Merge branch 'for-linus-4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs

Pull btrfs updates from Chris Mason:
 "We have a lot of subvolume quota improvements in here, along with big
  piles of cleanups from Dave Sterba and Anand Jain and others.

  Josef pitched in a batch of allocator fixes based on production use
  here at FB.  We found that mount -o ssd_spread greatly improved our
  performance on hardware raid5/6, but it exposed some CPU bottlenecks
  in the allocator.  These patches make a huge difference"

* 'for-linus-4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (100 commits)
  Btrfs: fix hole punching when using the no-holes feature
  Btrfs: find_free_extent: Do not erroneously skip LOOP_CACHING_WAIT state
  btrfs: Fix a data space underflow warning
  btrfs: qgroup: Fix a rebase bug which will cause qgroup double free
  btrfs: qgroup: Fix a race in delayed_ref which leads to abort trans
  btrfs: clear PF_NOFREEZE in cleaner_kthread()
  btrfs: qgroup: Don't copy extent buffer to do qgroup rescan
  btrfs: add balance filters limits, stripes and usage to supported mask
  btrfs: extend balance filter usage to take minimum and maximum
  btrfs: add balance filter for stripes
  btrfs: extend balance filter limit to take minimum and maximum
  btrfs: fix use after free iterating extrefs
  btrfs: check unsupported filters in balance arguments
  Btrfs: fix regression running delayed references when using qgroups
  Btrfs: fix regression when running delayed references
  Btrfs: don't do extra bitmap search in one bit case
  Btrfs: keep track of largest extent in bitmaps
  Btrfs: don't keep trying to build clusters if we are fragmented
  Btrfs: cut down on loops through the allocator
  Btrfs: don't continue setting up space cache when enospc
  ...

fs/btrfs/backref.c

@@ -362,6 +362,12 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
 		goto out;
 	}
 
+	if (btrfs_test_is_dummy_root(root)) {
+		srcu_read_unlock(&fs_info->subvol_srcu, index);
+		ret = -ENOENT;
+		goto out;
+	}
+
 	if (path->search_commit_root)
 		root_level = btrfs_header_level(root->commit_root);
 	else if (time_seq == (u64)-1)

fs/btrfs/check-integrity.c

@@ -667,7 +667,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
 	selected_super = kzalloc(sizeof(*selected_super), GFP_NOFS);
 	if (NULL == selected_super) {
 		printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
-		return -1;
+		return -ENOMEM;
 	}
 
 	list_for_each_entry(device, dev_head, dev_list) {
@@ -845,8 +845,8 @@ static int btrfsic_process_superblock_dev_mirror(
 		superblock_tmp->never_written = 0;
 		superblock_tmp->mirror_num = 1 + superblock_mirror_num;
 		if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE)
-			printk_in_rcu(KERN_INFO "New initial S-block (bdev %p, %s)"
-				     " @%llu (%s/%llu/%d)\n",
+			btrfs_info_in_rcu(device->dev_root->fs_info,
+				"new initial S-block (bdev %p, %s) @%llu (%s/%llu/%d)",
 				     superblock_bdev,
 				     rcu_str_deref(device->name), dev_bytenr,
 				     dev_state->name, dev_bytenr,
@@ -1660,7 +1660,7 @@ static int btrfsic_read_block(struct btrfsic_state *state,
 					  sizeof(*block_ctx->pagev)) *
 					 num_pages, GFP_NOFS);
 	if (!block_ctx->mem_to_free)
-		return -1;
+		return -ENOMEM;
 	block_ctx->datav = block_ctx->mem_to_free;
 	block_ctx->pagev = (struct page **)(block_ctx->datav + num_pages);
 	for (i = 0; i < num_pages; i++) {

fs/btrfs/compression.c

@@ -745,11 +745,13 @@ out:
 	return ret;
 }
 
-static struct list_head comp_idle_workspace[BTRFS_COMPRESS_TYPES];
-static spinlock_t comp_workspace_lock[BTRFS_COMPRESS_TYPES];
-static int comp_num_workspace[BTRFS_COMPRESS_TYPES];
-static atomic_t comp_alloc_workspace[BTRFS_COMPRESS_TYPES];
-static wait_queue_head_t comp_workspace_wait[BTRFS_COMPRESS_TYPES];
+static struct {
+	struct list_head idle_ws;
+	spinlock_t ws_lock;
+	int num_ws;
+	atomic_t alloc_ws;
+	wait_queue_head_t ws_wait;
+} btrfs_comp_ws[BTRFS_COMPRESS_TYPES];
 
 static const struct btrfs_compress_op * const btrfs_compress_op[] = {
 	&btrfs_zlib_compress,
@@ -761,10 +763,10 @@ void __init btrfs_init_compress(void)
 	int i;
 
 	for (i = 0; i < BTRFS_COMPRESS_TYPES; i++) {
-		INIT_LIST_HEAD(&comp_idle_workspace[i]);
-		spin_lock_init(&comp_workspace_lock[i]);
-		atomic_set(&comp_alloc_workspace[i], 0);
-		init_waitqueue_head(&comp_workspace_wait[i]);
+		INIT_LIST_HEAD(&btrfs_comp_ws[i].idle_ws);
+		spin_lock_init(&btrfs_comp_ws[i].ws_lock);
+		atomic_set(&btrfs_comp_ws[i].alloc_ws, 0);
+		init_waitqueue_head(&btrfs_comp_ws[i].ws_wait);
 	}
 }
 
@@ -778,38 +780,38 @@ static struct list_head *find_workspace(int type)
 	int cpus = num_online_cpus();
 	int idx = type - 1;
 
-	struct list_head *idle_workspace	= &comp_idle_workspace[idx];
-	spinlock_t *workspace_lock		= &comp_workspace_lock[idx];
-	atomic_t *alloc_workspace		= &comp_alloc_workspace[idx];
-	wait_queue_head_t *workspace_wait	= &comp_workspace_wait[idx];
-	int *num_workspace			= &comp_num_workspace[idx];
+	struct list_head *idle_ws	= &btrfs_comp_ws[idx].idle_ws;
+	spinlock_t *ws_lock		= &btrfs_comp_ws[idx].ws_lock;
+	atomic_t *alloc_ws		= &btrfs_comp_ws[idx].alloc_ws;
+	wait_queue_head_t *ws_wait	= &btrfs_comp_ws[idx].ws_wait;
+	int *num_ws			= &btrfs_comp_ws[idx].num_ws;
 again:
-	spin_lock(workspace_lock);
-	if (!list_empty(idle_workspace)) {
-		workspace = idle_workspace->next;
+	spin_lock(ws_lock);
+	if (!list_empty(idle_ws)) {
+		workspace = idle_ws->next;
 		list_del(workspace);
-		(*num_workspace)--;
-		spin_unlock(workspace_lock);
+		(*num_ws)--;
+		spin_unlock(ws_lock);
 		return workspace;
 
 	}
-	if (atomic_read(alloc_workspace) > cpus) {
+	if (atomic_read(alloc_ws) > cpus) {
 		DEFINE_WAIT(wait);
 
-		spin_unlock(workspace_lock);
-		prepare_to_wait(workspace_wait, &wait, TASK_UNINTERRUPTIBLE);
-		if (atomic_read(alloc_workspace) > cpus && !*num_workspace)
+		spin_unlock(ws_lock);
+		prepare_to_wait(ws_wait, &wait, TASK_UNINTERRUPTIBLE);
+		if (atomic_read(alloc_ws) > cpus && !*num_ws)
 			schedule();
-		finish_wait(workspace_wait, &wait);
+		finish_wait(ws_wait, &wait);
 		goto again;
 	}
-	atomic_inc(alloc_workspace);
-	spin_unlock(workspace_lock);
+	atomic_inc(alloc_ws);
+	spin_unlock(ws_lock);
 
 	workspace = btrfs_compress_op[idx]->alloc_workspace();
 	if (IS_ERR(workspace)) {
-		atomic_dec(alloc_workspace);
-		wake_up(workspace_wait);
+		atomic_dec(alloc_ws);
+		wake_up(ws_wait);
 	}
 	return workspace;
 }
@@ -821,27 +823,30 @@ again:
 static void free_workspace(int type, struct list_head *workspace)
 {
 	int idx = type - 1;
-	struct list_head *idle_workspace	= &comp_idle_workspace[idx];
-	spinlock_t *workspace_lock		= &comp_workspace_lock[idx];
-	atomic_t *alloc_workspace		= &comp_alloc_workspace[idx];
-	wait_queue_head_t *workspace_wait	= &comp_workspace_wait[idx];
-	int *num_workspace			= &comp_num_workspace[idx];
-
-	spin_lock(workspace_lock);
-	if (*num_workspace < num_online_cpus()) {
-		list_add(workspace, idle_workspace);
-		(*num_workspace)++;
-		spin_unlock(workspace_lock);
+	struct list_head *idle_ws	= &btrfs_comp_ws[idx].idle_ws;
+	spinlock_t *ws_lock		= &btrfs_comp_ws[idx].ws_lock;
+	atomic_t *alloc_ws		= &btrfs_comp_ws[idx].alloc_ws;
+	wait_queue_head_t *ws_wait	= &btrfs_comp_ws[idx].ws_wait;
+	int *num_ws			= &btrfs_comp_ws[idx].num_ws;
+
+	spin_lock(ws_lock);
+	if (*num_ws < num_online_cpus()) {
+		list_add(workspace, idle_ws);
+		(*num_ws)++;
+		spin_unlock(ws_lock);
 		goto wake;
 	}
-	spin_unlock(workspace_lock);
+	spin_unlock(ws_lock);
 
 	btrfs_compress_op[idx]->free_workspace(workspace);
-	atomic_dec(alloc_workspace);
+	atomic_dec(alloc_ws);
 wake:
+	/*
+	 * Make sure counter is updated before we wake up waiters.
+	 */
 	smp_mb();
-	if (waitqueue_active(workspace_wait))
-		wake_up(workspace_wait);
+	if (waitqueue_active(ws_wait))
+		wake_up(ws_wait);
 }
 
 /*
@@ -853,11 +858,11 @@ static void free_workspaces(void)
 	int i;
 
 	for (i = 0; i < BTRFS_COMPRESS_TYPES; i++) {
-		while (!list_empty(&comp_idle_workspace[i])) {
-			workspace = comp_idle_workspace[i].next;
+		while (!list_empty(&btrfs_comp_ws[i].idle_ws)) {
+			workspace = btrfs_comp_ws[i].idle_ws.next;
 			list_del(workspace);
 			btrfs_compress_op[i]->free_workspace(workspace);
-			atomic_dec(&comp_alloc_workspace[i]);
+			atomic_dec(&btrfs_comp_ws[i].alloc_ws);
 		}
 	}
 }

fs/btrfs/ctree.c

@@ -1011,7 +1011,7 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 			return ret;
 		if (refs == 0) {
 			ret = -EROFS;
-			btrfs_std_error(root->fs_info, ret);
+			btrfs_std_error(root->fs_info, ret, NULL);
 			return ret;
 		}
 	} else {
@@ -1927,7 +1927,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 		child = read_node_slot(root, mid, 0);
 		if (!child) {
 			ret = -EROFS;
-			btrfs_std_error(root->fs_info, ret);
+			btrfs_std_error(root->fs_info, ret, NULL);
 			goto enospc;
 		}
 
