, &slot);
+		ret = key_search(b, key, level, &prev_cmp, &slot);
 
 		if (level != 0) {
 			int dec = 0;
@@ -2820,7 +2818,11 @@ again:
 				btrfs_clear_path_blocking(p, b,
 							  BTRFS_READ_LOCK);
 			}
-			b = tree_mod_log_rewind(root->fs_info, b, time_seq);
+			b = tree_mod_log_rewind(root->fs_info, p, b, time_seq);
+			if (!b) {
+				ret = -ENOMEM;
+				goto done;
+			}
 			p->locks[level] = BTRFS_READ_LOCK;
 			p->nodes[level] = b;
 		} else {
@@ -3143,13 +3145,11 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
 	btrfs_set_header_backref_rev(c, BTRFS_MIXED_BACKREF_REV);
 	btrfs_set_header_owner(c, root->root_key.objectid);
 
-	write_extent_buffer(c, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(c),
+	write_extent_buffer(c, root->fs_info->fsid, btrfs_header_fsid(c),
 			    BTRFS_FSID_SIZE);
 
 	write_extent_buffer(c, root->fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(c),
-			    BTRFS_UUID_SIZE);
+			    btrfs_header_chunk_tree_uuid(c), BTRFS_UUID_SIZE);
 
 	btrfs_set_node_key(c, &lower_key, 0);
 	btrfs_set_node_blockptr(c, 0, lower->start);
@@ -3208,7 +3208,7 @@ static void insert_ptr(struct btrfs_trans_handle *trans,
 	}
 	if (level) {
 		ret = tree_mod_log_insert_key(root->fs_info, lower, slot,
-					      MOD_LOG_KEY_ADD);
+					      MOD_LOG_KEY_ADD, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 	btrfs_set_node_key(lower, key, slot);
@@ -3284,10 +3284,9 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 	btrfs_set_header_backref_rev(split, BTRFS_MIXED_BACKREF_REV);
 	btrfs_set_header_owner(split, root->root_key.objectid);
 	write_extent_buffer(split, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(split),
-			    BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(split), BTRFS_FSID_SIZE);
 	write_extent_buffer(split, root->fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(split),
+			    btrfs_header_chunk_tree_uuid(split),
 			    BTRFS_UUID_SIZE);
 
 	tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid);
@@ -4040,11 +4039,10 @@ again:
 	btrfs_set_header_owner(right, root->root_key.objectid);
 	btrfs_set_header_level(right, 0);
 	write_extent_buffer(right, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(right),
-			    BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(right), BTRFS_FSID_SIZE);
 
 	write_extent_buffer(right, root->fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(right),
+			    btrfs_header_chunk_tree_uuid(right),
 			    BTRFS_UUID_SIZE);
 
 	if (split == 0) {
@@ -4642,7 +4640,7 @@ static void del_ptr(struct btrfs_root *root, struct btrfs_path *path,
 			      (nritems - slot - 1));
 	} else if (level) {
 		ret = tree_mod_log_insert_key(root->fs_info, parent, slot,
-					      MOD_LOG_KEY_REMOVE);
+					      MOD_LOG_KEY_REMOVE, GFP_NOFS);
 		BUG_ON(ret < 0);
 	}
 
@@ -4814,7 +4812,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
  * This may release the path, and so you may lose any locks held at the
  * time you call it.
  */
-int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
+static int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
 {
 	struct btrfs_key key;
 	struct btrfs_disk_key found_key;
@@ -5329,19 +5327,20 @@ int btrfs_compare_trees(struct btrfs_root *left_root,
 					goto out;
 				advance_right = ADVANCE;
 			} else {
+				enum btrfs_compare_tree_result cmp;
+
 				WARN_ON(!extent_buffer_uptodate(left_path->nodes[0]));
 				ret = tree_compare_item(left_root, left_path,
 						right_path, tmp_buf);
-				if (ret) {
-					WARN_ON(!extent_buffer_uptodate(left_path->nodes[0]));
-					ret = changed_cb(left_root, right_root,
-						left_path, right_path,
-						&left_key,
-						BTRFS_COMPARE_TREE_CHANGED,
-						ctx);
-					if (ret < 0)
-						goto out;
-				}
+				if (ret)
+					cmp = BTRFS_COMPARE_TREE_CHANGED;
+				else
+					cmp = BTRFS_COMPARE_TREE_SAME;
+				ret = changed_cb(left_root, right_root,
+						 left_path, right_path,
+						 &left_key, cmp, ctx);
+				if (ret < 0)
+					goto out;
 				advance_left = ADVANCE;
 				advance_right = ADVANCE;
 			}

fs/btrfs/ctree.h

@@ -23,6 +23,7 @@
 #include <linux/highmem.h>
 #include <linux/fs.h>
 #include <linux/rwsem.h>
+#include <linux/semaphore.h>
 #include <linux/completion.h>
 #include <linux/backing-dev.h>
 #include <linux/wait.h>
@@ -91,6 +92,9 @@ struct btrfs_ordered_sum;
 /* holds quota configuration and tracking */
 #define BTRFS_QUOTA_TREE_OBJECTID 8ULL
 
+/* for storing items that use the BTRFS_UUID_KEY* types */
+#define BTRFS_UUID_TREE_OBJECTID 9ULL
+
 /* for storing balance parameters in the root tree */
 #define BTRFS_BALANCE_OBJECTID -4ULL
 
@@ -142,7 +146,7 @@ struct btrfs_ordered_sum;
 
 #define BTRFS_EMPTY_SUBVOL_DIR_OBJECTID 2
 
-#define BTRFS_DEV_REPLACE_DEVID 0
+#define BTRFS_DEV_REPLACE_DEVID 0ULL
 
 /*
  * the max metadata block size.  This limit is somewhat artificial,
@@ -478,9 +482,10 @@ struct btrfs_super_block {
 	char label[BTRFS_LABEL_SIZE];
 
 	__le64 cache_generation;
+	__le64 uuid_tree_generation;
 
 	/* future expansion */
-	__le64 reserved[31];
+	__le64 reserved[30];
 	u8 sys_chunk_array[BTRFS_SYSTEM_CHUNK_ARRAY_SIZE];
 	struct btrfs_root_backup super_roots[BTRFS_NUM_BACKUP_ROOTS];
 } __attribute__ ((__packed__));
@@ -1188,6 +1193,7 @@ enum btrfs_caching_type {
 	BTRFS_CACHE_STARTED	= 1,
 	BTRFS_CACHE_FAST	= 2,
 	BTRFS_CACHE_FINISHED	= 3,
+	BTRFS_CACHE_ERROR	= 4,
 };
 
 enum btrfs_disk_cache_state {
@@ -1302,6 +1308,7 @@ struct btrfs_fs_info {
 	struct btrfs_root *fs_root;
 	struct btrfs_root *csum_root;
 	struct btrfs_root *quota_root;
+	struct btrfs_root *uuid_root;
 
 	/* the log root tree is a directory of all the other log roots */
 	struct btrfs_root *log_root_tree;
@@ -1350,6 +1357,7 @@ struct btrfs_fs_info {
 	u64 last_trans_log_full_commit;
 	unsigned long mount_opt;
 	unsigned long compress_type:4;
+	int commit_interval;
 	/*
 	 * It is a suggestive number, the read side is safe even it gets a
 	 * wrong number because we will write out the data into a regular
@@ -1411,6 +1419,13 @@ struct btrfs_fs_info {
 	 * before jumping into the main commit.
 	 */
 	struct mutex ordered_operations_mutex;
+
+	/*
+	 * Same as ordered_operations_mutex except this is for ordered extents
+	 * and not the operations.
+	 */
+	struct mutex ordered_extent_flush_mutex;
+
 	struct rw_semaphore extent_commit_sem;
 
 	struct rw_semaphore cleanup_work_sem;
@@ -1641,6 +1656,9 @@ struct btrfs_fs_info {
 	struct btrfs_dev_replace dev_replace;
 
 	atomic_t mutually_exclusive_operation_running;
+
+	struct semaphore uuid_tree_rescan_sem;
+	unsigned int update_uuid_tree_gen:1;
 };
 
 /*
@@ -1933,6 +1951,19 @@ struct btrfs_ioctl_defrag_range_args {
  */
 #define BTRFS_DEV_REPLACE_KEY	250
 
+/*
+ * Stores items that allow to quickly map UUIDs to something else.
+ * These items are part of the filesystem UUID tree.
+ * The key is built like this:
+ * (UUID_upper_64_bits, BTRFS_UUID_KEY*, UUID_lower_64_bits).
+ */
+#if BTRFS_UUID_SIZE != 16
+#error "UUID items require BTRFS_UUID_SIZE == 16!"
+#endif
+#define BTRFS_UUID_KEY_SUBVOL	251	/* for UUIDs assigned to subvols */
+#define BTRFS_UUID_KEY_RECEIVED_SUBVOL	252	/* for UUIDs assigned to
+						 * received subvols */
+
 /*
  * string items are for debugging.  They just store a short string of
  * data in the FS
@@ -1967,6 +1998,9 @@ struct btrfs_ioctl_defrag_range_args {
 #define BTRFS_MOUNT_CHECK_INTEGRITY	(1 << 20)
 #define BTRFS_MOUNT_CHECK_INTEGRITY_INCLUDING_EXTENT_DATA (1 << 21)
 #define BTRFS_MOUNT_PANIC_ON_FATAL_ERROR	(1 << 22)
+#define BTRFS_MOUNT_RESCAN_UUID_TREE	(1 << 23)
+
+#define BTRFS_DEFAULT_COMMIT_INTERVAL	(30)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
@@ -2130,14 +2164,14 @@ BTRFS_SETGET_STACK_FUNCS(stack_device_bandwidth, struct btrfs_dev_item,
 BTRFS_SETGET_STACK_FUNCS(stack_device_generation, struct btrfs_dev_item,
 			 generation, 64);
 
-static inline char *btrfs_device_uuid(struct btrfs_dev_item *d)
+static inline unsigned long btrfs_device_uuid(struct btrfs_dev_item *d)
 {
-	return (char *)d + offsetof(struct btrfs_dev_item, uuid);
+	return (unsigned long)d + offsetof(struct btrfs_dev_item, uuid);
 }
 
-static inline char *btrfs_device_fsid(struct btrfs_dev_item *d)
+static inline unsigned long btrfs_device_fsid(struct btrfs_dev_item *d)
 {
-	return (char *)d + offsetof(struct btrfs_dev_item, fsid);
+	return (unsigned long)d + offsetof(struct btrfs_dev_item, fsid);
 }
 
 BTRFS_SETGET_FUNCS(chunk_length, struct btrfs_chunk, length, 64);
@@ -2240,6 +2274,23 @@ BTRFS_SETGET_FUNCS(inode_gid, struct btrfs_inode_item, gid, 32);
 BTRFS_SETGET_FUNCS(inode_mode, struct btrfs_inode_item, mode, 32);
 BTRFS_SETGET_FUNCS(inode_rdev, struct btrfs_inode_item, rdev, 64);
 BTRFS_SETGET_FUNCS(inode_flags, struct btrfs_inode_item, flags, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_generation, struct btrfs_inode_item,
+			 generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_sequence, struct btrfs_inode_item,
+			 sequence, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_transid, struct btrfs_inode_item,
+			 transid, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_size, struct btrfs_inode_item, size, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_nbytes, struct btrfs_inode_item,
+			 nbytes, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_block_group, struct btrfs_inode_item,
+			 block_group, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_nlink, struct btrfs_inode_item, nlink, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_uid, struct btrfs_inode_item, uid, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_gid, struct btrfs_inode_item, gid, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_mode, struct btrfs_inode_item, mode, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_rdev, struct btrfs_inode_item, rdev, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_inode_flags, struct btrfs_inode_item, flags, 64);
 
 static inline struct btrfs_timespec *
 btrfs_inode_atime(struct btrfs_inode_item *inode_item)
@@ -2267,6 +2318,8 @@ btrfs_inode_ctime(struct btrfs_inode_item *inode_item)
 
 BTRFS_SETGET_FUNCS(timespec_sec, struct btrfs_timespec, sec, 64);
 BTRFS_SETGET_FUNCS(timespec_nsec, struct btrfs_timespec, nsec, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_timespec_sec, struct btrfs_timespec, sec, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_timespec_nsec, struct btrfs_timespec, nsec, 32);
 
 /* struct btrfs_dev_extent */
 BTRFS_SETGET_FUNCS(dev_extent_chunk_tree, struct btrfs_dev_extent,
@@ -2277,10 +2330,10 @@ BTRFS_SETGET_FUNCS(dev_extent_chunk_offset, struct btrfs_dev_extent,
 		   chunk_offset, 64);
 BTRFS_SETGET_FUNCS(dev_extent_length, struct btrfs_dev_extent, length, 64);
 
-static inline u8 *btrfs_dev_extent_chunk_tree_uuid(struct btrfs_dev_extent *dev)
+static inline unsigned long btrfs_dev_extent_chunk_tree_uuid(struct btrfs_dev_extent *dev)
 {
 	unsigned long ptr = offsetof(struct btrfs_dev_extent, chunk_tree_uuid);
-	return (u8 *)((unsigned long)dev + ptr);
+	return (unsigned long)dev + ptr;
 }
 
 BTRFS_SETGET_FUNCS(extent_refs, struct btrfs_extent_item, refs, 64);
@@ -2348,6 +2401,10 @@ BTRFS_SETGET_FUNCS(ref_count_v0, struct btrfs_extent_ref_v0, count, 32);
 /* struct btrfs_node */
 BTRFS_SETGET_FUNCS(key_blockptr, struct btrfs_key_ptr, blockptr, 64);
 BTRFS_SETGET_FUNCS(key_generation, struct btrfs_key_ptr, generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_key_blockptr, struct btrfs_key_ptr,
+			 blockptr, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_key_generation, struct btrfs_key_ptr,
+			 generation, 64);
 
 static inline u64 btrfs_node_blockptr(struct extent_buffer *eb, int nr)
 {
@@ -2404,6 +2461,8 @@ static inline void btrfs_set_node_key(struct extent_buffer *eb,
 /* struct btrfs_item */
 BTRFS_SETGET_FUNCS(item_offset, struct btrfs_item, offset, 32);
 BTRFS_SETGET_FUNCS(item_size, struct btrfs_item, size, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_item_offset, struct btrfs_item, offset, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_item_size, struct btrfs_item, size, 32);
 
 static inline unsigned long btrfs_item_nr_offset(int nr)
 {
@@ -2466,6 +2525,13 @@ BTRFS_SETGET_FUNCS(dir_data_len, struct btrfs_dir_item, data_len, 16);
 BTRFS_SETGET_FUNCS(dir_type, struct btrfs_dir_item, type, 8);
 BTRFS_SETGET_FUNCS(dir_name_len, struct btrfs_dir_item, name_len, 16);
 BTRFS_SETGET_FUNCS(dir_transid, struct btrfs_dir_item, transid, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_type, struct btrfs_dir_item, type, 8);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_data_len, struct btrfs_dir_item,
+			 data_len, 16);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_name_len, struct btrfs_dir_item,
+			 name_len, 16);
+BTRFS_SETGET_STACK_FUNCS(stack_dir_transid, struct btrfs_dir_item,
+			 transid, 64);
 
 static inline void btrfs_dir_item_key(struct extent_buffer *eb,
 				      struct btrfs_dir_item *item,
@@ -2568,6 +2634,12 @@ BTRFS_SETGET_HEADER_FUNCS(header_owner, struct btrfs_header, owner, 64);
 BTRFS_SETGET_HEADER_FUNCS(header_nritems, struct btrfs_header, nritems, 32);
 BTRFS_SETGET_HEADER_FUNCS(header_flags, struct btrfs_header, flags, 64);
 BTRFS_SETGET_HEADER_FUNCS(header_level, struct btrfs_header, level, 8);
+BTRFS_SETGET_STACK_FUNCS(stack_header_generation, struct btrfs_header,
+			 generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_header_owner, struct btrfs_header, owner, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_header_nritems, struct btrfs_header,
+			 nritems, 32);
+BTRFS_SETGET_STACK_FUNCS(stack_header_bytenr, struct btrfs_header, bytenr, 64);
 
 static inline int btrfs_header_flag(struct extent_buffer *eb, u64 flag)
 {
@@ -2603,16 +2675,14 @@ static inline void btrfs_set_header_backref_rev(struct extent_buffer *eb,
 	btrfs_set_header_flags(eb, flags);
 }
 
-static inline u8 *btrfs_header_fsid(struct extent_buffer *eb)
+static inline unsigned long btrfs_header_fsid(struct extent_buffer *eb)
 {
-	unsigned long ptr = offsetof(struct btrfs_header, fsid);
-	return (u8 *)ptr;
+	return offsetof(struct btrfs_header, fsid);
 }
 
-static inline u8 *btrfs_header_chunk_tree_uuid(struct extent_buffer *eb)
+static inline unsigned long btrfs_header_chunk_tree_uuid(struct extent_buffer *eb)
 {
-	unsigned long ptr = offsetof(struct btrfs_header, chunk_tree_uuid);
-	return (u8 *)ptr;
+	return offsetof(struct btrfs_header, chunk_tree_uuid);
 }
 
 static inline int btrfs_is_leaf(struct extent_buffer *eb)
@@ -2830,6 +2900,9 @@ BTRFS_SETGET_STACK_FUNCS(super_csum_type, struct btrfs_super_block,
 			 csum_type, 16);
 BTRFS_SETGET_STACK_FUNCS(super_cache_generation, struct btrfs_super_block,
 			 cache_generation, 64);
+BTRFS_SETGET_STACK_FUNCS(super_magic, struct btrfs_super_block, magic, 64);
+BTRFS_SETGET_STACK_FUNCS(super_uuid_tree_generation, struct btrfs_super_block,
+			 uuid_tree_generation, 64);
 
 static inline int btrfs_super_csum_size(struct btrfs_super_block *s)
 {
@@ -2847,6 +2920,14 @@ static inline unsigned long btrfs_leaf_data(struct extent_buffer *l)
 
 /* struct btrfs_file_extent_item */
 BTRFS_SETGET_FUNCS(file_extent_type, struct btrfs_file_extent_item, type, 8);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_disk_bytenr,
+			 struct btrfs_file_extent_item, disk_bytenr, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_offset,
+			 struct btrfs_file_extent_item, offset, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_generation,
+			 struct btrfs_file_extent_item, generation, 64);
+BTRFS_SETGET_STACK_FUNCS(stack_file_extent_num_bytes,
+			 struct btrfs_file_extent_item, num_bytes, 64);
 
 static inline unsigned long
 btrfs_file_extent_inline_start(struct btrfs_file_extent_item *e)
@@ -3107,11 +3188,9 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   u64 root_objectid, u64 owner, u64 offset,
 				   struct btrfs_key *ins);
-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, int is_data);
+int btrfs_reserve_extent(struct btrfs_root *root, u64 num_bytes,
+			 u64 min_alloc_size, u64 empty_size, u64 hint_byte,
+			 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,
@@ -3175,7 +3254,7 @@ void btrfs_orphan_release_metadata(struct inode *inode);
 int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 				     struct btrfs_block_rsv *rsv,
 				     int nitems,
-				     u64 *qgroup_reserved);
+				     u64 *qgroup_reserved, bool use_global_rsv);
 void btrfs_subvolume_release_metadata(struct btrfs_root *root,
 				      struct btrfs_block_rsv *rsv,
 				      u64 qgroup_reserved);
@@ -3245,6 +3324,7 @@ enum btrfs_compare_tree_result {
 	BTRFS_COMPARE_TREE_NEW,
 	BTRFS_COMPARE_TREE_DELETED,
 	BTRFS_COMPARE_TREE_CHANGED,
+	BTRFS_COMPARE_TREE_SAME,
 };
 typedef int (*btrfs_changed_cb_t)(struct btrfs_root *left_root,
 				  struct btrfs_root *right_root,
@@ -3380,6 +3460,7 @@ static inline void free_fs_info(struct btrfs_fs_info *fs_info)
 	kfree(fs_info->dev_root);
 	kfree(fs_info->csum_root);
 	kfree(fs_info->quota_root);
+	kfree(fs_info->uuid_root);
 	kfree(fs_info->super_copy);
 	kfree(fs_info->super_for_commit);
 	kfree(fs_info);
@@ -3414,8 +3495,6 @@ 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 extent_buffer *eb, int slot,
-			  struct btrfs_root_item *item);
 int btrfs_find_root(struct btrfs_root *root, struct btrfs_key *search_key,
 		    struct btrfs_path *path, struct btrfs_root_item *root_item,
 		    struct btrfs_key *root_key);
@@ -3426,6 +3505,17 @@ void btrfs_check_and_init_root_item(struct btrfs_root_item *item);
 void btrfs_update_root_times(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root);
 
+/* uuid-tree.c */
+int btrfs_uuid_tree_add(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid);
+int btrfs_uuid_tree_rem(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid);
+int btrfs_uuid_tree_iterate(struct btrfs_fs_info *fs_info,
+			    int (*check_func)(struct btrfs_fs_info *, u8 *, u8,
+					      u64));
+
 /* dir-item.c */
 int btrfs_check_dir_item_collision(struct btrfs_root *root, u64 dir,
 			  const char *name, int name_len);
@@ -3509,12 +3599,14 @@ int btrfs_find_name_in_ext_backref(struct btrfs_path *path,
 				   struct btrfs_inode_extref **extref_ret);
 
 /* file-item.c */
+struct btrfs_dio_private;
 int btrfs_del_csums(struct btrfs_trans_handle *trans,
 		    struct btrfs_root *root, u64 bytenr, u64 len);
 int btrfs_lookup_bio_sums(struct btrfs_root *root, struct inode *inode,
 			  struct bio *bio, u32 *dst);
 int btrfs_lookup_bio_sums_dio(struct btrfs_root *root, struct inode *inode,
-			      struct bio *bio, u64 logical_offset);
+			      struct btrfs_dio_private *dip, struct bio *bio,
+			      u64 logical_offset);
 int btrfs_insert_file_extent(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root,
 			     u64 objectid, u64 pos,
@@ -3552,8 +3644,7 @@ void btrfs_wait_and_free_delalloc_work(struct btrfs_delalloc_work *work);
 struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *page,
 					   size_t pg_offset, u64 start, u64 len,
 					   int create);
-noinline int can_nocow_extent(struct btrfs_trans_handle *trans,
-			      struct inode *inode, u64 offset, u64 *len,
+noinline int can_nocow_extent(struct inode *inode, u64 offset, u64 *len,
 			      u64 *orig_start, u64 *orig_block_len,
 			      u64 *ram_bytes);
 
@@ -3643,11 +3734,15 @@ extern const struct dentry_operations btrfs_dentry_operations;
 long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
 void btrfs_update_iflags(struct inode *inode);
 void btrfs_inherit_iflags(struct inode *inode, struct inode *dir);
+int btrfs_is_empty_uuid(u8 *uuid);
 int btrfs_defrag_file(struct inode *inode, struct file *file,
 		      struct btrfs_ioctl_defrag_range_args *range,
 		      u64 newer_than, unsigned long max_pages);
 void btrfs_get_block_group_info(struct list_head *groups_list,
 				struct btrfs_ioctl_space_info *space);
+void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
+			       struct btrfs_ioctl_balance_args *bargs);
+
 
 /* file.c */
 int btrfs_auto_defrag_init(void);
@@ -3720,6 +3815,22 @@ void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...)
 #define btrfs_debug(fs_info, fmt, args...) \
 	btrfs_printk(fs_info, KERN_DEBUG fmt, ##args)
 
+#ifdef CONFIG_BTRFS_ASSERT
+
+static inline void assfail(char *expr, char *file, int line)
+{
+	printk(KERN_ERR "BTRFS assertion failed: %s, file: %s, line: %d",
+	       expr, file, line);
+	BUG();
+}
+
+#define ASSERT(expr)	\
+	(likely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__))
+#else
+#define ASSERT(expr)	((void)0)
+#endif
+
+#define btrfs_assert()
 __printf(5, 6)
 void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		     unsigned int line, int errno, const char *fmt, ...);

fs/btrfs/delayed-inode.c

@@ -21,6 +21,7 @@
 #include "delayed-inode.h"
 #include "disk-io.h"
 #include "transaction.h"
+#include "ctree.h"
 
 #define BTRFS_DELAYED_WRITEBACK		512
 #define BTRFS_DELAYED_BACKGROUND	128
@@ -1453,10 +1454,10 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans,
 
 	dir_item = (struct btrfs_dir_item *)delayed_item->data;
 	dir_item->location = *disk_key;
-	dir_item->transid = cpu_to_le64(trans->transid);
-	dir_item->data_len = 0;
-	dir_item->name_len = cpu_to_le16(name_len);
-	dir_item->type = type;
+	btrfs_set_stack_dir_transid(dir_item, trans->transid);
+	btrfs_set_stack_dir_data_len(dir_item, 0);
+	btrfs_set_stack_dir_name_len(dir_item, name_len);
+	btrfs_set_stack_dir_type(dir_item, type);
 	memcpy((char *)(dir_item + 1), name, name_len);
 
 	ret = btrfs_delayed_item_reserve_metadata(trans, root, delayed_item);
@@ -1470,13 +1471,11 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans,
 	mutex_lock(&delayed_node->mutex);
 	ret = __btrfs_add_delayed_insertion_item(delayed_node, delayed_item);
 	if (unlikely(ret)) {
-		printk(KERN_ERR "err add delayed dir index item(name: %s) into "
-				"the insertion tree of the delayed node"
+		printk(KERN_ERR "err add delayed dir index item(name: %.*s) "
+				"into the insertion tree of the delayed node"
 				"(root id: %llu, inode id: %llu, errno: %d)\n",
-				name,
-				(unsigned long long)delayed_node->root->objectid,
-				(unsigned long long)delayed_node->inode_id,
-				ret);
+				name_len, name, delayed_node->root->objectid,
+				delayed_node->inode_id, ret);
 		BUG();
 	}
 	mutex_unlock(&delayed_node->mutex);
@@ -1547,9 +1546,7 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans,
 		printk(KERN_ERR "err add delayed dir index item(index: %llu) "
 				"into the deletion tree of the delayed node"
 				"(root id: %llu, inode id: %llu, errno: %d)\n",
-				(unsigned long long)index,
-				(unsigned long long)node->root->objectid,
-				(unsigned long long)node->inode_id,
+				index, node->root->objectid, node->inode_id,
 				ret);
 		BUG();
 	}
@@ -1699,7 +1696,7 @@ int btrfs_readdir_delayed_dir_index(struct dir_context *ctx,
 
 		di = (struct btrfs_dir_item *)curr->data;
 		name = (char *)(di + 1);
-		name_len = le16_to_cpu(di->name_len);
+		name_len = btrfs_stack_dir_name_len(di);
 
 		d_type = btrfs_filetype_table[di->type];
 		btrfs_disk_key_to_cpu(&location, &di->location);
@@ -1716,27 +1713,6 @@ int btrfs_readdir_delayed_dir_index(struct dir_context *ctx,
 	return 0;
 }
 
