Merge branch 'for-linus-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs

Pull btrfs fixes from Chris Mason:
 "We've queued up a few different fixes in here.  These range from
  enospc corners to fsync and quota fixes, and a few targeted at error
  handling for corrupt metadata/fuzzing"

* 'for-linus-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: fix lockdep warning on deadlock against an inode's log mutex
  Btrfs: detect corruption when non-root leaf has zero item
  Btrfs: check btree node's nritems
  btrfs: don't create or leak aliased root while cleaning up orphans
  Btrfs: fix em leak in find_first_block_group
  btrfs: do not background blkdev_put()
  Btrfs: clarify do_chunk_alloc()'s return value
  btrfs: fix fsfreeze hang caused by delayed iputs deal
  btrfs: update btrfs_space_info's bytes_may_use timely
  btrfs: divide btrfs_update_reserved_bytes() into two functions
  btrfs: use correct offset for reloc_inode in prealloc_file_extent_cluster()
  btrfs: qgroup: Fix qgroup incorrectness caused by log replay
  btrfs: relocation: Fix leaking qgroups numbers on data extents
  btrfs: qgroup: Refactor btrfs_qgroup_insert_dirty_extent()
  btrfs: waiting on qgroup rescan should not always be interruptible
  btrfs: properly track when rescan worker is running
  btrfs: flush_space: treat return value of do_chunk_alloc properly
  Btrfs: add ASSERT for block group's memory leak
  btrfs: backref: Fix soft lockup in __merge_refs function
  Btrfs: fix memory leak of reloc_root

fs/btrfs/backref.c

@@ -589,6 +589,7 @@ static void __merge_refs(struct list_head *head, int mode)
 
 			list_del(&ref2->list);
 			kmem_cache_free(btrfs_prelim_ref_cache, ref2);
+			cond_resched();
 		}
 
 	}

fs/btrfs/ctree.h

@@ -1028,6 +1028,7 @@ struct btrfs_fs_info {
 	struct btrfs_workqueue *qgroup_rescan_workers;
 	struct completion qgroup_rescan_completion;
 	struct btrfs_work qgroup_rescan_work;
+	bool qgroup_rescan_running;	/* protected by qgroup_rescan_lock */
 
 	/* filesystem state */
 	unsigned long fs_state;
@@ -1079,6 +1080,8 @@ struct btrfs_fs_info {
 	struct list_head pinned_chunks;
 
 	int creating_free_space_tree;
+	/* Used to record internally whether fs has been frozen */
+	int fs_frozen;
 };
 
 struct btrfs_subvolume_writers {
@@ -2578,7 +2581,7 @@ 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_root *root, u64 num_bytes,
+int btrfs_reserve_extent(struct btrfs_root *root, u64 ram_bytes, u64 num_bytes,
 			 u64 min_alloc_size, u64 empty_size, u64 hint_byte,
 			 struct btrfs_key *ins, int is_data, int delalloc);
 int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,

fs/btrfs/delayed-ref.c

@@ -541,7 +541,6 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	struct btrfs_delayed_ref_head *existing;
 	struct btrfs_delayed_ref_head *head_ref = NULL;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	struct btrfs_qgroup_extent_record *qexisting;
 	int count_mod = 1;
 	int must_insert_reserved = 0;
 
@@ -606,10 +605,8 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 		qrecord->num_bytes = num_bytes;
 		qrecord->old_roots = NULL;
 
-		qexisting = btrfs_qgroup_insert_dirty_extent(fs_info,
-							     delayed_refs,
-							     qrecord);
-		if (qexisting)
+		if(btrfs_qgroup_insert_dirty_extent_nolock(fs_info,
+					delayed_refs, qrecord))
 			kfree(qrecord);
 	}
 

fs/btrfs/disk-io.c

@@ -559,8 +559,29 @@ static noinline int check_leaf(struct btrfs_root *root,
 	u32 nritems = btrfs_header_nritems(leaf);
 	int slot;
 
-	if (nritems == 0)
+	if (nritems == 0) {
+		struct btrfs_root *check_root;
+
+		key.objectid = btrfs_header_owner(leaf);
+		key.type = BTRFS_ROOT_ITEM_KEY;
+		key.offset = (u64)-1;
+
+		check_root = btrfs_get_fs_root(root->fs_info, &key, false);
+		/*
+		 * The only reason we also check NULL here is that during
+		 * open_ctree() some roots has not yet been set up.
+		 */
+		if (!IS_ERR_OR_NULL(check_root)) {
+			/* if leaf is the root, then it's fine */
+			if (leaf->start !=
+			    btrfs_root_bytenr(&check_root->root_item)) {
+				CORRUPT("non-root leaf's nritems is 0",
+					leaf, root, 0);
+				return -EIO;
+			}
+		}
 		return 0;
+	}
 
 	/* Check the 0 item */
 	if (btrfs_item_offset_nr(leaf, 0) + btrfs_item_size_nr(leaf, 0) !=
@@ -612,6 +633,19 @@ static noinline int check_leaf(struct btrfs_root *root,
 	return 0;
 }
 
+static int check_node(struct btrfs_root *root, struct extent_buffer *node)
+{
+	unsigned long nr = btrfs_header_nritems(node);
+
+	if (nr == 0 || nr > BTRFS_NODEPTRS_PER_BLOCK(root)) {
+		btrfs_crit(root->fs_info,
+			   "corrupt node: block %llu root %llu nritems %lu",
+			   node->start, root->objectid, nr);
+		return -EIO;
+	}
+	return 0;
+}
+
 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)
@@ -682,6 +716,9 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 		ret = -EIO;
 	}
 
+	if (found_level > 0 && check_node(root, eb))
+		ret = -EIO;
+
 	if (!ret)
 		set_extent_buffer_uptodate(eb);
 err:
@@ -1618,8 +1655,8 @@ fail:
 	return ret;
 }
 
-static struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
-					       u64 root_id)
+struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
+					u64 root_id)
 {
 	struct btrfs_root *root;
 
@@ -2298,6 +2335,7 @@ static void btrfs_init_qgroup(struct btrfs_fs_info *fs_info)
 	fs_info->quota_enabled = 0;
 	fs_info->pending_quota_state = 0;
 	fs_info->qgroup_ulist = NULL;
+	fs_info->qgroup_rescan_running = false;
 	mutex_init(&fs_info->qgroup_rescan_lock);
 }
 
@@ -2624,6 +2662,7 @@ int open_ctree(struct super_block *sb,
 	atomic_set(&fs_info->qgroup_op_seq, 0);
 	atomic_set(&fs_info->reada_works_cnt, 0);
 	atomic64_set(&fs_info->tree_mod_seq, 0);
+	fs_info->fs_frozen = 0;
 	fs_info->sb = sb;
 	fs_info->max_inline = BTRFS_DEFAULT_MAX_INLINE;
 	fs_info->metadata_ratio = 0;
@@ -3739,8 +3778,15 @@ void btrfs_drop_and_free_fs_root(struct btrfs_fs_info *fs_info,
 	if (btrfs_root_refs(&root->root_item) == 0)
 		synchronize_srcu(&fs_info->subvol_srcu);
 