@@ -2030,7 +2030,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 		 */
 		if (!left) {
 			ret = -EROFS;
-			btrfs_std_error(root->fs_info, ret);
+			btrfs_std_error(root->fs_info, ret, NULL);
 			goto enospc;
 		}
 		wret = balance_node_right(trans, root, mid, left);
@@ -4940,8 +4940,8 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 {
 	struct extent_buffer *leaf;
 	struct btrfs_item *item;
-	int last_off;
-	int dsize = 0;
+	u32 last_off;
+	u32 dsize = 0;
 	int ret = 0;
 	int wret;
 	int i;

fs/btrfs/ctree.h

@@ -823,8 +823,18 @@ struct btrfs_disk_balance_args {
 	 */
 	__le64 profiles;
 
-	/* usage filter */
-	__le64 usage;
+	/*
+	 * usage filter
+	 * BTRFS_BALANCE_ARGS_USAGE with a single value means '0..N'
+	 * BTRFS_BALANCE_ARGS_USAGE_RANGE - range syntax, min..max
+	 */
+	union {
+		__le64 usage;
+		struct {
+			__le32 usage_min;
+			__le32 usage_max;
+		};
+	};
 
 	/* devid filter */
 	__le64 devid;
@@ -846,10 +856,27 @@ struct btrfs_disk_balance_args {
 	/* BTRFS_BALANCE_ARGS_* */
 	__le64 flags;
 
-	/* BTRFS_BALANCE_ARGS_LIMIT value */
-	__le64 limit;
+	/*
+	 * BTRFS_BALANCE_ARGS_LIMIT with value 'limit'
+	 * BTRFS_BALANCE_ARGS_LIMIT_RANGE - the extend version can use minimum
+	 * and maximum
+	 */
+	union {
+		__le64 limit;
+		struct {
+			__le32 limit_min;
+			__le32 limit_max;
+		};
+	};
 
-	__le64 unused[7];
+	/*
+	 * Process chunks that cross stripes_min..stripes_max devices,
+	 * BTRFS_BALANCE_ARGS_STRIPES_RANGE
+	 */
+	__le32 stripes_min;
+	__le32 stripes_max;
+
+	__le64 unused[6];
 } __attribute__ ((__packed__));
 
 /*
@@ -1154,6 +1181,10 @@ struct btrfs_space_info {
 				   delalloc/allocations */
 	u64 bytes_readonly;	/* total bytes that are read only */
 
+	u64 max_extent_size;	/* This will hold the maximum extent size of
+				   the space info if we had an ENOSPC in the
+				   allocator. */
+
 	unsigned int full:1;	/* indicates that we cannot allocate any more
 				   chunks for this space */
 	unsigned int chunk_alloc:1;	/* set if we are allocating a chunk */
@@ -1228,6 +1259,9 @@ struct btrfs_free_cluster {
 	/* first extent starting offset */
 	u64 window_start;
 
+	/* We did a full search and couldn't create a cluster */
+	bool fragmented;
+
 	struct btrfs_block_group_cache *block_group;
 	/*
 	 * when a cluster is allocated from a block group, we put the
@@ -1943,6 +1977,9 @@ struct btrfs_root {
 	int send_in_progress;
 	struct btrfs_subvolume_writers *subv_writers;
 	atomic_t will_be_snapshoted;
+
+	/* For qgroup metadata space reserve */
+	atomic_t qgroup_meta_rsv;
 };
 
 struct btrfs_ioctl_defrag_range_args {
@@ -2145,6 +2182,8 @@ struct btrfs_ioctl_defrag_range_args {
 #define BTRFS_MOUNT_CHECK_INTEGRITY_INCLUDING_EXTENT_DATA (1 << 21)
 #define BTRFS_MOUNT_PANIC_ON_FATAL_ERROR	(1 << 22)
 #define BTRFS_MOUNT_RESCAN_UUID_TREE	(1 << 23)
+#define BTRFS_MOUNT_FRAGMENT_DATA	(1 << 24)
+#define BTRFS_MOUNT_FRAGMENT_METADATA	(1 << 25)
 
 #define BTRFS_DEFAULT_COMMIT_INTERVAL	(30)
 #define BTRFS_DEFAULT_MAX_INLINE	(8192)
@@ -2169,6 +2208,18 @@ struct btrfs_ioctl_defrag_range_args {
 	btrfs_clear_opt(root->fs_info->mount_opt, opt);			\
 }
 
+#ifdef CONFIG_BTRFS_DEBUG
+static inline int
+btrfs_should_fragment_free_space(struct btrfs_root *root,
+				 struct btrfs_block_group_cache *block_group)
+{
+	return (btrfs_test_opt(root, FRAGMENT_METADATA) &&
+		block_group->flags & BTRFS_BLOCK_GROUP_METADATA) ||
+	       (btrfs_test_opt(root, FRAGMENT_DATA) &&
+		block_group->flags &  BTRFS_BLOCK_GROUP_DATA);
+}
+#endif
+
 /*
  * Requests for changes that need to be done during transaction commit.
  *
@@ -3379,7 +3430,8 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root,
 				     u64 root_objectid, u64 owner,
-				     u64 offset, struct btrfs_key *ins);
+				     u64 offset, u64 ram_bytes,
+				     struct btrfs_key *ins);
 int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root,
 				   u64 root_objectid, u64 owner, u64 offset,
@@ -3398,7 +3450,7 @@ int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans,
 int btrfs_free_extent(struct btrfs_trans_handle *trans,
 		      struct btrfs_root *root,
 		      u64 bytenr, u64 num_bytes, u64 parent, u64 root_objectid,
-		      u64 owner, u64 offset, int no_quota);
+		      u64 owner, u64 offset);
 
 int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len,
 			       int delalloc);
@@ -3411,7 +3463,7 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans,
 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 			 struct btrfs_root *root,
 			 u64 bytenr, u64 num_bytes, u64 parent,
-			 u64 root_objectid, u64 owner, u64 offset, int no_quota);
+			 u64 root_objectid, u64 owner, u64 offset);
 
 int btrfs_start_dirty_block_groups(struct btrfs_trans_handle *trans,
 				   struct btrfs_root *root);
@@ -3449,8 +3501,11 @@ enum btrfs_reserve_flush_enum {
 	BTRFS_RESERVE_FLUSH_ALL,
 };
 
-int btrfs_check_data_free_space(struct inode *inode, u64 bytes, u64 write_bytes);
-void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes);
+int btrfs_check_data_free_space(struct inode *inode, u64 start, u64 len);
+int btrfs_alloc_data_chunk_ondemand(struct inode *inode, u64 bytes);
+void btrfs_free_reserved_data_space(struct inode *inode, u64 start, u64 len);
+void btrfs_free_reserved_data_space_noquota(struct inode *inode, u64 start,
+					    u64 len);
 void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root);
 void btrfs_trans_release_chunk_metadata(struct btrfs_trans_handle *trans);
@@ -3466,8 +3521,8 @@ void btrfs_subvolume_release_metadata(struct btrfs_root *root,
 				      u64 qgroup_reserved);
 int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes);
 void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes);
-int btrfs_delalloc_reserve_space(struct inode *inode, u64 num_bytes);
-void btrfs_delalloc_release_space(struct inode *inode, u64 num_bytes);
+int btrfs_delalloc_reserve_space(struct inode *inode, u64 start, u64 len);
+void btrfs_delalloc_release_space(struct inode *inode, u64 start, u64 len);
 void btrfs_init_block_rsv(struct btrfs_block_rsv *rsv, unsigned short type);
 struct btrfs_block_rsv *btrfs_alloc_block_rsv(struct btrfs_root *root,
 					      unsigned short type);
@@ -4004,8 +4059,8 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
 /* sysfs.c */
 int btrfs_init_sysfs(void);
 void btrfs_exit_sysfs(void);
-int btrfs_sysfs_add_one(struct btrfs_fs_info *fs_info);
-void btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info);
+int btrfs_sysfs_add_mounted(struct btrfs_fs_info *fs_info);
+void btrfs_sysfs_remove_mounted(struct btrfs_fs_info *fs_info);
 
 /* xattr.c */
 ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size);
@@ -4039,14 +4094,102 @@ void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...)
 #define btrfs_info(fs_info, fmt, args...) \
 	btrfs_printk(fs_info, KERN_INFO fmt, ##args)
 
+/*
+ * Wrappers that use printk_in_rcu
+ */
+#define btrfs_emerg_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_EMERG fmt, ##args)
+#define btrfs_alert_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_ALERT fmt, ##args)
+#define btrfs_crit_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_CRIT fmt, ##args)
+#define btrfs_err_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_ERR fmt, ##args)
+#define btrfs_warn_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_WARNING fmt, ##args)
+#define btrfs_notice_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_NOTICE fmt, ##args)
+#define btrfs_info_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_INFO fmt, ##args)
+
+/*
+ * Wrappers that use a ratelimited printk_in_rcu
+ */
+#define btrfs_emerg_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_EMERG fmt, ##args)
+#define btrfs_alert_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_ALERT fmt, ##args)
+#define btrfs_crit_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_CRIT fmt, ##args)
+#define btrfs_err_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_ERR fmt, ##args)
+#define btrfs_warn_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_WARNING fmt, ##args)
+#define btrfs_notice_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_NOTICE fmt, ##args)
+#define btrfs_info_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_INFO fmt, ##args)
+
+/*
+ * Wrappers that use a ratelimited printk
+ */
+#define btrfs_emerg_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_EMERG fmt, ##args)
+#define btrfs_alert_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_ALERT fmt, ##args)
+#define btrfs_crit_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_CRIT fmt, ##args)
+#define btrfs_err_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_ERR fmt, ##args)
+#define btrfs_warn_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_WARNING fmt, ##args)
+#define btrfs_notice_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_NOTICE fmt, ##args)
+#define btrfs_info_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_INFO fmt, ##args)
 #ifdef DEBUG
 #define btrfs_debug(fs_info, fmt, args...) \
 	btrfs_printk(fs_info, KERN_DEBUG fmt, ##args)
+#define btrfs_debug_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_in_rcu(fs_info, KERN_DEBUG fmt, ##args)
+#define btrfs_debug_rl_in_rcu(fs_info, fmt, args...) \
+	btrfs_printk_rl_in_rcu(fs_info, KERN_DEBUG fmt, ##args)
+#define btrfs_debug_rl(fs_info, fmt, args...) \
+	btrfs_printk_ratelimited(fs_info, KERN_DEBUG fmt, ##args)
 #else
 #define btrfs_debug(fs_info, fmt, args...) \
     no_printk(KERN_DEBUG fmt, ##args)
+#define btrfs_debug_in_rcu(fs_info, fmt, args...) \
+	no_printk(KERN_DEBUG fmt, ##args)
+#define btrfs_debug_rl_in_rcu(fs_info, fmt, args...) \
+	no_printk(KERN_DEBUG fmt, ##args)
+#define btrfs_debug_rl(fs_info, fmt, args...) \
+	no_printk(KERN_DEBUG fmt, ##args)
 #endif
 