-BTRFS_SETGET_STACK_FUNCS(stack_inode_generation, struct btrfs_inode_item,
-			 generation, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_sequence, struct btrfs_inode_item,
-			 sequence, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_transid, struct btrfs_inode_item,
-			 transid, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_size, struct btrfs_inode_item, size, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_nbytes, struct btrfs_inode_item,
-			 nbytes, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_block_group, struct btrfs_inode_item,
-			 block_group, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_nlink, struct btrfs_inode_item, nlink, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_uid, struct btrfs_inode_item, uid, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_gid, struct btrfs_inode_item, gid, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_mode, struct btrfs_inode_item, mode, 32);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_rdev, struct btrfs_inode_item, rdev, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_inode_flags, struct btrfs_inode_item, flags, 64);
-
-BTRFS_SETGET_STACK_FUNCS(stack_timespec_sec, struct btrfs_timespec, sec, 64);
-BTRFS_SETGET_STACK_FUNCS(stack_timespec_nsec, struct btrfs_timespec, nsec, 32);
-
 static void fill_stack_inode_item(struct btrfs_trans_handle *trans,
 				  struct btrfs_inode_item *inode_item,
 				  struct inode *inode)

fs/btrfs/delayed-ref.c

@@ -241,7 +241,7 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-static void inline drop_delayed_ref(struct btrfs_trans_handle *trans,
+static inline void drop_delayed_ref(struct btrfs_trans_handle *trans,
 				    struct btrfs_delayed_ref_root *delayed_refs,
 				    struct btrfs_delayed_ref_node *ref)
 {
@@ -600,7 +600,7 @@ static noinline void add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	INIT_LIST_HEAD(&head_ref->cluster);
 	mutex_init(&head_ref->mutex);
 
-	trace_btrfs_delayed_ref_head(ref, head_ref, action);
+	trace_add_delayed_ref_head(ref, head_ref, action);
 
 	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
 
@@ -661,7 +661,7 @@ static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 		ref->type = BTRFS_TREE_BLOCK_REF_KEY;
 	full_ref->level = level;
 
-	trace_btrfs_delayed_tree_ref(ref, full_ref, action);
+	trace_add_delayed_tree_ref(ref, full_ref, action);
 
 	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
 
@@ -722,7 +722,7 @@ static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	full_ref->objectid = owner;
 	full_ref->offset = offset;
 
-	trace_btrfs_delayed_data_ref(ref, full_ref, action);
+	trace_add_delayed_data_ref(ref, full_ref, action);
 
 	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
 

fs/btrfs/dev-replace.c

@@ -148,13 +148,13 @@ no_valid_dev_replace_entry_found:
 		    !btrfs_test_opt(dev_root, DEGRADED)) {
 			ret = -EIO;
 			pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "srcdev (devid %llu) is missing, need to run 'btrfs dev scan'?\n",
-				(unsigned long long)src_devid);
+				src_devid);
 		}
 		if (!dev_replace->tgtdev &&
 		    !btrfs_test_opt(dev_root, DEGRADED)) {
 			ret = -EIO;
 			pr_warn("btrfs: cannot mount because device replace operation is ongoing and\n" "tgtdev (devid %llu) is missing, need to run btrfs dev scan?\n",
-				(unsigned long long)BTRFS_DEV_REPLACE_DEVID);
+				BTRFS_DEV_REPLACE_DEVID);
 		}
 		if (dev_replace->tgtdev) {
 			if (dev_replace->srcdev) {

fs/btrfs/disk-io.c

@@ -31,6 +31,7 @@
 #include <linux/migrate.h>
 #include <linux/ratelimit.h>
 #include <linux/uuid.h>
+#include <linux/semaphore.h>
 #include <asm/unaligned.h>
 #include "compat.h"
 #include "ctree.h"
@@ -302,9 +303,8 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
 			printk_ratelimited(KERN_INFO "btrfs: %s checksum verify "
 				       "failed on %llu wanted %X found %X "
 				       "level %d\n",
-				       root->fs_info->sb->s_id,
-				       (unsigned long long)buf->start, val, found,
-				       btrfs_header_level(buf));
+				       root->fs_info->sb->s_id, buf->start,
+				       val, found, btrfs_header_level(buf));
 			if (result != (char *)&inline_result)
 				kfree(result);
 			return 1;
@@ -345,9 +345,7 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 	}
 	printk_ratelimited("parent transid verify failed on %llu wanted %llu "
 		       "found %llu\n",
-		       (unsigned long long)eb->start,
-		       (unsigned long long)parent_transid,
-		       (unsigned long long)btrfs_header_generation(eb));
+		       eb->start, parent_transid, btrfs_header_generation(eb));
 	ret = 1;
 	clear_extent_buffer_uptodate(eb);
 out:
@@ -497,8 +495,7 @@ static int check_tree_block_fsid(struct btrfs_root *root,
 	u8 fsid[BTRFS_UUID_SIZE];
 	int ret = 1;
 
-	read_extent_buffer(eb, fsid, (unsigned long)btrfs_header_fsid(eb),
-			   BTRFS_FSID_SIZE);
+	read_extent_buffer(eb, fsid, btrfs_header_fsid(eb), BTRFS_FSID_SIZE);
 	while (fs_devices) {
 		if (!memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE)) {
 			ret = 0;
@@ -512,8 +509,7 @@ static int check_tree_block_fsid(struct btrfs_root *root,
 #define CORRUPT(reason, eb, root, slot)				\
 	printk(KERN_CRIT "btrfs: corrupt leaf, %s: block=%llu,"	\
 	       "root=%llu, slot=%d\n", reason,			\
-	       (unsigned long long)btrfs_header_bytenr(eb),	\
-	       (unsigned long long)root->objectid, slot)
+	       btrfs_header_bytenr(eb),	root->objectid, slot)
 
 static noinline int check_leaf(struct btrfs_root *root,
 			       struct extent_buffer *leaf)
@@ -576,8 +572,9 @@ static noinline int check_leaf(struct btrfs_root *root,
 	return 0;
 }
 
-static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
-			       struct extent_state *state, int mirror)
+static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
+				      u64 phy_offset, struct page *page,
+				      u64 start, u64 end, int mirror)
 {
 	struct extent_io_tree *tree;
 	u64 found_start;
@@ -612,14 +609,13 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 	if (found_start != eb->start) {
 		printk_ratelimited(KERN_INFO "btrfs bad tree block start "
 			       "%llu %llu\n",
-			       (unsigned long long)found_start,
-			       (unsigned long long)eb->start);
+			       found_start, eb->start);
 		ret = -EIO;
 		goto err;
 	}
 	if (check_tree_block_fsid(root, eb)) {
 		printk_ratelimited(KERN_INFO "btrfs bad fsid on block %llu\n",
-			       (unsigned long long)eb->start);
+			       eb->start);
 		ret = -EIO;
 		goto err;
 	}
@@ -1148,6 +1144,10 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
 		return NULL;
 
 	ret = btree_read_extent_buffer_pages(root, buf, 0, parent_transid);
+	if (ret) {
+		free_extent_buffer(buf);
+		return NULL;
+	}
 	return buf;
 
 }
@@ -1291,11 +1291,10 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans,
 	btrfs_set_header_owner(leaf, objectid);
 	root->node = leaf;
 
-	write_extent_buffer(leaf, fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(leaf),
+	write_extent_buffer(leaf, fs_info->fsid, btrfs_header_fsid(leaf),
 			    BTRFS_FSID_SIZE);
 	write_extent_buffer(leaf, fs_info->chunk_tree_uuid,
-			    (unsigned long)btrfs_header_chunk_tree_uuid(leaf),
+			    btrfs_header_chunk_tree_uuid(leaf),
 			    BTRFS_UUID_SIZE);
 	btrfs_mark_buffer_dirty(leaf);
 
@@ -1379,8 +1378,7 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 	root->node = leaf;
 
 	write_extent_buffer(root->node, root->fs_info->fsid,
-			    (unsigned long)btrfs_header_fsid(root->node),
-			    BTRFS_FSID_SIZE);
+			    btrfs_header_fsid(root->node), BTRFS_FSID_SIZE);
 	btrfs_mark_buffer_dirty(root->node);
 	btrfs_tree_unlock(root->node);
 	return root;
@@ -1413,11 +1411,11 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 	log_root->root_key.offset = root->root_key.objectid;
 
 	inode_item = &log_root->root_item.inode;
-	inode_item->generation = cpu_to_le64(1);
-	inode_item->size = cpu_to_le64(3);
-	inode_item->nlink = cpu_to_le32(1);
-	inode_item->nbytes = cpu_to_le64(root->leafsize);
-	inode_item->mode = cpu_to_le32(S_IFDIR | 0755);
+	btrfs_set_stack_inode_generation(inode_item, 1);
+	btrfs_set_stack_inode_size(inode_item, 3);
+	btrfs_set_stack_inode_nlink(inode_item, 1);
+	btrfs_set_stack_inode_nbytes(inode_item, root->leafsize);
+	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
 
 	btrfs_set_root_node(&log_root->root_item, log_root->node);
 
@@ -1428,8 +1426,8 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
-					struct btrfs_key *key)
+static struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root,
+					       struct btrfs_key *key)
 {
 	struct btrfs_root *root;
 	struct btrfs_fs_info *fs_info = tree_root->fs_info;
@@ -1529,8 +1527,8 @@ fail:
 	return ret;
 }
 
-struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
-					u64 root_id)
+static struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
+					       u64 root_id)
 {
 	struct btrfs_root *root;
 
@@ -1581,10 +1579,16 @@ struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info,
 	if (location->objectid == BTRFS_QUOTA_TREE_OBJECTID)
 		return fs_info->quota_root ? fs_info->quota_root :
 					     ERR_PTR(-ENOENT);
+	if (location->objectid == BTRFS_UUID_TREE_OBJECTID)
+		return fs_info->uuid_root ? fs_info->uuid_root :
+					    ERR_PTR(-ENOENT);
 again:
 	root = btrfs_lookup_fs_root(fs_info, location->objectid);
-	if (root)
+	if (root) {
+		if (btrfs_root_refs(&root->root_item) == 0)
+			return ERR_PTR(-ENOENT);
 		return root;
+	}
 
 	root = btrfs_read_fs_root(fs_info->tree_root, location);
 	if (IS_ERR(root))
@@ -1737,7 +1741,7 @@ static int transaction_kthread(void *arg)
 
 	do {
 		cannot_commit = false;
-		delay = HZ * 30;
+		delay = HZ * root->fs_info->commit_interval;
 		mutex_lock(&root->fs_info->transaction_kthread_mutex);
 
 		spin_lock(&root->fs_info->trans_lock);
@@ -1749,7 +1753,8 @@ static int transaction_kthread(void *arg)
 
 		now = get_seconds();
 		if (cur->state < TRANS_STATE_BLOCKED &&
-		    (now < cur->start_time || now - cur->start_time < 30)) {
+		    (now < cur->start_time ||
+		     now - cur->start_time < root->fs_info->commit_interval)) {
 			spin_unlock(&root->fs_info->trans_lock);
 			delay = HZ * 5;
 			goto sleep;
@@ -2038,6 +2043,12 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
 		info->quota_root->node = NULL;
 		info->quota_root->commit_root = NULL;
 	}
+	if (info->uuid_root) {
+		free_extent_buffer(info->uuid_root->node);
+		free_extent_buffer(info->uuid_root->commit_root);
+		info->uuid_root->node = NULL;
+		info->uuid_root->commit_root = NULL;
+	}
 	if (chunk_root) {
 		free_extent_buffer(info->chunk_root->node);
 		free_extent_buffer(info->chunk_root->commit_root);
@@ -2098,11 +2109,14 @@ int open_ctree(struct super_block *sb,
 	struct btrfs_root *chunk_root;
 	struct btrfs_root *dev_root;
 	struct btrfs_root *quota_root;
+	struct btrfs_root *uuid_root;
 	struct btrfs_root *log_tree_root;
 	int ret;
 	int err = -EINVAL;
 	int num_backups_tried = 0;
 	int backup_index = 0;
+	bool create_uuid_tree;
+	bool check_uuid_tree;
 
 	tree_root = fs_info->tree_root = btrfs_alloc_root(fs_info);
 	chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info);
@@ -2189,6 +2203,7 @@ int open_ctree(struct super_block *sb,
 	fs_info->defrag_inodes = RB_ROOT;
 	fs_info->free_chunk_space = 0;
 	fs_info->tree_mod_log = RB_ROOT;
+	fs_info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
 
 	/* readahead state */
 	INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT);
@@ -2270,6 +2285,7 @@ int open_ctree(struct super_block *sb,
 
 
 	mutex_init(&fs_info->ordered_operations_mutex);
+	mutex_init(&fs_info->ordered_extent_flush_mutex);
 	mutex_init(&fs_info->tree_log_mutex);
 	mutex_init(&fs_info->chunk_mutex);
 	mutex_init(&fs_info->transaction_kthread_mutex);
@@ -2278,6 +2294,7 @@ int open_ctree(struct super_block *sb,
 	init_rwsem(&fs_info->extent_commit_sem);
 	init_rwsem(&fs_info->cleanup_work_sem);
 	init_rwsem(&fs_info->subvol_sem);
+	sema_init(&fs_info->uuid_tree_rescan_sem, 1);
 	fs_info->dev_replace.lock_owner = 0;
 	atomic_set(&fs_info->dev_replace.nesting_level, 0);
 	mutex_init(&fs_info->dev_replace.lock_finishing_cancel_unmount);
@@ -2383,7 +2400,7 @@ int open_ctree(struct super_block *sb,
 	if (features) {
 		printk(KERN_ERR "BTRFS: couldn't mount because of "
 		       "unsupported optional features (%Lx).\n",
-		       (unsigned long long)features);
+		       features);
 		err = -EINVAL;
 		goto fail_alloc;
 	}
@@ -2453,7 +2470,7 @@ int open_ctree(struct super_block *sb,
 	if (!(sb->s_flags & MS_RDONLY) && features) {
 		printk(KERN_ERR "BTRFS: couldn't mount RDWR because of "
 		       "unsupported option features (%Lx).\n",
-		       (unsigned long long)features);
+		       features);
 		err = -EINVAL;
 		goto fail_alloc;
 	}
@@ -2466,20 +2483,17 @@ int open_ctree(struct super_block *sb,
 			   &fs_info->generic_worker);
 
 	btrfs_init_workers(&fs_info->delalloc_workers, "delalloc",
-			   fs_info->thread_pool_size,
-			   &fs_info->generic_worker);
+			   fs_info->thread_pool_size, NULL);
 
 	btrfs_init_workers(&fs_info->flush_workers, "flush_delalloc",
-			   fs_info->thread_pool_size,
-			   &fs_info->generic_worker);
+			   fs_info->thread_pool_size, NULL);
 
 	btrfs_init_workers(&fs_info->submit_workers, "submit",
 			   min_t(u64, fs_devices->num_devices,
-			   fs_info->thread_pool_size),
-			   &fs_info->generic_worker);
+			   fs_info->thread_pool_size), NULL);
 
 	btrfs_init_workers(&fs_info->caching_workers, "cache",
-			   2, &fs_info->generic_worker);
+			   fs_info->thread_pool_size, NULL);
 
 	/* a higher idle thresh on the submit workers makes it much more
 	 * likely that bios will be send down in a sane order to the
@@ -2575,7 +2589,7 @@ int open_ctree(struct super_block *sb,
 	sb->s_blocksize = sectorsize;
 	sb->s_blocksize_bits = blksize_bits(sectorsize);
 
-	if (disk_super->magic != cpu_to_le64(BTRFS_MAGIC)) {
+	if (btrfs_super_magic(disk_super) != BTRFS_MAGIC) {
 		printk(KERN_INFO "btrfs: valid FS not found on %s\n", sb->s_id);
 		goto fail_sb_buffer;
 	}
@@ -2615,8 +2629,7 @@ int open_ctree(struct super_block *sb,
 	chunk_root->commit_root = btrfs_root_node(chunk_root);
 
 	read_extent_buffer(chunk_root->node, fs_info->chunk_tree_uuid,
-	   (unsigned long)btrfs_header_chunk_tree_uuid(chunk_root->node),
-	   BTRFS_UUID_SIZE);
+	   btrfs_header_chunk_tree_uuid(chunk_root->node), BTRFS_UUID_SIZE);
 
 	ret = btrfs_read_chunk_tree(chunk_root);
 	if (ret) {
@@ -2696,6 +2709,22 @@ retry_root_backup:
 		fs_info->quota_root = quota_root;
 	}
 
+	location.objectid = BTRFS_UUID_TREE_OBJECTID;
+	uuid_root = btrfs_read_tree_root(tree_root, &location);
+	if (IS_ERR(uuid_root)) {
+		ret = PTR_ERR(uuid_root);
+		if (ret != -ENOENT)
+			goto recovery_tree_root;
+		create_uuid_tree = true;
+		check_uuid_tree = false;
+	} else {
+		uuid_root->track_dirty = 1;
+		fs_info->uuid_root = uuid_root;
+		create_uuid_tree = false;
+		check_uuid_tree =
+		    generation != btrfs_super_uuid_tree_generation(disk_super);
+	}
+
 	fs_info->generation = generation;
 	fs_info->last_trans_committed = generation;
 
@@ -2882,6 +2911,29 @@ retry_root_backup:
 
 	btrfs_qgroup_rescan_resume(fs_info);
 
+	if (create_uuid_tree) {
+		pr_info("btrfs: creating UUID tree\n");
+		ret = btrfs_create_uuid_tree(fs_info);
+		if (ret) {
+			pr_warn("btrfs: failed to create the UUID tree %d\n",
+				ret);
+			close_ctree(tree_root);
+			return ret;
+		}
+	} else if (check_uuid_tree ||
+		   btrfs_test_opt(tree_root, RESCAN_UUID_TREE)) {
+		pr_info("btrfs: checking UUID tree\n");
+		ret = btrfs_check_uuid_tree(fs_info);
+		if (ret) {
+			pr_warn("btrfs: failed to check the UUID tree %d\n",
+				ret);
+			close_ctree(tree_root);
+			return ret;
+		}
+	} else {
+		fs_info->update_uuid_tree_gen = 1;
+	}
+
 	return 0;
 
 fail_qgroup:
@@ -2983,15 +3035,17 @@ struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
 	 */
 	for (i = 0; i < 1; i++) {
 		bytenr = btrfs_sb_offset(i);
-		if (bytenr + 4096 >= i_size_read(bdev->bd_inode))
+		if (bytenr + BTRFS_SUPER_INFO_SIZE >=
+					i_size_read(bdev->bd_inode))
 			break;
-		bh = __bread(bdev, bytenr / 4096, 4096);
+		bh = __bread(bdev, bytenr / 4096,
+					BTRFS_SUPER_INFO_SIZE);
 		if (!bh)
 			continue;
 
 		super = (struct btrfs_super_block *)bh->b_data;
 		if (btrfs_super_bytenr(super) != bytenr ||
-		    super->magic != cpu_to_le64(BTRFS_MAGIC)) {
+		    btrfs_super_magic(super) != BTRFS_MAGIC) {
 			brelse(bh);
 			continue;
 		}
@@ -3311,7 +3365,6 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 	int total_errors = 0;
 	u64 flags;
 
-	max_errors = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
 	do_barriers = !btrfs_test_opt(root, NOBARRIER);
 	backup_super_roots(root->fs_info);
 
@@ -3320,6 +3373,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	head = &root->fs_info->fs_devices->devices;
+	max_errors = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
 
 	if (do_barriers) {
 		ret = barrier_all_devices(root->fs_info);
@@ -3362,8 +3416,10 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 		printk(KERN_ERR "btrfs: %d errors while writing supers\n",
 		       total_errors);
 
-		/* This shouldn't happen. FUA is masked off if unsupported */
-		BUG();
+		/* FUA is masked off if unsupported and can't be the reason */
+		btrfs_error(root->fs_info, -EIO,
+			    "%d errors while writing supers", total_errors);
+		return -EIO;
 	}
 
 	total_errors = 0;
@@ -3421,6 +3477,8 @@ static void free_fs_root(struct btrfs_root *root)
 {
 	iput(root->cache_inode);
 	WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
+	btrfs_free_block_rsv(root, root->orphan_block_rsv);
+	root->orphan_block_rsv = NULL;
 	if (root->anon_dev)
 		free_anon_bdev(root->anon_dev);
 	free_extent_buffer(root->node);
@@ -3510,6 +3568,11 @@ int close_ctree(struct btrfs_root *root)
 	fs_info->closing = 1;
 	smp_mb();
 
+	/* wait for the uuid_scan task to finish */
+	down(&fs_info->uuid_tree_rescan_sem);
+	/* avoid complains from lockdep et al., set sem back to initial state */
+	up(&fs_info->uuid_tree_rescan_sem);
+
 	/* pause restriper - we want to resume on mount */
 	btrfs_pause_balance(fs_info);
 
@@ -3573,6 +3636,9 @@ int close_ctree(struct btrfs_root *root)
 
 	btrfs_free_stripe_hash_table(fs_info);
 
+	btrfs_free_block_rsv(root, root->orphan_block_rsv);
+	root->orphan_block_rsv = NULL;
+
 	return 0;
 }
 
@@ -3608,9 +3674,7 @@ void btrfs_mark_buffer_dirty(struct extent_buffer *buf)
 	if (transid != root->fs_info->generation)
 		WARN(1, KERN_CRIT "btrfs transid mismatch buffer %llu, "
 		       "found %llu running %llu\n",
-			(unsigned long long)buf->start,
-			(unsigned long long)transid,
-			(unsigned long long)root->fs_info->generation);
+			buf->start, transid, root->fs_info->generation);
 	was_dirty = set_extent_buffer_dirty(buf);
 	if (!was_dirty)
 		__percpu_counter_add(&root->fs_info->dirty_metadata_bytes,
@@ -3744,8 +3808,8 @@ static void btrfs_destroy_all_ordered_extents(struct btrfs_fs_info *fs_info)
 	spin_unlock(&fs_info->ordered_root_lock);
 }
 
-int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
-			       struct btrfs_root *root)
+static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
+				      struct btrfs_root *root)
 {
 	struct rb_node *node;
 	struct btrfs_delayed_ref_root *delayed_refs;

fs/btrfs/extent-tree.c

@@ -113,7 +113,8 @@ static noinline int
 block_group_cache_done(struct btrfs_block_group_cache *cache)
 {
 	smp_mb();
-	return cache->cached == BTRFS_CACHE_FINISHED;
+	return cache->cached == BTRFS_CACHE_FINISHED ||
+		cache->cached == BTRFS_CACHE_ERROR;
 }
 
 static int block_group_bits(struct btrfs_block_group_cache *cache, u64 bits)
@@ -389,7 +390,7 @@ static noinline void caching_thread(struct btrfs_work *work)
 	u64 total_found = 0;
 	u64 last = 0;
 	u32 nritems;
-	int ret = 0;
+	int ret = -ENOMEM;
 
 	caching_ctl = container_of(work, struct btrfs_caching_control, work);
 	block_group = caching_ctl->block_group;
@@ -420,6 +421,7 @@ again:
 	/* need to make sure the commit_root doesn't disappear */
 	down_read(&fs_info->extent_commit_sem);
 
+next:
 	ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
 	if (ret < 0)
 		goto err;
@@ -459,6 +461,16 @@ again:
 			continue;
 		}
 
+		if (key.objectid < last) {
+			key.objectid = last;
+			key.offset = 0;
+			key.type = BTRFS_EXTENT_ITEM_KEY;
+
+			caching_ctl->progress = last;
+			btrfs_release_path(path);
+			goto next;
+		}
+
 		if (key.objectid < block_group->key.objectid) {
 			path->slots[0]++;
 			continue;
@@ -506,6 +518,12 @@ err:
 
 	mutex_unlock(&caching_ctl->mutex);
 out:
+	if (ret) {
+		spin_lock(&block_group->lock);
+		block_group->caching_ctl = NULL;
+		block_group->cached = BTRFS_CACHE_ERROR;
+		spin_unlock(&block_group->lock);
+	}
 	wake_up(&caching_ctl->wait);
 
 	put_caching_control(caching_ctl);
@@ -771,10 +789,23 @@ again:
 		goto out_free;
 
 	if (ret > 0 && metadata && key.type == BTRFS_METADATA_ITEM_KEY) {
-		key.type = BTRFS_EXTENT_ITEM_KEY;
-		key.offset = root->leafsize;
-		btrfs_release_path(path);
-		goto again;
+		metadata = 0;
+		if (path->slots[0]) {
+			path->slots[0]--;
+			btrfs_item_key_to_cpu(path->nodes[0], &key,
+					      path->slots[0]);
+			if (key.objectid == bytenr &&
+			    key.type == BTRFS_EXTENT_ITEM_KEY &&
+			    key.offset == root->leafsize)
+				ret = 0;
+		}
+		if (ret) {
+			key.objectid = bytenr;
+			key.type = BTRFS_EXTENT_ITEM_KEY;
+			key.offset = root->leafsize;
+			btrfs_release_path(path);
+			goto again;
+		}
 	}
 
 	if (ret == 0) {
@@ -2011,6 +2042,8 @@ static int run_delayed_data_ref(struct btrfs_trans_handle *trans,
 	ins.type = BTRFS_EXTENT_ITEM_KEY;
 
 	ref = btrfs_delayed_node_to_data_ref(node);
+	trace_run_delayed_data_ref(node, ref, node->action);
+
 	if (node->type == BTRFS_SHARED_DATA_REF_KEY)
 		parent = ref->parent;
 	else
@@ -2154,6 +2187,8 @@ static int run_delayed_tree_ref(struct btrfs_trans_handle *trans,
 						 SKINNY_METADATA);
 
 	ref = btrfs_delayed_node_to_tree_ref(node);
+	trace_run_delayed_tree_ref(node, ref, node->action);
+
 	if (node->type == BTRFS_SHARED_BLOCK_REF_KEY)
 		parent = ref->parent;
 	else
@@ -2212,6 +2247,8 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
 		 */
 		BUG_ON(extent_op);
 		head = btrfs_delayed_node_to_head(node);
+		trace_run_delayed_ref_head(node, head, node->action);
+
 		if (insert_reserved) {
 			btrfs_pin_extent(root, node->bytenr,
 					 node->num_bytes, 1);
@@ -2403,6 +2440,8 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 			default:
 				WARN_ON(1);
 			}
+		} else {
+			list_del_init(&locked_ref->cluster);
 		}
 		spin_unlock(&delayed_refs->lock);
 