-	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state))
+	if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) {
 		btrfs_free_log(NULL, root);
+		if (root->reloc_root) {
+			free_extent_buffer(root->reloc_root->node);
+			free_extent_buffer(root->reloc_root->commit_root);
+			btrfs_put_fs_root(root->reloc_root);
+			root->reloc_root = NULL;
+		}
+	}
 
 	if (root->free_ino_pinned)
 		__btrfs_remove_free_space_cache(root->free_ino_pinned);
@@ -3851,7 +3897,7 @@ void close_ctree(struct btrfs_root *root)
 	smp_mb();
 
 	/* wait for the qgroup rescan worker to stop */
-	btrfs_qgroup_wait_for_completion(fs_info);
+	btrfs_qgroup_wait_for_completion(fs_info, false);
 
 	/* wait for the uuid_scan task to finish */
 	down(&fs_info->uuid_tree_rescan_sem);

fs/btrfs/disk-io.h

@@ -68,6 +68,8 @@ struct extent_buffer *btrfs_find_tree_block(struct btrfs_fs_info *fs_info,
 struct btrfs_root *btrfs_read_fs_root(struct btrfs_root *tree_root,
 				      struct btrfs_key *location);
 int btrfs_init_fs_root(struct btrfs_root *root);
+struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
+					u64 root_id);
 int btrfs_insert_fs_root(struct btrfs_fs_info *fs_info,
 			 struct btrfs_root *root);
 void btrfs_free_fs_roots(struct btrfs_fs_info *fs_info);

fs/btrfs/extent-tree.c

@@ -60,21 +60,6 @@ enum {
 	CHUNK_ALLOC_FORCE = 2,
 };
 
-/*
- * Control how reservations are dealt with.
- *
- * RESERVE_FREE - freeing a reservation.
- * RESERVE_ALLOC - allocating space and we need to update bytes_may_use for
- *   ENOSPC accounting
- * RESERVE_ALLOC_NO_ACCOUNT - allocating space and we should not update
- *   bytes_may_use as the ENOSPC accounting is done elsewhere
- */
-enum {
-	RESERVE_FREE = 0,
-	RESERVE_ALLOC = 1,
-	RESERVE_ALLOC_NO_ACCOUNT = 2,
-};
-
 static int update_block_group(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root, u64 bytenr,
 			      u64 num_bytes, int alloc);
@@ -104,9 +89,10 @@ static int find_next_key(struct btrfs_path *path, int level,
 			 struct btrfs_key *key);
 static void dump_space_info(struct btrfs_space_info *info, u64 bytes,
 			    int dump_block_groups);
-static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache,
-				       u64 num_bytes, int reserve,
-				       int delalloc);
+static int btrfs_add_reserved_bytes(struct btrfs_block_group_cache *cache,
+				    u64 ram_bytes, u64 num_bytes, int delalloc);
+static int btrfs_free_reserved_bytes(struct btrfs_block_group_cache *cache,
+				     u64 num_bytes, int delalloc);
 static int block_rsv_use_bytes(struct btrfs_block_rsv *block_rsv,
 			       u64 num_bytes);
 int btrfs_pin_extent(struct btrfs_root *root,
@@ -3501,7 +3487,6 @@ again:
 		dcs = BTRFS_DC_SETUP;
 	else if (ret == -ENOSPC)
 		set_bit(BTRFS_TRANS_CACHE_ENOSPC, &trans->transaction->flags);
-	btrfs_free_reserved_data_space(inode, 0, num_pages);
 
 out_put:
 	iput(inode);
@@ -4472,6 +4457,15 @@ void check_system_chunk(struct btrfs_trans_handle *trans,
 	}
 }
 
+/*
+ * If force is CHUNK_ALLOC_FORCE:
+ *    - return 1 if it successfully allocates a chunk,
+ *    - return errors including -ENOSPC otherwise.
+ * If force is NOT CHUNK_ALLOC_FORCE:
+ *    - return 0 if it doesn't need to allocate a new chunk,
+ *    - return 1 if it successfully allocates a chunk,
+ *    - return errors including -ENOSPC otherwise.
+ */
 static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 			  struct btrfs_root *extent_root, u64 flags, int force)
 {
@@ -4882,7 +4876,7 @@ static int flush_space(struct btrfs_root *root,
 				     btrfs_get_alloc_profile(root, 0),
 				     CHUNK_ALLOC_NO_FORCE);
 		btrfs_end_transaction(trans, root);
-		if (ret == -ENOSPC)
+		if (ret > 0 || ret == -ENOSPC)
 			ret = 0;
 		break;
 	case COMMIT_TRANS:
@@ -6497,19 +6491,15 @@ void btrfs_wait_block_group_reservations(struct btrfs_block_group_cache *bg)
 }
 
 /**
- * btrfs_update_reserved_bytes - update the block_group and space info counters
+ * btrfs_add_reserved_bytes - update the block_group and space info counters
  * @cache:	The cache we are manipulating
+ * @ram_bytes:  The number of bytes of file content, and will be same to
+ *              @num_bytes except for the compress path.
  * @num_bytes:	The number of bytes in question
- * @reserve:	One of the reservation enums
  * @delalloc:   The blocks are allocated for the delalloc write
  *
- * This is called by the allocator when it reserves space, or by somebody who is
- * freeing space that was never actually used on disk.  For example if you
- * reserve some space for a new leaf in transaction A and before transaction A
- * commits you free that leaf, you call this with reserve set to 0 in order to
- * clear the reservation.
- *
- * Metadata reservations should be called with RESERVE_ALLOC so we do the proper
+ * This is called by the allocator when it reserves space. Metadata
+ * reservations should be called with RESERVE_ALLOC so we do the proper
  * ENOSPC accounting.  For data we handle the reservation through clearing the
  * delalloc bits in the io_tree.  We have to do this since we could end up
  * allocating less disk space for the amount of data we have reserved in the
@@ -6519,44 +6509,63 @@ void btrfs_wait_block_group_reservations(struct btrfs_block_group_cache *bg)
  * make the reservation and return -EAGAIN, otherwise this function always
  * succeeds.
  */
-static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache,
-				       u64 num_bytes, int reserve, int delalloc)
+static int btrfs_add_reserved_bytes(struct btrfs_block_group_cache *cache,
+				    u64 ram_bytes, u64 num_bytes, int delalloc)
 {
 	struct btrfs_space_info *space_info = cache->space_info;
 	int ret = 0;
 
 	spin_lock(&space_info->lock);
 	spin_lock(&cache->lock);
-	if (reserve != RESERVE_FREE) {
-		if (cache->ro) {
-			ret = -EAGAIN;
-		} else {
-			cache->reserved += num_bytes;
-			space_info->bytes_reserved += num_bytes;
-			if (reserve == RESERVE_ALLOC) {
-				trace_btrfs_space_reservation(cache->fs_info,
-						"space_info", space_info->flags,
-						num_bytes, 0);
-				space_info->bytes_may_use -= num_bytes;
-			}
-
-			if (delalloc)
-				cache->delalloc_bytes += num_bytes;
-		}
+	if (cache->ro) {
+		ret = -EAGAIN;
 	} else {
-		if (cache->ro)
-			space_info->bytes_readonly += num_bytes;
-		cache->reserved -= num_bytes;
-		space_info->bytes_reserved -= num_bytes;
+		cache->reserved += num_bytes;
+		space_info->bytes_reserved += num_bytes;
 