+#define btrfs_printk_in_rcu(fs_info, fmt, args...)	\
+do {							\
+	rcu_read_lock();				\
+	btrfs_printk(fs_info, fmt, ##args);		\
+	rcu_read_unlock();				\
+} while (0)
+
+#define btrfs_printk_ratelimited(fs_info, fmt, args...)		\
+do {								\
+	static DEFINE_RATELIMIT_STATE(_rs,			\
+		DEFAULT_RATELIMIT_INTERVAL,			\
+		DEFAULT_RATELIMIT_BURST);       		\
+	if (__ratelimit(&_rs))					\
+		btrfs_printk(fs_info, fmt, ##args);		\
+} while (0)
+
+#define btrfs_printk_rl_in_rcu(fs_info, fmt, args...)		\
+do {								\
+	rcu_read_lock();					\
+	btrfs_printk_ratelimited(fs_info, fmt, ##args);		\
+	rcu_read_unlock();					\
+} while (0)
+
 #ifdef CONFIG_BTRFS_ASSERT
 
 __cold
@@ -4127,14 +4270,7 @@ do {								\
 				  __LINE__, (errno));		\
 } while (0)
 
-#define btrfs_std_error(fs_info, errno)				\
-do {								\
-	if ((errno))						\
-		__btrfs_std_error((fs_info), __func__,		\
-				   __LINE__, (errno), NULL);	\
-} while (0)
-
-#define btrfs_error(fs_info, errno, fmt, args...)		\
+#define btrfs_std_error(fs_info, errno, fmt, args...)		\
 do {								\
 	__btrfs_std_error((fs_info), __func__, __LINE__,	\
 			  (errno), fmt, ##args);		\

fs/btrfs/delayed-inode.c

@@ -463,6 +463,10 @@ static int __btrfs_add_delayed_deletion_item(struct btrfs_delayed_node *node,
 static void finish_one_item(struct btrfs_delayed_root *delayed_root)
 {
 	int seq = atomic_inc_return(&delayed_root->items_seq);
+
+	/*
+	 * atomic_dec_return implies a barrier for waitqueue_active
+	 */
 	if ((atomic_dec_return(&delayed_root->items) <
 	    BTRFS_DELAYED_BACKGROUND || seq % BTRFS_DELAYED_BATCH == 0) &&
 	    waitqueue_active(&delayed_root->wait))

fs/btrfs/delayed-ref.c

@@ -197,6 +197,119 @@ static inline void drop_delayed_ref(struct btrfs_trans_handle *trans,
 		trans->delayed_ref_updates--;
 }
 
+static bool merge_ref(struct btrfs_trans_handle *trans,
+		      struct btrfs_delayed_ref_root *delayed_refs,
+		      struct btrfs_delayed_ref_head *head,
+		      struct btrfs_delayed_ref_node *ref,
+		      u64 seq)
+{
+	struct btrfs_delayed_ref_node *next;
+	bool done = false;
+
+	next = list_first_entry(&head->ref_list, struct btrfs_delayed_ref_node,
+				list);
+	while (!done && &next->list != &head->ref_list) {
+		int mod;
+		struct btrfs_delayed_ref_node *next2;
+
+		next2 = list_next_entry(next, list);
+
+		if (next == ref)
+			goto next;
+
+		if (seq && next->seq >= seq)
+			goto next;
+
+		if (next->type != ref->type)
+			goto next;
+
+		if ((ref->type == BTRFS_TREE_BLOCK_REF_KEY ||
+		     ref->type == BTRFS_SHARED_BLOCK_REF_KEY) &&
+		    comp_tree_refs(btrfs_delayed_node_to_tree_ref(ref),
+				   btrfs_delayed_node_to_tree_ref(next),
+				   ref->type))
+			goto next;
+		if ((ref->type == BTRFS_EXTENT_DATA_REF_KEY ||
+		     ref->type == BTRFS_SHARED_DATA_REF_KEY) &&
+		    comp_data_refs(btrfs_delayed_node_to_data_ref(ref),
+				   btrfs_delayed_node_to_data_ref(next)))
+			goto next;
+
+		if (ref->action == next->action) {
+			mod = next->ref_mod;
+		} else {
+			if (ref->ref_mod < next->ref_mod) {
+				swap(ref, next);
+				done = true;
+			}
+			mod = -next->ref_mod;
+		}
+
+		drop_delayed_ref(trans, delayed_refs, head, next);
+		ref->ref_mod += mod;
+		if (ref->ref_mod == 0) {
+			drop_delayed_ref(trans, delayed_refs, head, ref);
+			done = true;
+		} else {
+			/*
+			 * Can't have multiples of the same ref on a tree block.
+			 */
+			WARN_ON(ref->type == BTRFS_TREE_BLOCK_REF_KEY ||
+				ref->type == BTRFS_SHARED_BLOCK_REF_KEY);
+		}
+next:
+		next = next2;
+	}
+
+	return done;
+}
+
+void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans,
+			      struct btrfs_fs_info *fs_info,
+			      struct btrfs_delayed_ref_root *delayed_refs,
+			      struct btrfs_delayed_ref_head *head)
+{
+	struct btrfs_delayed_ref_node *ref;
+	u64 seq = 0;
+
+	assert_spin_locked(&head->lock);
+
+	if (list_empty(&head->ref_list))
+		return;
+
+	/* We don't have too many refs to merge for data. */
+	if (head->is_data)
+		return;
+
+	spin_lock(&fs_info->tree_mod_seq_lock);
+	if (!list_empty(&fs_info->tree_mod_seq_list)) {
+		struct seq_list *elem;
+
+		elem = list_first_entry(&fs_info->tree_mod_seq_list,
+					struct seq_list, list);
+		seq = elem->seq;
+	}
+	spin_unlock(&fs_info->tree_mod_seq_lock);
+
+	ref = list_first_entry(&head->ref_list, struct btrfs_delayed_ref_node,
+			       list);
+	while (&ref->list != &head->ref_list) {
+		if (seq && ref->seq >= seq)
+			goto next;
+
+		if (merge_ref(trans, delayed_refs, head, ref, seq)) {
+			if (list_empty(&head->ref_list))
+				break;
+			ref = list_first_entry(&head->ref_list,
+					       struct btrfs_delayed_ref_node,
+					       list);
+			continue;
+		}
+next:
+		ref = list_next_entry(ref, list);
+	}
+}
+
 int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,
 			    struct btrfs_delayed_ref_root *delayed_refs,
 			    u64 seq)
@@ -292,8 +405,7 @@ add_delayed_ref_tail_merge(struct btrfs_trans_handle *trans,
 	exist = list_entry(href->ref_list.prev, struct btrfs_delayed_ref_node,
 			   list);
 	/* No need to compare bytenr nor is_head */
-	if (exist->type != ref->type || exist->no_quota != ref->no_quota ||
-	    exist->seq != ref->seq)
+	if (exist->type != ref->type || exist->seq != ref->seq)
 		goto add_tail;
 
 	if ((exist->type == BTRFS_TREE_BLOCK_REF_KEY ||
@@ -423,7 +535,8 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 		     struct btrfs_trans_handle *trans,
 		     struct btrfs_delayed_ref_node *ref,
 		     struct btrfs_qgroup_extent_record *qrecord,
-		     u64 bytenr, u64 num_bytes, int action, int is_data)
+		     u64 bytenr, u64 num_bytes, u64 ref_root, u64 reserved,
+		     int action, int is_data)
 {
 	struct btrfs_delayed_ref_head *existing;
 	struct btrfs_delayed_ref_head *head_ref = NULL;
@@ -432,6 +545,9 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	int count_mod = 1;
 	int must_insert_reserved = 0;
 
+	/* If reserved is provided, it must be a data extent. */
+	BUG_ON(!is_data && reserved);
+
 	/*
 	 * the head node stores the sum of all the mods, so dropping a ref
 	 * should drop the sum in the head node by one.
@@ -476,9 +592,16 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	INIT_LIST_HEAD(&head_ref->ref_list);
 	head_ref->processing = 0;
 	head_ref->total_ref_mod = count_mod;
+	head_ref->qgroup_reserved = 0;
+	head_ref->qgroup_ref_root = 0;
 
 	/* Record qgroup extent info if provided */
 	if (qrecord) {
+		if (ref_root && reserved) {
+			head_ref->qgroup_ref_root = ref_root;
+			head_ref->qgroup_reserved = reserved;
+		}
+
 		qrecord->bytenr = bytenr;
 		qrecord->num_bytes = num_bytes;
 		qrecord->old_roots = NULL;
@@ -497,6 +620,8 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	existing = htree_insert(&delayed_refs->href_root,
 				&head_ref->href_node);
 	if (existing) {
+		WARN_ON(ref_root && reserved && existing->qgroup_ref_root
+			&& existing->qgroup_reserved);
 		update_existing_head_ref(delayed_refs, &existing->node, ref);
 		/*
 		 * we've updated the existing ref, free the newly
@@ -524,7 +649,7 @@ add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 		     struct btrfs_delayed_ref_head *head_ref,
 		     struct btrfs_delayed_ref_node *ref, u64 bytenr,
 		     u64 num_bytes, u64 parent, u64 ref_root, int level,
-		     int action, int no_quota)
+		     int action)
 {
 	struct btrfs_delayed_tree_ref *full_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
@@ -546,7 +671,6 @@ add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 	ref->action = action;
 	ref->is_head = 0;
 	ref->in_tree = 1;
-	ref->no_quota = no_quota;
 	ref->seq = seq;
 
 	full_ref = btrfs_delayed_node_to_tree_ref(ref);
@@ -579,7 +703,7 @@ add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 		     struct btrfs_delayed_ref_head *head_ref,
 		     struct btrfs_delayed_ref_node *ref, u64 bytenr,
 		     u64 num_bytes, u64 parent, u64 ref_root, u64 owner,
-		     u64 offset, int action, int no_quota)
+		     u64 offset, int action)
 {
 	struct btrfs_delayed_data_ref *full_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
@@ -602,7 +726,6 @@ add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	ref->action = action;
 	ref->is_head = 0;
 	ref->in_tree = 1;
-	ref->no_quota = no_quota;
 	ref->seq = seq;
 
 	full_ref = btrfs_delayed_node_to_data_ref(ref);
@@ -633,17 +756,13 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes, u64 parent,
 			       u64 ref_root,  int level, int action,
-			       struct btrfs_delayed_extent_op *extent_op,
-			       int no_quota)
+			       struct btrfs_delayed_extent_op *extent_op)
 {
 	struct btrfs_delayed_tree_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
 	struct btrfs_qgroup_extent_record *record = NULL;
 