@@ -2425,7 +2464,6 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
 		 * list before we release it.
 		 */
 		if (btrfs_delayed_ref_is_head(ref)) {
-			list_del_init(&locked_ref->cluster);
 			btrfs_delayed_ref_unlock(locked_ref);
 			locked_ref = NULL;
 		}
@@ -3799,8 +3837,12 @@ again:
 	if (force < space_info->force_alloc)
 		force = space_info->force_alloc;
 	if (space_info->full) {
+		if (should_alloc_chunk(extent_root, space_info, force))
+			ret = -ENOSPC;
+		else
+			ret = 0;
 		spin_unlock(&space_info->lock);
-		return 0;
+		return ret;
 	}
 
 	if (!should_alloc_chunk(extent_root, space_info, force)) {
@@ -4320,6 +4362,9 @@ static struct btrfs_block_rsv *get_block_rsv(
 	if (root == root->fs_info->csum_root && trans->adding_csums)
 		block_rsv = trans->block_rsv;
 
+	if (root == root->fs_info->uuid_root)
+		block_rsv = trans->block_rsv;
+
 	if (!block_rsv)
 		block_rsv = root->block_rsv;
 
@@ -4729,10 +4774,12 @@ void btrfs_orphan_release_metadata(struct inode *inode)
 int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 				     struct btrfs_block_rsv *rsv,
 				     int items,
-				     u64 *qgroup_reserved)
+				     u64 *qgroup_reserved,
+				     bool use_global_rsv)
 {
 	u64 num_bytes;
 	int ret;
+	struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv;
 
 	if (root->fs_info->quota_enabled) {
 		/* One for parent inode, two for dir entries */
@@ -4751,6 +4798,10 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 					    BTRFS_BLOCK_GROUP_METADATA);
 	ret = btrfs_block_rsv_add(root, rsv, num_bytes,
 				  BTRFS_RESERVE_FLUSH_ALL);
+
+	if (ret == -ENOSPC && use_global_rsv)
+		ret = btrfs_block_rsv_migrate(global_rsv, rsv, num_bytes);
+
 	if (ret) {
 		if (*qgroup_reserved)
 			btrfs_qgroup_free(root, *qgroup_reserved);
@@ -5668,7 +5719,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 
 			if (ret) {
 				btrfs_err(info, "umm, got %d back from search, was looking for %llu",
-					ret, (unsigned long long)bytenr);
+					ret, bytenr);
 				if (ret > 0)
 					btrfs_print_leaf(extent_root,
 							 path->nodes[0]);
@@ -5684,11 +5735,8 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 		WARN_ON(1);
 		btrfs_err(info,
 			"unable to find ref byte nr %llu parent %llu root %llu  owner %llu offset %llu",
-			(unsigned long long)bytenr,
-			(unsigned long long)parent,
-			(unsigned long long)root_objectid,
-			(unsigned long long)owner_objectid,
-			(unsigned long long)owner_offset);
+			bytenr, parent, root_objectid, owner_objectid,
+			owner_offset);
 	} else {
 		btrfs_abort_transaction(trans, extent_root, ret);
 		goto out;
@@ -5717,7 +5765,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 					-1, 1);
 		if (ret) {
 			btrfs_err(info, "umm, got %d back from search, was looking for %llu",
-				ret, (unsigned long long)bytenr);
+				ret, bytenr);
 			btrfs_print_leaf(extent_root, path->nodes[0]);
 		}
 		if (ret < 0) {
@@ -5999,8 +6047,11 @@ static u64 stripe_align(struct btrfs_root *root,
  * for our min num_bytes.  Another option is to have it go ahead
  * and look in the rbtree for a free extent of a given size, but this
  * is a good start.
+ *
+ * Callers of this must check if cache->cached == BTRFS_CACHE_ERROR before using
+ * any of the information in this block group.
  */
-static noinline int
+static noinline void
 wait_block_group_cache_progress(struct btrfs_block_group_cache *cache,
 				u64 num_bytes)
 {
@@ -6008,28 +6059,29 @@ wait_block_group_cache_progress(struct btrfs_block_group_cache *cache,
 
 	caching_ctl = get_caching_control(cache);
 	if (!caching_ctl)
-		return 0;
+		return;
 
 	wait_event(caching_ctl->wait, block_group_cache_done(cache) ||
 		   (cache->free_space_ctl->free_space >= num_bytes));
 
 	put_caching_control(caching_ctl);
-	return 0;
 }
 
 static noinline int
 wait_block_group_cache_done(struct btrfs_block_group_cache *cache)
 {
 	struct btrfs_caching_control *caching_ctl;
+	int ret = 0;
 
 	caching_ctl = get_caching_control(cache);
 	if (!caching_ctl)
-		return 0;
+		return (cache->cached == BTRFS_CACHE_ERROR) ? -EIO : 0;
 
 	wait_event(caching_ctl->wait, block_group_cache_done(cache));
-
+	if (cache->cached == BTRFS_CACHE_ERROR)
+		ret = -EIO;
 	put_caching_control(caching_ctl);
-	return 0;
+	return ret;
 }
 
 int __get_raid_index(u64 flags)
@@ -6070,8 +6122,7 @@ enum btrfs_loop_type {
  * ins->offset == number of blocks
  * Any available blocks before search_start are skipped.
  */
-static noinline int find_free_extent(struct btrfs_trans_handle *trans,
-				     struct btrfs_root *orig_root,
+static noinline int find_free_extent(struct btrfs_root *orig_root,
 				     u64 num_bytes, u64 empty_size,
 				     u64 hint_byte, struct btrfs_key *ins,
 				     u64 flags)
@@ -6212,6 +6263,8 @@ have_block_group:
 			ret = 0;
 		}
 
+		if (unlikely(block_group->cached == BTRFS_CACHE_ERROR))
+			goto loop;
 		if (unlikely(block_group->ro))
 			goto loop;
 
@@ -6292,10 +6345,10 @@ refill_cluster:
 					      block_group->full_stripe_len);
 
 			/* allocate a cluster in this block group */
-			ret = btrfs_find_space_cluster(trans, root,
-					       block_group, last_ptr,
-					       search_start, num_bytes,
-					       aligned_cluster);
+			ret = btrfs_find_space_cluster(root, block_group,
+						       last_ptr, search_start,
+						       num_bytes,
+						       aligned_cluster);
 			if (ret == 0) {
 				/*
 				 * now pull our allocation out of this
@@ -6426,17 +6479,28 @@ loop:
 		index = 0;
 		loop++;
 		if (loop == LOOP_ALLOC_CHUNK) {
+			struct btrfs_trans_handle *trans;
+
+			trans = btrfs_join_transaction(root);
+			if (IS_ERR(trans)) {
+				ret = PTR_ERR(trans);
+				goto out;
+			}
+
 			ret = do_chunk_alloc(trans, root, flags,
 					     CHUNK_ALLOC_FORCE);
 			/*
 			 * Do not bail out on ENOSPC since we
 			 * can do more things.
 			 */
-			if (ret < 0 && ret != -ENOSPC) {
+			if (ret < 0 && ret != -ENOSPC)
 				btrfs_abort_transaction(trans,
 							root, ret);
+			else
+				ret = 0;
+			btrfs_end_transaction(trans, root);
+			if (ret)
 				goto out;
-			}
 		}
 
 		if (loop == LOOP_NO_EMPTY_SIZE) {
@@ -6463,19 +6527,15 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes,
 
 	spin_lock(&info->lock);
 	printk(KERN_INFO "space_info %llu has %llu free, is %sfull\n",
-	       (unsigned long long)info->flags,
-	       (unsigned long long)(info->total_bytes - info->bytes_used -
-				    info->bytes_pinned - info->bytes_reserved -
-				    info->bytes_readonly),
+	       info->flags,
+	       info->total_bytes - info->bytes_used - info->bytes_pinned -
+	       info->bytes_reserved - info->bytes_readonly,
 	       (info->full) ? "" : "not ");
 	printk(KERN_INFO "space_info total=%llu, used=%llu, pinned=%llu, "
 	       "reserved=%llu, may_use=%llu, readonly=%llu\n",
-	       (unsigned long long)info->total_bytes,
-	       (unsigned long long)info->bytes_used,
-	       (unsigned long long)info->bytes_pinned,
-	       (unsigned long long)info->bytes_reserved,
-	       (unsigned long long)info->bytes_may_use,
-	       (unsigned long long)info->bytes_readonly);
+	       info->total_bytes, info->bytes_used, info->bytes_pinned,
+	       info->bytes_reserved, info->bytes_may_use,
+	       info->bytes_readonly);
 	spin_unlock(&info->lock);
 
 	if (!dump_block_groups)
@@ -6486,12 +6546,9 @@ again:
 	list_for_each_entry(cache, &info->block_groups[index], list) {
 		spin_lock(&cache->lock);
 		printk(KERN_INFO "block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %s\n",
-		       (unsigned long long)cache->key.objectid,
-		       (unsigned long long)cache->key.offset,
-		       (unsigned long long)btrfs_block_group_used(&cache->item),
-		       (unsigned long long)cache->pinned,
-		       (unsigned long long)cache->reserved,
-		       cache->ro ? "[readonly]" : "");
+		       cache->key.objectid, cache->key.offset,
+		       btrfs_block_group_used(&cache->item), cache->pinned,
+		       cache->reserved, cache->ro ? "[readonly]" : "");
 		btrfs_dump_free_space(cache, bytes);
 		spin_unlock(&cache->lock);
 	}
@@ -6500,8 +6557,7 @@ again:
 	up_read(&info->groups_sem);
 }
 
-int btrfs_reserve_extent(struct btrfs_trans_handle *trans,
-			 struct btrfs_root *root,
+int btrfs_reserve_extent(struct btrfs_root *root,
 			 u64 num_bytes, u64 min_alloc_size,
 			 u64 empty_size, u64 hint_byte,
 			 struct btrfs_key *ins, int is_data)
@@ -6513,8 +6569,8 @@ int btrfs_reserve_extent(struct btrfs_trans_handle *trans,
 	flags = btrfs_get_alloc_profile(root, is_data);
 again:
 	WARN_ON(num_bytes < root->sectorsize);
-	ret = find_free_extent(trans, root, num_bytes, empty_size,
-			       hint_byte, ins, flags);
+	ret = find_free_extent(root, num_bytes, empty_size, hint_byte, ins,
+			       flags);
 
 	if (ret == -ENOSPC) {
 		if (!final_tried) {
@@ -6529,8 +6585,7 @@ again:
 
 			sinfo = __find_space_info(root->fs_info, flags);
 			btrfs_err(root->fs_info, "allocation failed flags %llu, wanted %llu",
-				(unsigned long long)flags,
-				(unsigned long long)num_bytes);
+				flags, num_bytes);
 			if (sinfo)
 				dump_space_info(sinfo, num_bytes, 1);
 		}
@@ -6550,7 +6605,7 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root,
 	cache = btrfs_lookup_block_group(root->fs_info, start);
 	if (!cache) {
 		btrfs_err(root->fs_info, "Unable to find block group for %llu",
-			(unsigned long long)start);
+			start);
 		return -ENOSPC;
 	}
 
@@ -6646,8 +6701,7 @@ static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 	ret = update_block_group(root, ins->objectid, ins->offset, 1);
 	if (ret) { /* -ENOENT, logic error */
 		btrfs_err(fs_info, "update block group failed for %llu %llu",
-			(unsigned long long)ins->objectid,
-			(unsigned long long)ins->offset);
+			ins->objectid, ins->offset);
 		BUG();
 	}
 	return ret;
@@ -6719,8 +6773,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 	ret = update_block_group(root, ins->objectid, root->leafsize, 1);
 	if (ret) { /* -ENOENT, logic error */
 		btrfs_err(fs_info, "update block group failed for %llu %llu",
-			(unsigned long long)ins->objectid,
-			(unsigned long long)ins->offset);
+			ins->objectid, ins->offset);
 		BUG();
 	}
 	return ret;
@@ -6902,7 +6955,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
 	if (IS_ERR(block_rsv))
 		return ERR_CAST(block_rsv);
 
-	ret = btrfs_reserve_extent(trans, root, blocksize, blocksize,
+	ret = btrfs_reserve_extent(root, blocksize, blocksize,
 				   empty_size, hint, &ins, 0);
 	if (ret) {
 		unuse_block_rsv(root->fs_info, block_rsv, blocksize);
@@ -7173,6 +7226,8 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans,
 		next = btrfs_find_create_tree_block(root, bytenr, blocksize);
 		if (!next)
 			return -ENOMEM;
+		btrfs_set_buffer_lockdep_class(root->root_key.objectid, next,
+					       level - 1);
 		reada = 1;
 	}
 	btrfs_tree_lock(next);
@@ -7658,7 +7713,7 @@ out:
 	 * don't have it in the radix (like when we recover after a power fail
 	 * or unmount) so we don't leak memory.
 	 */
-	if (root_dropped == false)
+	if (!for_reloc && root_dropped == false)
 		btrfs_add_dead_root(root);
 	if (err)
 		btrfs_std_error(root->fs_info, err);
@@ -8192,7 +8247,8 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 		 * We haven't cached this block group, which means we could
 		 * possibly have excluded extents on this block group.
 		 */
-		if (block_group->cached == BTRFS_CACHE_NO)
+		if (block_group->cached == BTRFS_CACHE_NO ||
+		    block_group->cached == BTRFS_CACHE_ERROR)
 			free_excluded_extents(info->extent_root, block_group);
 
 		btrfs_remove_free_space_cache(block_group);
@@ -8409,9 +8465,13 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 		 * avoid allocating from un-mirrored block group if there are
 		 * mirrored block groups.
 		 */
-		list_for_each_entry(cache, &space_info->block_groups[3], list)
+		list_for_each_entry(cache,
+				&space_info->block_groups[BTRFS_RAID_RAID0],
+				list)
 			set_block_group_ro(cache, 1);
-		list_for_each_entry(cache, &space_info->block_groups[4], list)
+		list_for_each_entry(cache,
+				&space_info->block_groups[BTRFS_RAID_SINGLE],
+				list)
 			set_block_group_ro(cache, 1);
 	}
 

fs/btrfs/extent_io.h

@@ -29,6 +29,7 @@
  */
 #define EXTENT_BIO_COMPRESSED 1
 #define EXTENT_BIO_TREE_LOG 2
+#define EXTENT_BIO_PARENT_LOCKED 4
 #define EXTENT_BIO_FLAG_SHIFT 16
 
 /* these are bit numbers for test/set bit */
@@ -44,14 +45,11 @@
 #define EXTENT_BUFFER_DUMMY 9
 
 /* these are flags for extent_clear_unlock_delalloc */
-#define EXTENT_CLEAR_UNLOCK_PAGE 0x1
-#define EXTENT_CLEAR_UNLOCK	 0x2
-#define EXTENT_CLEAR_DELALLOC	 0x4
-#define EXTENT_CLEAR_DIRTY	 0x8
-#define EXTENT_SET_WRITEBACK	 0x10
-#define EXTENT_END_WRITEBACK	 0x20
-#define EXTENT_SET_PRIVATE2	 0x40
-#define EXTENT_CLEAR_ACCOUNTING  0x80
+#define PAGE_UNLOCK		(1 << 0)
+#define PAGE_CLEAR_DIRTY	(1 << 1)
+#define PAGE_SET_WRITEBACK	(1 << 2)
+#define PAGE_END_WRITEBACK	(1 << 3)
+#define PAGE_SET_PRIVATE2	(1 << 4)
 
 /*
  * page->private values.  Every page that is controlled by the extent
@@ -62,6 +60,7 @@
 
 struct extent_state;
 struct btrfs_root;
+struct btrfs_io_bio;
 
 typedef	int (extent_submit_bio_hook_t)(struct inode *inode, int rw,
 				       struct bio *bio, int mirror_num,
@@ -77,8 +76,9 @@ struct extent_io_ops {
 			      size_t size, struct bio *bio,
 			      unsigned long bio_flags);
 	int (*readpage_io_failed_hook)(struct page *page, int failed_mirror);
-	int (*readpage_end_io_hook)(struct page *page, u64 start, u64 end,
-				    struct extent_state *state, int mirror);
+	int (*readpage_end_io_hook)(struct btrfs_io_bio *io_bio, u64 phy_offset,
+				    struct page *page, u64 start, u64 end,
+				    int mirror);
 	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,
@@ -200,6 +200,8 @@ int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end,
 int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end);
 int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
 			  get_extent_t *get_extent, int mirror_num);
+int extent_read_full_page_nolock(struct extent_io_tree *tree, struct page *page,
+				 get_extent_t *get_extent, int mirror_num);
 int __init extent_io_init(void);
 void extent_io_exit(void);
 
@@ -261,11 +263,6 @@ int extent_readpages(struct extent_io_tree *tree,
 		     get_extent_t get_extent);
 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);
 
@@ -330,10 +327,10 @@ int map_private_extent_buffer(struct extent_buffer *eb, unsigned long offset,
 		      unsigned long *map_len);
 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,
-				struct extent_io_tree *tree,
-				u64 start, u64 end, struct page *locked_page,
-				unsigned long op);
+int extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end,
+				 struct page *locked_page,
+				 unsigned long bits_to_clear,
+				 unsigned long page_ops);
 struct bio *
 btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
 		gfp_t gfp_flags);

fs/btrfs/file-item.c

@@ -23,6 +23,7 @@
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
+#include "volumes.h"
 #include "print-tree.h"
 
 #define __MAX_CSUM_ITEMS(r, size) ((unsigned long)(((BTRFS_LEAF_DATA_SIZE(r) - \
@@ -152,28 +153,54 @@ int btrfs_lookup_file_extent(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
+static void btrfs_io_bio_endio_readpage(struct btrfs_io_bio *bio, int err)
+{
+	kfree(bio->csum_allocated);
+}
+
 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[16];
-	int len;
 	struct bio_vec *bvec = bio->bi_io_vec;
-	int bio_index = 0;
+	struct btrfs_io_bio *btrfs_bio = btrfs_io_bio(bio);
+	struct btrfs_csum_item *item = NULL;
+	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
+	struct btrfs_path *path;
+	u8 *csum;
 	u64 offset = 0;
 	u64 item_start_offset = 0;
 	u64 item_last_offset = 0;
 	u64 disk_bytenr;
 	u32 diff;
-	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
+	int nblocks;
+	int bio_index = 0;
 	int count;
-	struct btrfs_path *path;
-	struct btrfs_csum_item *item = NULL;
-	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
+	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
 
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+
+	nblocks = bio->bi_size >> inode->i_sb->s_blocksize_bits;
+	if (!dst) {
+		if (nblocks * csum_size > BTRFS_BIO_INLINE_CSUM_SIZE) {
+			btrfs_bio->csum_allocated = kmalloc(nblocks * csum_size,
+							    GFP_NOFS);
+			if (!btrfs_bio->csum_allocated) {
+				btrfs_free_path(path);
+				return -ENOMEM;
+			}
+			btrfs_bio->csum = btrfs_bio->csum_allocated;
+			btrfs_bio->end_io = btrfs_io_bio_endio_readpage;
+		} else {
+			btrfs_bio->csum = btrfs_bio->csum_inline;
+		}
+		csum = btrfs_bio->csum;
+	} else {
+		csum = (u8 *)dst;
+	}
+
 	if (bio->bi_size > PAGE_CACHE_SIZE * 8)
 		path->reada = 2;
 
@@ -194,11 +221,10 @@ 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;
-		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr, sum,
-					       len);
+		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr,
+					       (u32 *)csum, nblocks);
 		if (count)
 			goto found;
 
@@ -213,7 +239,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 						 path, disk_bytenr, 0);
 			if (IS_ERR(item)) {
 				count = 1;
-				sum[0] = 0;
+				memset(csum, 0, csum_size);
 				if (BTRFS_I(inode)->root->root_key.objectid ==
 				    BTRFS_DATA_RELOC_TREE_OBJECTID) {
 					set_extent_bits(io_tree, offset,
@@ -222,9 +248,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 				} else {
 					printk(KERN_INFO "btrfs no csum found "
 					       "for inode %llu start %llu\n",
-					       (unsigned long long)
-					       btrfs_ino(inode),
-					       (unsigned long long)offset);
+					       btrfs_ino(inode), offset);
 				}
 				item = NULL;
 				btrfs_release_path(path);
@@ -249,23 +273,14 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 		diff = disk_bytenr - item_start_offset;
 		diff = diff / root->sectorsize;
 		diff = diff * csum_size;
-		count = min_t(int, len, (item_last_offset - disk_bytenr) >>
-					inode->i_sb->s_blocksize_bits);
-		read_extent_buffer(path->nodes[0], sum,
+		count = min_t(int, nblocks, (item_last_offset - disk_bytenr) >>
+					    inode->i_sb->s_blocksize_bits);
+		read_extent_buffer(path->nodes[0], csum,
 				   ((unsigned long)item) + diff,
 				   csum_size * count);
 found:
-		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);
-		}
+		csum += count * csum_size;
+		nblocks -= count;
 		while (count--) {
 			disk_bytenr += bvec->bv_len;
 			offset += bvec->bv_len;
@@ -284,9 +299,19 @@ int btrfs_lookup_bio_sums(struct btrfs_root *root, struct inode *inode,
 }
 
 int btrfs_lookup_bio_sums_dio(struct btrfs_root *root, struct inode *inode,
-			      struct bio *bio, u64 offset)
+			      struct btrfs_dio_private *dip, struct bio *bio,
+			      u64 offset)
 {
-	return __btrfs_lookup_bio_sums(root, inode, bio, offset, NULL, 1);
+	int len = (bio->bi_sector << 9) - dip->disk_bytenr;
+	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
+	int ret;
+
+	len >>= inode->i_sb->s_blocksize_bits;
+	len *= csum_size;
+
+	ret = __btrfs_lookup_bio_sums(root, inode, bio, offset,
+				      (u32 *)(dip->csum + len), 1);
+	return ret;
 }
 
 int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,

fs/btrfs/file.c

@@ -1334,7 +1334,6 @@ fail:
 static noinline int check_can_nocow(struct inode *inode, loff_t pos,
 				    size_t *write_bytes)
 {
-	struct btrfs_trans_handle *trans;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_ordered_extent *ordered;
 	u64 lockstart, lockend;
@@ -1356,16 +1355,8 @@ static noinline int check_can_nocow(struct inode *inode, loff_t pos,
 		btrfs_put_ordered_extent(ordered);
 	}
 
-	trans = btrfs_join_transaction(root);
-	if (IS_ERR(trans)) {
-		unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
-		return PTR_ERR(trans);
-	}
-
 	num_bytes = lockend - lockstart + 1;
-	ret = can_nocow_extent(trans, inode, lockstart, &num_bytes, NULL, NULL,
-			       NULL);
-	btrfs_end_transaction(trans, root);
+	ret = can_nocow_extent(inode, lockstart, &num_bytes, NULL, NULL, NULL);
 	if (ret <= 0) {
 		ret = 0;
 	} else {

fs/btrfs/free-space-cache.c

@@ -308,7 +308,7 @@ static void io_ctl_unmap_page(struct io_ctl *io_ctl)
 
 static void io_ctl_map_page(struct io_ctl *io_ctl, int clear)
 {
-	BUG_ON(io_ctl->index >= io_ctl->num_pages);
+	ASSERT(io_ctl->index < io_ctl->num_pages);
 	io_ctl->page = io_ctl->pages[io_ctl->index++];
 	io_ctl->cur = kmap(io_ctl->page);
 	io_ctl->orig = io_ctl->cur;
@@ -673,8 +673,7 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 		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);
+			BTRFS_I(inode)->generation, generation);
 		return 0;
 	}
 
@@ -729,7 +728,7 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 				goto free_cache;
 			}
 		} else {
-			BUG_ON(!num_bitmaps);
+			ASSERT(num_bitmaps);
 			num_bitmaps--;
 			e->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS);
 			if (!e->bitmap) {
@@ -1029,7 +1028,7 @@ static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
 	leaf = path->nodes[0];
 	if (ret > 0) {
 		struct btrfs_key found_key;
-		BUG_ON(!path->slots[0]);
+		ASSERT(path->slots[0]);
 		path->slots[0]--;
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 		if (found_key.objectid != BTRFS_FREE_SPACE_OBJECTID ||
@@ -1117,7 +1116,7 @@ int btrfs_write_out_cache(struct btrfs_root *root,
 static inline unsigned long offset_to_bit(u64 bitmap_start, u32 unit,
 					  u64 offset)
 {
-	BUG_ON(offset < bitmap_start);
+	ASSERT(offset >= bitmap_start);
 	offset -= bitmap_start;
 	return (unsigned long)(div_u64(offset, unit));
 }
@@ -1272,7 +1271,7 @@ tree_search_offset(struct btrfs_free_space_ctl *ctl,
 		if (n) {
 			entry = rb_entry(n, struct btrfs_free_space,
 					offset_index);
-			BUG_ON(entry->offset > offset);
+			ASSERT(entry->offset <= offset);
 		} else {
 			if (fuzzy)
 				return entry;
@@ -1336,7 +1335,7 @@ static int link_free_space(struct btrfs_free_space_ctl *ctl,
 {
 	int ret = 0;
 
-	BUG_ON(!info->bitmap && !info->bytes);
+	ASSERT(info->bytes || info->bitmap);
 	ret = tree_insert_offset(&ctl->free_space_offset, info->offset,
 				 &info->offset_index, (info->bitmap != NULL));
 	if (ret)
@@ -1359,7 +1358,7 @@ static void recalculate_thresholds(struct btrfs_free_space_ctl *ctl)
 
 	max_bitmaps = max(max_bitmaps, 1);
 
-	BUG_ON(ctl->total_bitmaps > max_bitmaps);
+	ASSERT(ctl->total_bitmaps <= max_bitmaps);
 
 	/*
 	 * The goal is to keep the total amount of memory used per 1gb of space
@@ -1403,7 +1402,7 @@ static inline void __bitmap_clear_bits(struct btrfs_free_space_ctl *ctl,
 
 	start = offset_to_bit(info->offset, ctl->unit, offset);
 	count = bytes_to_bits(bytes, ctl->unit);
-	BUG_ON(start + count > BITS_PER_BITMAP);
+	ASSERT(start + count <= BITS_PER_BITMAP);
 
 	bitmap_clear(info->bitmap, start, count);
 
@@ -1426,7 +1425,7 @@ static void bitmap_set_bits(struct btrfs_free_space_ctl *ctl,
 
 	start = offset_to_bit(info->offset, ctl->unit, offset);
 	count = bytes_to_bits(bytes, ctl->unit);
-	BUG_ON(start + count > BITS_PER_BITMAP);
+	ASSERT(start + count <= BITS_PER_BITMAP);
 
 	bitmap_set(info->bitmap, start, count);
 
@@ -1742,7 +1741,7 @@ no_cluster_bitmap:
 	bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
 					 1, 0);
 	if (!bitmap_info) {
-		BUG_ON(added);
+		ASSERT(added == 0);
 		goto new_bitmap;
 	}
 