+		trace_btrfs_space_reservation(cache->fs_info,
+				"space_info", space_info->flags,
+				ram_bytes, 0);
+		space_info->bytes_may_use -= ram_bytes;
 		if (delalloc)
-			cache->delalloc_bytes -= num_bytes;
+			cache->delalloc_bytes += num_bytes;
 	}
 	spin_unlock(&cache->lock);
 	spin_unlock(&space_info->lock);
 	return ret;
 }
 
+/**
+ * btrfs_free_reserved_bytes - update the block_group and space info counters
+ * @cache:      The cache we are manipulating
+ * @num_bytes:  The number of bytes in question
+ * @delalloc:   The blocks are allocated for the delalloc write
+ *
+ * This is called by somebody who is freeing space that was never actually used
+ * on disk.  For example if you reserve some space for a new leaf in transaction
+ * A and before transaction A commits you free that leaf, you call this with
+ * reserve set to 0 in order to clear the reservation.
+ */
+
+static int btrfs_free_reserved_bytes(struct btrfs_block_group_cache *cache,
+				     u64 num_bytes, int delalloc)
+{
+	struct btrfs_space_info *space_info = cache->space_info;
+	int ret = 0;
+
+	spin_lock(&space_info->lock);
+	spin_lock(&cache->lock);
+	if (cache->ro)
+		space_info->bytes_readonly += num_bytes;
+	cache->reserved -= num_bytes;
+	space_info->bytes_reserved -= num_bytes;
+
+	if (delalloc)
+		cache->delalloc_bytes -= num_bytes;
+	spin_unlock(&cache->lock);
+	spin_unlock(&space_info->lock);
+	return ret;
+}
 void btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root)
 {
@@ -7191,7 +7200,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 		WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags));
 
 		btrfs_add_free_space(cache, buf->start, buf->len);
-		btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE, 0);
+		btrfs_free_reserved_bytes(cache, buf->len, 0);
 		btrfs_put_block_group(cache);
 		trace_btrfs_reserved_extent_free(root, buf->start, buf->len);
 		pin = 0;
@@ -7416,9 +7425,9 @@ btrfs_release_block_group(struct btrfs_block_group_cache *cache,
  * the free space extent currently.
  */
 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, int delalloc)
+				u64 ram_bytes, u64 num_bytes, u64 empty_size,
+				u64 hint_byte, struct btrfs_key *ins,
+				u64 flags, int delalloc)
 {
 	int ret = 0;
 	struct btrfs_root *root = orig_root->fs_info->extent_root;
@@ -7430,8 +7439,6 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	struct btrfs_space_info *space_info;
 	int loop = 0;
 	int index = __get_raid_index(flags);
-	int alloc_type = (flags & BTRFS_BLOCK_GROUP_DATA) ?
-		RESERVE_ALLOC_NO_ACCOUNT : RESERVE_ALLOC;
 	bool failed_cluster_refill = false;
 	bool failed_alloc = false;
 	bool use_cluster = true;
@@ -7763,8 +7770,8 @@ checks:
 					     search_start - offset);
 		BUG_ON(offset > search_start);
 
-		ret = btrfs_update_reserved_bytes(block_group, num_bytes,
-						  alloc_type, delalloc);
+		ret = btrfs_add_reserved_bytes(block_group, ram_bytes,
+				num_bytes, delalloc);
 		if (ret == -EAGAIN) {
 			btrfs_add_free_space(block_group, offset, num_bytes);
 			goto loop;
@@ -7936,7 +7943,7 @@ again:
 	up_read(&info->groups_sem);
 }
 
-int btrfs_reserve_extent(struct btrfs_root *root,
+int btrfs_reserve_extent(struct btrfs_root *root, u64 ram_bytes,
 			 u64 num_bytes, u64 min_alloc_size,
 			 u64 empty_size, u64 hint_byte,
 			 struct btrfs_key *ins, int is_data, int delalloc)
@@ -7948,8 +7955,8 @@ int btrfs_reserve_extent(struct btrfs_root *root,
 	flags = btrfs_get_alloc_profile(root, is_data);
 again:
 	WARN_ON(num_bytes < root->sectorsize);
-	ret = find_free_extent(root, num_bytes, empty_size, hint_byte, ins,
-			       flags, delalloc);
+	ret = find_free_extent(root, ram_bytes, num_bytes, empty_size,
+			       hint_byte, ins, flags, delalloc);
 	if (!ret && !is_data) {
 		btrfs_dec_block_group_reservations(root->fs_info,
 						   ins->objectid);
@@ -7958,6 +7965,7 @@ again:
 			num_bytes = min(num_bytes >> 1, ins->offset);
 			num_bytes = round_down(num_bytes, root->sectorsize);
 			num_bytes = max(num_bytes, min_alloc_size);
+			ram_bytes = num_bytes;
 			if (num_bytes == min_alloc_size)
 				final_tried = true;
 			goto again;
@@ -7995,7 +8003,7 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root,
 		if (btrfs_test_opt(root->fs_info, DISCARD))
 			ret = btrfs_discard_extent(root, start, len, NULL);
 		btrfs_add_free_space(cache, start, len);
-		btrfs_update_reserved_bytes(cache, len, RESERVE_FREE, delalloc);
+		btrfs_free_reserved_bytes(cache, len, delalloc);
 		trace_btrfs_reserved_extent_free(root, start, len);
 	}
 
@@ -8223,8 +8231,8 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
 	if (!block_group)
 		return -EINVAL;
 
-	ret = btrfs_update_reserved_bytes(block_group, ins->offset,
-					  RESERVE_ALLOC_NO_ACCOUNT, 0);
+	ret = btrfs_add_reserved_bytes(block_group, ins->offset,
+				       ins->offset, 0);
 	BUG_ON(ret); /* logic error */
 	ret = alloc_reserved_file_extent(trans, root, 0, root_objectid,
 					 0, owner, offset, ins, 1);
@@ -8368,7 +8376,7 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 	if (IS_ERR(block_rsv))
 		return ERR_CAST(block_rsv);
 
-	ret = btrfs_reserve_extent(root, blocksize, blocksize,
+	ret = btrfs_reserve_extent(root, blocksize, blocksize, blocksize,
 				   empty_size, hint, &ins, 0, 0);
 	if (ret)
 		goto out_unuse;
@@ -8521,35 +8529,6 @@ reada:
 	wc->reada_slot = slot;
 }
 