-	if (!is_fstree(ref_root) || !fs_info->quota_enabled)
-		no_quota = 0;
-
 	BUG_ON(extent_op && extent_op->is_data);
 	ref = kmem_cache_alloc(btrfs_delayed_tree_ref_cachep, GFP_NOFS);
 	if (!ref)
@@ -669,11 +788,10 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 	 * the spin lock
 	 */
 	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, record,
-					bytenr, num_bytes, action, 0);
+					bytenr, num_bytes, 0, 0, action, 0);
 
 	add_delayed_tree_ref(fs_info, trans, head_ref, &ref->node, bytenr,
-				   num_bytes, parent, ref_root, level, action,
-				   no_quota);
+			     num_bytes, parent, ref_root, level, action);
 	spin_unlock(&delayed_refs->lock);
 
 	return 0;
@@ -693,18 +811,14 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes,
 			       u64 parent, u64 ref_root,
-			       u64 owner, u64 offset, int action,
-			       struct btrfs_delayed_extent_op *extent_op,
-			       int no_quota)
+			       u64 owner, u64 offset, u64 reserved, int action,
+			       struct btrfs_delayed_extent_op *extent_op)
 {
 	struct btrfs_delayed_data_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
 	struct btrfs_qgroup_extent_record *record = NULL;
 
-	if (!is_fstree(ref_root) || !fs_info->quota_enabled)
-		no_quota = 0;
-
 	BUG_ON(extent_op && !extent_op->is_data);
 	ref = kmem_cache_alloc(btrfs_delayed_data_ref_cachep, GFP_NOFS);
 	if (!ref)
@@ -736,16 +850,44 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	 * the spin lock
 	 */
 	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, record,
-					bytenr, num_bytes, action, 1);
+					bytenr, num_bytes, ref_root, reserved,
+					action, 1);
 
 	add_delayed_data_ref(fs_info, trans, head_ref, &ref->node, bytenr,
 				   num_bytes, parent, ref_root, owner, offset,
-				   action, no_quota);
+				   action);
 	spin_unlock(&delayed_refs->lock);
 
 	return 0;
 }
 
+int btrfs_add_delayed_qgroup_reserve(struct btrfs_fs_info *fs_info,
+				     struct btrfs_trans_handle *trans,
+				     u64 ref_root, u64 bytenr, u64 num_bytes)
+{
+	struct btrfs_delayed_ref_root *delayed_refs;
+	struct btrfs_delayed_ref_head *ref_head;
+	int ret = 0;
+
+	if (!fs_info->quota_enabled || !is_fstree(ref_root))
+		return 0;
+
+	delayed_refs = &trans->transaction->delayed_refs;
+
+	spin_lock(&delayed_refs->lock);
+	ref_head = find_ref_head(&delayed_refs->href_root, bytenr, 0);
+	if (!ref_head) {
+		ret = -ENOENT;
+		goto out;
+	}
+	WARN_ON(ref_head->qgroup_reserved || ref_head->qgroup_ref_root);
+	ref_head->qgroup_ref_root = ref_root;
+	ref_head->qgroup_reserved = num_bytes;
+out:
+	spin_unlock(&delayed_refs->lock);
+	return ret;
+}
+
 int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 				struct btrfs_trans_handle *trans,
 				u64 bytenr, u64 num_bytes,
@@ -764,7 +906,7 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 	spin_lock(&delayed_refs->lock);
 
 	add_delayed_ref_head(fs_info, trans, &head_ref->node, NULL, bytenr,
-			     num_bytes, BTRFS_UPDATE_DELAYED_HEAD,
+			     num_bytes, 0, 0, BTRFS_UPDATE_DELAYED_HEAD,
 			     extent_op->is_data);
 
 	spin_unlock(&delayed_refs->lock);

fs/btrfs/delayed-ref.h

@@ -68,7 +68,6 @@ struct btrfs_delayed_ref_node {
 
 	unsigned int action:8;
 	unsigned int type:8;
-	unsigned int no_quota:1;
 	/* is this node still in the rbtree? */
 	unsigned int is_head:1;
 	unsigned int in_tree:1;
@@ -112,6 +111,17 @@ struct btrfs_delayed_ref_head {
 	 */
 	int total_ref_mod;
 
+	/*
+	 * For qgroup reserved space freeing.
+	 *
+	 * ref_root and reserved will be recorded after
+	 * BTRFS_ADD_DELAYED_EXTENT is called.
+	 * And will be used to free reserved qgroup space at
+	 * run_delayed_refs() time.
+	 */
+	u64 qgroup_ref_root;
+	u64 qgroup_reserved;
+
 	/*
 	 * when a new extent is allocated, it is just reserved in memory
 	 * The actual extent isn't inserted into the extent allocation tree
@@ -233,15 +243,16 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes, u64 parent,
 			       u64 ref_root, int level, int action,
-			       struct btrfs_delayed_extent_op *extent_op,
-			       int no_quota);
+			       struct btrfs_delayed_extent_op *extent_op);
 int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes,
 			       u64 parent, u64 ref_root,
-			       u64 owner, u64 offset, int action,
-			       struct btrfs_delayed_extent_op *extent_op,
-			       int no_quota);
+			       u64 owner, u64 offset, u64 reserved, int action,
+			       struct btrfs_delayed_extent_op *extent_op);
+int btrfs_add_delayed_qgroup_reserve(struct btrfs_fs_info *fs_info,
+				     struct btrfs_trans_handle *trans,
+				     u64 ref_root, u64 bytenr, u64 num_bytes);
 int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 				struct btrfs_trans_handle *trans,
 				u64 bytenr, u64 num_bytes,

fs/btrfs/dev-replace.c

@@ -327,19 +327,6 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	    args->start.tgtdev_name[0] == '\0')
 		return -EINVAL;
 
-	/*
-	 * Here we commit the transaction to make sure commit_total_bytes
-	 * of all the devices are updated.
-	 */
-	trans = btrfs_attach_transaction(root);
-	if (!IS_ERR(trans)) {
-		ret = btrfs_commit_transaction(trans, root);
-		if (ret)
-			return ret;
-	} else if (PTR_ERR(trans) != -ENOENT) {
-		return PTR_ERR(trans);
-	}
-
 	/* the disk copy procedure reuses the scrub code */
 	mutex_lock(&fs_info->volume_mutex);
 	ret = btrfs_dev_replace_find_srcdev(root, args->start.srcdevid,
@@ -356,6 +343,19 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	if (ret)
 		return ret;
 
+	/*
+	 * Here we commit the transaction to make sure commit_total_bytes
+	 * of all the devices are updated.
+	 */
+	trans = btrfs_attach_transaction(root);
+	if (!IS_ERR(trans)) {
+		ret = btrfs_commit_transaction(trans, root);
+		if (ret)
+			return ret;
+	} else if (PTR_ERR(trans) != -ENOENT) {
+		return PTR_ERR(trans);
+	}
+
 	btrfs_dev_replace_lock(dev_replace);
 	switch (dev_replace->replace_state) {
 	case BTRFS_IOCTL_DEV_REPLACE_STATE_NEVER_STARTED:
@@ -375,12 +375,8 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	WARN_ON(!tgt_device);
 	dev_replace->tgtdev = tgt_device;
 
-	ret = btrfs_kobj_add_device(tgt_device->fs_devices, tgt_device);
-	if (ret)
-		btrfs_err(root->fs_info, "kobj add dev failed %d\n", ret);
-
-	printk_in_rcu(KERN_INFO
-		      "BTRFS: dev_replace from %s (devid %llu) to %s started\n",
+	btrfs_info_in_rcu(root->fs_info,
+		      "dev_replace from %s (devid %llu) to %s started",
 		      src_device->missing ? "<missing disk>" :
 		        rcu_str_deref(src_device->name),
 		      src_device->devid,
@@ -401,6 +397,10 @@ int btrfs_dev_replace_start(struct btrfs_root *root,
 	args->result = BTRFS_IOCTL_DEV_REPLACE_RESULT_NO_ERROR;
 	btrfs_dev_replace_unlock(dev_replace);
 
+	ret = btrfs_sysfs_add_device_link(tgt_device->fs_devices, tgt_device);
+	if (ret)
+		btrfs_err(root->fs_info, "kobj add dev failed %d\n", ret);
+
 	btrfs_wait_ordered_roots(root->fs_info, -1);
 
 	/* force writing the updated state information to disk */
@@ -454,8 +454,7 @@ static void btrfs_rm_dev_replace_blocked(struct btrfs_fs_info *fs_info)
 static void btrfs_rm_dev_replace_unblocked(struct btrfs_fs_info *fs_info)
 {
 	clear_bit(BTRFS_FS_STATE_DEV_REPLACING, &fs_info->fs_state);
-	if (waitqueue_active(&fs_info->replace_wait))
-		wake_up(&fs_info->replace_wait);
+	wake_up(&fs_info->replace_wait);
 }
 
 static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
@@ -523,8 +522,8 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 								src_device,
 								tgt_device);
 	} else {
-		printk_in_rcu(KERN_ERR
-			      "BTRFS: btrfs_scrub_dev(%s, %llu, %s) failed %d\n",
+		btrfs_err_in_rcu(root->fs_info,
+			      "btrfs_scrub_dev(%s, %llu, %s) failed %d",
 			      src_device->missing ? "<missing disk>" :
 			        rcu_str_deref(src_device->name),
 			      src_device->devid,
@@ -540,8 +539,8 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 		return scrub_ret;
 	}
 
-	printk_in_rcu(KERN_INFO
-		      "BTRFS: dev_replace from %s (devid %llu) to %s finished\n",
+	btrfs_info_in_rcu(root->fs_info,
+		      "dev_replace from %s (devid %llu) to %s finished",
 		      src_device->missing ? "<missing disk>" :
 		        rcu_str_deref(src_device->name),
 		      src_device->devid,
@@ -586,7 +585,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 	mutex_unlock(&uuid_mutex);
 
 	/* replace the sysfs entry */
-	btrfs_kobj_rm_device(fs_info->fs_devices, src_device);
+	btrfs_sysfs_rm_device_link(fs_info->fs_devices, src_device);
 	btrfs_rm_dev_replace_free_srcdev(fs_info, src_device);
 