@@ -1882,7 +1881,7 @@ out:
 
 	if (ret) {
 		printk(KERN_CRIT "btrfs: unable to add free space :%d\n", ret);
-		BUG_ON(ret == -EEXIST);
+		ASSERT(ret != -EEXIST);
 	}
 
 	return ret;
@@ -1991,8 +1990,7 @@ void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
 		if (info->bytes >= bytes && !block_group->ro)
 			count++;
 		printk(KERN_CRIT "entry offset %llu, bytes %llu, bitmap %s\n",
-		       (unsigned long long)info->offset,
-		       (unsigned long long)info->bytes,
+		       info->offset, info->bytes,
 		       (info->bitmap) ? "yes" : "no");
 	}
 	printk(KERN_INFO "block group has cluster?: %s\n",
@@ -2371,7 +2369,7 @@ again:
 	rb_erase(&entry->offset_index, &ctl->free_space_offset);
 	ret = tree_insert_offset(&cluster->root, entry->offset,
 				 &entry->offset_index, 1);
-	BUG_ON(ret); /* -EEXIST; Logic error */
+	ASSERT(!ret); /* -EEXIST; Logic error */
 
 	trace_btrfs_setup_cluster(block_group, cluster,
 				  total_found * ctl->unit, 1);
@@ -2464,7 +2462,7 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
 		ret = tree_insert_offset(&cluster->root, entry->offset,
 					 &entry->offset_index, 0);
 		total_size += entry->bytes;
-		BUG_ON(ret); /* -EEXIST; Logic error */
+		ASSERT(!ret); /* -EEXIST; Logic error */
 	} while (node && entry != last);
 
 	cluster->max_size = max_extent;
@@ -2525,8 +2523,7 @@ setup_cluster_bitmap(struct btrfs_block_group_cache *block_group,
  * returns zero and sets up cluster if things worked out, otherwise
  * it returns -enospc
  */
-int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *root,
+int btrfs_find_space_cluster(struct btrfs_root *root,
 			     struct btrfs_block_group_cache *block_group,
 			     struct btrfs_free_cluster *cluster,
 			     u64 offset, u64 bytes, u64 empty_size)
@@ -2856,7 +2853,7 @@ u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root)
 
 		ret = search_bitmap(ctl, entry, &offset, &count);
 		/* Logic error; Should be empty if it can't find anything */
-		BUG_ON(ret);
+		ASSERT(!ret);
 
 		ino = offset;
 		bitmap_clear_bits(ctl, entry, offset, 1);
@@ -2973,33 +2970,68 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root,
 }
 
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-static struct btrfs_block_group_cache *init_test_block_group(void)
+/*
+ * 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.
+ */
+int test_add_free_space_entry(struct btrfs_block_group_cache *cache,
+			      u64 offset, u64 bytes, bool bitmap)
 {
-	struct btrfs_block_group_cache *cache;
+	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;
 
-	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;
+again:
+	if (!info) {
+		info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS);
+		if (!info)
+			return -ENOMEM;
 	}
 
-	cache->key.objectid = 0;
-	cache->key.offset = 1024 * 1024 * 1024;
-	cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
-	cache->sectorsize = 4096;
+	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;
+	}
 
-	spin_lock_init(&cache->lock);
-	INIT_LIST_HEAD(&cache->list);
-	INIT_LIST_HEAD(&cache->cluster_list);
-	INIT_LIST_HEAD(&cache->new_bg_list);
+	bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes);
+	bytes -= bytes_added;
+	offset += bytes_added;
+	spin_unlock(&ctl->tree_lock);
 
-	btrfs_init_free_space_ctl(cache);
+	if (bytes)
+		goto again;
 
-	return cache;
+	if (map)
+		kfree(map);
+	return 0;
 }
 
 /*
@@ -3007,8 +3039,8 @@ static struct btrfs_block_group_cache *init_test_block_group(void)
  * 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)
+int test_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;
@@ -3085,411 +3117,4 @@ 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;
-}
-
-#define test_msg(fmt, ...) printk(KERN_INFO "btrfs: selftest: " fmt, ##__VA_ARGS__)
-
-/*
- * This test just does basic sanity checking, making sure we can add an exten
- * entry and remove space from either end and the middle, and make sure we can
- * remove space that covers adjacent extent entries.
- */
-static int test_extents(struct btrfs_block_group_cache *cache)
-{
-	int ret = 0;
-
-	test_msg("Running extent only tests\n");
-
-	/* First just make sure we can remove an entire entry */
-	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error adding initial extents %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing extent %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 4 * 1024 * 1024)) {
-		test_msg("Full remove left some lingering space\n");
-		return -1;
-	}
-
-	/* Ok edge and middle cases now */
-	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error adding half extent %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing tail end %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing front end %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 2 * 1024 * 1024, 4096);
-	if (ret) {
-		test_msg("Error removing middle piece %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 1 * 1024 * 1024)) {
-		test_msg("Still have space at the front\n");
-		return -1;
-	}
-
-	if (check_exists(cache, 2 * 1024 * 1024, 4096)) {
-		test_msg("Still have space in the middle\n");
-		return -1;
-	}
-
-	if (check_exists(cache, 3 * 1024 * 1024, 1 * 1024 * 1024)) {
-		test_msg("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;
-
-	test_msg("Running bitmap only tests\n");
-
-	ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't create a bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Error removing bitmap full range %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 4 * 1024 * 1024)) {
-		test_msg("Left some space in bitmap\n");
-		return -1;
-	}
-
-	ret = add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
-	if (ret) {
-		test_msg("Couldn't add to our bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 2 * 1024 * 1024);
-	if (ret) {
-		test_msg("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) {
-		test_msg("Couldn't add space that straddles two bitmaps %d\n",
-				ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, next_bitmap_offset -
-				      (1 * 1024 * 1024), 2 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove overlapping space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, next_bitmap_offset - (1 * 1024 * 1024),
-			 2 * 1024 * 1024)) {
-		test_msg("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;
-
-	test_msg("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) {
-		test_msg("Couldn't create bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove extent entry %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 0, 1 * 1024 * 1024)) {
-		test_msg("Left remnants after our remove\n");
-		return -1;
-	}
-
-	/* Now to add back the extent entry and remove from the bitmap */
-	ret = add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't re-add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 4 * 1024 * 1024, 1 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove from bitmap %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 4 * 1024 * 1024, 1 * 1024 * 1024)) {
-		test_msg("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) {
-		test_msg("Couldn't add to a bitmap %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 512 * 1024, 3 * 1024 * 1024);
-	if (ret) {
-		test_msg("Couldn't remove overlapping space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 512 * 1024, 3 * 1024 * 1024)) {
-		test_msg("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) {
-		test_msg("Couldn't add space to the bitmap %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, 2 * 1024 * 1024, 2 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent to the cache %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 4 * 1024 * 1024);
-	if (ret) {
-		test_msg("Problem removing overlapping space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, 3 * 1024 * 1024, 4 * 1024 * 1024)) {
-		test_msg("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) {
-		test_msg("Couldn't add bitmap %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, bitmap_offset - 1 * 1024 * 1024,
-				   5 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, bitmap_offset + 1 * 1024 * 1024,
-				      5 * 1024 * 1024);
-	if (ret) {
-		test_msg("Failed to free our space %d\n", ret);
-		return ret;
-	}
-
-	if (check_exists(cache, bitmap_offset + 1 * 1024 * 1024,
-			 5 * 1024 * 1024)) {
-		test_msg("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) {
-		test_msg("Couldn't add bitmap entry %d\n", ret);
-		return ret;
-	}
-
-	ret = add_free_space_entry(cache, 3 * 1024 * 1024, 1 * 1024 * 1024, 0);
-	if (ret) {
-		test_msg("Couldn't add extent entry %d\n", ret);
-		return ret;
-	}
-
-	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 3 * 1024 * 1024);
-	if (ret) {
-		test_msg("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;
-
-	test_msg("Running btrfs free space cache tests\n");
-
-	cache = init_test_block_group();
-	if (!cache) {
-		test_msg("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);
-	test_msg("Free space cache tests finished\n");
-}
-#undef test_msg
-#else /* !CONFIG_BTRFS_FS_RUN_SANITY_TESTS */
-void btrfs_test_free_space_cache(void) {}
-#endif /* !CONFIG_BTRFS_FS_RUN_SANITY_TESTS */
+#endif /* CONFIG_BTRFS_FS_RUN_SANITY_TESTS */

fs/btrfs/free-space-cache.h

@@ -98,8 +98,7 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group,
 u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root);
 void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
 			   u64 bytes);
-int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *root,
+int btrfs_find_space_cluster(struct btrfs_root *root,
 			     struct btrfs_block_group_cache *block_group,
 			     struct btrfs_free_cluster *cluster,
 			     u64 offset, u64 bytes, u64 empty_size);
@@ -113,6 +112,12 @@ int btrfs_return_cluster_to_free_space(
 int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group,
 			   u64 *trimmed, u64 start, u64 end, u64 minlen);
 
-void btrfs_test_free_space_cache(void);
+/* Support functions for runnint our sanity tests */
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+int test_add_free_space_entry(struct btrfs_block_group_cache *cache,
+			      u64 offset, u64 bytes, bool bitmap);
+int test_check_exists(struct btrfs_block_group_cache *cache,
+		      u64 offset, u64 bytes);
+#endif
 
 #endif

fs/btrfs/lzo.c

@@ -207,8 +207,10 @@ static int lzo_compress_pages(struct list_head *ws,
 		}
 
 		/* we're making it bigger, give up */
-		if (tot_in > 8192 && tot_in < tot_out)
+		if (tot_in > 8192 && tot_in < tot_out) {
+			ret = -1;
 			goto out;
+		}
 
 		/* we're all done */
 		if (tot_in >= len)

fs/btrfs/ordered-data.c

@@ -67,7 +67,7 @@ static void ordered_data_tree_panic(struct inode *inode, int errno,
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	btrfs_panic(fs_info, errno, "Inconsistency in ordered tree at offset "
-		    "%llu\n", (unsigned long long)offset);
+		    "%llu\n", offset);
 }
 
 /*
@@ -205,6 +205,7 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset,
 	entry->bytes_left = len;
 	entry->inode = igrab(inode);
 	entry->compress_type = compress_type;
+	entry->truncated_len = (u64)-1;
 	if (type != BTRFS_ORDERED_IO_DONE && type != BTRFS_ORDERED_COMPLETE)
 		set_bit(type, &entry->flags);
 
@@ -336,14 +337,12 @@ int btrfs_dec_test_first_ordered_pending(struct inode *inode,
 	*file_offset = dec_end;
 	if (dec_start > dec_end) {
 		printk(KERN_CRIT "bad ordering dec_start %llu end %llu\n",
-		       (unsigned long long)dec_start,
-		       (unsigned long long)dec_end);
+		       dec_start, dec_end);
 	}
 	to_dec = dec_end - dec_start;
 	if (to_dec > entry->bytes_left) {
 		printk(KERN_CRIT "bad ordered accounting left %llu size %llu\n",
-		       (unsigned long long)entry->bytes_left,
-		       (unsigned long long)to_dec);
+		       entry->bytes_left, to_dec);
 	}
 	entry->bytes_left -= to_dec;
 	if (!uptodate)
@@ -403,8 +402,7 @@ have_entry:
 
 	if (io_size > entry->bytes_left) {
 		printk(KERN_CRIT "bad ordered accounting left %llu size %llu\n",
-		       (unsigned long long)entry->bytes_left,
-		       (unsigned long long)io_size);
+		       entry->bytes_left, io_size);
 	}
 	entry->bytes_left -= io_size;
 	if (!uptodate)
@@ -671,7 +669,7 @@ int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans,
 	INIT_LIST_HEAD(&splice);
 	INIT_LIST_HEAD(&works);
 
-	mutex_lock(&root->fs_info->ordered_operations_mutex);
+	mutex_lock(&root->fs_info->ordered_extent_flush_mutex);
 	spin_lock(&root->fs_info->ordered_root_lock);
 	list_splice_init(&cur_trans->ordered_operations, &splice);
 	while (!list_empty(&splice)) {
@@ -718,7 +716,7 @@ out:
 		list_del_init(&work->list);
 		btrfs_wait_and_free_delalloc_work(work);
 	}
-	mutex_unlock(&root->fs_info->ordered_operations_mutex);
+	mutex_unlock(&root->fs_info->ordered_extent_flush_mutex);
 	return ret;
 }
 
@@ -923,12 +921,16 @@ int btrfs_ordered_update_i_size(struct inode *inode, u64 offset,
 	struct btrfs_ordered_extent *test;
 	int ret = 1;
 
-	if (ordered)
+	spin_lock_irq(&tree->lock);
+	if (ordered) {
 		offset = entry_end(ordered);
-	else
+		if (test_bit(BTRFS_ORDERED_TRUNCATED, &ordered->flags))
+			offset = min(offset,
+				     ordered->file_offset +
+				     ordered->truncated_len);
+	} else {
 		offset = ALIGN(offset, BTRFS_I(inode)->root->sectorsize);
-
-	spin_lock_irq(&tree->lock);
+	}
 	disk_i_size = BTRFS_I(inode)->disk_i_size;
 
 	/* truncate file */

fs/btrfs/ordered-data.h

@@ -69,6 +69,7 @@ struct btrfs_ordered_sum {
 				       * the isize. */
 #define BTRFS_ORDERED_LOGGED_CSUM 8 /* We've logged the csums on this ordered
 				       ordered extent */
+#define BTRFS_ORDERED_TRUNCATED 9 /* Set when we have to truncate an extent */
 
 struct btrfs_ordered_extent {
 	/* logical offset in the file */
@@ -96,6 +97,12 @@ struct btrfs_ordered_extent {
 	 */
 	u64 outstanding_isize;
 
+	/*
+	 * If we get truncated we need to adjust the file extent we enter for
+	 * this ordered extent so that we do not expose stale data.
+	 */
+	u64 truncated_len;
+
 	/* flags (described above) */
 	unsigned long flags;
 

fs/btrfs/print-tree.c

@@ -26,14 +26,12 @@ static void print_chunk(struct extent_buffer *eb, struct btrfs_chunk *chunk)
 	int i;
 	printk(KERN_INFO "\t\tchunk length %llu owner %llu type %llu "
 	       "num_stripes %d\n",
-	       (unsigned long long)btrfs_chunk_length(eb, chunk),
-	       (unsigned long long)btrfs_chunk_owner(eb, chunk),
-	       (unsigned long long)btrfs_chunk_type(eb, chunk),
-	       num_stripes);
+	       btrfs_chunk_length(eb, chunk), btrfs_chunk_owner(eb, chunk),
+	       btrfs_chunk_type(eb, chunk), num_stripes);
 	for (i = 0 ; i < num_stripes ; i++) {
 		printk(KERN_INFO "\t\t\tstripe %d devid %llu offset %llu\n", i,
-		      (unsigned long long)btrfs_stripe_devid_nr(eb, chunk, i),
-		      (unsigned long long)btrfs_stripe_offset_nr(eb, chunk, i));
+		      btrfs_stripe_devid_nr(eb, chunk, i),
+		      btrfs_stripe_offset_nr(eb, chunk, i));
 	}
 }
 static void print_dev_item(struct extent_buffer *eb,
@@ -41,18 +39,18 @@ static void print_dev_item(struct extent_buffer *eb,
 {
 	printk(KERN_INFO "\t\tdev item devid %llu "
 	       "total_bytes %llu bytes used %llu\n",
-	       (unsigned long long)btrfs_device_id(eb, dev_item),
-	       (unsigned long long)btrfs_device_total_bytes(eb, dev_item),
-	       (unsigned long long)btrfs_device_bytes_used(eb, dev_item));
+	       btrfs_device_id(eb, dev_item),
+	       btrfs_device_total_bytes(eb, dev_item),
+	       btrfs_device_bytes_used(eb, dev_item));
 }
 static void print_extent_data_ref(struct extent_buffer *eb,
 				  struct btrfs_extent_data_ref *ref)
 {
 	printk(KERN_INFO "\t\textent data backref root %llu "
 	       "objectid %llu offset %llu count %u\n",
-	       (unsigned long long)btrfs_extent_data_ref_root(eb, ref),
-	       (unsigned long long)btrfs_extent_data_ref_objectid(eb, ref),
-	       (unsigned long long)btrfs_extent_data_ref_offset(eb, ref),
+	       btrfs_extent_data_ref_root(eb, ref),
+	       btrfs_extent_data_ref_objectid(eb, ref),
+	       btrfs_extent_data_ref_offset(eb, ref),
 	       btrfs_extent_data_ref_count(eb, ref));
 }
 
@@ -87,19 +85,17 @@ static void print_extent_item(struct extent_buffer *eb, int slot)
 	flags = btrfs_extent_flags(eb, ei);
 
 	printk(KERN_INFO "\t\textent refs %llu gen %llu flags %llu\n",
-	       (unsigned long long)btrfs_extent_refs(eb, ei),
-	       (unsigned long long)btrfs_extent_generation(eb, ei),
-	       (unsigned long long)flags);
+	       btrfs_extent_refs(eb, ei), btrfs_extent_generation(eb, ei),
+	       flags);
 
 	if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
 		struct btrfs_tree_block_info *info;
 		info = (struct btrfs_tree_block_info *)(ei + 1);
 		btrfs_tree_block_key(eb, info, &key);
-		printk(KERN_INFO "\t\ttree block key (%llu %x %llu) "
+		printk(KERN_INFO "\t\ttree block key (%llu %u %llu) "
 		       "level %d\n",
-		       (unsigned long long)btrfs_disk_key_objectid(&key),
-		       key.type,
-		       (unsigned long long)btrfs_disk_key_offset(&key),
+		       btrfs_disk_key_objectid(&key), key.type,
+		       btrfs_disk_key_offset(&key),
 		       btrfs_tree_block_level(eb, info));
 		iref = (struct btrfs_extent_inline_ref *)(info + 1);
 	} else {
@@ -115,11 +111,11 @@ static void print_extent_item(struct extent_buffer *eb, int slot)
 		switch (type) {
 		case BTRFS_TREE_BLOCK_REF_KEY:
 			printk(KERN_INFO "\t\ttree block backref "
-				"root %llu\n", (unsigned long long)offset);
+				"root %llu\n", offset);
 			break;
 		case BTRFS_SHARED_BLOCK_REF_KEY:
 			printk(KERN_INFO "\t\tshared block backref "
-				"parent %llu\n", (unsigned long long)offset);
+				"parent %llu\n", offset);
 			break;
 		case BTRFS_EXTENT_DATA_REF_KEY:
 			dref = (struct btrfs_extent_data_ref *)(&iref->offset);
@@ -129,8 +125,7 @@ static void print_extent_item(struct extent_buffer *eb, int slot)
 			sref = (struct btrfs_shared_data_ref *)(iref + 1);
 			printk(KERN_INFO "\t\tshared data backref "
 			       "parent %llu count %u\n",
-			       (unsigned long long)offset,
-			       btrfs_shared_data_ref_count(eb, sref));
+			       offset, btrfs_shared_data_ref_count(eb, sref));
 			break;
 		default:
 			BUG();
@@ -148,13 +143,32 @@ static void print_extent_ref_v0(struct extent_buffer *eb, int slot)
 	ref0 = btrfs_item_ptr(eb, slot, struct btrfs_extent_ref_v0);
 	printk("\t\textent back ref root %llu gen %llu "
 		"owner %llu num_refs %lu\n",
-		(unsigned long long)btrfs_ref_root_v0(eb, ref0),
-		(unsigned long long)btrfs_ref_generation_v0(eb, ref0),
-		(unsigned long long)btrfs_ref_objectid_v0(eb, ref0),
+		btrfs_ref_root_v0(eb, ref0),
+		btrfs_ref_generation_v0(eb, ref0),
+		btrfs_ref_objectid_v0(eb, ref0),
 		(unsigned long)btrfs_ref_count_v0(eb, ref0));
 }
 #endif
 
+static void print_uuid_item(struct extent_buffer *l, unsigned long offset,
+			    u32 item_size)
+{
+	if (!IS_ALIGNED(item_size, sizeof(u64))) {
+		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+			(unsigned long)item_size);
+		return;
+	}
+	while (item_size) {
+		__le64 subvol_id;
+
+		read_extent_buffer(l, &subvol_id, offset, sizeof(subvol_id));
+		printk(KERN_INFO "\t\tsubvol_id %llu\n",
+		       (unsigned long long)le64_to_cpu(subvol_id));
+		item_size -= sizeof(u64);
+		offset += sizeof(u64);
+	}
+}
+
 void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 {
 	int i;
@@ -177,39 +191,34 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 	nr = btrfs_header_nritems(l);
 
 	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));
+		   btrfs_header_bytenr(l), nr, btrfs_leaf_free_space(root, l));
 	for (i = 0 ; i < nr ; i++) {
 		item = btrfs_item_nr(l, i);
 		btrfs_item_key_to_cpu(l, &key, i);
 		type = btrfs_key_type(&key);
-		printk(KERN_INFO "\titem %d key (%llu %x %llu) itemoff %d "
+		printk(KERN_INFO "\titem %d key (%llu %u %llu) itemoff %d "
 		       "itemsize %d\n",
-			i,
-			(unsigned long long)key.objectid, type,
-			(unsigned long long)key.offset,
+			i, key.objectid, type, key.offset,
 			btrfs_item_offset(l, item), btrfs_item_size(l, item));
 		switch (type) {
 		case BTRFS_INODE_ITEM_KEY:
 			ii = btrfs_item_ptr(l, i, struct btrfs_inode_item);
 			printk(KERN_INFO "\t\tinode generation %llu size %llu "
 			       "mode %o\n",
-			       (unsigned long long)
 			       btrfs_inode_generation(l, ii),
-			      (unsigned long long)btrfs_inode_size(l, ii),
+			       btrfs_inode_size(l, ii),
 			       btrfs_inode_mode(l, ii));
 			break;
 		case BTRFS_DIR_ITEM_KEY:
 			di = btrfs_item_ptr(l, i, struct btrfs_dir_item);
 			btrfs_dir_item_key_to_cpu(l, di, &found_key);
 			printk(KERN_INFO "\t\tdir oid %llu type %u\n",
-				(unsigned long long)found_key.objectid,
+				found_key.objectid,
 				btrfs_dir_type(l, di));
 			break;
 		case BTRFS_ROOT_ITEM_KEY:
 			ri = btrfs_item_ptr(l, i, struct btrfs_root_item);
 			printk(KERN_INFO "\t\troot data bytenr %llu refs %u\n",
-				(unsigned long long)
 				btrfs_disk_root_bytenr(l, ri),
 				btrfs_disk_root_refs(l, ri));
 			break;
@@ -245,17 +254,12 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 			}
 			printk(KERN_INFO "\t\textent data disk bytenr %llu "
 			       "nr %llu\n",
-			       (unsigned long long)
 			       btrfs_file_extent_disk_bytenr(l, fi),
-			       (unsigned long long)
 			       btrfs_file_extent_disk_num_bytes(l, fi));
 			printk(KERN_INFO "\t\textent data offset %llu "
 			       "nr %llu ram %llu\n",
-			       (unsigned long long)
 			       btrfs_file_extent_offset(l, fi),
-			       (unsigned long long)
 			       btrfs_file_extent_num_bytes(l, fi),
-			       (unsigned long long)
 			       btrfs_file_extent_ram_bytes(l, fi));
 			break;
 		case BTRFS_EXTENT_REF_V0_KEY:
@@ -269,7 +273,6 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 			bi = btrfs_item_ptr(l, i,
 					    struct btrfs_block_group_item);
 			printk(KERN_INFO "\t\tblock group used %llu\n",
-			       (unsigned long long)
 			       btrfs_disk_block_group_used(l, bi));
 			break;
 		case BTRFS_CHUNK_ITEM_KEY:
@@ -286,13 +289,9 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 			printk(KERN_INFO "\t\tdev extent chunk_tree %llu\n"
 			       "\t\tchunk objectid %llu chunk offset %llu "
 			       "length %llu\n",
-			       (unsigned long long)
 			       btrfs_dev_extent_chunk_tree(l, dev_extent),
-			       (unsigned long long)
 			       btrfs_dev_extent_chunk_objectid(l, dev_extent),
-			       (unsigned long long)
 			       btrfs_dev_extent_chunk_offset(l, dev_extent),
-			       (unsigned long long)
 			       btrfs_dev_extent_length(l, dev_extent));
 			break;
 		case BTRFS_DEV_STATS_KEY:
@@ -301,6 +300,11 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l)
 		case BTRFS_DEV_REPLACE_KEY:
 			printk(KERN_INFO "\t\tdev replace\n");
 			break;
+		case BTRFS_UUID_KEY_SUBVOL:
+		case BTRFS_UUID_KEY_RECEIVED_SUBVOL:
+			print_uuid_item(l, btrfs_item_ptr_offset(l, i),
+					btrfs_item_size_nr(l, i));
+			break;
 		};
 	}
 }
@@ -320,16 +324,13 @@ void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *c)
 		return;
 	}
 	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);
+		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",
-		       i,
-		       (unsigned long long)key.objectid,
-		       key.type,
-		       (unsigned long long)key.offset,
-		       (unsigned long long)btrfs_node_blockptr(c, i));
+		       i, key.objectid, key.type, key.offset,
+		       btrfs_node_blockptr(c, i));
 	}
 	for (i = 0; i < nr; i++) {
 		struct extent_buffer *next = read_tree_block(root,

fs/btrfs/qgroup.c

@@ -157,18 +157,11 @@ static struct btrfs_qgroup *add_qgroup_rb(struct btrfs_fs_info *fs_info,
 	return qgroup;
 }
 
-/* must be called with qgroup_lock held */
-static int del_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid)
+static void __del_qgroup_rb(struct btrfs_qgroup *qgroup)
 {
-	struct btrfs_qgroup *qgroup = find_qgroup_rb(fs_info, qgroupid);
 	struct btrfs_qgroup_list *list;
 
-	if (!qgroup)
-		return -ENOENT;
-
-	rb_erase(&qgroup->node, &fs_info->qgroup_tree);
 	list_del(&qgroup->dirty);
-
 	while (!list_empty(&qgroup->groups)) {
 		list = list_first_entry(&qgroup->groups,
 					struct btrfs_qgroup_list, next_group);
@@ -185,7 +178,18 @@ static int del_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid)
 		kfree(list);
 	}
 	kfree(qgroup);
+}
 
+/* must be called with qgroup_lock held */
+static int del_qgroup_rb(struct btrfs_fs_info *fs_info, u64 qgroupid)
+{
+	struct btrfs_qgroup *qgroup = find_qgroup_rb(fs_info, qgroupid);
+
+	if (!qgroup)
+		return -ENOENT;
+
+	rb_erase(&qgroup->node, &fs_info->qgroup_tree);
+	__del_qgroup_rb(qgroup);
 	return 0;
 }
 
@@ -394,8 +398,7 @@ next1:
 		if (ret == -ENOENT) {
 			printk(KERN_WARNING
 				"btrfs: orphan qgroup relation 0x%llx->0x%llx\n",
-				(unsigned long long)found_key.objectid,
-				(unsigned long long)found_key.offset);
+				found_key.objectid, found_key.offset);
 			ret = 0;	/* ignore the error */
 		}
 		if (ret)
@@ -428,39 +431,28 @@ out:
 }
 