-/*
- * These may not be seen by the usual inc/dec ref code so we have to
- * add them here.
- */
-static int record_one_subtree_extent(struct btrfs_trans_handle *trans,
-				     struct btrfs_root *root, u64 bytenr,
-				     u64 num_bytes)
-{
-	struct btrfs_qgroup_extent_record *qrecord;
-	struct btrfs_delayed_ref_root *delayed_refs;
-
-	qrecord = kmalloc(sizeof(*qrecord), GFP_NOFS);
-	if (!qrecord)
-		return -ENOMEM;
-
-	qrecord->bytenr = bytenr;
-	qrecord->num_bytes = num_bytes;
-	qrecord->old_roots = NULL;
-
-	delayed_refs = &trans->transaction->delayed_refs;
-	spin_lock(&delayed_refs->lock);
-	if (btrfs_qgroup_insert_dirty_extent(trans->fs_info,
-					     delayed_refs, qrecord))
-		kfree(qrecord);
-	spin_unlock(&delayed_refs->lock);
-
-	return 0;
-}
-
 static int account_leaf_items(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root,
 			      struct extent_buffer *eb)
@@ -8583,7 +8562,8 @@ static int account_leaf_items(struct btrfs_trans_handle *trans,
 
 		num_bytes = btrfs_file_extent_disk_num_bytes(eb, fi);
 
-		ret = record_one_subtree_extent(trans, root, bytenr, num_bytes);
+		ret = btrfs_qgroup_insert_dirty_extent(trans, root->fs_info,
+				bytenr, num_bytes, GFP_NOFS);
 		if (ret)
 			return ret;
 	}
@@ -8732,8 +8712,9 @@ walk_down:
 			btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK);
 			path->locks[level] = BTRFS_READ_LOCK_BLOCKING;
 
-			ret = record_one_subtree_extent(trans, root, child_bytenr,
-							root->nodesize);
+			ret = btrfs_qgroup_insert_dirty_extent(trans,
+					root->fs_info, child_bytenr,
+					root->nodesize, GFP_NOFS);
 			if (ret)
 				goto out;
 		}
@@ -9906,6 +9887,7 @@ static int find_first_block_group(struct btrfs_root *root,
 			} else {
 				ret = 0;
 			}
+			free_extent_map(em);
 			goto out;
 		}
 		path->slots[0]++;
@@ -9942,6 +9924,7 @@ void btrfs_put_block_group_cache(struct btrfs_fs_info *info)
 		block_group->iref = 0;
 		block_group->inode = NULL;
 		spin_unlock(&block_group->lock);
+		ASSERT(block_group->io_ctl.inode == NULL);
 		iput(inode);
 		last = block_group->key.objectid + block_group->key.offset;
 		btrfs_put_block_group(block_group);
@@ -9999,6 +9982,10 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 			free_excluded_extents(info->extent_root, block_group);
 
 		btrfs_remove_free_space_cache(block_group);
+		ASSERT(list_empty(&block_group->dirty_list));
+		ASSERT(list_empty(&block_group->io_list));
+		ASSERT(list_empty(&block_group->bg_list));
+		ASSERT(atomic_read(&block_group->count) == 1);
 		btrfs_put_block_group(block_group);
 
 		spin_lock(&info->block_group_cache_lock);

fs/btrfs/extent_io.h

@@ -20,6 +20,7 @@
 #define EXTENT_DAMAGED		(1U << 14)
 #define EXTENT_NORESERVE	(1U << 15)
 #define EXTENT_QGROUP_RESERVED	(1U << 16)
+#define EXTENT_CLEAR_DATA_RESV	(1U << 17)
 #define EXTENT_IOBITS		(EXTENT_LOCKED | EXTENT_WRITEBACK)
 #define EXTENT_CTLBITS		(EXTENT_DO_ACCOUNTING | EXTENT_FIRST_DELALLOC)
 

fs/btrfs/file.c

@@ -2070,7 +2070,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	}
 	trans->sync = true;
 
-	btrfs_init_log_ctx(&ctx);
+	btrfs_init_log_ctx(&ctx, inode);
 
 	ret = btrfs_log_dentry_safe(trans, root, dentry, start, end, &ctx);
 	if (ret < 0) {
@@ -2675,6 +2675,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 
 	alloc_start = round_down(offset, blocksize);
 	alloc_end = round_up(offset + len, blocksize);
+	cur_offset = alloc_start;
 
 	/* Make sure we aren't being give some crap mode */
 	if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
@@ -2767,7 +2768,6 @@ static long btrfs_fallocate(struct file *file, int mode,
 
 	/* First, check if we exceed the qgroup limit */
 	INIT_LIST_HEAD(&reserve_list);
-	cur_offset = alloc_start;
 	while (1) {
 		em = btrfs_get_extent(inode, NULL, 0, cur_offset,
 				      alloc_end - cur_offset, 0);
@@ -2794,6 +2794,14 @@ static long btrfs_fallocate(struct file *file, int mode,
 					last_byte - cur_offset);
 			if (ret < 0)
 				break;
+		} else {
+			/*
+			 * Do not need to reserve unwritten extent for this
+			 * range, free reserved data space first, otherwise
+			 * it'll result in false ENOSPC error.
+			 */
+			btrfs_free_reserved_data_space(inode, cur_offset,
+				last_byte - cur_offset);
 		}
 		free_extent_map(em);
 		cur_offset = last_byte;
@@ -2811,6 +2819,9 @@ static long btrfs_fallocate(struct file *file, int mode,
 					range->start,
 					range->len, 1 << inode->i_blkbits,
 					offset + len, &alloc_hint);
+		else
+			btrfs_free_reserved_data_space(inode, range->start,
+						       range->len);
 		list_del(&range->list);
 		kfree(range);
 	}
@@ -2845,18 +2856,11 @@ out_unlock:
 	unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
 			     &cached_state, GFP_KERNEL);
 out:
-	/*
-	 * As we waited the extent range, the data_rsv_map must be empty
-	 * in the range, as written data range will be released from it.
-	 * And for prealloacted extent, it will also be released when
-	 * its metadata is written.
-	 * So this is completely used as cleanup.
-	 */
-	btrfs_qgroup_free_data(inode, alloc_start, alloc_end - alloc_start);
 	inode_unlock(inode);
 	/* Let go of our reservation. */
-	btrfs_free_reserved_data_space(inode, alloc_start,
-				       alloc_end - alloc_start);
+	if (ret != 0)
+		btrfs_free_reserved_data_space(inode, alloc_start,
+				       alloc_end - cur_offset);
 	return ret;
 }
 

fs/btrfs/inode-map.c

@@ -495,10 +495,9 @@ again:
 	ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, prealloc,
 					      prealloc, prealloc, &alloc_hint);
 	if (ret) {
-		btrfs_delalloc_release_space(inode, 0, prealloc);
+		btrfs_delalloc_release_metadata(inode, prealloc);
 		goto out_put;
 	}
-	btrfs_free_reserved_data_space(inode, 0, prealloc);
 
 	ret = btrfs_write_out_ino_cache(root, trans, path, inode);
 out_put:

fs/btrfs/inode.c

@@ -566,6 +566,8 @@ cont:
 						     PAGE_SET_WRITEBACK |
 						     page_error_op |
 						     PAGE_END_WRITEBACK);
+			btrfs_free_reserved_data_space_noquota(inode, start,
+						end - start + 1);
 			goto free_pages_out;
 		}
 	}
@@ -742,7 +744,7 @@ retry:
 		lock_extent(io_tree, async_extent->start,
 			    async_extent->start + async_extent->ram_size - 1);
 