 	/* write back the superblocks */
@@ -809,8 +808,8 @@ static int btrfs_dev_replace_kthread(void *data)
 		progress = status_args->status.progress_1000;
 		kfree(status_args);
 		progress = div_u64(progress, 10);
-		printk_in_rcu(KERN_INFO
-			"BTRFS: continuing dev_replace from %s (devid %llu) to %s @%u%%\n",
+		btrfs_info_in_rcu(fs_info,
+			"continuing dev_replace from %s (devid %llu) to %s @%u%%",
 			dev_replace->srcdev->missing ? "<missing disk>" :
 			rcu_str_deref(dev_replace->srcdev->name),
 			dev_replace->srcdev->devid,

fs/btrfs/disk-io.c

@@ -319,9 +319,9 @@ static int csum_tree_block(struct btrfs_fs_info *fs_info,
 			memcpy(&found, result, csum_size);
 
 			read_extent_buffer(buf, &val, 0, csum_size);
-			printk_ratelimited(KERN_WARNING
-				"BTRFS: %s checksum verify failed on %llu wanted %X found %X "
-				"level %d\n",
+			btrfs_warn_rl(fs_info,
+				"%s checksum verify failed on %llu wanted %X found %X "
+				"level %d",
 				fs_info->sb->s_id, buf->start,
 				val, found, btrfs_header_level(buf));
 			if (result != (char *)&inline_result)
@@ -368,9 +368,9 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 		ret = 0;
 		goto out;
 	}
-	printk_ratelimited(KERN_ERR
-	    "BTRFS (device %s): parent transid verify failed on %llu wanted %llu found %llu\n",
-			eb->fs_info->sb->s_id, eb->start,
+	btrfs_err_rl(eb->fs_info,
+		"parent transid verify failed on %llu wanted %llu found %llu",
+			eb->start,
 			parent_transid, btrfs_header_generation(eb));
 	ret = 1;
 
@@ -629,15 +629,14 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 
 	found_start = btrfs_header_bytenr(eb);
 	if (found_start != eb->start) {
-		printk_ratelimited(KERN_ERR "BTRFS (device %s): bad tree block start "
-			       "%llu %llu\n",
-			       eb->fs_info->sb->s_id, found_start, eb->start);
+		btrfs_err_rl(eb->fs_info, "bad tree block start %llu %llu",
+			       found_start, eb->start);
 		ret = -EIO;
 		goto err;
 	}
 	if (check_tree_block_fsid(root->fs_info, eb)) {
-		printk_ratelimited(KERN_ERR "BTRFS (device %s): bad fsid on block %llu\n",
-			       eb->fs_info->sb->s_id, eb->start);
+		btrfs_err_rl(eb->fs_info, "bad fsid on block %llu",
+			       eb->start);
 		ret = -EIO;
 		goto err;
 	}
@@ -802,6 +801,9 @@ static void run_one_async_done(struct btrfs_work *work)
 	limit = btrfs_async_submit_limit(fs_info);
 	limit = limit * 2 / 3;
 
+	/*
+	 * atomic_dec_return implies a barrier for waitqueue_active
+	 */
 	if (atomic_dec_return(&fs_info->nr_async_submits) < limit &&
 	    waitqueue_active(&fs_info->async_submit_wait))
 		wake_up(&fs_info->async_submit_wait);
@@ -1265,6 +1267,7 @@ static void __setup_root(u32 nodesize, u32 sectorsize, u32 stripesize,
 	atomic_set(&root->orphan_inodes, 0);
 	atomic_set(&root->refs, 1);
 	atomic_set(&root->will_be_snapshoted, 0);
+	atomic_set(&root->qgroup_meta_rsv, 0);
 	root->log_transid = 0;
 	root->log_transid_committed = -1;
 	root->last_log_commit = 0;
@@ -1759,6 +1762,7 @@ static int cleaner_kthread(void *arg)
 	int again;
 	struct btrfs_trans_handle *trans;
 
+	set_freezable();
 	do {
 		again = 0;
 
@@ -2348,8 +2352,7 @@ static int btrfs_replay_log(struct btrfs_fs_info *fs_info,
 	u64 bytenr = btrfs_super_log_root(disk_super);
 
 	if (fs_devices->rw_devices == 0) {
-		printk(KERN_WARNING "BTRFS: log replay required "
-		       "on RO media\n");
+		btrfs_warn(fs_info, "log replay required on RO media");
 		return -EIO;
 	}
 
@@ -2364,12 +2367,12 @@ static int btrfs_replay_log(struct btrfs_fs_info *fs_info,
 	log_tree_root->node = read_tree_block(tree_root, bytenr,
 			fs_info->generation + 1);
 	if (IS_ERR(log_tree_root->node)) {
-		printk(KERN_ERR "BTRFS: failed to read log tree\n");
+		btrfs_warn(fs_info, "failed to read log tree");
 		ret = PTR_ERR(log_tree_root->node);
 		kfree(log_tree_root);
 		return ret;
 	} else if (!extent_buffer_uptodate(log_tree_root->node)) {
-		printk(KERN_ERR "BTRFS: failed to read log tree\n");
+		btrfs_err(fs_info, "failed to read log tree");
 		free_extent_buffer(log_tree_root->node);
 		kfree(log_tree_root);
 		return -EIO;
@@ -2377,7 +2380,7 @@ static int btrfs_replay_log(struct btrfs_fs_info *fs_info,
 	/* returns with log_tree_root freed on success */
 	ret = btrfs_recover_log_trees(log_tree_root);
 	if (ret) {
-		btrfs_error(tree_root->fs_info, ret,
+		btrfs_std_error(tree_root->fs_info, ret,
 			    "Failed to recover log tree");
 		free_extent_buffer(log_tree_root->node);
 		kfree(log_tree_root);
@@ -2653,8 +2656,8 @@ int open_ctree(struct super_block *sb,
 	 * Read super block and check the signature bytes only
 	 */
 	bh = btrfs_read_dev_super(fs_devices->latest_bdev);
-	if (!bh) {
-		err = -EINVAL;
+	if (IS_ERR(bh)) {
+		err = PTR_ERR(bh);
 		goto fail_alloc;
 	}
 
@@ -2937,7 +2940,7 @@ retry_root_backup:
 		goto fail_fsdev_sysfs;
 	}
 
-	ret = btrfs_sysfs_add_one(fs_info);
+	ret = btrfs_sysfs_add_mounted(fs_info);
 	if (ret) {
 		pr_err("BTRFS: failed to init sysfs interface: %d\n", ret);
 		goto fail_fsdev_sysfs;
@@ -3117,7 +3120,7 @@ fail_cleaner:
 	filemap_write_and_wait(fs_info->btree_inode->i_mapping);
 
 fail_sysfs:
-	btrfs_sysfs_remove_one(fs_info);
+	btrfs_sysfs_remove_mounted(fs_info);
 
 fail_fsdev_sysfs:
 	btrfs_sysfs_remove_fsid(fs_info->fs_devices);
@@ -3179,8 +3182,8 @@ static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate)
 		struct btrfs_device *device = (struct btrfs_device *)
 			bh->b_private;
 
-		printk_ratelimited_in_rcu(KERN_WARNING "BTRFS: lost page write due to "
-					  "I/O error on %s\n",
+		btrfs_warn_rl_in_rcu(device->dev_root->fs_info,
+				"lost page write due to IO error on %s",
 					  rcu_str_deref(device->name));
 		/* note, we dont' set_buffer_write_io_error because we have
 		 * our own ways of dealing with the IO errors
@@ -3192,6 +3195,37 @@ static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate)
 	put_bh(bh);
 }
 
+int btrfs_read_dev_one_super(struct block_device *bdev, int copy_num,
+			struct buffer_head **bh_ret)
+{
+	struct buffer_head *bh;
+	struct btrfs_super_block *super;
+	u64 bytenr;
+
+	bytenr = btrfs_sb_offset(copy_num);
+	if (bytenr + BTRFS_SUPER_INFO_SIZE >= i_size_read(bdev->bd_inode))
+		return -EINVAL;
+
+	bh = __bread(bdev, bytenr / 4096, BTRFS_SUPER_INFO_SIZE);
+	/*
+	 * If we fail to read from the underlying devices, as of now
+	 * the best option we have is to mark it EIO.
+	 */
+	if (!bh)
+		return -EIO;
+
+	super = (struct btrfs_super_block *)bh->b_data;
+	if (btrfs_super_bytenr(super) != bytenr ||
+		    btrfs_super_magic(super) != BTRFS_MAGIC) {
+		brelse(bh);
+		return -EINVAL;
+	}
+
+	*bh_ret = bh;
+	return 0;
+}
+
+
 struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
 {
 	struct buffer_head *bh;
@@ -3199,7 +3233,7 @@ struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
 	struct btrfs_super_block *super;
 	int i;
 	u64 transid = 0;
-	u64 bytenr;
+	int ret = -EINVAL;
 
 	/* we would like to check all the supers, but that would make
 	 * a btrfs mount succeed after a mkfs from a different FS.
@@ -3207,21 +3241,11 @@ struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
 	 * later supers, using BTRFS_SUPER_MIRROR_MAX instead
 	 */
 	for (i = 0; i < 1; i++) {
-		bytenr = btrfs_sb_offset(i);
-		if (bytenr + BTRFS_SUPER_INFO_SIZE >=
-					i_size_read(bdev->bd_inode))
-			break;
-		bh = __bread(bdev, bytenr / 4096,
-					BTRFS_SUPER_INFO_SIZE);
-		if (!bh)
+		ret = btrfs_read_dev_one_super(bdev, i, &bh);
+		if (ret)
 			continue;
 
 		super = (struct btrfs_super_block *)bh->b_data;
-		if (btrfs_super_bytenr(super) != bytenr ||
-		    btrfs_super_magic(super) != BTRFS_MAGIC) {
-			brelse(bh);
-			continue;
-		}
 
 		if (!latest || btrfs_super_generation(super) > transid) {
 			brelse(latest);
@@ -3231,6 +3255,10 @@ struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
 			brelse(bh);
 		}
 	}
+
+	if (!latest)
+		return ERR_PTR(ret);
+
 	return latest;
 }
 
@@ -3299,8 +3327,9 @@ static int write_dev_supers(struct btrfs_device *device,
 			bh = __getblk(device->bdev, bytenr / 4096,
 				      BTRFS_SUPER_INFO_SIZE);
 			if (!bh) {
-				printk(KERN_ERR "BTRFS: couldn't get super "
-				       "buffer head for bytenr %Lu\n", bytenr);
+				btrfs_err(device->dev_root->fs_info,
+				    "couldn't get super buffer head for bytenr %llu",
+				    bytenr);
 				errors++;
 				continue;
 			}
@@ -3449,22 +3478,31 @@ static int barrier_all_devices(struct btrfs_fs_info *info)
 
 int btrfs_get_num_tolerated_disk_barrier_failures(u64 flags)
 {
-	if ((flags & (BTRFS_BLOCK_GROUP_DUP |
-		      BTRFS_BLOCK_GROUP_RAID0 |
-		      BTRFS_AVAIL_ALLOC_BIT_SINGLE)) ||
-	    ((flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0))
-		return 0;
+	int raid_type;
+	int min_tolerated = INT_MAX;
 