 /*
- * This is only called from close_ctree() or open_ctree(), both in single-
- * treaded paths. Clean up the in-memory structures. No locking needed.
+ * This is called from close_ctree() or open_ctree() or btrfs_quota_disable(),
+ * first two are in single-threaded paths.And for the third one, we have set
+ * quota_root to be null with qgroup_lock held before, so it is safe to clean
+ * up the in-memory structures without qgroup_lock held.
  */
 void btrfs_free_qgroup_config(struct btrfs_fs_info *fs_info)
 {
 	struct rb_node *n;
 	struct btrfs_qgroup *qgroup;
-	struct btrfs_qgroup_list *list;
 
 	while ((n = rb_first(&fs_info->qgroup_tree))) {
 		qgroup = rb_entry(n, struct btrfs_qgroup, node);
 		rb_erase(n, &fs_info->qgroup_tree);
-
-		while (!list_empty(&qgroup->groups)) {
-			list = list_first_entry(&qgroup->groups,
-						struct btrfs_qgroup_list,
-						next_group);
-			list_del(&list->next_group);
-			list_del(&list->next_member);
-			kfree(list);
-		}
-
-		while (!list_empty(&qgroup->members)) {
-			list = list_first_entry(&qgroup->members,
-						struct btrfs_qgroup_list,
-						next_member);
-			list_del(&list->next_group);
-			list_del(&list->next_member);
-			kfree(list);
-		}
-		kfree(qgroup);
+		__del_qgroup_rb(qgroup);
 	}
+	/*
+	 * we call btrfs_free_qgroup_config() when umounting
+	 * filesystem and disabling quota, so we set qgroup_ulit
+	 * to be null here to avoid double free.
+	 */
 	ulist_free(fs_info->qgroup_ulist);
+	fs_info->qgroup_ulist = NULL;
 }
 
 static int add_qgroup_relation_item(struct btrfs_trans_handle *trans,
@@ -946,13 +938,9 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 	fs_info->pending_quota_state = 0;
 	quota_root = fs_info->quota_root;
 	fs_info->quota_root = NULL;
-	btrfs_free_qgroup_config(fs_info);
 	spin_unlock(&fs_info->qgroup_lock);
 
-	if (!quota_root) {
-		ret = -EINVAL;
-		goto out;
-	}
+	btrfs_free_qgroup_config(fs_info);
 
 	ret = btrfs_clean_quota_tree(trans, quota_root);
 	if (ret)
@@ -1174,7 +1162,7 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans,
 	if (ret) {
 		fs_info->qgroup_flags |= BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
 		printk(KERN_INFO "unable to update quota limit for %llu\n",
-		       (unsigned long long)qgroupid);
+		       qgroupid);
 	}
 
 	spin_lock(&fs_info->qgroup_lock);
@@ -1884,10 +1872,9 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 					 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.objectid,
 		 fs_info->qgroup_rescan_progress.type,
-		 (unsigned long long)fs_info->qgroup_rescan_progress.offset,
-		 ret);
+		 fs_info->qgroup_rescan_progress.offset, ret);
 
 	if (ret) {
 		/*

fs/btrfs/raid56.c

@@ -1540,8 +1540,10 @@ static int full_stripe_write(struct btrfs_raid_bio *rbio)
 	int ret;
 
 	ret = alloc_rbio_parity_pages(rbio);
-	if (ret)
+	if (ret) {
+		__free_raid_bio(rbio);
 		return ret;
+	}
 
 	ret = lock_stripe_add(rbio);
 	if (ret == 0)
@@ -1687,11 +1689,8 @@ int raid56_parity_write(struct btrfs_root *root, struct bio *bio,
 	struct blk_plug_cb *cb;
 
 	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
-	if (IS_ERR(rbio)) {
-		kfree(raid_map);
-		kfree(bbio);
+	if (IS_ERR(rbio))
 		return PTR_ERR(rbio);
-	}
 	bio_list_add(&rbio->bio_list, bio);
 	rbio->bio_list_bytes = bio->bi_size;
 
@@ -2041,9 +2040,8 @@ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio,
 	int ret;
 
 	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
-	if (IS_ERR(rbio)) {
+	if (IS_ERR(rbio))
 		return PTR_ERR(rbio);
-	}
 
 	rbio->read_rebuild = 1;
 	bio_list_add(&rbio->bio_list, bio);
@@ -2052,6 +2050,8 @@ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio,
 	rbio->faila = find_logical_bio_stripe(rbio, bio);
 	if (rbio->faila == -1) {
 		BUG();
+		kfree(raid_map);
+		kfree(bbio);
 		kfree(rbio);
 		return -EIO;
 	}

fs/btrfs/relocation.c

@@ -335,7 +335,7 @@ static void backref_tree_panic(struct rb_node *rb_node, int errno, u64 bytenr)
 	if (bnode->root)
 		fs_info = bnode->root->fs_info;
 	btrfs_panic(fs_info, errno, "Inconsistency in backref cache "
-		    "found at offset %llu\n", (unsigned long long)bytenr);
+		    "found at offset %llu\n", bytenr);
 }
 
 /*
@@ -641,6 +641,11 @@ int find_inline_backref(struct extent_buffer *leaf, int slot,
 		WARN_ON(item_size < sizeof(*ei) + sizeof(*bi));
 		return 1;
 	}
+	if (key.type == BTRFS_METADATA_ITEM_KEY &&
+	    item_size <= sizeof(*ei)) {
+		WARN_ON(item_size < sizeof(*ei));
+		return 1;
+	}
 
 	if (key.type == BTRFS_EXTENT_ITEM_KEY) {
 		bi = (struct btrfs_tree_block_info *)(ei + 1);
@@ -691,6 +696,7 @@ struct backref_node *build_backref_tree(struct reloc_control *rc,
 	int cowonly;
 	int ret;
 	int err = 0;
+	bool need_check = true;
 
 	path1 = btrfs_alloc_path();
 	path2 = btrfs_alloc_path();
@@ -914,6 +920,7 @@ again:
 			cur->bytenr);
 
 		lower = cur;
+		need_check = true;
 		for (; level < BTRFS_MAX_LEVEL; level++) {
 			if (!path2->nodes[level]) {
 				BUG_ON(btrfs_root_bytenr(&root->root_item) !=
@@ -957,14 +964,12 @@ again:
 
 				/*
 				 * add the block to pending list if we
-				 * need check its backrefs. only block
-				 * at 'cur->level + 1' is added to the
-				 * tail of pending list. this guarantees
-				 * we check backrefs from lower level
-				 * blocks to upper level blocks.
+				 * need check its backrefs, we only do this once
+				 * while walking up a tree as we will catch
+				 * anything else later on.
 				 */
-				if (!upper->checked &&
-				    level == cur->level + 1) {
+				if (!upper->checked && need_check) {
+					need_check = false;
 					list_add_tail(&edge->list[UPPER],
 						      &list);
 				} else
@@ -2314,8 +2319,13 @@ again:
 			BUG_ON(root->reloc_root != reloc_root);
 
 			ret = merge_reloc_root(rc, root);
-			if (ret)
+			if (ret) {
+				__update_reloc_root(reloc_root, 1);
+				free_extent_buffer(reloc_root->node);
+				free_extent_buffer(reloc_root->commit_root);
+				kfree(reloc_root);
 				goto out;
+			}
 		} else {
 			list_del_init(&reloc_root->root_list);
 		}
@@ -2344,9 +2354,6 @@ again:
 			if (IS_ERR(root))
 				continue;
 
-			if (btrfs_root_refs(&root->root_item) == 0)
-				continue;
-
 			trans = btrfs_join_transaction(root);
 			BUG_ON(IS_ERR(trans));
 
@@ -3628,7 +3635,7 @@ int add_data_references(struct reloc_control *rc,
 	unsigned long ptr;
 	unsigned long end;
 	u32 blocksize = btrfs_level_size(rc->extent_root, 0);
-	int ret;
+	int ret = 0;
 	int err = 0;
 
 	eb = path->nodes[0];
@@ -3655,6 +3662,10 @@ int add_data_references(struct reloc_control *rc,
 		} else {
 			BUG();
 		}
+		if (ret) {
+			err = ret;
+			goto out;
+		}
 		ptr += btrfs_extent_inline_ref_size(key.type);
 	}
 	WARN_ON(ptr > end);
@@ -3700,6 +3711,7 @@ int add_data_references(struct reloc_control *rc,
 		}
 		path->slots[0]++;
 	}
+out:
 	btrfs_release_path(path);
 	if (err)
 		free_block_list(blocks);
@@ -4219,8 +4231,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 	}
 
 	printk(KERN_INFO "btrfs: relocating block group %llu flags %llu\n",
-	       (unsigned long long)rc->block_group->key.objectid,
-	       (unsigned long long)rc->block_group->flags);
+	       rc->block_group->key.objectid, rc->block_group->flags);
 
 	ret = btrfs_start_all_delalloc_inodes(fs_info, 0);
 	if (ret < 0) {
@@ -4242,7 +4253,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start)
 			break;
 
 		printk(KERN_INFO "btrfs: found %llu extents\n",
-			(unsigned long long)rc->extents_found);
+			rc->extents_found);
 
 		if (rc->stage == MOVE_DATA_EXTENTS && rc->found_file_extent) {
 			btrfs_wait_ordered_range(rc->data_inode, 0, (u64)-1);

fs/btrfs/root-tree.c

@@ -29,8 +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 extent_buffer *eb, int slot,
-			  struct btrfs_root_item *item)
+static void btrfs_read_root_item(struct extent_buffer *eb, int slot,
+				struct btrfs_root_item *item)
 {
 	uuid_le uuid;
 	int len;
@@ -155,8 +155,7 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root
 	if (ret != 0) {
 		btrfs_print_leaf(root, path->nodes[0]);
 		printk(KERN_CRIT "unable to update root key %llu %u %llu\n",
-		       (unsigned long long)key->objectid, key->type,
-		       (unsigned long long)key->offset);
+		       key->objectid, key->type, key->offset);
 		BUG_ON(1);
 	}
 
@@ -490,13 +489,13 @@ again:
  */
 void btrfs_check_and_init_root_item(struct btrfs_root_item *root_item)
 {
-	u64 inode_flags = le64_to_cpu(root_item->inode.flags);
+	u64 inode_flags = btrfs_stack_inode_flags(&root_item->inode);
 
 	if (!(inode_flags & BTRFS_INODE_ROOT_ITEM_INIT)) {
 		inode_flags |= BTRFS_INODE_ROOT_ITEM_INIT;
-		root_item->inode.flags = cpu_to_le64(inode_flags);
-		root_item->flags = 0;
-		root_item->byte_limit = 0;
+		btrfs_set_stack_inode_flags(&root_item->inode, inode_flags);
+		btrfs_set_root_flags(root_item, 0);
+		btrfs_set_root_limit(root_item, 0);
 	}
 }
 
@@ -507,8 +506,8 @@ void btrfs_update_root_times(struct btrfs_trans_handle *trans,
 	struct timespec ct = CURRENT_TIME;
 
 	spin_lock(&root->root_item_lock);
-	item->ctransid = cpu_to_le64(trans->transid);
-	item->ctime.sec = cpu_to_le64(ct.tv_sec);
-	item->ctime.nsec = cpu_to_le32(ct.tv_nsec);
+	btrfs_set_root_ctransid(item, trans->transid);
+	btrfs_set_stack_timespec_sec(&item->ctime, ct.tv_sec);
+	btrfs_set_stack_timespec_nsec(&item->ctime, ct.tv_nsec);
 	spin_unlock(&root->root_item_lock);
 }

fs/btrfs/scrub.c

@@ -754,8 +754,7 @@ out:
 			num_uncorrectable_read_errors);
 		printk_ratelimited_in_rcu(KERN_ERR
 			"btrfs: unable to fixup (nodatasum) error at logical %llu on dev %s\n",
-			(unsigned long long)fixup->logical,
-			rcu_str_deref(fixup->dev->name));
+			fixup->logical, rcu_str_deref(fixup->dev->name));
 	}
 
 	btrfs_free_path(path);
@@ -1154,8 +1153,7 @@ corrected_error:
 			spin_unlock(&sctx->stat_lock);
 			printk_ratelimited_in_rcu(KERN_ERR
 				"btrfs: fixed up error at logical %llu on dev %s\n",
-				(unsigned long long)logical,
-				rcu_str_deref(dev->name));
+				logical, rcu_str_deref(dev->name));
 		}
 	} else {
 did_not_correct_error:
@@ -1164,8 +1162,7 @@ did_not_correct_error:
 		spin_unlock(&sctx->stat_lock);
 		printk_ratelimited_in_rcu(KERN_ERR
 			"btrfs: unable to fixup (regular) error at logical %llu on dev %s\n",
-			(unsigned long long)logical,
-			rcu_str_deref(dev->name));
+			logical, rcu_str_deref(dev->name));
 	}
 
 out:
@@ -1345,12 +1342,12 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info,
 		mapped_buffer = kmap_atomic(sblock->pagev[0]->page);
 		h = (struct btrfs_header *)mapped_buffer;
 
-		if (sblock->pagev[0]->logical != le64_to_cpu(h->bytenr) ||
+		if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h) ||
 		    memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE) ||
 		    memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid,
 			   BTRFS_UUID_SIZE)) {
 			sblock->header_error = 1;
-		} else if (generation != le64_to_cpu(h->generation)) {
+		} else if (generation != btrfs_stack_header_generation(h)) {
 			sblock->header_error = 1;
 			sblock->generation_error = 1;
 		}
@@ -1720,10 +1717,10 @@ static int scrub_checksum_tree_block(struct scrub_block *sblock)
 	 * b) the page is already kmapped
 	 */
 
-	if (sblock->pagev[0]->logical != le64_to_cpu(h->bytenr))
+	if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h))
 		++fail;
 
-	if (sblock->pagev[0]->generation != le64_to_cpu(h->generation))
+	if (sblock->pagev[0]->generation != btrfs_stack_header_generation(h))
 		++fail;
 
 	if (memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
@@ -1786,10 +1783,10 @@ static int scrub_checksum_super(struct scrub_block *sblock)
 	s = (struct btrfs_super_block *)mapped_buffer;
 	memcpy(on_disk_csum, s->csum, sctx->csum_size);
 
-	if (sblock->pagev[0]->logical != le64_to_cpu(s->bytenr))
+	if (sblock->pagev[0]->logical != btrfs_super_bytenr(s))
 		++fail_cor;
 
-	if (sblock->pagev[0]->generation != le64_to_cpu(s->generation))
+	if (sblock->pagev[0]->generation != btrfs_super_generation(s))
 		++fail_gen;
 
 	if (memcmp(s->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
@@ -2455,8 +2452,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 				printk(KERN_ERR
 				       "btrfs scrub: tree block %llu spanning "
 				       "stripes, ignored. logical=%llu\n",
-				       (unsigned long long)key.objectid,
-				       (unsigned long long)logical);
+				       key.objectid, logical);
 				goto next;
 			}
 
@@ -2863,9 +2859,8 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start,
 	if (fs_info->chunk_root->sectorsize != PAGE_SIZE) {
 		/* not supported for data w/o checksums */
 		printk(KERN_ERR
-		       "btrfs_scrub: size assumption sectorsize != PAGE_SIZE (%d != %lld) fails\n",
-		       fs_info->chunk_root->sectorsize,
-		       (unsigned long long)PAGE_SIZE);
+		       "btrfs_scrub: size assumption sectorsize != PAGE_SIZE (%d != %lu) fails\n",
+		       fs_info->chunk_root->sectorsize, PAGE_SIZE);
 		return -EINVAL;
 	}
 
@@ -3175,11 +3170,9 @@ static void copy_nocow_pages_worker(struct btrfs_work *work)
 					  copy_nocow_pages_for_inode,
 					  nocow_ctx);
 	if (ret != 0 && ret != -ENOENT) {
-		pr_warn("iterate_inodes_from_logical() failed: log %llu, phys %llu, len %llu, mir %llu, ret %d\n",
-			(unsigned long long)logical,
-			(unsigned long long)physical_for_dev_replace,
-			(unsigned long long)len,
-			(unsigned long long)mirror_num, ret);
+		pr_warn("iterate_inodes_from_logical() failed: log %llu, phys %llu, len %llu, mir %u, ret %d\n",
+			logical, physical_for_dev_replace, len, mirror_num,
+			ret);
 		not_written = 1;
 		goto out;
 	}
@@ -3224,11 +3217,6 @@ static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root, void *ctx)
 		return PTR_ERR(local_root);
 	}
 
-	if (btrfs_root_refs(&local_root->root_item) == 0) {
-		srcu_read_unlock(&fs_info->subvol_srcu, srcu_index);
-		return -ENOENT;
-	}
-
 	key.type = BTRFS_INODE_ITEM_KEY;
 	key.objectid = inum;
 	key.offset = 0;

fs/btrfs/send.c

@@ -26,6 +26,7 @@
 #include <linux/radix-tree.h>
 #include <linux/crc32c.h>
 #include <linux/vmalloc.h>
+#include <linux/string.h>
 
 #include "send.h"
 #include "backref.h"
@@ -54,8 +55,8 @@ struct fs_path {
 
 			char *buf;
 			int buf_len;
-			int reversed:1;
-			int virtual_mem:1;
+			unsigned int reversed:1;
+			unsigned int virtual_mem:1;
 			char inline_buf[];
 		};
 		char pad[PAGE_SIZE];
@@ -1668,6 +1669,7 @@ static int will_overwrite_ref(struct send_ctx *sctx, u64 dir, u64 dir_gen,
 			      u64 *who_ino, u64 *who_gen)
 {
 	int ret = 0;
+	u64 gen;
 	u64 other_inode = 0;
 	u8 other_type = 0;
 
@@ -1678,6 +1680,24 @@ static int will_overwrite_ref(struct send_ctx *sctx, u64 dir, u64 dir_gen,
 	if (ret <= 0)
 		goto out;
 
+	/*
+	 * If we have a parent root we need to verify that the parent dir was
+	 * not delted and then re-created, if it was then we have no overwrite
+	 * and we can just unlink this entry.
+	 */
+	if (sctx->parent_root) {
+		ret = get_inode_info(sctx->parent_root, dir, NULL, &gen, NULL,
+				     NULL, NULL, NULL);
+		if (ret < 0 && ret != -ENOENT)
+			goto out;
+		if (ret) {
+			ret = 0;
+			goto out;
+		}
+		if (gen != dir_gen)
+			goto out;
+	}
+
 	ret = lookup_dir_item_inode(sctx->parent_root, dir, name, name_len,
 			&other_inode, &other_type);
 	if (ret < 0 && ret != -ENOENT)
@@ -2519,7 +2539,8 @@ static int did_create_dir(struct send_ctx *sctx, u64 dir)
 		di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
 		btrfs_dir_item_key_to_cpu(eb, di, &di_key);
 
-		if (di_key.objectid < sctx->send_progress) {
+		if (di_key.type != BTRFS_ROOT_ITEM_KEY &&
+		    di_key.objectid < sctx->send_progress) {
 			ret = 1;
 			goto out;
 		}
@@ -2581,7 +2602,6 @@ static int record_ref(struct list_head *head, u64 dir,
 		      u64 dir_gen, struct fs_path *path)
 {
 	struct recorded_ref *ref;
-	char *tmp;
 
 	ref = kmalloc(sizeof(*ref), GFP_NOFS);
 	if (!ref)
@@ -2591,25 +2611,35 @@ static int record_ref(struct list_head *head, u64 dir,
 	ref->dir_gen = dir_gen;
 	ref->full_path = path;
 
-	tmp = strrchr(ref->full_path->start, '/');
-	if (!tmp) {
-		ref->name_len = ref->full_path->end - ref->full_path->start;
-		ref->name = ref->full_path->start;
+	ref->name = (char *)kbasename(ref->full_path->start);
+	ref->name_len = ref->full_path->end - ref->name;
+	ref->dir_path = ref->full_path->start;
+	if (ref->name == ref->full_path->start)
 		ref->dir_path_len = 0;
-		ref->dir_path = ref->full_path->start;
-	} else {
-		tmp++;
-		ref->name_len = ref->full_path->end - tmp;
-		ref->name = tmp;
-		ref->dir_path = ref->full_path->start;
+	else
 		ref->dir_path_len = ref->full_path->end -
 				ref->full_path->start - 1 - ref->name_len;
-	}
 
 	list_add_tail(&ref->list, head);
 	return 0;
 }
 
+static int dup_ref(struct recorded_ref *ref, struct list_head *list)
+{
+	struct recorded_ref *new;
+
+	new = kmalloc(sizeof(*ref), GFP_NOFS);
+	if (!new)
+		return -ENOMEM;
+
+	new->dir = ref->dir;
+	new->dir_gen = ref->dir_gen;
+	new->full_path = NULL;
+	INIT_LIST_HEAD(&new->list);
+	list_add_tail(&new->list, list);
+	return 0;
+}
+
 static void __free_recorded_refs(struct list_head *head)
 {
 	struct recorded_ref *cur;
@@ -2724,9 +2754,7 @@ static int process_recorded_refs(struct send_ctx *sctx)
 	int ret = 0;
 	struct recorded_ref *cur;
 	struct recorded_ref *cur2;
-	struct ulist *check_dirs = NULL;
-	struct ulist_iterator uit;
-	struct ulist_node *un;
+	struct list_head check_dirs;
 	struct fs_path *valid_path = NULL;
 	u64 ow_inode = 0;
 	u64 ow_gen;
@@ -2740,6 +2768,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 	 * which is always '..'
 	 */
 	BUG_ON(sctx->cur_ino <= BTRFS_FIRST_FREE_OBJECTID);
+	INIT_LIST_HEAD(&check_dirs);
 
 	valid_path = fs_path_alloc();
 	if (!valid_path) {
@@ -2747,12 +2776,6 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 		goto out;
 	}
 
-	check_dirs = ulist_alloc(GFP_NOFS);
-	if (!check_dirs) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
 	/*
 	 * First, check if the first ref of the current inode was overwritten
 	 * before. If yes, we know that the current inode was already orphanized
@@ -2889,8 +2912,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 					goto out;
 			}
 		}
-		ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-				GFP_NOFS);
+		ret = dup_ref(cur, &check_dirs);
 		if (ret < 0)
 			goto out;
 	}
@@ -2918,8 +2940,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 		}
 
 		list_for_each_entry(cur, &sctx->deleted_refs, list) {
-			ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-					GFP_NOFS);
+			ret = dup_ref(cur, &check_dirs);
 			if (ret < 0)
 				goto out;
 		}
@@ -2930,8 +2951,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 		 */
 		cur = list_entry(sctx->deleted_refs.next, struct recorded_ref,
 				list);
-		ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-				GFP_NOFS);
+		ret = dup_ref(cur, &check_dirs);
 		if (ret < 0)
 			goto out;
 	} else if (!S_ISDIR(sctx->cur_inode_mode)) {
@@ -2951,12 +2971,10 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 				if (ret < 0)
 					goto out;
 			}
-			ret = ulist_add(check_dirs, cur->dir, cur->dir_gen,
-					GFP_NOFS);
+			ret = dup_ref(cur, &check_dirs);
 			if (ret < 0)
 				goto out;
 		}
-
 		/*
 		 * If the inode is still orphan, unlink the orphan. This may
 		 * happen when a previous inode did overwrite the first ref
@@ -2978,33 +2996,32 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 	 * deletion and if it's finally possible to perform the rmdir now.
 	 * We also update the inode stats of the parent dirs here.
 	 */
-	ULIST_ITER_INIT(&uit);
-	while ((un = ulist_next(check_dirs, &uit))) {
+	list_for_each_entry(cur, &check_dirs, list) {
 		/*
 		 * In case we had refs into dirs that were not processed yet,
 		 * we don't need to do the utime and rmdir logic for these dirs.
 		 * The dir will be processed later.
 		 */
-		if (un->val > sctx->cur_ino)
+		if (cur->dir > sctx->cur_ino)
 			continue;
 
-		ret = get_cur_inode_state(sctx, un->val, un->aux);
+		ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
 		if (ret < 0)
 			goto out;
 
 		if (ret == inode_state_did_create ||
 		    ret == inode_state_no_change) {
 			/* TODO delayed utimes */
-			ret = send_utimes(sctx, un->val, un->aux);
+			ret = send_utimes(sctx, cur->dir, cur->dir_gen);
 			if (ret < 0)
 				goto out;
 		} else if (ret == inode_state_did_delete) {
-			ret = can_rmdir(sctx, un->val, sctx->cur_ino);
+			ret = can_rmdir(sctx, cur->dir, sctx->cur_ino);
 			if (ret < 0)
 				goto out;
 			if (ret) {
-				ret = get_cur_path(sctx, un->val, un->aux,
-						valid_path);
+				ret = get_cur_path(sctx, cur->dir,
+						   cur->dir_gen, valid_path);
 				if (ret < 0)
 					goto out;
 				ret = send_rmdir(sctx, valid_path);
@@ -3017,8 +3034,8 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);
 	ret = 0;
 
 out:
+	__free_recorded_refs(&check_dirs);
 	free_recorded_refs(sctx);
-	ulist_free(check_dirs);
 	fs_path_free(valid_path);
 	return ret;
 }
@@ -3119,6 +3136,8 @@ out:
 
 struct find_ref_ctx {
 	u64 dir;
+	u64 dir_gen;
+	struct btrfs_root *root;
 	struct fs_path *name;
 	int found_idx;
 };
@@ -3128,9 +3147,21 @@ static int __find_iref(int num, u64 dir, int index,
 		       void *ctx_)
 {
 	struct find_ref_ctx *ctx = ctx_;
+	u64 dir_gen;
+	int ret;
 
 	if (dir == ctx->dir && fs_path_len(name) == fs_path_len(ctx->name) &&
 	    strncmp(name->start, ctx->name->start, fs_path_len(name)) == 0) {
+		/*
+		 * To avoid doing extra lookups we'll only do this if everything
+		 * else matches.
+		 */
+		ret = get_inode_info(ctx->root, dir, NULL, &dir_gen, NULL,
+				     NULL, NULL, NULL);
+		if (ret)
+			return ret;
+		if (dir_gen != ctx->dir_gen)
+			return 0;
 		ctx->found_idx = num;
 		return 1;
 	}
@@ -3140,14 +3171,16 @@ static int __find_iref(int num, u64 dir, int index,
 static int find_iref(struct btrfs_root *root,
 		     struct btrfs_path *path,
 		     struct btrfs_key *key,
-		     u64 dir, struct fs_path *name)
+		     u64 dir, u64 dir_gen, struct fs_path *name)
 {
 	int ret;
 	struct find_ref_ctx ctx;
 
 	ctx.dir = dir;
 	ctx.name = name;
+	ctx.dir_gen = dir_gen;
 	ctx.found_idx = -1;
+	ctx.root = root;
 
 	ret = iterate_inode_ref(root, path, key, 0, __find_iref, &ctx);
 	if (ret < 0)
@@ -3163,11 +3196,17 @@ static int __record_changed_new_ref(int num, u64 dir, int index,
 				    struct fs_path *name,
 				    void *ctx)
 {
+	u64 dir_gen;
 	int ret;
 	struct send_ctx *sctx = ctx;
 