-		ret = btrfs_reserve_extent(root,
+		ret = btrfs_reserve_extent(root, async_extent->ram_size,
 					   async_extent->compressed_size,
 					   async_extent->compressed_size,
 					   0, alloc_hint, &ins, 1, 1);
@@ -969,7 +971,8 @@ static noinline int cow_file_range(struct inode *inode,
 				     EXTENT_DEFRAG, PAGE_UNLOCK |
 				     PAGE_CLEAR_DIRTY | PAGE_SET_WRITEBACK |
 				     PAGE_END_WRITEBACK);
-
+			btrfs_free_reserved_data_space_noquota(inode, start,
+						end - start + 1);
 			*nr_written = *nr_written +
 			     (end - start + PAGE_SIZE) / PAGE_SIZE;
 			*page_started = 1;
@@ -989,7 +992,7 @@ static noinline int cow_file_range(struct inode *inode,
 		unsigned long op;
 
 		cur_alloc_size = disk_num_bytes;
-		ret = btrfs_reserve_extent(root, cur_alloc_size,
+		ret = btrfs_reserve_extent(root, cur_alloc_size, cur_alloc_size,
 					   root->sectorsize, 0, alloc_hint,
 					   &ins, 1, 1);
 		if (ret < 0)
@@ -1489,8 +1492,10 @@ out_check:
 		extent_clear_unlock_delalloc(inode, cur_offset,
 					     cur_offset + num_bytes - 1,
 					     locked_page, EXTENT_LOCKED |
-					     EXTENT_DELALLOC, PAGE_UNLOCK |
-					     PAGE_SET_PRIVATE2);
+					     EXTENT_DELALLOC |
+					     EXTENT_CLEAR_DATA_RESV,
+					     PAGE_UNLOCK | PAGE_SET_PRIVATE2);
+
 		if (!nolock && nocow)
 			btrfs_end_write_no_snapshoting(root);
 		cur_offset = extent_end;
@@ -1807,7 +1812,9 @@ static void btrfs_clear_bit_hook(struct inode *inode,
 			return;
 
 		if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID
-		    && do_list && !(state->state & EXTENT_NORESERVE))
+		    && do_list && !(state->state & EXTENT_NORESERVE)
+		    && (*bits & (EXTENT_DO_ACCOUNTING |
+		    EXTENT_CLEAR_DATA_RESV)))
 			btrfs_free_reserved_data_space_noquota(inode,
 					state->start, len);
 
@@ -7251,7 +7258,7 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode,
 	int ret;
 
 	alloc_hint = get_extent_allocation_hint(inode, start, len);
-	ret = btrfs_reserve_extent(root, len, root->sectorsize, 0,
+	ret = btrfs_reserve_extent(root, len, len, root->sectorsize, 0,
 				   alloc_hint, &ins, 1, 1);
 	if (ret)
 		return ERR_PTR(ret);
@@ -7751,6 +7758,13 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 				ret = PTR_ERR(em2);
 				goto unlock_err;
 			}
+			/*
+			 * For inode marked NODATACOW or extent marked PREALLOC,
+			 * use the existing or preallocated extent, so does not
+			 * need to adjust btrfs_space_info's bytes_may_use.
+			 */
+			btrfs_free_reserved_data_space_noquota(inode,
+					start, len);
 			goto unlock;
 		}
 	}
@@ -7785,7 +7799,6 @@ unlock:
 			i_size_write(inode, start + len);
 
 		adjust_dio_outstanding_extents(inode, dio_data, len);
-		btrfs_free_reserved_data_space(inode, start, len);
 		WARN_ON(dio_data->reserve < len);
 		dio_data->reserve -= len;
 		dio_data->unsubmitted_oe_range_end = start + len;
@@ -10306,6 +10319,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 	u64 last_alloc = (u64)-1;
 	int ret = 0;
 	bool own_trans = true;
+	u64 end = start + num_bytes - 1;
 
 	if (trans)
 		own_trans = false;
@@ -10327,8 +10341,8 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 		 * sized chunks.
 		 */
 		cur_bytes = min(cur_bytes, last_alloc);
-		ret = btrfs_reserve_extent(root, cur_bytes, min_size, 0,
-					   *alloc_hint, &ins, 1, 0);
+		ret = btrfs_reserve_extent(root, cur_bytes, cur_bytes,
+				min_size, 0, *alloc_hint, &ins, 1, 0);
 		if (ret) {
 			if (own_trans)
 				btrfs_end_transaction(trans, root);
@@ -10414,6 +10428,9 @@ next:
 		if (own_trans)
 			btrfs_end_transaction(trans, root);
 	}
+	if (cur_offset < end)
+		btrfs_free_reserved_data_space(inode, cur_offset,
+			end - cur_offset + 1);
 	return ret;
 }
 

fs/btrfs/ioctl.c

@@ -5084,7 +5084,7 @@ static long btrfs_ioctl_quota_rescan_wait(struct file *file, void __user *arg)
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	return btrfs_qgroup_wait_for_completion(root->fs_info);
+	return btrfs_qgroup_wait_for_completion(root->fs_info, true);
 }
 
 static long _btrfs_ioctl_set_received_subvol(struct file *file,

fs/btrfs/qgroup.c

@@ -995,7 +995,7 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 		goto out;
 	fs_info->quota_enabled = 0;
 	fs_info->pending_quota_state = 0;
-	btrfs_qgroup_wait_for_completion(fs_info);
+	btrfs_qgroup_wait_for_completion(fs_info, false);
 	spin_lock(&fs_info->qgroup_lock);
 	quota_root = fs_info->quota_root;
 	fs_info->quota_root = NULL;
@@ -1453,10 +1453,9 @@ int btrfs_qgroup_prepare_account_extents(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
-struct btrfs_qgroup_extent_record *
-btrfs_qgroup_insert_dirty_extent(struct btrfs_fs_info *fs_info,
-				 struct btrfs_delayed_ref_root *delayed_refs,
-				 struct btrfs_qgroup_extent_record *record)
+int btrfs_qgroup_insert_dirty_extent_nolock(struct btrfs_fs_info *fs_info,
+				struct btrfs_delayed_ref_root *delayed_refs,
+				struct btrfs_qgroup_extent_record *record)
 {
 	struct rb_node **p = &delayed_refs->dirty_extent_root.rb_node;
 	struct rb_node *parent_node = NULL;
@@ -1475,12 +1474,42 @@ btrfs_qgroup_insert_dirty_extent(struct btrfs_fs_info *fs_info,
 		else if (bytenr > entry->bytenr)
 			p = &(*p)->rb_right;
 		else
-			return entry;
+			return 1;
 	}
 