-	if (flags & (BTRFS_BLOCK_GROUP_RAID1 |
-		     BTRFS_BLOCK_GROUP_RAID5 |
-		     BTRFS_BLOCK_GROUP_RAID10))
-		return 1;
+	if ((flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0 ||
+	    (flags & BTRFS_AVAIL_ALLOC_BIT_SINGLE))
+		min_tolerated = min(min_tolerated,
+				    btrfs_raid_array[BTRFS_RAID_SINGLE].
+				    tolerated_failures);
 
-	if (flags & BTRFS_BLOCK_GROUP_RAID6)
-		return 2;
+	for (raid_type = 0; raid_type < BTRFS_NR_RAID_TYPES; raid_type++) {
+		if (raid_type == BTRFS_RAID_SINGLE)
+			continue;
+		if (!(flags & btrfs_raid_group[raid_type]))
+			continue;
+		min_tolerated = min(min_tolerated,
+				    btrfs_raid_array[raid_type].
+				    tolerated_failures);
+	}
 
-	pr_warn("BTRFS: unknown raid type: %llu\n", flags);
-	return 0;
+	if (min_tolerated == INT_MAX) {
+		pr_warn("BTRFS: unknown raid flag: %llu\n", flags);
+		min_tolerated = 0;
+	}
+
+	return min_tolerated;
 }
 
 int btrfs_calc_num_tolerated_disk_barrier_failures(
@@ -3548,7 +3586,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 		if (ret) {
 			mutex_unlock(
 				&root->fs_info->fs_devices->device_list_mutex);
-			btrfs_error(root->fs_info, ret,
+			btrfs_std_error(root->fs_info, ret,
 				    "errors while submitting device barriers.");
 			return ret;
 		}
@@ -3588,7 +3626,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 		mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 
 		/* FUA is masked off if unsupported and can't be the reason */
-		btrfs_error(root->fs_info, -EIO,
+		btrfs_std_error(root->fs_info, -EIO,
 			    "%d errors while writing supers", total_errors);
 		return -EIO;
 	}
@@ -3606,7 +3644,7 @@ static int write_all_supers(struct btrfs_root *root, int max_mirrors)
 	}
 	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
 	if (total_errors > max_errors) {
-		btrfs_error(root->fs_info, -EIO,
+		btrfs_std_error(root->fs_info, -EIO,
 			    "%d errors while writing supers", total_errors);
 		return -EIO;
 	}
@@ -3792,7 +3830,7 @@ void close_ctree(struct btrfs_root *root)
 		       percpu_counter_sum(&fs_info->delalloc_bytes));
 	}
 
-	btrfs_sysfs_remove_one(fs_info);
+	btrfs_sysfs_remove_mounted(fs_info);
 	btrfs_sysfs_remove_fsid(fs_info->fs_devices);
 
 	btrfs_free_fs_roots(fs_info);
@@ -4290,25 +4328,6 @@ again:
 	return 0;
 }
 
-static void btrfs_free_pending_ordered(struct btrfs_transaction *cur_trans,
-				       struct btrfs_fs_info *fs_info)
-{
-	struct btrfs_ordered_extent *ordered;
-
-	spin_lock(&fs_info->trans_lock);
-	while (!list_empty(&cur_trans->pending_ordered)) {
-		ordered = list_first_entry(&cur_trans->pending_ordered,
-					   struct btrfs_ordered_extent,
-					   trans_list);
-		list_del_init(&ordered->trans_list);
-		spin_unlock(&fs_info->trans_lock);
-
-		btrfs_put_ordered_extent(ordered);
-		spin_lock(&fs_info->trans_lock);
-	}
-	spin_unlock(&fs_info->trans_lock);
-}
-
 void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 				   struct btrfs_root *root)
 {
@@ -4320,7 +4339,6 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 	cur_trans->state = TRANS_STATE_UNBLOCKED;
 	wake_up(&root->fs_info->transaction_wait);
 
-	btrfs_free_pending_ordered(cur_trans, root->fs_info);
 	btrfs_destroy_delayed_inodes(root);
 	btrfs_assert_delayed_root_empty(root);
 

fs/btrfs/disk-io.h

@@ -60,6 +60,8 @@ void close_ctree(struct btrfs_root *root);
 int write_ctree_super(struct btrfs_trans_handle *trans,
 		      struct btrfs_root *root, int max_mirrors);
 struct buffer_head *btrfs_read_dev_super(struct block_device *bdev);
+int btrfs_read_dev_one_super(struct block_device *bdev, int copy_num,
+			struct buffer_head **bh_ret);
 int btrfs_commit_super(struct btrfs_root *root);
 struct extent_buffer *btrfs_find_tree_block(struct btrfs_fs_info *fs_info,
 					    u64 bytenr);

fs/btrfs/extent-tree.c

@@ -95,8 +95,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root,
 				     u64 parent, u64 root_objectid,
 				     u64 flags, struct btrfs_disk_key *key,
-				     int level, struct btrfs_key *ins,
-				     int no_quota);
+				     int level, struct btrfs_key *ins);
 static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 			  struct btrfs_root *extent_root, u64 flags,
 			  int force);
@@ -332,6 +331,27 @@ static void put_caching_control(struct btrfs_caching_control *ctl)
 		kfree(ctl);
 }
 
+#ifdef CONFIG_BTRFS_DEBUG
+static void fragment_free_space(struct btrfs_root *root,
+				struct btrfs_block_group_cache *block_group)
+{
+	u64 start = block_group->key.objectid;
+	u64 len = block_group->key.offset;
+	u64 chunk = block_group->flags & BTRFS_BLOCK_GROUP_METADATA ?
+		root->nodesize : root->sectorsize;
+	u64 step = chunk << 1;
+
+	while (len > chunk) {
+		btrfs_remove_free_space(block_group, start, chunk);
+		start += step;
+		if (len < step)
+			len = 0;
+		else
+			len -= step;
+	}
+}
+#endif
+
 /*
  * this is only called by cache_block_group, since we could have freed extents
  * we need to check the pinned_extents for any extents that can't be used yet
@@ -388,6 +408,7 @@ static noinline void caching_thread(struct btrfs_work *work)
 	u64 last = 0;
 	u32 nritems;
 	int ret = -ENOMEM;
+	bool wakeup = true;
 
 	caching_ctl = container_of(work, struct btrfs_caching_control, work);
 	block_group = caching_ctl->block_group;
@@ -400,6 +421,15 @@ static noinline void caching_thread(struct btrfs_work *work)
 
 	last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET);
 
+#ifdef CONFIG_BTRFS_DEBUG
+	/*
+	 * If we're fragmenting we don't want to make anybody think we can
+	 * allocate from this block group until we've had a chance to fragment
+	 * the free space.
+	 */
+	if (btrfs_should_fragment_free_space(extent_root, block_group))
+		wakeup = false;
+#endif
 	/*
 	 * We don't want to deadlock with somebody trying to allocate a new
 	 * extent for the extent root while also trying to search the extent
@@ -441,7 +471,8 @@ next:
 
 			if (need_resched() ||
 			    rwsem_is_contended(&fs_info->commit_root_sem)) {
-				caching_ctl->progress = last;
+				if (wakeup)
+					caching_ctl->progress = last;
 				btrfs_release_path(path);
 				up_read(&fs_info->commit_root_sem);
 				mutex_unlock(&caching_ctl->mutex);
@@ -464,7 +495,8 @@ next:
 			key.offset = 0;
 			key.type = BTRFS_EXTENT_ITEM_KEY;
 
-			caching_ctl->progress = last;
+			if (wakeup)
+				caching_ctl->progress = last;
 			btrfs_release_path(path);
 			goto next;
 		}
@@ -491,7 +523,8 @@ next:
 
 			if (total_found > (1024 * 1024 * 2)) {
 				total_found = 0;
-				wake_up(&caching_ctl->wait);
+				if (wakeup)
+					wake_up(&caching_ctl->wait);
 			}
 		}
 		path->slots[0]++;
@@ -501,13 +534,27 @@ next:
 	total_found += add_new_free_space(block_group, fs_info, last,
 					  block_group->key.objectid +
 					  block_group->key.offset);
-	caching_ctl->progress = (u64)-1;
-
 	spin_lock(&block_group->lock);
 	block_group->caching_ctl = NULL;
 	block_group->cached = BTRFS_CACHE_FINISHED;
 	spin_unlock(&block_group->lock);
 
+#ifdef CONFIG_BTRFS_DEBUG
+	if (btrfs_should_fragment_free_space(extent_root, block_group)) {
+		u64 bytes_used;
+
+		spin_lock(&block_group->space_info->lock);
+		spin_lock(&block_group->lock);
+		bytes_used = block_group->key.offset -
+			btrfs_block_group_used(&block_group->item);
+		block_group->space_info->bytes_used += bytes_used >> 1;
+		spin_unlock(&block_group->lock);
+		spin_unlock(&block_group->space_info->lock);
+		fragment_free_space(extent_root, block_group);
+	}
+#endif
+
+	caching_ctl->progress = (u64)-1;
 err:
 	btrfs_free_path(path);
 	up_read(&fs_info->commit_root_sem);
@@ -607,6 +654,22 @@ static int cache_block_group(struct btrfs_block_group_cache *cache,
 			}
 		}
 		spin_unlock(&cache->lock);
+#ifdef CONFIG_BTRFS_DEBUG
+		if (ret == 1 &&
+		    btrfs_should_fragment_free_space(fs_info->extent_root,
+						     cache)) {
+			u64 bytes_used;
+
+			spin_lock(&cache->space_info->lock);
+			spin_lock(&cache->lock);
+			bytes_used = cache->key.offset -
+				btrfs_block_group_used(&cache->item);
+			cache->space_info->bytes_used += bytes_used >> 1;
+			spin_unlock(&cache->lock);
+			spin_unlock(&cache->space_info->lock);
+			fragment_free_space(fs_info->extent_root, cache);
+		}
+#endif
 		mutex_unlock(&caching_ctl->mutex);
 
 		wake_up(&caching_ctl->wait);
@@ -2009,8 +2072,7 @@ int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr,
 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 			 struct btrfs_root *root,
 			 u64 bytenr, u64 num_bytes, u64 parent,
-			 u64 root_objectid, u64 owner, u64 offset,
-			 int no_quota)
+			 u64 root_objectid, u64 owner, u64 offset)
 {
 	int ret;
 	struct btrfs_fs_info *fs_info = root->fs_info;
@@ -2022,12 +2084,12 @@ int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 		ret = btrfs_add_delayed_tree_ref(fs_info, trans, bytenr,
 					num_bytes,
 					parent, root_objectid, (int)owner,
-					BTRFS_ADD_DELAYED_REF, NULL, no_quota);
+					BTRFS_ADD_DELAYED_REF, NULL);
 	} else {
 		ret = btrfs_add_delayed_data_ref(fs_info, trans, bytenr,
-					num_bytes,
-					parent, root_objectid, owner, offset,
-					BTRFS_ADD_DELAYED_REF, NULL, no_quota);
+					num_bytes, parent, root_objectid,
+					owner, offset, 0,
+					BTRFS_ADD_DELAYED_REF, NULL);
 	}
 	return ret;
 }
@@ -2048,15 +2110,11 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 	u64 num_bytes = node->num_bytes;
 	u64 refs;
 	int ret;
-	int no_quota = node->no_quota;
 