+	ret = get_inode_info(sctx->send_root, dir, NULL, &dir_gen, NULL,
+			     NULL, NULL, NULL);
+	if (ret)
+		return ret;
+
 	ret = find_iref(sctx->parent_root, sctx->right_path,
-			sctx->cmp_key, dir, name);
+			sctx->cmp_key, dir, dir_gen, name);
 	if (ret == -ENOENT)
 		ret = __record_new_ref(num, dir, index, name, sctx);
 	else if (ret > 0)
@@ -3180,11 +3219,17 @@ static int __record_changed_deleted_ref(int num, u64 dir, int index,
 					struct fs_path *name,
 					void *ctx)
 {
+	u64 dir_gen;
 	int ret;
 	struct send_ctx *sctx = ctx;
 
+	ret = get_inode_info(sctx->parent_root, dir, NULL, &dir_gen, NULL,
+			     NULL, NULL, NULL);
+	if (ret)
+		return ret;
+
 	ret = find_iref(sctx->send_root, sctx->left_path, sctx->cmp_key,
-			dir, name);
+			dir, dir_gen, name);
 	if (ret == -ENOENT)
 		ret = __record_deleted_ref(num, dir, index, name, sctx);
 	else if (ret > 0)
@@ -3869,7 +3914,8 @@ static int is_extent_unchanged(struct send_ctx *sctx,
 	btrfs_item_key_to_cpu(eb, &found_key, slot);
 	if (found_key.objectid != key.objectid ||
 	    found_key.type != key.type) {
-		ret = 0;
+		/* If we're a hole then just pretend nothing changed */
+		ret = (left_disknr) ? 0 : 1;
 		goto out;
 	}
 
@@ -3895,7 +3941,8 @@ static int is_extent_unchanged(struct send_ctx *sctx,
 		 * This may only happen on the first iteration.
 		 */
 		if (found_key.offset + right_len <= ekey->offset) {
-			ret = 0;
+			/* If we're a hole just pretend nothing changed */
+			ret = (left_disknr) ? 0 : 1;
 			goto out;
 		}
 
@@ -3960,8 +4007,8 @@ static int process_extent(struct send_ctx *sctx,
 			  struct btrfs_path *path,
 			  struct btrfs_key *key)
 {
-	int ret = 0;
 	struct clone_root *found_clone = NULL;
+	int ret = 0;
 
 	if (S_ISLNK(sctx->cur_inode_mode))
 		return 0;
@@ -3974,6 +4021,32 @@ static int process_extent(struct send_ctx *sctx,
 			ret = 0;
 			goto out;
 		}
+	} else {
+		struct btrfs_file_extent_item *ei;
+		u8 type;
+
+		ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
+				    struct btrfs_file_extent_item);
+		type = btrfs_file_extent_type(path->nodes[0], ei);
+		if (type == BTRFS_FILE_EXTENT_PREALLOC ||
+		    type == BTRFS_FILE_EXTENT_REG) {
+			/*
+			 * The send spec does not have a prealloc command yet,
+			 * so just leave a hole for prealloc'ed extents until
+			 * we have enough commands queued up to justify rev'ing
+			 * the send spec.
+			 */
+			if (type == BTRFS_FILE_EXTENT_PREALLOC) {
+				ret = 0;
+				goto out;
+			}
+
+			/* Have a hole, just skip it. */
+			if (btrfs_file_extent_disk_bytenr(path->nodes[0], ei) == 0) {
+				ret = 0;
+				goto out;
+			}
+		}
 	}
 
 	ret = find_extent_clone(sctx, path, key->objectid, key->offset,
@@ -4361,6 +4434,64 @@ static int changed_extent(struct send_ctx *sctx,
 	return ret;
 }
 
+static int dir_changed(struct send_ctx *sctx, u64 dir)
+{
+	u64 orig_gen, new_gen;
+	int ret;
+
+	ret = get_inode_info(sctx->send_root, dir, NULL, &new_gen, NULL, NULL,
+			     NULL, NULL);
+	if (ret)
+		return ret;
+
+	ret = get_inode_info(sctx->parent_root, dir, NULL, &orig_gen, NULL,
+			     NULL, NULL, NULL);
+	if (ret)
+		return ret;
+
+	return (orig_gen != new_gen) ? 1 : 0;
+}
+
+static int compare_refs(struct send_ctx *sctx, struct btrfs_path *path,
+			struct btrfs_key *key)
+{
+	struct btrfs_inode_extref *extref;
+	struct extent_buffer *leaf;
+	u64 dirid = 0, last_dirid = 0;
+	unsigned long ptr;
+	u32 item_size;
+	u32 cur_offset = 0;
+	int ref_name_len;
+	int ret = 0;
+
+	/* Easy case, just check this one dirid */
+	if (key->type == BTRFS_INODE_REF_KEY) {
+		dirid = key->offset;
+
+		ret = dir_changed(sctx, dirid);
+		goto out;
+	}
+
+	leaf = path->nodes[0];
+	item_size = btrfs_item_size_nr(leaf, path->slots[0]);
+	ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
+	while (cur_offset < item_size) {
+		extref = (struct btrfs_inode_extref *)(ptr +
+						       cur_offset);
+		dirid = btrfs_inode_extref_parent(leaf, extref);
+		ref_name_len = btrfs_inode_extref_name_len(leaf, extref);
+		cur_offset += ref_name_len + sizeof(*extref);
+		if (dirid == last_dirid)
+			continue;
+		ret = dir_changed(sctx, dirid);
+		if (ret)
+			break;
+		last_dirid = dirid;
+	}
+out:
+	return ret;
+}
+
 /*
  * Updates compare related fields in sctx and simply forwards to the actual
  * changed_xxx functions.
@@ -4376,6 +4507,19 @@ static int changed_cb(struct btrfs_root *left_root,
 	int ret = 0;
 	struct send_ctx *sctx = ctx;
 
+	if (result == BTRFS_COMPARE_TREE_SAME) {
+		if (key->type != BTRFS_INODE_REF_KEY &&
+		    key->type != BTRFS_INODE_EXTREF_KEY)
+			return 0;
+		ret = compare_refs(sctx, left_path, key);
+		if (!ret)
+			return 0;
+		if (ret < 0)
+			return ret;
+		result = BTRFS_COMPARE_TREE_CHANGED;
+		ret = 0;
+	}
+
 	sctx->left_path = left_path;
 	sctx->right_path = right_path;
 	sctx->cmp_key = key;

fs/btrfs/super.c

@@ -56,6 +56,8 @@
 #include "rcu-string.h"
 #include "dev-replace.h"
 #include "free-space-cache.h"
+#include "backref.h"
+#include "tests/btrfs-tests.h"
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/btrfs.h>
@@ -320,14 +322,15 @@ enum {
 	Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, Opt_inode_cache,
 	Opt_no_space_cache, Opt_recovery, Opt_skip_balance,
 	Opt_check_integrity, Opt_check_integrity_including_extent_data,
-	Opt_check_integrity_print_mask, Opt_fatal_errors,
+	Opt_check_integrity_print_mask, Opt_fatal_errors, Opt_rescan_uuid_tree,
+	Opt_commit_interval,
 	Opt_err,
 };
 
 static match_table_t tokens = {
 	{Opt_degraded, "degraded"},
 	{Opt_subvol, "subvol=%s"},
-	{Opt_subvolid, "subvolid=%d"},
+	{Opt_subvolid, "subvolid=%s"},
 	{Opt_device, "device=%s"},
 	{Opt_nodatasum, "nodatasum"},
 	{Opt_nodatacow, "nodatacow"},
@@ -360,7 +363,9 @@ static match_table_t tokens = {
 	{Opt_check_integrity, "check_int"},
 	{Opt_check_integrity_including_extent_data, "check_int_data"},
 	{Opt_check_integrity_print_mask, "check_int_print_mask=%d"},
+	{Opt_rescan_uuid_tree, "rescan_uuid_tree"},
 	{Opt_fatal_errors, "fatal_errors=%s"},
+	{Opt_commit_interval, "commit=%d"},
 	{Opt_err, NULL},
 };
 
@@ -496,10 +501,15 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			btrfs_set_opt(info->mount_opt, NOBARRIER);
 			break;
 		case Opt_thread_pool:
-			intarg = 0;
-			match_int(&args[0], &intarg);
-			if (intarg)
+			ret = match_int(&args[0], &intarg);
+			if (ret) {
+				goto out;
+			} else if (intarg > 0) {
 				info->thread_pool_size = intarg;
+			} else {
+				ret = -EINVAL;
+				goto out;
+			}
 			break;
 		case Opt_max_inline:
 			num = match_strdup(&args[0]);
@@ -513,7 +523,10 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 						root->sectorsize);
 				}
 				printk(KERN_INFO "btrfs: max_inline at %llu\n",
-					(unsigned long long)info->max_inline);
+					info->max_inline);
+			} else {
+				ret = -ENOMEM;
+				goto out;
 			}
 			break;
 		case Opt_alloc_start:
@@ -525,7 +538,10 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				kfree(num);
 				printk(KERN_INFO
 					"btrfs: allocations start at %llu\n",
-					(unsigned long long)info->alloc_start);
+					info->alloc_start);
+			} else {
+				ret = -ENOMEM;
+				goto out;
 			}
 			break;
 		case Opt_noacl:
@@ -540,12 +556,16 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			btrfs_set_opt(info->mount_opt, FLUSHONCOMMIT);
 			break;
 		case Opt_ratio:
-			intarg = 0;
-			match_int(&args[0], &intarg);
-			if (intarg) {
+			ret = match_int(&args[0], &intarg);
+			if (ret) {
+				goto out;
+			} else if (intarg >= 0) {
 				info->metadata_ratio = intarg;
 				printk(KERN_INFO "btrfs: metadata ratio %d\n",
 				       info->metadata_ratio);
+			} else {
+				ret = -EINVAL;
+				goto out;
 			}
 			break;
 		case Opt_discard:
@@ -554,6 +574,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 		case Opt_space_cache:
 			btrfs_set_opt(info->mount_opt, SPACE_CACHE);
 			break;
+		case Opt_rescan_uuid_tree:
+			btrfs_set_opt(info->mount_opt, RESCAN_UUID_TREE);
+			break;
 		case Opt_no_space_cache:
 			printk(KERN_INFO "btrfs: disabling disk space caching\n");
 			btrfs_clear_opt(info->mount_opt, SPACE_CACHE);
@@ -596,13 +619,17 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY);
 			break;
 		case Opt_check_integrity_print_mask:
-			intarg = 0;
-			match_int(&args[0], &intarg);
-			if (intarg) {
+			ret = match_int(&args[0], &intarg);
+			if (ret) {
+				goto out;
+			} else if (intarg >= 0) {
 				info->check_integrity_print_mask = intarg;
 				printk(KERN_INFO "btrfs:"
 				       " check_integrity_print_mask 0x%x\n",
 				       info->check_integrity_print_mask);
+			} else {
+				ret = -EINVAL;
+				goto out;
 			}
 			break;
 #else
@@ -626,6 +653,29 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				goto out;
 			}
 			break;
+		case Opt_commit_interval:
+			intarg = 0;
+			ret = match_int(&args[0], &intarg);
+			if (ret < 0) {
+				printk(KERN_ERR
+					"btrfs: invalid commit interval\n");
+				ret = -EINVAL;
+				goto out;
+			}
+			if (intarg > 0) {
+				if (intarg > 300) {
+					printk(KERN_WARNING
+					    "btrfs: excessive commit interval %d\n",
+							intarg);
+				}
+				info->commit_interval = intarg;
+			} else {
+				printk(KERN_INFO
+				    "btrfs: using default commit interval %ds\n",
+				    BTRFS_DEFAULT_COMMIT_INTERVAL);
+				info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
+			}
+			break;
 		case Opt_err:
 			printk(KERN_INFO "btrfs: unrecognized mount option "
 			       "'%s'\n", p);
@@ -654,8 +704,8 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags,
 {
 	substring_t args[MAX_OPT_ARGS];
 	char *device_name, *opts, *orig, *p;
+	char *num = NULL;
 	int error = 0;
-	int intarg;
 
 	if (!options)
 		return 0;
@@ -679,17 +729,23 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags,
 		case Opt_subvol:
 			kfree(*subvol_name);
 			*subvol_name = match_strdup(&args[0]);
+			if (!*subvol_name) {
+				error = -ENOMEM;
+				goto out;
+			}
 			break;
 		case Opt_subvolid:
-			intarg = 0;
-			error = match_int(&args[0], &intarg);
-			if (!error) {
+			num = match_strdup(&args[0]);
+			if (num) {
+				*subvol_objectid = memparse(num, NULL);
+				kfree(num);
 				/* we want the original fs_tree */
-				if (!intarg)
+				if (!*subvol_objectid)
 					*subvol_objectid =
 						BTRFS_FS_TREE_OBJECTID;
-				else
-					*subvol_objectid = intarg;
+			} else {
+				error = -EINVAL;
+				goto out;
 			}
 			break;
 		case Opt_subvolrootid:
@@ -892,11 +948,9 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 	if (btrfs_test_opt(root, NOBARRIER))
 		seq_puts(seq, ",nobarrier");
 	if (info->max_inline != 8192 * 1024)
-		seq_printf(seq, ",max_inline=%llu",
-			   (unsigned long long)info->max_inline);
+		seq_printf(seq, ",max_inline=%llu", info->max_inline);
 	if (info->alloc_start != 0)
-		seq_printf(seq, ",alloc_start=%llu",
-			   (unsigned long long)info->alloc_start);
+		seq_printf(seq, ",alloc_start=%llu", info->alloc_start);
 	if (info->thread_pool_size !=  min_t(unsigned long,
 					     num_online_cpus() + 2, 8))
 		seq_printf(seq, ",thread_pool=%d", info->thread_pool_size);
@@ -928,6 +982,8 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",space_cache");
 	else
 		seq_puts(seq, ",nospace_cache");
+	if (btrfs_test_opt(root, RESCAN_UUID_TREE))
+		seq_puts(seq, ",rescan_uuid_tree");
 	if (btrfs_test_opt(root, CLEAR_CACHE))
 		seq_puts(seq, ",clear_cache");
 	if (btrfs_test_opt(root, USER_SUBVOL_RM_ALLOWED))
@@ -940,8 +996,24 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",inode_cache");
 	if (btrfs_test_opt(root, SKIP_BALANCE))
 		seq_puts(seq, ",skip_balance");
+	if (btrfs_test_opt(root, RECOVERY))
+		seq_puts(seq, ",recovery");
+#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
+	if (btrfs_test_opt(root, CHECK_INTEGRITY_INCLUDING_EXTENT_DATA))
+		seq_puts(seq, ",check_int_data");
+	else if (btrfs_test_opt(root, CHECK_INTEGRITY))
+		seq_puts(seq, ",check_int");
+	if (info->check_integrity_print_mask)
+		seq_printf(seq, ",check_int_print_mask=%d",
+				info->check_integrity_print_mask);
+#endif
+	if (info->metadata_ratio)
+		seq_printf(seq, ",metadata_ratio=%d",
+				info->metadata_ratio);
 	if (btrfs_test_opt(root, PANIC_ON_FATAL_ERROR))
 		seq_puts(seq, ",fatal_errors=panic");
+	if (info->commit_interval != BTRFS_DEFAULT_COMMIT_INTERVAL)
+		seq_printf(seq, ",commit=%d", info->commit_interval);
 	return 0;
 }
 
@@ -1696,6 +1768,11 @@ static void btrfs_print_info(void)
 			"\n");
 }
 
+static int btrfs_run_sanity_tests(void)
+{
+	return btrfs_test_free_space_cache();
+}
+
 static int __init init_btrfs_fs(void)
 {
 	int err;
@@ -1734,23 +1811,32 @@ static int __init init_btrfs_fs(void)
 	if (err)
 		goto free_auto_defrag;
 
-	err = btrfs_interface_init();
+	err = btrfs_prelim_ref_init();
 	if (err)
-		goto free_delayed_ref;
+		goto free_prelim_ref;
 
-	err = register_filesystem(&btrfs_fs_type);
+	err = btrfs_interface_init();
 	if (err)
-		goto unregister_ioctl;
+		goto free_delayed_ref;
 
 	btrfs_init_lockdep();
 
 	btrfs_print_info();
-	btrfs_test_free_space_cache();
+
+	err = btrfs_run_sanity_tests();
+	if (err)
+		goto unregister_ioctl;
+
+	err = register_filesystem(&btrfs_fs_type);
+	if (err)
+		goto unregister_ioctl;
 
 	return 0;
 
 unregister_ioctl:
 	btrfs_interface_exit();
+free_prelim_ref:
+	btrfs_prelim_ref_exit();
 free_delayed_ref:
 	btrfs_delayed_ref_exit();
 free_auto_defrag:
@@ -1777,6 +1863,7 @@ static void __exit exit_btrfs_fs(void)
 	btrfs_delayed_ref_exit();
 	btrfs_auto_defrag_exit();
 	btrfs_delayed_inode_exit();
+	btrfs_prelim_ref_exit();
 	ordered_data_exit();
 	extent_map_exit();
 	extent_io_exit();

fs/btrfs/tests/btrfs-tests.h

@@ -0,0 +1,34 @@
+/*
+ * Copyright (C) 2013 Fusion IO.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#ifndef __BTRFS_TESTS
+#define __BTRFS_TESTS
+
+#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
+
+#define test_msg(fmt, ...) pr_info("btrfs: selftest: " fmt, ##__VA_ARGS__)
+
+int btrfs_test_free_space_cache(void);
+#else
+static inline int btrfs_test_free_space_cache(void)
+{
+	return 0;
+}
+#endif
+
+#endif

fs/btrfs/tests/free-space-tests.c

@@ -0,0 +1,395 @@
+/*
+ * Copyright (C) 2013 Fusion IO.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+#include <linux/slab.h>
+#include "btrfs-tests.h"
+#include "../ctree.h"
+#include "../free-space-cache.h"
+
+#define BITS_PER_BITMAP		(PAGE_CACHE_SIZE * 8)
+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;
+}
+
+/*
+ * 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;
+
+	test_msg("Running extent only tests\n");
+
+	/* First just make sure we can remove an entire entry */
+	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error adding initial extents %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing extent %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 4 * 1024 * 1024)) {
+		test_msg("Full remove left some lingering space\n");
+		return -1;
+	}
+
+	/* Ok edge and middle cases now */
+	ret = btrfs_add_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error adding half extent %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing tail end %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing front end %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 2 * 1024 * 1024, 4096);
+	if (ret) {
+		test_msg("Error removing middle peice %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 1 * 1024 * 1024)) {
+		test_msg("Still have space at the front\n");
+		return -1;
+	}
+
+	if (test_check_exists(cache, 2 * 1024 * 1024, 4096)) {
+		test_msg("Still have space in the middle\n");
+		return -1;
+	}
+
+	if (test_check_exists(cache, 3 * 1024 * 1024, 1 * 1024 * 1024)) {
+		test_msg("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;
+
+	test_msg("Running bitmap only tests\n");
+
+	ret = test_add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't create a bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing bitmap full range %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 4 * 1024 * 1024)) {
+		test_msg("Left some space in bitmap\n");
+		return -1;
+	}
+
+	ret = test_add_free_space_entry(cache, 0, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add to our bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 2 * 1024 * 1024);
+	if (ret) {
+		test_msg("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 = test_add_free_space_entry(cache, next_bitmap_offset -
+				   (2 * 1024 * 1024), 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add space that straddles two bitmaps %d\n",
+				ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, next_bitmap_offset -
+				      (1 * 1024 * 1024), 2 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, next_bitmap_offset - (1 * 1024 * 1024),
+			 2 * 1024 * 1024)) {
+		test_msg("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;
+
+	test_msg("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 = test_add_free_space_entry(cache, 4 * 1024 * 1024, 1 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't create bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 0, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove extent entry %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 0, 1 * 1024 * 1024)) {
+		test_msg("Left remnants after our remove\n");
+		return -1;
+	}
+
+	/* Now to add back the extent entry and remove from the bitmap */
+	ret = test_add_free_space_entry(cache, 0, 1 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't re-add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 4 * 1024 * 1024, 1 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove from bitmap %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 4 * 1024 * 1024, 1 * 1024 * 1024)) {
+		test_msg("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 = test_add_free_space_entry(cache, 1 * 1024 * 1024, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add to a bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 512 * 1024, 3 * 1024 * 1024);
+	if (ret) {
+		test_msg("Couldn't remove overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 512 * 1024, 3 * 1024 * 1024)) {
+		test_msg("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 = test_add_free_space_entry(cache, 4 * 1024 * 1024, 4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add space to the bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, 2 * 1024 * 1024, 2 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent to the cache %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 3 * 1024 * 1024, 4 * 1024 * 1024);
+	if (ret) {
+		test_msg("Problem removing overlapping space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, 3 * 1024 * 1024, 4 * 1024 * 1024)) {
+		test_msg("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 = test_add_free_space_entry(cache, bitmap_offset + 4 * 1024 * 1024,
+				   4 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add bitmap %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, bitmap_offset - 1 * 1024 * 1024,
+				   5 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, bitmap_offset + 1 * 1024 * 1024,
+				      5 * 1024 * 1024);
+	if (ret) {
+		test_msg("Failed to free our space %d\n", ret);
+		return ret;
+	}
+
+	if (test_check_exists(cache, bitmap_offset + 1 * 1024 * 1024,
+			 5 * 1024 * 1024)) {
+		test_msg("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 = test_add_free_space_entry(cache, 1 * 1024 * 1024, 2 * 1024 * 1024, 1);
+	if (ret) {
+		test_msg("Couldn't add bitmap entry %d\n", ret);
+		return ret;
+	}
+
+	ret = test_add_free_space_entry(cache, 3 * 1024 * 1024, 1 * 1024 * 1024, 0);
+	if (ret) {
+		test_msg("Couldn't add extent entry %d\n", ret);
+		return ret;
+	}
+
+	ret = btrfs_remove_free_space(cache, 1 * 1024 * 1024, 3 * 1024 * 1024);
+	if (ret) {
+		test_msg("Error removing bitmap and extent overlapping %d\n", ret);
+		return ret;
+	}
+
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	return 0;
+}
+
+int btrfs_test_free_space_cache(void)
+{
+	struct btrfs_block_group_cache *cache;
+	int ret;
+
+	test_msg("Running btrfs free space cache tests\n");
+
+	cache = init_test_block_group();
+	if (!cache) {
+		test_msg("Couldn't run the tests\n");
+		return 0;
+	}
+
+	ret = test_extents(cache);
+	if (ret)
+		goto out;
+	ret = test_bitmaps(cache);
+	if (ret)
+		goto out;
+	ret = test_bitmaps_and_extents(cache);
+	if (ret)
+		goto out;
+out:
+	__btrfs_remove_free_space_cache(cache->free_space_ctl);
+	kfree(cache->free_space_ctl);
+	kfree(cache);
+	test_msg("Free space cache tests finished\n");
+	return ret;
+}

fs/btrfs/transaction.c

@@ -837,7 +837,7 @@ int btrfs_wait_marked_extents(struct btrfs_root *root,
  * them in one of two extent_io trees.  This is used to make sure all of
  * those extents are on disk for transaction or log commit
  */
-int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
+static int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
 				struct extent_io_tree *dirty_pages, int mark)
 {
 	int ret;
@@ -1225,8 +1225,8 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 		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_stack_timespec_sec(&new_root_item->otime, cur_time.tv_sec);
+	btrfs_set_stack_timespec_nsec(&new_root_item->otime, cur_time.tv_nsec);
 	btrfs_set_root_otransid(new_root_item, trans->transid);
 
 	old = btrfs_lock_root_node(root);
@@ -1311,8 +1311,26 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 					 dentry->d_name.len * 2);
 	parent_inode->i_mtime = parent_inode->i_ctime = CURRENT_TIME;
 	ret = btrfs_update_inode_fallback(trans, parent_root, parent_inode);
-	if (ret)
+	if (ret) {
+		btrfs_abort_transaction(trans, root, ret);
+		goto fail;
+	}
+	ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root, new_uuid.b,
+				  BTRFS_UUID_KEY_SUBVOL, objectid);
+	if (ret) {
 		btrfs_abort_transaction(trans, root, ret);
+		goto fail;
+	}
+	if (!btrfs_is_empty_uuid(new_root_item->received_uuid)) {
+		ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
+					  new_root_item->received_uuid,
+					  BTRFS_UUID_KEY_RECEIVED_SUBVOL,
+					  objectid);
+		if (ret && ret != -EEXIST) {
+			btrfs_abort_transaction(trans, root, ret);
+			goto fail;
+		}
+	}
 fail:
 	pending->error = ret;
 dir_item_existed:
@@ -1362,6 +1380,8 @@ static void update_super_roots(struct btrfs_root *root)
 	super->root_level = root_item->level;
 	if (btrfs_test_opt(root, SPACE_CACHE))
 		super->cache_generation = root_item->generation;
+	if (root->fs_info->update_uuid_tree_gen)
+		super->uuid_tree_generation = root_item->generation;
 }
 
 int btrfs_transaction_in_commit(struct btrfs_fs_info *info)
@@ -1928,8 +1948,7 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 	list_del_init(&root->root_list);
 	spin_unlock(&fs_info->trans_lock);
 
-	pr_debug("btrfs: cleaner removing %llu\n",
-			(unsigned long long)root->objectid);
+	pr_debug("btrfs: cleaner removing %llu\n", root->objectid);
 
 	btrfs_kill_all_delayed_nodes(root);
 