 	rb_link_node(&record->node, parent_node, p);
 	rb_insert_color(&record->node, &delayed_refs->dirty_extent_root);
-	return NULL;
+	return 0;
+}
+
+int btrfs_qgroup_insert_dirty_extent(struct btrfs_trans_handle *trans,
+		struct btrfs_fs_info *fs_info, u64 bytenr, u64 num_bytes,
+		gfp_t gfp_flag)
+{
+	struct btrfs_qgroup_extent_record *record;
+	struct btrfs_delayed_ref_root *delayed_refs;
+	int ret;
+
+	if (!fs_info->quota_enabled || bytenr == 0 || num_bytes == 0)
+		return 0;
+	if (WARN_ON(trans == NULL))
+		return -EINVAL;
+	record = kmalloc(sizeof(*record), gfp_flag);
+	if (!record)
+		return -ENOMEM;
+
+	delayed_refs = &trans->transaction->delayed_refs;
+	record->bytenr = bytenr;
+	record->num_bytes = num_bytes;
+	record->old_roots = NULL;
+
+	spin_lock(&delayed_refs->lock);
+	ret = btrfs_qgroup_insert_dirty_extent_nolock(fs_info, delayed_refs,
+						      record);
+	spin_unlock(&delayed_refs->lock);
+	if (ret > 0)
+		kfree(record);
+	return 0;
 }
 
 #define UPDATE_NEW	0
@@ -2303,6 +2332,10 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
 	int err = -ENOMEM;
 	int ret = 0;
 
+	mutex_lock(&fs_info->qgroup_rescan_lock);
+	fs_info->qgroup_rescan_running = true;
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
+
 	path = btrfs_alloc_path();
 	if (!path)
 		goto out;
@@ -2369,6 +2402,9 @@ out:
 	}
 
 done:
+	mutex_lock(&fs_info->qgroup_rescan_lock);
+	fs_info->qgroup_rescan_running = false;
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
 	complete_all(&fs_info->qgroup_rescan_completion);
 }
 
@@ -2487,20 +2523,26 @@ btrfs_qgroup_rescan(struct btrfs_fs_info *fs_info)
 	return 0;
 }
 
-int btrfs_qgroup_wait_for_completion(struct btrfs_fs_info *fs_info)
+int btrfs_qgroup_wait_for_completion(struct btrfs_fs_info *fs_info,
+				     bool interruptible)
 {
 	int running;
 	int ret = 0;
 
 	mutex_lock(&fs_info->qgroup_rescan_lock);
 	spin_lock(&fs_info->qgroup_lock);
-	running = fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN;
+	running = fs_info->qgroup_rescan_running;
 	spin_unlock(&fs_info->qgroup_lock);
 	mutex_unlock(&fs_info->qgroup_rescan_lock);
 
-	if (running)
+	if (!running)
+		return 0;
+
+	if (interruptible)
 		ret = wait_for_completion_interruptible(
 					&fs_info->qgroup_rescan_completion);
+	else
+		wait_for_completion(&fs_info->qgroup_rescan_completion);
 
 	return ret;
 }

fs/btrfs/qgroup.h

@@ -46,7 +46,8 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 			struct btrfs_fs_info *fs_info);
 int btrfs_qgroup_rescan(struct btrfs_fs_info *fs_info);
 void btrfs_qgroup_rescan_resume(struct btrfs_fs_info *fs_info);
-int btrfs_qgroup_wait_for_completion(struct btrfs_fs_info *fs_info);
+int btrfs_qgroup_wait_for_completion(struct btrfs_fs_info *fs_info,
+				     bool interruptible);
 int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans,
 			      struct btrfs_fs_info *fs_info, u64 src, u64 dst);
 int btrfs_del_qgroup_relation(struct btrfs_trans_handle *trans,
@@ -63,10 +64,35 @@ void btrfs_free_qgroup_config(struct btrfs_fs_info *fs_info);
 struct btrfs_delayed_extent_op;
 int btrfs_qgroup_prepare_account_extents(struct btrfs_trans_handle *trans,
 					 struct btrfs_fs_info *fs_info);
-struct btrfs_qgroup_extent_record *
-btrfs_qgroup_insert_dirty_extent(struct btrfs_fs_info *fs_info,
-				 struct btrfs_delayed_ref_root *delayed_refs,
-				 struct btrfs_qgroup_extent_record *record);
+/*
+ * Insert one dirty extent record into @delayed_refs, informing qgroup to
+ * account that extent at commit trans time.
+ *
+ * No lock version, caller must acquire delayed ref lock and allocate memory.
+ *
+ * Return 0 for success insert
+ * Return >0 for existing record, caller can free @record safely.
+ * Error is not possible
+ */
+int btrfs_qgroup_insert_dirty_extent_nolock(
+		struct btrfs_fs_info *fs_info,
+		struct btrfs_delayed_ref_root *delayed_refs,
+		struct btrfs_qgroup_extent_record *record);
+
+/*
+ * Insert one dirty extent record into @delayed_refs, informing qgroup to
+ * account that extent at commit trans time.
+ *
+ * Better encapsulated version.
+ *
+ * Return 0 if the operation is done.
+ * Return <0 for error, like memory allocation failure or invalid parameter
+ * (NULL trans)
+ */
+int btrfs_qgroup_insert_dirty_extent(struct btrfs_trans_handle *trans,
+		struct btrfs_fs_info *fs_info, u64 bytenr, u64 num_bytes,
+		gfp_t gfp_flag);
+
 int
 btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans,
 			    struct btrfs_fs_info *fs_info,

fs/btrfs/relocation.c

@@ -31,6 +31,7 @@
 #include "async-thread.h"
 #include "free-space-cache.h"
 #include "inode-map.h"
+#include "qgroup.h"
 
 /*
  * backref_node, mapping_node and tree_block start with this
@@ -3037,15 +3038,19 @@ int prealloc_file_extent_cluster(struct inode *inode,
 	u64 num_bytes;
 	int nr = 0;
 	int ret = 0;
+	u64 prealloc_start = cluster->start - offset;
+	u64 prealloc_end = cluster->end - offset;
+	u64 cur_offset;
 
 	BUG_ON(cluster->start != cluster->boundary[0]);
 	inode_lock(inode);
 
-	ret = btrfs_check_data_free_space(inode, cluster->start,
-					  cluster->end + 1 - cluster->start);
+	ret = btrfs_check_data_free_space(inode, prealloc_start,
+					  prealloc_end + 1 - prealloc_start);
 	if (ret)
 		goto out;
 