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
 
-	if (!is_fstree(root_objectid) || !root->fs_info->quota_enabled)
-		no_quota = 1;
-
 	path->reada = 1;
 	path->leave_spinning = 1;
 	/* this will setup the path even if it fails to insert the back ref */
@@ -2291,8 +2349,7 @@ static int run_delayed_tree_ref(struct btrfs_trans_handle *trans,
 						parent, ref_root,
 						extent_op->flags_to_set,
 						&extent_op->key,
-						ref->level, &ins,
-						node->no_quota);
+						ref->level, &ins);
 	} else if (node->action == BTRFS_ADD_DELAYED_REF) {
 		ret = __btrfs_inc_extent_ref(trans, root, node,
 					     parent, ref_root,
@@ -2345,6 +2402,11 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
 						      node->num_bytes);
 			}
 		}
+
+		/* Also free its reserved qgroup space */
+		btrfs_qgroup_free_delayed_ref(root->fs_info,
+					      head->qgroup_ref_root,
+					      head->qgroup_reserved);
 		return ret;
 	}
 
@@ -2433,7 +2495,21 @@ static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 			}
 		}
 
+		/*
+		 * We need to try and merge add/drops of the same ref since we
+		 * can run into issues with relocate dropping the implicit ref
+		 * and then it being added back again before the drop can
+		 * finish.  If we merged anything we need to re-loop so we can
+		 * get a good ref.
+		 * Or we can get node references of the same type that weren't
+		 * merged when created due to bumps in the tree mod seq, and
+		 * we need to merge them to prevent adding an inline extent
+		 * backref before dropping it (triggering a BUG_ON at
+		 * insert_inline_extent_backref()).
+		 */
 		spin_lock(&locked_ref->lock);
+		btrfs_merge_delayed_refs(trans, fs_info, delayed_refs,
+					 locked_ref);
 
 		/*
 		 * locked_ref is the head node, so we have to go one
@@ -3109,7 +3185,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
 	int level;
 	int ret = 0;
 	int (*process_func)(struct btrfs_trans_handle *, struct btrfs_root *,
-			    u64, u64, u64, u64, u64, u64, int);
+			    u64, u64, u64, u64, u64, u64);
 
 
 	if (btrfs_test_is_dummy_root(root))
@@ -3150,15 +3226,14 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
 			key.offset -= btrfs_file_extent_offset(buf, fi);
 			ret = process_func(trans, root, bytenr, num_bytes,
 					   parent, ref_root, key.objectid,
-					   key.offset, 1);
+					   key.offset);
 			if (ret)
 				goto fail;
 		} else {
 			bytenr = btrfs_node_blockptr(buf, i);
 			num_bytes = root->nodesize;
 			ret = process_func(trans, root, bytenr, num_bytes,
-					   parent, ref_root, level - 1, 0,
-					   1);
+					   parent, ref_root, level - 1, 0);
 			if (ret)
 				goto fail;
 		}
@@ -3338,6 +3413,15 @@ again:
 	}
 	spin_unlock(&block_group->lock);
 
+	/*
+	 * We hit an ENOSPC when setting up the cache in this transaction, just
+	 * skip doing the setup, we've already cleared the cache so we're safe.
+	 */
+	if (test_bit(BTRFS_TRANS_CACHE_ENOSPC, &trans->transaction->flags)) {
+		ret = -ENOSPC;
+		goto out_put;
+	}
+
 	/*
 	 * Try to preallocate enough space based on how big the block group is.
 	 * Keep in mind this has to include any pinned space which could end up
@@ -3351,16 +3435,26 @@ again:
 	num_pages *= 16;
 	num_pages *= PAGE_CACHE_SIZE;
 
-	ret = btrfs_check_data_free_space(inode, num_pages, num_pages);
+	ret = btrfs_check_data_free_space(inode, 0, num_pages);
 	if (ret)
 		goto out_put;
 
 	ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, num_pages,
 					      num_pages, num_pages,
 					      &alloc_hint);
+	/*
+	 * Our cache requires contiguous chunks so that we don't modify a bunch
+	 * of metadata or split extents when writing the cache out, which means
+	 * we can enospc if we are heavily fragmented in addition to just normal
+	 * out of space conditions.  So if we hit this just skip setting up any
+	 * other block groups for this transaction, maybe we'll unpin enough
+	 * space the next time around.
+	 */
 	if (!ret)
 		dcs = BTRFS_DC_SETUP;
-	btrfs_free_reserved_data_space(inode, num_pages);
+	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);
@@ -3746,6 +3840,7 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags,
 	found->bytes_readonly = 0;
 	found->bytes_may_use = 0;
 	found->full = 0;
+	found->max_extent_size = 0;
 	found->force_alloc = CHUNK_ALLOC_NO_FORCE;
 	found->chunk_alloc = 0;
 	found->flush = 0;
@@ -3822,7 +3917,8 @@ static u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags)
 {
 	u64 num_devices = root->fs_info->fs_devices->rw_devices;
 	u64 target;
-	u64 tmp;
+	u64 raid_type;
+	u64 allowed = 0;
 
 	/*
 	 * see if restripe for this chunk_type is in progress, if so
@@ -3840,31 +3936,26 @@ static u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags)
 	spin_unlock(&root->fs_info->balance_lock);
 
 	/* First, mask out the RAID levels which aren't possible */
-	if (num_devices == 1)
-		flags &= ~(BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID0 |
-			   BTRFS_BLOCK_GROUP_RAID5);
-	if (num_devices < 3)
-		flags &= ~BTRFS_BLOCK_GROUP_RAID6;
-	if (num_devices < 4)
-		flags &= ~BTRFS_BLOCK_GROUP_RAID10;
-
-	tmp = flags & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID0 |
-		       BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID5 |
-		       BTRFS_BLOCK_GROUP_RAID6 | BTRFS_BLOCK_GROUP_RAID10);
-	flags &= ~tmp;
-
-	if (tmp & BTRFS_BLOCK_GROUP_RAID6)
-		tmp = BTRFS_BLOCK_GROUP_RAID6;
-	else if (tmp & BTRFS_BLOCK_GROUP_RAID5)
-		tmp = BTRFS_BLOCK_GROUP_RAID5;
-	else if (tmp & BTRFS_BLOCK_GROUP_RAID10)
-		tmp = BTRFS_BLOCK_GROUP_RAID10;
-	else if (tmp & BTRFS_BLOCK_GROUP_RAID1)
-		tmp = BTRFS_BLOCK_GROUP_RAID1;
-	else if (tmp & BTRFS_BLOCK_GROUP_RAID0)
-		tmp = BTRFS_BLOCK_GROUP_RAID0;
-
-	return extended_to_chunk(flags | tmp);
+	for (raid_type = 0; raid_type < BTRFS_NR_RAID_TYPES; raid_type++) {
+		if (num_devices >= btrfs_raid_array[raid_type].devs_min)
+			allowed |= btrfs_raid_group[raid_type];
+	}
+	allowed &= flags;
+
+	if (allowed & BTRFS_BLOCK_GROUP_RAID6)
+		allowed = BTRFS_BLOCK_GROUP_RAID6;
+	else if (allowed & BTRFS_BLOCK_GROUP_RAID5)
+		allowed = BTRFS_BLOCK_GROUP_RAID5;
+	else if (allowed & BTRFS_BLOCK_GROUP_RAID10)
+		allowed = BTRFS_BLOCK_GROUP_RAID10;
+	else if (allowed & BTRFS_BLOCK_GROUP_RAID1)
+		allowed = BTRFS_BLOCK_GROUP_RAID1;
+	else if (allowed & BTRFS_BLOCK_GROUP_RAID0)
+		allowed = BTRFS_BLOCK_GROUP_RAID0;
+
+	flags &= ~BTRFS_BLOCK_GROUP_PROFILE_MASK;
+
+	return extended_to_chunk(flags | allowed);
 }
 
 static u64 get_alloc_profile(struct btrfs_root *root, u64 orig_flags)
@@ -3903,11 +3994,7 @@ u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data)
 	return ret;
 }
 
-/*
- * This will check the space that the inode allocates from to make sure we have
- * enough space for bytes.
- */
-int btrfs_check_data_free_space(struct inode *inode, u64 bytes, u64 write_bytes)
+int btrfs_alloc_data_chunk_ondemand(struct inode *inode, u64 bytes)
 {
 	struct btrfs_space_info *data_sinfo;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
@@ -4006,7 +4093,8 @@ commit_trans:
 			if (IS_ERR(trans))
 				return PTR_ERR(trans);
 			if (have_pinned_space >= 0 ||
-			    trans->transaction->have_free_bgs ||
+			    test_bit(BTRFS_TRANS_HAVE_FREE_BGS,
+				     &trans->transaction->flags) ||
 			    need_commit > 0) {
 				ret = btrfs_commit_transaction(trans, root);
 				if (ret)
@@ -4028,38 +4116,86 @@ commit_trans:
 					      data_sinfo->flags, bytes, 1);
 		return -ENOSPC;
 	}
-	ret = btrfs_qgroup_reserve(root, write_bytes);
-	if (ret)
-		goto out;
 	data_sinfo->bytes_may_use += bytes;
 	trace_btrfs_space_reservation(root->fs_info, "space_info",
 				      data_sinfo->flags, bytes, 1);
-out:
 	spin_unlock(&data_sinfo->lock);
 
 	return ret;
 }
 
 /*
- * Called if we need to clear a data reservation for this inode.
+ * New check_data_free_space() with ability for precious data reservation
+ * Will replace old btrfs_check_data_free_space(), but for patch split,
+ * add a new function first and then replace it.
+ */
+int btrfs_check_data_free_space(struct inode *inode, u64 start, u64 len)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	int ret;
+
+	/* align the range */
+	len = round_up(start + len, root->sectorsize) -
+	      round_down(start, root->sectorsize);
+	start = round_down(start, root->sectorsize);
+
+	ret = btrfs_alloc_data_chunk_ondemand(inode, len);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Use new btrfs_qgroup_reserve_data to reserve precious data space
+	 *
+	 * TODO: Find a good method to avoid reserve data space for NOCOW
+	 * range, but don't impact performance on quota disable case.
+	 */
+	ret = btrfs_qgroup_reserve_data(inode, start, len);
+	return ret;
+}
+
+/*
+ * Called if we need to clear a data reservation for this inode
+ * Normally in a error case.
+ *
+ * This one will *NOT* use accurate qgroup reserved space API, just for case
+ * which we can't sleep and is sure it won't affect qgroup reserved space.
+ * Like clear_bit_hook().
  */
-void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes)
+void btrfs_free_reserved_data_space_noquota(struct inode *inode, u64 start,
+					    u64 len)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_space_info *data_sinfo;
 