@@ -1942,6 +1961,5 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 	 * 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;
+	return (ret < 0) ? 0 : 1;
 }

fs/btrfs/transaction.h

@@ -160,8 +160,6 @@ int btrfs_should_end_transaction(struct btrfs_trans_handle *trans,
 void btrfs_throttle(struct btrfs_root *root);
 int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root);
-int btrfs_write_and_wait_marked_extents(struct btrfs_root *root,
-				struct extent_io_tree *dirty_pages, int mark);
 int btrfs_write_marked_extents(struct btrfs_root *root,
 				struct extent_io_tree *dirty_pages, int mark);
 int btrfs_wait_marked_extents(struct btrfs_root *root,

fs/btrfs/tree-log.c

@@ -747,7 +747,8 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
 	ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
 	if (ret)
 		goto out;
-	btrfs_run_delayed_items(trans, root);
+	else
+		ret = btrfs_run_delayed_items(trans, root);
 out:
 	kfree(name);
 	iput(inode);
@@ -923,7 +924,9 @@ again:
 				kfree(victim_name);
 				if (ret)
 					return ret;
-				btrfs_run_delayed_items(trans, root);
+				ret = btrfs_run_delayed_items(trans, root);
+				if (ret)
+					return ret;
 				*search_done = 1;
 				goto again;
 			}
@@ -990,7 +993,9 @@ again:
 								 inode,
 								 victim_name,
 								 victim_name_len);
-					btrfs_run_delayed_items(trans, root);
+					if (!ret)
+						ret = btrfs_run_delayed_items(
+								  trans, root);
 				}
 				iput(victim_parent);
 				kfree(victim_name);
@@ -1536,8 +1541,10 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans,
 
 	name_len = btrfs_dir_name_len(eb, di);
 	name = kmalloc(name_len, GFP_NOFS);
-	if (!name)
-		return -ENOMEM;
+	if (!name) {
+		ret = -ENOMEM;
+		goto out;
+	}
 
 	log_type = btrfs_dir_type(eb, di);
 	read_extent_buffer(eb, name, (unsigned long)(di + 1),
@@ -1810,7 +1817,7 @@ again:
 			ret = btrfs_unlink_inode(trans, root, dir, inode,
 						 name, name_len);
 			if (!ret)
-				btrfs_run_delayed_items(trans, root);
+				ret = btrfs_run_delayed_items(trans, root);
 			kfree(name);
 			iput(inode);
 			if (ret)

fs/btrfs/uuid-tree.c

@@ -0,0 +1,358 @@
+/*
+ * Copyright (C) STRATO AG 2013.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+#include <linux/uuid.h>
+#include <asm/unaligned.h>
+#include "ctree.h"
+#include "transaction.h"
+#include "disk-io.h"
+#include "print-tree.h"
+
+
+static void btrfs_uuid_to_key(u8 *uuid, u8 type, struct btrfs_key *key)
+{
+	key->type = type;
+	key->objectid = get_unaligned_le64(uuid);
+	key->offset = get_unaligned_le64(uuid + sizeof(u64));
+}
+
+/* return -ENOENT for !found, < 0 for errors, or 0 if an item was found */
+static int btrfs_uuid_tree_lookup(struct btrfs_root *uuid_root, u8 *uuid,
+				  u8 type, u64 subid)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct extent_buffer *eb;
+	int slot;
+	u32 item_size;
+	unsigned long offset;
+	struct btrfs_key key;
+
+	if (WARN_ON_ONCE(!uuid_root)) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, type, &key);
+	ret = btrfs_search_slot(NULL, uuid_root, &key, path, 0, 0);
+	if (ret < 0) {
+		goto out;
+	} else if (ret > 0) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	eb = path->nodes[0];
+	slot = path->slots[0];
+	item_size = btrfs_item_size_nr(eb, slot);
+	offset = btrfs_item_ptr_offset(eb, slot);
+	ret = -ENOENT;
+
+	if (!IS_ALIGNED(item_size, sizeof(u64))) {
+		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+			(unsigned long)item_size);
+		goto out;
+	}
+	while (item_size) {
+		__le64 data;
+
+		read_extent_buffer(eb, &data, offset, sizeof(data));
+		if (le64_to_cpu(data) == subid) {
+			ret = 0;
+			break;
+		}
+		offset += sizeof(data);
+		item_size -= sizeof(data);
+	}
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+int btrfs_uuid_tree_add(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid_cpu)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct btrfs_key key;
+	struct extent_buffer *eb;
+	int slot;
+	unsigned long offset;
+	__le64 subid_le;
+
+	ret = btrfs_uuid_tree_lookup(uuid_root, uuid, type, subid_cpu);
+	if (ret != -ENOENT)
+		return ret;
+
+	if (WARN_ON_ONCE(!uuid_root)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, type, &key);
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = btrfs_insert_empty_item(trans, uuid_root, path, &key,
+				      sizeof(subid_le));
+	if (ret >= 0) {
+		/* Add an item for the type for the first time */
+		eb = path->nodes[0];
+		slot = path->slots[0];
+		offset = btrfs_item_ptr_offset(eb, slot);
+	} else if (ret == -EEXIST) {
+		/*
+		 * An item with that type already exists.
+		 * Extend the item and store the new subid at the end.
+		 */
+		btrfs_extend_item(uuid_root, path, sizeof(subid_le));
+		eb = path->nodes[0];
+		slot = path->slots[0];
+		offset = btrfs_item_ptr_offset(eb, slot);
+		offset += btrfs_item_size_nr(eb, slot) - sizeof(subid_le);
+	} else if (ret < 0) {
+		pr_warn("btrfs: insert uuid item failed %d (0x%016llx, 0x%016llx) type %u!\n",
+			ret, (unsigned long long)key.objectid,
+			(unsigned long long)key.offset, type);
+		goto out;
+	}
+
+	ret = 0;
+	subid_le = cpu_to_le64(subid_cpu);
+	write_extent_buffer(eb, &subid_le, offset, sizeof(subid_le));
+	btrfs_mark_buffer_dirty(eb);
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+int btrfs_uuid_tree_rem(struct btrfs_trans_handle *trans,
+			struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			u64 subid)
+{
+	int ret;
+	struct btrfs_path *path = NULL;
+	struct btrfs_key key;
+	struct extent_buffer *eb;
+	int slot;
+	unsigned long offset;
+	u32 item_size;
+	unsigned long move_dst;
+	unsigned long move_src;
+	unsigned long move_len;
+
+	if (WARN_ON_ONCE(!uuid_root)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	btrfs_uuid_to_key(uuid, type, &key);
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = btrfs_search_slot(trans, uuid_root, &key, path, -1, 1);
+	if (ret < 0) {
+		pr_warn("btrfs: error %d while searching for uuid item!\n",
+			ret);
+		goto out;
+	}
+	if (ret > 0) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	eb = path->nodes[0];
+	slot = path->slots[0];
+	offset = btrfs_item_ptr_offset(eb, slot);
+	item_size = btrfs_item_size_nr(eb, slot);
+	if (!IS_ALIGNED(item_size, sizeof(u64))) {
+		pr_warn("btrfs: uuid item with illegal size %lu!\n",
+			(unsigned long)item_size);
+		ret = -ENOENT;
+		goto out;
+	}
+	while (item_size) {
+		__le64 read_subid;
+
+		read_extent_buffer(eb, &read_subid, offset, sizeof(read_subid));
+		if (le64_to_cpu(read_subid) == subid)
+			break;
+		offset += sizeof(read_subid);
+		item_size -= sizeof(read_subid);
+	}
+
+	if (!item_size) {
+		ret = -ENOENT;
+		goto out;
+	}
+
+	item_size = btrfs_item_size_nr(eb, slot);
+	if (item_size == sizeof(subid)) {
+		ret = btrfs_del_item(trans, uuid_root, path);
+		goto out;
+	}
+
+	move_dst = offset;
+	move_src = offset + sizeof(subid);
+	move_len = item_size - (move_src - btrfs_item_ptr_offset(eb, slot));
+	memmove_extent_buffer(eb, move_dst, move_src, move_len);
+	btrfs_truncate_item(uuid_root, path, item_size - sizeof(subid), 1);
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
+static int btrfs_uuid_iter_rem(struct btrfs_root *uuid_root, u8 *uuid, u8 type,
+			       u64 subid)
+{
+	struct btrfs_trans_handle *trans;
+	int ret;
+
+	/* 1 - for the uuid item */
+	trans = btrfs_start_transaction(uuid_root, 1);
+	if (IS_ERR(trans)) {
+		ret = PTR_ERR(trans);
+		goto out;
+	}
+
+	ret = btrfs_uuid_tree_rem(trans, uuid_root, uuid, type, subid);
+	btrfs_end_transaction(trans, uuid_root);
+
+out:
+	return ret;
+}
+
+int btrfs_uuid_tree_iterate(struct btrfs_fs_info *fs_info,
+			    int (*check_func)(struct btrfs_fs_info *, u8 *, u8,
+					      u64))
+{
+	struct btrfs_root *root = fs_info->uuid_root;
+	struct btrfs_key key;
+	struct btrfs_key max_key;
+	struct btrfs_path *path;
+	int ret = 0;
+	struct extent_buffer *leaf;
+	int slot;
+	u32 item_size;
+	unsigned long offset;
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	key.objectid = 0;
+	key.type = 0;
+	key.offset = 0;
+	max_key.objectid = (u64)-1;
+	max_key.type = (u8)-1;
+	max_key.offset = (u64)-1;
+
+again_search_slot:
+	path->keep_locks = 1;
+	ret = btrfs_search_forward(root, &key, &max_key, path, 0);
+	if (ret) {
+		if (ret > 0)
+			ret = 0;
+		goto out;
+	}
+
+	while (1) {
+		cond_resched();
+		leaf = path->nodes[0];
+		slot = path->slots[0];
+		btrfs_item_key_to_cpu(leaf, &key, slot);
+
+		if (key.type != BTRFS_UUID_KEY_SUBVOL &&
+		    key.type != BTRFS_UUID_KEY_RECEIVED_SUBVOL)
+			goto skip;
+
+		offset = btrfs_item_ptr_offset(leaf, slot);
+		item_size = btrfs_item_size_nr(leaf, slot);
+		if (!IS_ALIGNED(item_size, sizeof(u64))) {
+			pr_warn("btrfs: uuid item with illegal size %lu!\n",
+				(unsigned long)item_size);
+			goto skip;
+		}
+		while (item_size) {
+			u8 uuid[BTRFS_UUID_SIZE];
+			__le64 subid_le;
+			u64 subid_cpu;
+
+			put_unaligned_le64(key.objectid, uuid);
+			put_unaligned_le64(key.offset, uuid + sizeof(u64));
+			read_extent_buffer(leaf, &subid_le, offset,
+					   sizeof(subid_le));
+			subid_cpu = le64_to_cpu(subid_le);
+			ret = check_func(fs_info, uuid, key.type, subid_cpu);
+			if (ret < 0)
+				goto out;
+			if (ret > 0) {
+				btrfs_release_path(path);
+				ret = btrfs_uuid_iter_rem(root, uuid, key.type,
+							  subid_cpu);
+				if (ret == 0) {
+					/*
+					 * this might look inefficient, but the
+					 * justification is that it is an
+					 * exception that check_func returns 1,
+					 * and that in the regular case only one
+					 * entry per UUID exists.
+					 */
+					goto again_search_slot;
+				}
+				if (ret < 0 && ret != -ENOENT)
+					goto out;
+			}
+			item_size -= sizeof(subid_le);
+			offset += sizeof(subid_le);
+		}
+
+skip:
+		ret = btrfs_next_item(root, path);
+		if (ret == 0)
+			continue;
+		else if (ret > 0)
+			ret = 0;
+		break;
+	}
+
+out:
+	btrfs_free_path(path);
+	if (ret)
+		pr_warn("btrfs: btrfs_uuid_tree_iterate failed %d\n", ret);
+	return 0;
+}

fs/btrfs/volumes.c

@@ -26,6 +26,7 @@
 #include <linux/ratelimit.h>
 #include <linux/kthread.h>
 #include <linux/raid/pq.h>
+#include <linux/semaphore.h>
 #include <asm/div64.h>
 #include "compat.h"
 #include "ctree.h"
@@ -62,6 +63,48 @@ static void unlock_chunks(struct btrfs_root *root)
 	mutex_unlock(&root->fs_info->chunk_mutex);
 }
 
+static struct btrfs_fs_devices *__alloc_fs_devices(void)
+{
+	struct btrfs_fs_devices *fs_devs;
+
+	fs_devs = kzalloc(sizeof(*fs_devs), GFP_NOFS);
+	if (!fs_devs)
+		return ERR_PTR(-ENOMEM);
+
+	mutex_init(&fs_devs->device_list_mutex);
+
+	INIT_LIST_HEAD(&fs_devs->devices);
+	INIT_LIST_HEAD(&fs_devs->alloc_list);
+	INIT_LIST_HEAD(&fs_devs->list);
+
+	return fs_devs;
+}
+
+/**
+ * alloc_fs_devices - allocate struct btrfs_fs_devices
+ * @fsid:	a pointer to UUID for this FS.  If NULL a new UUID is
+ *		generated.
+ *
+ * Return: a pointer to a new &struct btrfs_fs_devices on success;
+ * ERR_PTR() on error.  Returned struct is not linked onto any lists and
+ * can be destroyed with kfree() right away.
+ */
+static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid)
+{
+	struct btrfs_fs_devices *fs_devs;
+
+	fs_devs = __alloc_fs_devices();
+	if (IS_ERR(fs_devs))
+		return fs_devs;
+
+	if (fsid)
+		memcpy(fs_devs->fsid, fsid, BTRFS_FSID_SIZE);
+	else
+		generate_random_uuid(fs_devs->fsid);
+
+	return fs_devs;
+}
+
 static void free_fs_devices(struct btrfs_fs_devices *fs_devices)
 {
 	struct btrfs_device *device;
@@ -101,6 +144,27 @@ void btrfs_cleanup_fs_uuids(void)
 	}
 }
 
+static struct btrfs_device *__alloc_device(void)
+{
+	struct btrfs_device *dev;
+
+	dev = kzalloc(sizeof(*dev), GFP_NOFS);
+	if (!dev)
+		return ERR_PTR(-ENOMEM);
+
+	INIT_LIST_HEAD(&dev->dev_list);
+	INIT_LIST_HEAD(&dev->dev_alloc_list);
+
+	spin_lock_init(&dev->io_lock);
+
+	spin_lock_init(&dev->reada_lock);
+	atomic_set(&dev->reada_in_flight, 0);
+	INIT_RADIX_TREE(&dev->reada_zones, GFP_NOFS & ~__GFP_WAIT);
+	INIT_RADIX_TREE(&dev->reada_extents, GFP_NOFS & ~__GFP_WAIT);
+
+	return dev;
+}
+
 static noinline struct btrfs_device *__find_device(struct list_head *head,
 						   u64 devid, u8 *uuid)
 {
@@ -395,16 +459,14 @@ static noinline int device_list_add(const char *path,
 
 	fs_devices = find_fsid(disk_super->fsid);
 	if (!fs_devices) {
-		fs_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
-		if (!fs_devices)
-			return -ENOMEM;
-		INIT_LIST_HEAD(&fs_devices->devices);
-		INIT_LIST_HEAD(&fs_devices->alloc_list);
+		fs_devices = alloc_fs_devices(disk_super->fsid);
+		if (IS_ERR(fs_devices))
+			return PTR_ERR(fs_devices);
+
 		list_add(&fs_devices->list, &fs_uuids);
-		memcpy(fs_devices->fsid, disk_super->fsid, BTRFS_FSID_SIZE);
 		fs_devices->latest_devid = devid;
 		fs_devices->latest_trans = found_transid;
-		mutex_init(&fs_devices->device_list_mutex);
+
 		device = NULL;
 	} else {
 		device = __find_device(&fs_devices->devices, devid,
@@ -414,17 +476,12 @@ static noinline int device_list_add(const char *path,
 		if (fs_devices->opened)
 			return -EBUSY;
 
-		device = kzalloc(sizeof(*device), GFP_NOFS);
-		if (!device) {
+		device = btrfs_alloc_device(NULL, &devid,
+					    disk_super->dev_item.uuid);
+		if (IS_ERR(device)) {
 			/* we can safely leave the fs_devices entry around */
-			return -ENOMEM;
+			return PTR_ERR(device);
 		}
-		device->devid = devid;
-		device->dev_stats_valid = 0;
-		device->work.func = pending_bios_fn;
-		memcpy(device->uuid, disk_super->dev_item.uuid,
-		       BTRFS_UUID_SIZE);
-		spin_lock_init(&device->io_lock);
 
 		name = rcu_string_strdup(path, GFP_NOFS);
 		if (!name) {
@@ -432,22 +489,13 @@ static noinline int device_list_add(const char *path,
 			return -ENOMEM;
 		}
 		rcu_assign_pointer(device->name, name);
-		INIT_LIST_HEAD(&device->dev_alloc_list);
-
-		/* init readahead state */
-		spin_lock_init(&device->reada_lock);
-		device->reada_curr_zone = NULL;
-		atomic_set(&device->reada_in_flight, 0);
-		device->reada_next = 0;
-		INIT_RADIX_TREE(&device->reada_zones, GFP_NOFS & ~__GFP_WAIT);
-		INIT_RADIX_TREE(&device->reada_extents, GFP_NOFS & ~__GFP_WAIT);
 
 		mutex_lock(&fs_devices->device_list_mutex);
 		list_add_rcu(&device->dev_list, &fs_devices->devices);
+		fs_devices->num_devices++;
 		mutex_unlock(&fs_devices->device_list_mutex);
 
 		device->fs_devices = fs_devices;
-		fs_devices->num_devices++;
 	} else if (!device->name || strcmp(device->name->str, path)) {
 		name = rcu_string_strdup(path, GFP_NOFS);
 		if (!name)
@@ -474,25 +522,21 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 	struct btrfs_device *device;
 	struct btrfs_device *orig_dev;
 
-	fs_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
-	if (!fs_devices)
-		return ERR_PTR(-ENOMEM);
+	fs_devices = alloc_fs_devices(orig->fsid);
+	if (IS_ERR(fs_devices))
+		return fs_devices;
 
-	INIT_LIST_HEAD(&fs_devices->devices);
-	INIT_LIST_HEAD(&fs_devices->alloc_list);
-	INIT_LIST_HEAD(&fs_devices->list);
-	mutex_init(&fs_devices->device_list_mutex);
 	fs_devices->latest_devid = orig->latest_devid;
 	fs_devices->latest_trans = orig->latest_trans;
 	fs_devices->total_devices = orig->total_devices;
-	memcpy(fs_devices->fsid, orig->fsid, sizeof(fs_devices->fsid));
 
 	/* We have held the volume lock, it is safe to get the devices. */
 	list_for_each_entry(orig_dev, &orig->devices, dev_list) {
 		struct rcu_string *name;
 
-		device = kzalloc(sizeof(*device), GFP_NOFS);
-		if (!device)
+		device = btrfs_alloc_device(NULL, &orig_dev->devid,
+					    orig_dev->uuid);
+		if (IS_ERR(device))
 			goto error;
 
 		/*
@@ -506,13 +550,6 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 		}
 		rcu_assign_pointer(device->name, name);
 
-		device->devid = orig_dev->devid;
-		device->work.func = pending_bios_fn;
-		memcpy(device->uuid, orig_dev->uuid, sizeof(device->uuid));
-		spin_lock_init(&device->io_lock);
-		INIT_LIST_HEAD(&device->dev_list);
-		INIT_LIST_HEAD(&device->dev_alloc_list);
-
 		list_add(&device->dev_list, &fs_devices->devices);
 		device->fs_devices = fs_devices;
 		fs_devices->num_devices++;
@@ -636,23 +673,22 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 
 		if (device->can_discard)
 			fs_devices->num_can_discard--;
+		if (device->missing)
+			fs_devices->missing_devices--;
 
-		new_device = kmalloc(sizeof(*new_device), GFP_NOFS);
-		BUG_ON(!new_device); /* -ENOMEM */
-		memcpy(new_device, device, sizeof(*new_device));
+		new_device = btrfs_alloc_device(NULL, &device->devid,
+						device->uuid);
+		BUG_ON(IS_ERR(new_device)); /* -ENOMEM */
 
 		/* Safe because we are under uuid_mutex */
 		if (device->name) {
 			name = rcu_string_strdup(device->name->str, GFP_NOFS);
-			BUG_ON(device->name && !name); /* -ENOMEM */
+			BUG_ON(!name); /* -ENOMEM */
 			rcu_assign_pointer(new_device->name, name);
 		}
-		new_device->bdev = NULL;
-		new_device->writeable = 0;
-		new_device->in_fs_metadata = 0;
-		new_device->can_discard = 0;
-		spin_lock_init(&new_device->io_lock);
+
 		list_replace_rcu(&device->dev_list, &new_device->dev_list);
+		new_device->fs_devices = device->fs_devices;
 
 		call_rcu(&device->rcu, free_device);
 	}
@@ -865,7 +901,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 	disk_super = p + (bytenr & ~PAGE_CACHE_MASK);
 
 	if (btrfs_super_bytenr(disk_super) != bytenr ||
-	    disk_super->magic != cpu_to_le64(BTRFS_MAGIC))
+	    btrfs_super_magic(disk_super) != BTRFS_MAGIC)
 		goto error_unmap;
 
 	devid = btrfs_stack_device_id(&disk_super->dev_item);
@@ -880,8 +916,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 		printk(KERN_INFO "device fsid %pU ", disk_super->fsid);
 	}
 
-	printk(KERN_CONT "devid %llu transid %llu %s\n",
-	       (unsigned long long)devid, (unsigned long long)transid, path);
+	printk(KERN_CONT "devid %llu transid %llu %s\n", devid, transid, path);
 
 	ret = device_list_add(path, disk_super, devid, fs_devices_ret);
 	if (!ret && fs_devices_ret)
@@ -1278,8 +1313,7 @@ static int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
 	btrfs_set_dev_extent_chunk_offset(leaf, extent, chunk_offset);
 
 	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
-		    (unsigned long)btrfs_dev_extent_chunk_tree_uuid(extent),
-		    BTRFS_UUID_SIZE);
+		    btrfs_dev_extent_chunk_tree_uuid(extent), BTRFS_UUID_SIZE);
 
 	btrfs_set_dev_extent_length(leaf, extent, num_bytes);
 	btrfs_mark_buffer_dirty(leaf);
@@ -1307,15 +1341,14 @@ static u64 find_next_chunk(struct btrfs_fs_info *fs_info)
 	return ret;
 }
 
-static noinline int find_next_devid(struct btrfs_root *root, u64 *objectid)
+static noinline int find_next_devid(struct btrfs_fs_info *fs_info,
+				    u64 *devid_ret)
 {
 	int ret;
 	struct btrfs_key key;
 	struct btrfs_key found_key;
 	struct btrfs_path *path;
 
-	root = root->fs_info->chunk_root;
-
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -1324,20 +1357,21 @@ static noinline int find_next_devid(struct btrfs_root *root, u64 *objectid)
 	key.type = BTRFS_DEV_ITEM_KEY;
 	key.offset = (u64)-1;
 
-	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	ret = btrfs_search_slot(NULL, fs_info->chunk_root, &key, path, 0, 0);
 	if (ret < 0)
 		goto error;
 
 	BUG_ON(ret == 0); /* Corruption */
 
-	ret = btrfs_previous_item(root, path, BTRFS_DEV_ITEMS_OBJECTID,
+	ret = btrfs_previous_item(fs_info->chunk_root, path,
+				  BTRFS_DEV_ITEMS_OBJECTID,
 				  BTRFS_DEV_ITEM_KEY);
 	if (ret) {
-		*objectid = 1;
+		*devid_ret = 1;
 	} else {
 		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
 				      path->slots[0]);
-		*objectid = found_key.offset + 1;
+		*devid_ret = found_key.offset + 1;
 	}
 	ret = 0;
 error:
@@ -1391,9 +1425,9 @@ static int btrfs_add_device(struct btrfs_trans_handle *trans,
 	btrfs_set_device_bandwidth(leaf, dev_item, 0);
 	btrfs_set_device_start_offset(leaf, dev_item, 0);
 
-	ptr = (unsigned long)btrfs_device_uuid(dev_item);
+	ptr = btrfs_device_uuid(dev_item);
 	write_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
-	ptr = (unsigned long)btrfs_device_fsid(dev_item);
+	ptr = btrfs_device_fsid(dev_item);
 	write_extent_buffer(leaf, root->fs_info->fsid, ptr, BTRFS_UUID_SIZE);
 	btrfs_mark_buffer_dirty(leaf);
 
@@ -1562,7 +1596,9 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 		clear_super = true;
 	}
 
+	mutex_unlock(&uuid_mutex);
 	ret = btrfs_shrink_device(device, 0);
+	mutex_lock(&uuid_mutex);
 	if (ret)
 		goto error_undo;
 
@@ -1586,7 +1622,11 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	/*
 	 * the device list mutex makes sure that we don't change
 	 * the device list while someone else is writing out all
-	 * the device supers.
+	 * the device supers. Whoever is writing all supers, should
+	 * lock the device list mutex before getting the number of
+	 * devices in the super block (super_copy). Conversely,
+	 * whoever updates the number of devices in the super block
+	 * (super_copy) should hold the device list mutex.
 	 */
 
 	cur_devices = device->fs_devices;
@@ -1610,10 +1650,10 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 		device->fs_devices->open_devices--;
 
 	call_rcu(&device->rcu, free_device);
-	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	num_devices = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
 	btrfs_set_super_num_devices(root->fs_info->super_copy, num_devices);
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	if (cur_devices->open_devices == 0) {
 		struct btrfs_fs_devices *fs_devices;
@@ -1793,9 +1833,9 @@ static int btrfs_prepare_sprout(struct btrfs_root *root)
 	if (!fs_devices->seeding)
 		return -EINVAL;
 
-	seed_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
-	if (!seed_devices)
-		return -ENOMEM;
+	seed_devices = __alloc_fs_devices();
+	if (IS_ERR(seed_devices))
+		return PTR_ERR(seed_devices);
 
 	old_devices = clone_fs_devices(fs_devices);
 	if (IS_ERR(old_devices)) {
@@ -1814,7 +1854,6 @@ static int btrfs_prepare_sprout(struct btrfs_root *root)
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	list_splice_init_rcu(&fs_devices->devices, &seed_devices->devices,
 			      synchronize_rcu);
-	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 	list_splice_init(&fs_devices->alloc_list, &seed_devices->alloc_list);
 	list_for_each_entry(device, &seed_devices->devices, dev_list) {
@@ -1830,6 +1869,8 @@ static int btrfs_prepare_sprout(struct btrfs_root *root)
 	generate_random_uuid(fs_devices->fsid);
 	memcpy(root->fs_info->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
 	memcpy(disk_super->fsid, fs_devices->fsid, BTRFS_FSID_SIZE);
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+
 	super_flags = btrfs_super_flags(disk_super) &
 		      ~BTRFS_SUPER_FLAG_SEEDING;
 	btrfs_set_super_flags(disk_super, super_flags);
@@ -1889,11 +1930,9 @@ next_slot:
 		dev_item = btrfs_item_ptr(leaf, path->slots[0],
 					  struct btrfs_dev_item);
 		devid = btrfs_device_id(leaf, dev_item);
-		read_extent_buffer(leaf, dev_uuid,
-				   (unsigned long)btrfs_device_uuid(dev_item),
+		read_extent_buffer(leaf, dev_uuid, btrfs_device_uuid(dev_item),
 				   BTRFS_UUID_SIZE);
-		read_extent_buffer(leaf, fs_uuid,
-				   (unsigned long)btrfs_device_fsid(dev_item),
+		read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
 				   BTRFS_UUID_SIZE);
 		device = btrfs_find_device(root->fs_info, devid, dev_uuid,
 					   fs_uuid);
@@ -1956,10 +1995,10 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	}
 	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
-	device = kzalloc(sizeof(*device), GFP_NOFS);
-	if (!device) {
+	device = btrfs_alloc_device(root->fs_info, NULL, NULL);
+	if (IS_ERR(device)) {
 		/* we can safely leave the fs_devices entry around */
-		ret = -ENOMEM;
+		ret = PTR_ERR(device);
 		goto error;
 	}
 