+	cur_offset = prealloc_start;
 	while (nr < cluster->nr) {
 		start = cluster->boundary[nr] - offset;
 		if (nr + 1 < cluster->nr)
@@ -3055,16 +3060,21 @@ int prealloc_file_extent_cluster(struct inode *inode,
 
 		lock_extent(&BTRFS_I(inode)->io_tree, start, end);
 		num_bytes = end + 1 - start;
+		if (cur_offset < start)
+			btrfs_free_reserved_data_space(inode, cur_offset,
+					start - cur_offset);
 		ret = btrfs_prealloc_file_range(inode, 0, start,
 						num_bytes, num_bytes,
 						end + 1, &alloc_hint);
+		cur_offset = end + 1;
 		unlock_extent(&BTRFS_I(inode)->io_tree, start, end);
 		if (ret)
 			break;
 		nr++;
 	}
-	btrfs_free_reserved_data_space(inode, cluster->start,
-				       cluster->end + 1 - cluster->start);
+	if (cur_offset < prealloc_end)
+		btrfs_free_reserved_data_space(inode, cur_offset,
+				       prealloc_end + 1 - cur_offset);
 out:
 	inode_unlock(inode);
 	return ret;
@@ -3916,6 +3926,90 @@ int prepare_to_relocate(struct reloc_control *rc)
 	return 0;
 }
 
+/*
+ * Qgroup fixer for data chunk relocation.
+ * The data relocation is done in the following steps
+ * 1) Copy data extents into data reloc tree
+ * 2) Create tree reloc tree(special snapshot) for related subvolumes
+ * 3) Modify file extents in tree reloc tree
+ * 4) Merge tree reloc tree with original fs tree, by swapping tree blocks
+ *
+ * The problem is, data and tree reloc tree are not accounted to qgroup,
+ * and 4) will only info qgroup to track tree blocks change, not file extents
+ * in the tree blocks.
+ *
+ * The good news is, related data extents are all in data reloc tree, so we
+ * only need to info qgroup to track all file extents in data reloc tree
+ * before commit trans.
+ */
+static int qgroup_fix_relocated_data_extents(struct btrfs_trans_handle *trans,
+					     struct reloc_control *rc)
+{
+	struct btrfs_fs_info *fs_info = rc->extent_root->fs_info;
+	struct inode *inode = rc->data_inode;
+	struct btrfs_root *data_reloc_root = BTRFS_I(inode)->root;
+	struct btrfs_path *path;
+	struct btrfs_key key;
+	int ret = 0;
+
+	if (!fs_info->quota_enabled)
+		return 0;
+
+	/*
+	 * Only for stage where we update data pointers the qgroup fix is
+	 * valid.
+	 * For MOVING_DATA stage, we will miss the timing of swapping tree
+	 * blocks, and won't fix it.
+	 */
+	if (!(rc->stage == UPDATE_DATA_PTRS && rc->extents_found))
+		return 0;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+	key.objectid = btrfs_ino(inode);
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = 0;
+
+	ret = btrfs_search_slot(NULL, data_reloc_root, &key, path, 0, 0);
+	if (ret < 0)
+		goto out;
+
+	lock_extent(&BTRFS_I(inode)->io_tree, 0, (u64)-1);
+	while (1) {
+		struct btrfs_file_extent_item *fi;
+
+		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
+		if (key.objectid > btrfs_ino(inode))
+			break;
+		if (key.type != BTRFS_EXTENT_DATA_KEY)
+			goto next;
+		fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
+				    struct btrfs_file_extent_item);
+		if (btrfs_file_extent_type(path->nodes[0], fi) !=
+				BTRFS_FILE_EXTENT_REG)
+			goto next;
+		ret = btrfs_qgroup_insert_dirty_extent(trans, fs_info,
+			btrfs_file_extent_disk_bytenr(path->nodes[0], fi),
+			btrfs_file_extent_disk_num_bytes(path->nodes[0], fi),
+			GFP_NOFS);
+		if (ret < 0)
+			break;
+next:
+		ret = btrfs_next_item(data_reloc_root, path);
+		if (ret < 0)
+			break;
+		if (ret > 0) {
+			ret = 0;
+			break;
+		}
+	}
+	unlock_extent(&BTRFS_I(inode)->io_tree, 0 , (u64)-1);
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
 static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 {
 	struct rb_root blocks = RB_ROOT;
@@ -4102,10 +4196,16 @@ restart:
 
 	/* get rid of pinned extents */
 	trans = btrfs_join_transaction(rc->extent_root);
-	if (IS_ERR(trans))
+	if (IS_ERR(trans)) {
 		err = PTR_ERR(trans);
-	else
-		btrfs_commit_transaction(trans, rc->extent_root);
+		goto out_free;
+	}
+	err = qgroup_fix_relocated_data_extents(trans, rc);
+	if (err < 0) {
+		btrfs_abort_transaction(trans, err);
+		goto out_free;
+	}
+	btrfs_commit_transaction(trans, rc->extent_root);
 out_free:
 	btrfs_free_block_rsv(rc->extent_root, rc->block_rsv);
 	btrfs_free_path(path);
@@ -4468,10 +4568,16 @@ int btrfs_recover_relocation(struct btrfs_root *root)
 	unset_reloc_control(rc);
 
 	trans = btrfs_join_transaction(rc->extent_root);
-	if (IS_ERR(trans))
+	if (IS_ERR(trans)) {
 		err = PTR_ERR(trans);
-	else
-		err = btrfs_commit_transaction(trans, rc->extent_root);
+		goto out_free;
+	}
+	err = qgroup_fix_relocated_data_extents(trans, rc);
+	if (err < 0) {
+		btrfs_abort_transaction(trans, err);
+		goto out_free;
+	}
+	err = btrfs_commit_transaction(trans, rc->extent_root);
 out_free:
 	kfree(rc);
 out:

fs/btrfs/root-tree.c

@@ -272,6 +272,23 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root)
 		root_key.objectid = key.offset;
 		key.offset++;
 
+		/*
+		 * The root might have been inserted already, as before we look
+		 * for orphan roots, log replay might have happened, which
+		 * triggers a transaction commit and qgroup accounting, which
+		 * in turn reads and inserts fs roots while doing backref
+		 * walking.
+		 */
+		root = btrfs_lookup_fs_root(tree_root->fs_info,
+					    root_key.objectid);
+		if (root) {
+			WARN_ON(!test_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED,
+					  &root->state));
+			if (btrfs_root_refs(&root->root_item) == 0)
+				btrfs_add_dead_root(root);
+			continue;
+		}
+
 		root = btrfs_read_fs_root(tree_root, &root_key);
 		err = PTR_ERR_OR_ZERO(root);
 		if (err && err != -ENOENT) {
@@ -310,16 +327,8 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root)
 		set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state);
 
 		err = btrfs_insert_fs_root(root->fs_info, root);
-		/*
-		 * The root might have been inserted already, as before we look
-		 * for orphan roots, log replay might have happened, which
-		 * triggers a transaction commit and qgroup accounting, which
-		 * in turn reads and inserts fs roots while doing backref
-		 * walking.
-		 */
-		if (err == -EEXIST)
-			err = 0;
 		if (err) {
+			BUG_ON(err == -EEXIST);
 			btrfs_free_fs_root(root);
 			break;
 		}

fs/btrfs/super.c

@@ -2241,6 +2241,13 @@ static int btrfs_freeze(struct super_block *sb)
 	struct btrfs_trans_handle *trans;
 	struct btrfs_root *root = btrfs_sb(sb)->tree_root;
 
+	root->fs_info->fs_frozen = 1;
+	/*
+	 * We don't need a barrier here, we'll wait for any transaction that
+	 * could be in progress on other threads (and do delayed iputs that
+	 * we want to avoid on a frozen filesystem), or do the commit
+	 * ourselves.
+	 */
 	trans = btrfs_attach_transaction_barrier(root);
 	if (IS_ERR(trans)) {
 		/* no transaction, don't bother */
@@ -2251,6 +2258,14 @@ static int btrfs_freeze(struct super_block *sb)
 	return btrfs_commit_transaction(trans, root);
 }
 
+static int btrfs_unfreeze(struct super_block *sb)
+{
+	struct btrfs_root *root = btrfs_sb(sb)->tree_root;
+
+	root->fs_info->fs_frozen = 0;
+	return 0;
+}
+
 static int btrfs_show_devname(struct seq_file *m, struct dentry *root)
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(root->d_sb);
@@ -2299,6 +2314,7 @@ static const struct super_operations btrfs_super_ops = {
 	.statfs		= btrfs_statfs,
 	.remount_fs	= btrfs_remount,
 	.freeze_fs	= btrfs_freeze,
+	.unfreeze_fs	= btrfs_unfreeze,
 };
 
 static const struct file_operations btrfs_ctl_fops = {

fs/btrfs/transaction.c

@@ -2278,8 +2278,13 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	kmem_cache_free(btrfs_trans_handle_cachep, trans);
 
+	/*
+	 * If fs has been frozen, we can not handle delayed iputs, otherwise
+	 * it'll result in deadlock about SB_FREEZE_FS.
+	 */
 	if (current != root->fs_info->transaction_kthread &&
-	    current != root->fs_info->cleaner_kthread)
+	    current != root->fs_info->cleaner_kthread &&
+	    !root->fs_info->fs_frozen)
 		btrfs_run_delayed_iputs(root);
 
 	return ret;

fs/btrfs/tree-log.c

@@ -27,6 +27,7 @@
 #include "backref.h"
 #include "hash.h"
 #include "compression.h"
+#include "qgroup.h"
 
 /* magic values for the inode_only field in btrfs_log_inode:
  *
@@ -680,6 +681,21 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 		ins.type = BTRFS_EXTENT_ITEM_KEY;
 		offset = key->offset - btrfs_file_extent_offset(eb, item);
 
+		/*
+		 * Manually record dirty extent, as here we did a shallow
+		 * file extent item copy and skip normal backref update,
+		 * but modifying extent tree all by ourselves.
+		 * So need to manually record dirty extent for qgroup,
+		 * as the owner of the file extent changed from log tree
+		 * (doesn't affect qgroup) to fs/file tree(affects qgroup)
+		 */
+		ret = btrfs_qgroup_insert_dirty_extent(trans, root->fs_info,
+				btrfs_file_extent_disk_bytenr(eb, item),
+				btrfs_file_extent_disk_num_bytes(eb, item),
+				GFP_NOFS);
+		if (ret < 0)
+			goto out;
+
 		if (ins.objectid > 0) {
 			u64 csum_start;
 			u64 csum_end;
@@ -2807,7 +2823,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	 */
 	mutex_unlock(&root->log_mutex);
 
-	btrfs_init_log_ctx(&root_log_ctx);
+	btrfs_init_log_ctx(&root_log_ctx, NULL);
 
 	mutex_lock(&log_root_tree->log_mutex);
 	atomic_inc(&log_root_tree->log_batch);
@@ -4741,7 +4757,8 @@ again:
 			if (ret < 0) {
 				err = ret;
 				goto out_unlock;
-			} else if (ret > 0) {
+			} else if (ret > 0 && ctx &&
+				   other_ino != btrfs_ino(ctx->inode)) {
 				struct btrfs_key inode_key;
 				struct inode *other_inode;
 

fs/btrfs/tree-log.h

@@ -30,15 +30,18 @@ struct btrfs_log_ctx {
 	int log_transid;
 	int io_err;
 	bool log_new_dentries;
+	struct inode *inode;
 	struct list_head list;
 };
 
-static inline void btrfs_init_log_ctx(struct btrfs_log_ctx *ctx)
+static inline void btrfs_init_log_ctx(struct btrfs_log_ctx *ctx,
+				      struct inode *inode)
 {
 	ctx->log_ret = 0;
 	ctx->log_transid = 0;
 	ctx->io_err = 0;
 	ctx->log_new_dentries = false;
+	ctx->inode = inode;
 	INIT_LIST_HEAD(&ctx->list);
 }
 

fs/btrfs/volumes.c

@@ -834,10 +834,6 @@ static void __free_device(struct work_struct *work)
 	struct btrfs_device *device;
 
 	device = container_of(work, struct btrfs_device, rcu_work);
-
-	if (device->bdev)
-		blkdev_put(device->bdev, device->mode);
-
 	rcu_string_free(device->name);
 	kfree(device);
 }
@@ -852,6 +848,17 @@ static void free_device(struct rcu_head *head)
 	schedule_work(&device->rcu_work);
 }
 
+static void btrfs_close_bdev(struct btrfs_device *device)
+{
+	if (device->bdev && device->writeable) {
+		sync_blockdev(device->bdev);
+		invalidate_bdev(device->bdev);
+	}
+
+	if (device->bdev)
+		blkdev_put(device->bdev, device->mode);
+}
+
 static void btrfs_close_one_device(struct btrfs_device *device)
 {
 	struct btrfs_fs_devices *fs_devices = device->fs_devices;
@@ -870,10 +877,7 @@ static void btrfs_close_one_device(struct btrfs_device *device)
 	if (device->missing)
 		fs_devices->missing_devices--;
 
-	if (device->bdev && device->writeable) {
-		sync_blockdev(device->bdev);
-		invalidate_bdev(device->bdev);
-	}
+	btrfs_close_bdev(device);
 
 	new_device = btrfs_alloc_device(NULL, &device->devid,
 					device->uuid);
@@ -1932,6 +1936,8 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path, u64 devid)
 		btrfs_sysfs_rm_device_link(root->fs_info->fs_devices, device);
 	}
 
+	btrfs_close_bdev(device);
+
 	call_rcu(&device->rcu, free_device);
 
 	num_devices = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
@@ -2025,6 +2031,9 @@ void btrfs_rm_dev_replace_free_srcdev(struct btrfs_fs_info *fs_info,
 		/* zero out the old super if it is writable */
 		btrfs_scratch_superblocks(srcdev->bdev, srcdev->name->str);
 	}
+
+	btrfs_close_bdev(srcdev);
+
 	call_rcu(&srcdev->rcu, free_device);
 
 	/*
@@ -2080,6 +2089,8 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	 * the device_list_mutex lock.
 	 */
 	btrfs_scratch_superblocks(tgtdev->bdev, tgtdev->name->str);
+
+	btrfs_close_bdev(tgtdev);
 	call_rcu(&tgtdev->rcu, free_device);
 }