@@ -1971,13 +2010,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	}
 	rcu_assign_pointer(device->name, name);
 
-	ret = find_next_devid(root, &device->devid);
-	if (ret) {
-		rcu_string_free(device->name);
-		kfree(device);
-		goto error;
-	}
-
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		rcu_string_free(device->name);
@@ -1992,9 +2024,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	if (blk_queue_discard(q))
 		device->can_discard = 1;
 	device->writeable = 1;
-	device->work.func = pending_bios_fn;
-	generate_random_uuid(device->uuid);
-	spin_lock_init(&device->io_lock);
 	device->generation = trans->transid;
 	device->io_width = root->sectorsize;
 	device->io_align = root->sectorsize;
@@ -2121,6 +2150,7 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct list_head *devices;
 	struct rcu_string *name;
+	u64 devid = BTRFS_DEV_REPLACE_DEVID;
 	int ret = 0;
 
 	*device_out = NULL;
@@ -2142,9 +2172,9 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 		}
 	}
 
-	device = kzalloc(sizeof(*device), GFP_NOFS);
-	if (!device) {
-		ret = -ENOMEM;
+	device = btrfs_alloc_device(NULL, &devid, NULL);
+	if (IS_ERR(device)) {
+		ret = PTR_ERR(device);
 		goto error;
 	}
 
@@ -2161,10 +2191,6 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path,
 		device->can_discard = 1;
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	device->writeable = 1;
-	device->work.func = pending_bios_fn;
-	generate_random_uuid(device->uuid);
-	device->devid = BTRFS_DEV_REPLACE_DEVID;
-	spin_lock_init(&device->io_lock);
 	device->generation = 0;
 	device->io_width = root->sectorsize;
 	device->io_align = root->sectorsize;
@@ -2971,10 +2997,6 @@ again:
 		if (found_key.objectid != key.objectid)
 			break;
 
-		/* chunk zero is special */
-		if (found_key.offset == 0)
-			break;
-
 		chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
 
 		if (!counting) {
@@ -3010,6 +3032,8 @@ again:
 			spin_unlock(&fs_info->balance_lock);
 		}
 loop:
+		if (found_key.offset == 0)
+			break;
 		key.offset = found_key.offset - 1;
 	}
 
@@ -3074,9 +3098,6 @@ static void __cancel_balance(struct btrfs_fs_info *fs_info)
 	atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
 }
 
-void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
-			       struct btrfs_ioctl_balance_args *bargs);
-
 /*
  * Should be called with both balance and volume mutexes held
  */
@@ -3139,7 +3160,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	     (bctl->data.target & ~allowed))) {
 		printk(KERN_ERR "btrfs: unable to start balance with target "
 		       "data profile %llu\n",
-		       (unsigned long long)bctl->data.target);
+		       bctl->data.target);
 		ret = -EINVAL;
 		goto out;
 	}
@@ -3148,7 +3169,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	     (bctl->meta.target & ~allowed))) {
 		printk(KERN_ERR "btrfs: unable to start balance with target "
 		       "metadata profile %llu\n",
-		       (unsigned long long)bctl->meta.target);
+		       bctl->meta.target);
 		ret = -EINVAL;
 		goto out;
 	}
@@ -3157,7 +3178,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	     (bctl->sys.target & ~allowed))) {
 		printk(KERN_ERR "btrfs: unable to start balance with target "
 		       "system profile %llu\n",
-		       (unsigned long long)bctl->sys.target);
+		       bctl->sys.target);
 		ret = -EINVAL;
 		goto out;
 	}
@@ -3430,6 +3451,264 @@ int btrfs_cancel_balance(struct btrfs_fs_info *fs_info)
 	return 0;
 }
 
+static int btrfs_uuid_scan_kthread(void *data)
+{
+	struct btrfs_fs_info *fs_info = data;
+	struct btrfs_root *root = fs_info->tree_root;
+	struct btrfs_key key;
+	struct btrfs_key max_key;
+	struct btrfs_path *path = NULL;
+	int ret = 0;
+	struct extent_buffer *eb;
+	int slot;
+	struct btrfs_root_item root_item;
+	u32 item_size;
+	struct btrfs_trans_handle *trans = NULL;
+
+	path = btrfs_alloc_path();
+	if (!path) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	key.objectid = 0;
+	key.type = BTRFS_ROOT_ITEM_KEY;
+	key.offset = 0;
+
+	max_key.objectid = (u64)-1;
+	max_key.type = BTRFS_ROOT_ITEM_KEY;
+	max_key.offset = (u64)-1;
+
+	path->keep_locks = 1;
+
+	while (1) {
+		ret = btrfs_search_forward(root, &key, &max_key, path, 0);
+		if (ret) {
+			if (ret > 0)
+				ret = 0;
+			break;
+		}
+
+		if (key.type != BTRFS_ROOT_ITEM_KEY ||
+		    (key.objectid < BTRFS_FIRST_FREE_OBJECTID &&
+		     key.objectid != BTRFS_FS_TREE_OBJECTID) ||
+		    key.objectid > BTRFS_LAST_FREE_OBJECTID)
+			goto skip;
+
+		eb = path->nodes[0];
+		slot = path->slots[0];
+		item_size = btrfs_item_size_nr(eb, slot);
+		if (item_size < sizeof(root_item))
+			goto skip;
+
+		read_extent_buffer(eb, &root_item,
+				   btrfs_item_ptr_offset(eb, slot),
+				   (int)sizeof(root_item));
+		if (btrfs_root_refs(&root_item) == 0)
+			goto skip;
+
+		if (!btrfs_is_empty_uuid(root_item.uuid) ||
+		    !btrfs_is_empty_uuid(root_item.received_uuid)) {
+			if (trans)
+				goto update_tree;
+
+			btrfs_release_path(path);
+			/*
+			 * 1 - subvol uuid item
+			 * 1 - received_subvol uuid item
+			 */
+			trans = btrfs_start_transaction(fs_info->uuid_root, 2);
+			if (IS_ERR(trans)) {
+				ret = PTR_ERR(trans);
+				break;
+			}
+			continue;
+		} else {
+			goto skip;
+		}
+update_tree:
+		if (!btrfs_is_empty_uuid(root_item.uuid)) {
+			ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
+						  root_item.uuid,
+						  BTRFS_UUID_KEY_SUBVOL,
+						  key.objectid);
+			if (ret < 0) {
+				pr_warn("btrfs: uuid_tree_add failed %d\n",
+					ret);
+				break;
+			}
+		}
+
+		if (!btrfs_is_empty_uuid(root_item.received_uuid)) {
+			ret = btrfs_uuid_tree_add(trans, fs_info->uuid_root,
+						  root_item.received_uuid,
+						 BTRFS_UUID_KEY_RECEIVED_SUBVOL,
+						  key.objectid);
+			if (ret < 0) {
+				pr_warn("btrfs: uuid_tree_add failed %d\n",
+					ret);
+				break;
+			}
+		}
+
+skip:
+		if (trans) {
+			ret = btrfs_end_transaction(trans, fs_info->uuid_root);
+			trans = NULL;
+			if (ret)
+				break;
+		}
+
+		btrfs_release_path(path);
+		if (key.offset < (u64)-1) {
+			key.offset++;
+		} else if (key.type < BTRFS_ROOT_ITEM_KEY) {
+			key.offset = 0;
+			key.type = BTRFS_ROOT_ITEM_KEY;
+		} else if (key.objectid < (u64)-1) {
+			key.offset = 0;
+			key.type = BTRFS_ROOT_ITEM_KEY;
+			key.objectid++;
+		} else {
+			break;
+		}
+		cond_resched();
+	}
+
+out:
+	btrfs_free_path(path);
+	if (trans && !IS_ERR(trans))
+		btrfs_end_transaction(trans, fs_info->uuid_root);
+	if (ret)
+		pr_warn("btrfs: btrfs_uuid_scan_kthread failed %d\n", ret);
+	else
+		fs_info->update_uuid_tree_gen = 1;
+	up(&fs_info->uuid_tree_rescan_sem);
+	return 0;
+}
+
+/*
+ * Callback for btrfs_uuid_tree_iterate().
+ * returns:
+ * 0	check succeeded, the entry is not outdated.
+ * < 0	if an error occured.
+ * > 0	if the check failed, which means the caller shall remove the entry.
+ */
+static int btrfs_check_uuid_tree_entry(struct btrfs_fs_info *fs_info,
+				       u8 *uuid, u8 type, u64 subid)
+{
+	struct btrfs_key key;
+	int ret = 0;
+	struct btrfs_root *subvol_root;
+
+	if (type != BTRFS_UUID_KEY_SUBVOL &&
+	    type != BTRFS_UUID_KEY_RECEIVED_SUBVOL)
+		goto out;
+
+	key.objectid = subid;
+	key.type = BTRFS_ROOT_ITEM_KEY;
+	key.offset = (u64)-1;
+	subvol_root = btrfs_read_fs_root_no_name(fs_info, &key);
+	if (IS_ERR(subvol_root)) {
+		ret = PTR_ERR(subvol_root);
+		if (ret == -ENOENT)
+			ret = 1;
+		goto out;
+	}
+
+	switch (type) {
+	case BTRFS_UUID_KEY_SUBVOL:
+		if (memcmp(uuid, subvol_root->root_item.uuid, BTRFS_UUID_SIZE))
+			ret = 1;
+		break;
+	case BTRFS_UUID_KEY_RECEIVED_SUBVOL:
+		if (memcmp(uuid, subvol_root->root_item.received_uuid,
+			   BTRFS_UUID_SIZE))
+			ret = 1;
+		break;
+	}
+
+out:
+	return ret;
+}
+
+static int btrfs_uuid_rescan_kthread(void *data)
+{
+	struct btrfs_fs_info *fs_info = (struct btrfs_fs_info *)data;
+	int ret;
+
+	/*
+	 * 1st step is to iterate through the existing UUID tree and
+	 * to delete all entries that contain outdated data.
+	 * 2nd step is to add all missing entries to the UUID tree.
+	 */
+	ret = btrfs_uuid_tree_iterate(fs_info, btrfs_check_uuid_tree_entry);
+	if (ret < 0) {
+		pr_warn("btrfs: iterating uuid_tree failed %d\n", ret);
+		up(&fs_info->uuid_tree_rescan_sem);
+		return ret;
+	}
+	return btrfs_uuid_scan_kthread(data);
+}
+
+int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_trans_handle *trans;
+	struct btrfs_root *tree_root = fs_info->tree_root;
+	struct btrfs_root *uuid_root;
+	struct task_struct *task;
+	int ret;
+
+	/*
+	 * 1 - root node
+	 * 1 - root item
+	 */
+	trans = btrfs_start_transaction(tree_root, 2);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+
+	uuid_root = btrfs_create_tree(trans, fs_info,
+				      BTRFS_UUID_TREE_OBJECTID);
+	if (IS_ERR(uuid_root)) {
+		btrfs_abort_transaction(trans, tree_root,
+					PTR_ERR(uuid_root));
+		return PTR_ERR(uuid_root);
+	}
+
+	fs_info->uuid_root = uuid_root;
+
+	ret = btrfs_commit_transaction(trans, tree_root);
+	if (ret)
+		return ret;
+
+	down(&fs_info->uuid_tree_rescan_sem);
+	task = kthread_run(btrfs_uuid_scan_kthread, fs_info, "btrfs-uuid");
+	if (IS_ERR(task)) {
+		/* fs_info->update_uuid_tree_gen remains 0 in all error case */
+		pr_warn("btrfs: failed to start uuid_scan task\n");
+		up(&fs_info->uuid_tree_rescan_sem);
+		return PTR_ERR(task);
+	}
+
+	return 0;
+}
+
+int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info)
+{
+	struct task_struct *task;
+
+	down(&fs_info->uuid_tree_rescan_sem);
+	task = kthread_run(btrfs_uuid_rescan_kthread, fs_info, "btrfs-uuid");
+	if (IS_ERR(task)) {
+		/* fs_info->update_uuid_tree_gen remains 0 in all error case */
+		pr_warn("btrfs: failed to start uuid_rescan task\n");
+		up(&fs_info->uuid_tree_rescan_sem);
+		return PTR_ERR(task);
+	}
+
+	return 0;
+}
+
 /*
  * shrinking a device means finding all of the device extents past
  * the new size, and then following the back refs to the chunks.
@@ -4194,13 +4473,13 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len)
 	 * 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,
+		btrfs_crit(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 "
+		btrfs_crit(fs_info, "Invalid mapping for %Lu-%Lu, got "
 			    "%Lu-%Lu\n", logical, logical+len, em->start,
 			    em->start + em->len);
 		return 1;
@@ -4375,8 +4654,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 
 	if (!em) {
 		btrfs_crit(fs_info, "unable to find logical %llu len %llu",
-			(unsigned long long)logical,
-			(unsigned long long)*length);
+			logical, *length);
 		return -EINVAL;
 	}
 
@@ -4671,6 +4949,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
 	}
 	bbio = kzalloc(btrfs_bio_size(num_alloc_stripes), GFP_NOFS);
 	if (!bbio) {
+		kfree(raid_map);
 		ret = -ENOMEM;
 		goto out;
 	}
@@ -5246,9 +5525,7 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
 
 	if (map_length < 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);
+			logical, length, map_length);
 		BUG();
 	}
 
@@ -5314,23 +5591,72 @@ static struct btrfs_device *add_missing_dev(struct btrfs_root *root,
 	struct btrfs_device *device;
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 
-	device = kzalloc(sizeof(*device), GFP_NOFS);
-	if (!device)
+	device = btrfs_alloc_device(NULL, &devid, dev_uuid);
+	if (IS_ERR(device))
 		return NULL;
-	list_add(&device->dev_list,
-		 &fs_devices->devices);
-	device->devid = devid;
-	device->work.func = pending_bios_fn;
+
+	list_add(&device->dev_list, &fs_devices->devices);
 	device->fs_devices = fs_devices;
-	device->missing = 1;
 	fs_devices->num_devices++;
+
+	device->missing = 1;
 	fs_devices->missing_devices++;
-	spin_lock_init(&device->io_lock);
-	INIT_LIST_HEAD(&device->dev_alloc_list);
-	memcpy(device->uuid, dev_uuid, BTRFS_UUID_SIZE);
+
 	return device;
 }
 
+/**
+ * btrfs_alloc_device - allocate struct btrfs_device
+ * @fs_info:	used only for generating a new devid, can be NULL if
+ *		devid is provided (i.e. @devid != NULL).
+ * @devid:	a pointer to devid for this device.  If NULL a new devid
+ *		is generated.
+ * @uuid:	a pointer to UUID for this device.  If NULL a new UUID
+ *		is generated.
+ *
+ * Return: a pointer to a new &struct btrfs_device on success; ERR_PTR()
+ * on error.  Returned struct is not linked onto any lists and can be
+ * destroyed with kfree() right away.
+ */
+struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
+					const u64 *devid,
+					const u8 *uuid)
+{
+	struct btrfs_device *dev;
+	u64 tmp;
+
+	if (!devid && !fs_info) {
+		WARN_ON(1);
+		return ERR_PTR(-EINVAL);
+	}
+
+	dev = __alloc_device();
+	if (IS_ERR(dev))
+		return dev;
+
+	if (devid)
+		tmp = *devid;
+	else {
+		int ret;
+
+		ret = find_next_devid(fs_info, &tmp);
+		if (ret) {
+			kfree(dev);
+			return ERR_PTR(ret);
+		}
+	}
+	dev->devid = tmp;
+
+	if (uuid)
+		memcpy(dev->uuid, uuid, BTRFS_UUID_SIZE);
+	else
+		generate_random_uuid(dev->uuid);
+
+	dev->work.func = pending_bios_fn;
+
+	return dev;
+}
+
 static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
 			  struct extent_buffer *leaf,
 			  struct btrfs_chunk *chunk)
@@ -5437,7 +5763,7 @@ static void fill_device_from_item(struct extent_buffer *leaf,
 	WARN_ON(device->devid == BTRFS_DEV_REPLACE_DEVID);
 	device->is_tgtdev_for_dev_replace = 0;
 
-	ptr = (unsigned long)btrfs_device_uuid(dev_item);
+	ptr = btrfs_device_uuid(dev_item);
 	read_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
 }
 
@@ -5500,11 +5826,9 @@ static int read_one_dev(struct btrfs_root *root,
 	u8 dev_uuid[BTRFS_UUID_SIZE];
 
 	devid = btrfs_device_id(leaf, dev_item);
-	read_extent_buffer(leaf, dev_uuid,
-			   (unsigned long)btrfs_device_uuid(dev_item),
+	read_extent_buffer(leaf, dev_uuid, btrfs_device_uuid(dev_item),
 			   BTRFS_UUID_SIZE);
-	read_extent_buffer(leaf, fs_uuid,
-			   (unsigned long)btrfs_device_fsid(dev_item),
+	read_extent_buffer(leaf, fs_uuid, btrfs_device_fsid(dev_item),
 			   BTRFS_UUID_SIZE);
 
 	if (memcmp(fs_uuid, root->fs_info->fsid, BTRFS_UUID_SIZE)) {
@@ -5519,8 +5843,7 @@ static int read_one_dev(struct btrfs_root *root,
 			return -EIO;
 
 		if (!device) {
-			btrfs_warn(root->fs_info, "devid %llu missing",
-				(unsigned long long)devid);
+			btrfs_warn(root->fs_info, "devid %llu missing", devid);
 			device = add_missing_dev(root, devid, dev_uuid);
 			if (!device)
 				return -ENOMEM;
@@ -5644,14 +5967,15 @@ int btrfs_read_chunk_tree(struct btrfs_root *root)
 	mutex_lock(&uuid_mutex);
 	lock_chunks(root);
 
-	/* first we search for all of the device items, and then we
-	 * read in all of the chunk items.  This way we can create chunk
-	 * mappings that reference all of the devices that are afound
+	/*
+	 * Read all device items, and then all the chunk items. All
+	 * device items are found before any chunk item (their object id
+	 * is smaller than the lowest possible object id for a chunk
+	 * item - BTRFS_FIRST_CHUNK_TREE_OBJECTID).
 	 */
 	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
 	key.offset = 0;
 	key.type = 0;
-again:
 	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 	if (ret < 0)
 		goto error;
@@ -5667,17 +5991,13 @@ again:
 			break;
 		}
 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
-		if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
-			if (found_key.objectid != BTRFS_DEV_ITEMS_OBJECTID)
-				break;
-			if (found_key.type == BTRFS_DEV_ITEM_KEY) {
-				struct btrfs_dev_item *dev_item;
-				dev_item = btrfs_item_ptr(leaf, slot,
+		if (found_key.type == BTRFS_DEV_ITEM_KEY) {
+			struct btrfs_dev_item *dev_item;
+			dev_item = btrfs_item_ptr(leaf, slot,
 						  struct btrfs_dev_item);
-				ret = read_one_dev(root, leaf, dev_item);
-				if (ret)
-					goto error;
-			}
+			ret = read_one_dev(root, leaf, dev_item);
+			if (ret)
+				goto error;
 		} else if (found_key.type == BTRFS_CHUNK_ITEM_KEY) {
 			struct btrfs_chunk *chunk;
 			chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
@@ -5687,11 +6007,6 @@ again:
 		}
 		path->slots[0]++;
 	}
-	if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
-		key.objectid = 0;
-		btrfs_release_path(path);
-		goto again;
-	}
 	ret = 0;
 error:
 	unlock_chunks(root);

fs/btrfs/volumes.h

@@ -152,6 +152,8 @@ struct btrfs_fs_devices {
 	int rotating;
 };
 
+#define BTRFS_BIO_INLINE_CSUM_SIZE	64
+
 /*
  * we need the mirror number and stripe index to be passed around
  * the call chain while we are processing end_io (especially errors).
@@ -161,9 +163,14 @@ struct btrfs_fs_devices {
  * we allocate are actually btrfs_io_bios.  We'll cram as much of
  * struct btrfs_bio as we can into this over time.
  */
+typedef void (btrfs_io_bio_end_io_t) (struct btrfs_io_bio *bio, int err);
 struct btrfs_io_bio {
 	unsigned long mirror_num;
 	unsigned long stripe_index;
+	u8 *csum;
+	u8 csum_inline[BTRFS_BIO_INLINE_CSUM_SIZE];
+	u8 *csum_allocated;
+	btrfs_io_bio_end_io_t *end_io;
 	struct bio bio;
 };
 
@@ -298,6 +305,9 @@ 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);
+struct btrfs_device *btrfs_alloc_device(struct btrfs_fs_info *fs_info,
+					const u64 *devid,
+					const u8 *uuid);
 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);
@@ -315,6 +325,8 @@ int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info);
 int btrfs_recover_balance(struct btrfs_fs_info *fs_info);
 int btrfs_pause_balance(struct btrfs_fs_info *fs_info);
 int btrfs_cancel_balance(struct btrfs_fs_info *fs_info);
+int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info);
+int btrfs_check_uuid_tree(struct btrfs_fs_info *fs_info);
 int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset);
 int find_free_dev_extent(struct btrfs_trans_handle *trans,
 			 struct btrfs_device *device, u64 num_bytes,

include/trace/events/btrfs.h

@@ -439,7 +439,7 @@ TRACE_EVENT(btrfs_sync_fs,
 		{ BTRFS_UPDATE_DELAYED_HEAD, "UPDATE_DELAYED_HEAD" })
 			
 
-TRACE_EVENT(btrfs_delayed_tree_ref,
+DECLARE_EVENT_CLASS(btrfs_delayed_tree_ref,
 
 	TP_PROTO(struct btrfs_delayed_ref_node *ref,
 		 struct btrfs_delayed_tree_ref *full_ref,
@@ -481,7 +481,25 @@ TRACE_EVENT(btrfs_delayed_tree_ref,
 		  (unsigned long long)__entry->seq)
 );
 
-TRACE_EVENT(btrfs_delayed_data_ref,
+DEFINE_EVENT(btrfs_delayed_tree_ref,  add_delayed_tree_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_tree_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DEFINE_EVENT(btrfs_delayed_tree_ref,  run_delayed_tree_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_tree_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DECLARE_EVENT_CLASS(btrfs_delayed_data_ref,
 
 	TP_PROTO(struct btrfs_delayed_ref_node *ref,
 		 struct btrfs_delayed_data_ref *full_ref,
@@ -527,7 +545,25 @@ TRACE_EVENT(btrfs_delayed_data_ref,
 		  (unsigned long long)__entry->seq)
 );
 
-TRACE_EVENT(btrfs_delayed_ref_head,
+DEFINE_EVENT(btrfs_delayed_data_ref,  add_delayed_data_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_data_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DEFINE_EVENT(btrfs_delayed_data_ref,  run_delayed_data_ref,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_data_ref *full_ref,
+		 int action),
+
+	TP_ARGS(ref, full_ref, action)
+);
+
+DECLARE_EVENT_CLASS(btrfs_delayed_ref_head,
 
 	TP_PROTO(struct btrfs_delayed_ref_node *ref,
 		 struct btrfs_delayed_ref_head *head_ref,
@@ -556,6 +592,24 @@ TRACE_EVENT(btrfs_delayed_ref_head,
 		  __entry->is_data)
 );
 
+DEFINE_EVENT(btrfs_delayed_ref_head,  add_delayed_ref_head,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_ref_head *head_ref,
+		 int action),
+
+	TP_ARGS(ref, head_ref, action)
+);
+
+DEFINE_EVENT(btrfs_delayed_ref_head,  run_delayed_ref_head,
+
+	TP_PROTO(struct btrfs_delayed_ref_node *ref,
+		 struct btrfs_delayed_ref_head *head_ref,
+		 int action),
+
+	TP_ARGS(ref, head_ref, action)
+);
+
 #define show_chunk_type(type)					\
 	__print_flags(type, "|",				\
 		{ BTRFS_BLOCK_GROUP_DATA, 	"DATA"	},	\

include/uapi/linux/btrfs.h

@@ -305,6 +305,31 @@ struct btrfs_ioctl_clone_range_args {
 #define BTRFS_DEFRAG_RANGE_COMPRESS 1
 #define BTRFS_DEFRAG_RANGE_START_IO 2
 
+#define BTRFS_SAME_DATA_DIFFERS	1
+/* For extent-same ioctl */
+struct btrfs_ioctl_same_extent_info {
+	__s64 fd;		/* in - destination file */
+	__u64 logical_offset;	/* in - start of extent in destination */
+	__u64 bytes_deduped;	/* out - total # of bytes we were able
+				 * to dedupe from this file */
+	/* status of this dedupe operation:
+	 * 0 if dedup succeeds
+	 * < 0 for error
+	 * == BTRFS_SAME_DATA_DIFFERS if data differs
+	 */
+	__s32 status;		/* out - see above description */
+	__u32 reserved;
+};
+
+struct btrfs_ioctl_same_args {
+	__u64 logical_offset;	/* in - start of extent in source */
+	__u64 length;		/* in - length of extent */
+	__u16 dest_count;	/* in - total elements in info array */
+	__u16 reserved1;
+	__u32 reserved2;
+	struct btrfs_ioctl_same_extent_info info[0];
+};
+
 struct btrfs_ioctl_space_info {
 	__u64 flags;
 	__u64 total_bytes;
@@ -524,7 +549,7 @@ static inline char *btrfs_err_str(enum btrfs_err_code err_code)
 				   struct btrfs_ioctl_search_args)
 #define BTRFS_IOC_INO_LOOKUP _IOWR(BTRFS_IOCTL_MAGIC, 18, \
 				   struct btrfs_ioctl_ino_lookup_args)
-#define BTRFS_IOC_DEFAULT_SUBVOL _IOW(BTRFS_IOCTL_MAGIC, 19, u64)
+#define BTRFS_IOC_DEFAULT_SUBVOL _IOW(BTRFS_IOCTL_MAGIC, 19, __u64)
 #define BTRFS_IOC_SPACE_INFO _IOWR(BTRFS_IOCTL_MAGIC, 20, \
 				    struct btrfs_ioctl_space_args)
 #define BTRFS_IOC_START_SYNC _IOR(BTRFS_IOCTL_MAGIC, 24, __u64)
@@ -579,4 +604,7 @@ static inline char *btrfs_err_str(enum btrfs_err_code err_code)
 				      struct btrfs_ioctl_get_dev_stats)
 #define BTRFS_IOC_DEV_REPLACE _IOWR(BTRFS_IOCTL_MAGIC, 53, \
 				    struct btrfs_ioctl_dev_replace_args)
+#define BTRFS_IOC_FILE_EXTENT_SAME _IOWR(BTRFS_IOCTL_MAGIC, 54, \
+					 struct btrfs_ioctl_same_args)
+
 #endif /* _UAPI_LINUX_BTRFS_H */