-	/* make sure bytes are sectorsize aligned */
-	bytes = ALIGN(bytes, root->sectorsize);
+	/* Make sure the range is aligned to sectorsize */
+	len = round_up(start + len, root->sectorsize) -
+	      round_down(start, root->sectorsize);
+	start = round_down(start, root->sectorsize);
 
 	data_sinfo = root->fs_info->data_sinfo;
 	spin_lock(&data_sinfo->lock);
-	WARN_ON(data_sinfo->bytes_may_use < bytes);
-	data_sinfo->bytes_may_use -= bytes;
+	if (WARN_ON(data_sinfo->bytes_may_use < len))
+		data_sinfo->bytes_may_use = 0;
+	else
+		data_sinfo->bytes_may_use -= len;
 	trace_btrfs_space_reservation(root->fs_info, "space_info",
-				      data_sinfo->flags, bytes, 0);
+				      data_sinfo->flags, len, 0);
 	spin_unlock(&data_sinfo->lock);
 }
 
+/*
+ * Called if we need to clear a data reservation for this inode
+ * Normally in a error case.
+ *
+ * This one will handle the per-indoe data rsv map for accurate reserved
+ * space framework.
+ */
+void btrfs_free_reserved_data_space(struct inode *inode, u64 start, u64 len)
+{
+	btrfs_free_reserved_data_space_noquota(inode, start, len);
+	btrfs_qgroup_free_data(inode, start, len);
+}
+
 static void force_metadata_allocation(struct btrfs_fs_info *info)
 {
 	struct list_head *head = &info->space_info;
@@ -4891,13 +5027,9 @@ static struct btrfs_block_rsv *get_block_rsv(
 {
 	struct btrfs_block_rsv *block_rsv = NULL;
 
-	if (test_bit(BTRFS_ROOT_REF_COWS, &root->state))
-		block_rsv = trans->block_rsv;
-
-	if (root == root->fs_info->csum_root && trans->adding_csums)
-		block_rsv = trans->block_rsv;
-
-	if (root == root->fs_info->uuid_root)
+	if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) ||
+	    (root == root->fs_info->csum_root && trans->adding_csums) ||
+	     (root == root->fs_info->uuid_root))
 		block_rsv = trans->block_rsv;
 
 	if (!block_rsv)
@@ -5340,7 +5472,7 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 	if (root->fs_info->quota_enabled) {
 		/* One for parent inode, two for dir entries */
 		num_bytes = 3 * root->nodesize;
-		ret = btrfs_qgroup_reserve(root, num_bytes);
+		ret = btrfs_qgroup_reserve_meta(root, num_bytes);
 		if (ret)
 			return ret;
 	} else {
@@ -5358,10 +5490,8 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root,
 	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);
-	}
+	if (ret && *qgroup_reserved)
+		btrfs_qgroup_free_meta(root, *qgroup_reserved);
 
 	return ret;
 }
@@ -5522,15 +5652,15 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
 	spin_unlock(&BTRFS_I(inode)->lock);
 
 	if (root->fs_info->quota_enabled) {
-		ret = btrfs_qgroup_reserve(root, nr_extents * root->nodesize);
+		ret = btrfs_qgroup_reserve_meta(root,
+				nr_extents * root->nodesize);
 		if (ret)
 			goto out_fail;
 	}
 
 	ret = reserve_metadata_bytes(root, block_rsv, to_reserve, flush);
 	if (unlikely(ret)) {
-		if (root->fs_info->quota_enabled)
-			btrfs_qgroup_free(root, nr_extents * root->nodesize);
+		btrfs_qgroup_free_meta(root, nr_extents * root->nodesize);
 		goto out_fail;
 	}
 
@@ -5653,41 +5783,48 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes)
 }
 
 /**
- * btrfs_delalloc_reserve_space - reserve data and metadata space for delalloc
+ * btrfs_delalloc_reserve_space - reserve data and metadata space for
+ * delalloc
  * @inode: inode we're writing to
- * @num_bytes: the number of bytes we want to allocate
+ * @start: start range we are writing to
+ * @len: how long the range we are writing to
+ *
+ * TODO: This function will finally replace old btrfs_delalloc_reserve_space()
  *
  * This will do the following things
  *
- * o reserve space in the data space info for num_bytes
- * o reserve space in the metadata space info based on number of outstanding
+ * o reserve space in data space info for num bytes
+ *   and reserve precious corresponding qgroup space
+ *   (Done in check_data_free_space)
+ *
+ * o reserve space for metadata space, based on the number of outstanding
  *   extents and how much csums will be needed
- * o add to the inodes ->delalloc_bytes
+ *   also reserve metadata space in a per root over-reserve method.
+ * o add to the inodes->delalloc_bytes
  * o add it to the fs_info's delalloc inodes list.
+ *   (Above 3 all done in delalloc_reserve_metadata)
  *
- * This will return 0 for success and -ENOSPC if there is no space left.
+ * Return 0 for success
+ * Return <0 for error(-ENOSPC or -EQUOT)
  */
-int btrfs_delalloc_reserve_space(struct inode *inode, u64 num_bytes)
+int btrfs_delalloc_reserve_space(struct inode *inode, u64 start, u64 len)
 {
 	int ret;
 
-	ret = btrfs_check_data_free_space(inode, num_bytes, num_bytes);
-	if (ret)
-		return ret;
-
-	ret = btrfs_delalloc_reserve_metadata(inode, num_bytes);
-	if (ret) {
-		btrfs_free_reserved_data_space(inode, num_bytes);
+	ret = btrfs_check_data_free_space(inode, start, len);
+	if (ret < 0)
 		return ret;
-	}
-
-	return 0;
+	ret = btrfs_delalloc_reserve_metadata(inode, len);
+	if (ret < 0)
+		btrfs_free_reserved_data_space(inode, start, len);
+	return ret;
 }
 
 /**
  * btrfs_delalloc_release_space - release data and metadata space for delalloc
  * @inode: inode we're releasing space for
- * @num_bytes: the number of bytes we want to free up
+ * @start: start position of the space already reserved
+ * @len: the len of the space already reserved
  *
  * This must be matched with a call to btrfs_delalloc_reserve_space.  This is
  * called in the case that we don't need the metadata AND data reservations
@@ -5696,11 +5833,12 @@ int btrfs_delalloc_reserve_space(struct inode *inode, u64 num_bytes)
  * This function will release the metadata space that was not used and will
  * decrement ->delalloc_bytes and remove it from the fs_info delalloc_inodes
  * list if there are no delalloc bytes left.
+ * Also it will handle the qgroup reserved space.
  */
-void btrfs_delalloc_release_space(struct inode *inode, u64 num_bytes)
+void btrfs_delalloc_release_space(struct inode *inode, u64 start, u64 len)
 {
-	btrfs_delalloc_release_metadata(inode, num_bytes);
-	btrfs_free_reserved_data_space(inode, num_bytes);
+	btrfs_delalloc_release_metadata(inode, len);
+	btrfs_free_reserved_data_space(inode, start, len);
 }
 
 static int update_block_group(struct btrfs_trans_handle *trans,
@@ -6065,6 +6203,34 @@ void btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans,
 	update_global_block_rsv(fs_info);
 }
 
+/*
+ * Returns the free cluster for the given space info and sets empty_cluster to
+ * what it should be based on the mount options.
+ */
+static struct btrfs_free_cluster *
+fetch_cluster_info(struct btrfs_root *root, struct btrfs_space_info *space_info,
+		   u64 *empty_cluster)
+{
+	struct btrfs_free_cluster *ret = NULL;
+	bool ssd = btrfs_test_opt(root, SSD);
+
+	*empty_cluster = 0;
+	if (btrfs_mixed_space_info(space_info))
+		return ret;
+
+	if (ssd)
+		*empty_cluster = 2 * 1024 * 1024;
+	if (space_info->flags & BTRFS_BLOCK_GROUP_METADATA) {
+		ret = &root->fs_info->meta_alloc_cluster;
+		if (!ssd)
+			*empty_cluster = 64 * 1024;
+	} else if ((space_info->flags & BTRFS_BLOCK_GROUP_DATA) && ssd) {
+		ret = &root->fs_info->data_alloc_cluster;
+	}
+
+	return ret;
+}
+
 static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end,
 			      const bool return_free_space)
 {
@@ -6072,7 +6238,10 @@ static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end,
 	struct btrfs_block_group_cache *cache = NULL;
 	struct btrfs_space_info *space_info;
 	struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
+	struct btrfs_free_cluster *cluster = NULL;
 	u64 len;
+	u64 total_unpinned = 0;
+	u64 empty_cluster = 0;
 	bool readonly;
 
 	while (start <= end) {
@@ -6081,8 +6250,14 @@ static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end,
 		    start >= cache->key.objectid + cache->key.offset) {
 			if (cache)
 				btrfs_put_block_group(cache);
+			total_unpinned = 0;
 			cache = btrfs_lookup_block_group(fs_info, start);
 			BUG_ON(!cache); /* Logic error */
+
+			cluster = fetch_cluster_info(root,
+						     cache->space_info,
+						     &empty_cluster);
+			empty_cluster <<= 1;
 		}
 
 		len = cache->key.objectid + cache->key.offset - start;
@@ -6095,12 +6270,27 @@ static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end,
 		}
 
 		start += len;
+		total_unpinned += len;
 		space_info = cache->space_info;
 
+		/*
+		 * If this space cluster has been marked as fragmented and we've
+		 * unpinned enough in this block group to potentially allow a
+		 * cluster to be created inside of it go ahead and clear the
+		 * fragmented check.
+		 */
+		if (cluster && cluster->fragmented &&
+		    total_unpinned > empty_cluster) {
+			spin_lock(&cluster->lock);
+			cluster->fragmented = 0;
+			spin_unlock(&cluster->lock);
+		}
+
 		spin_lock(&space_info->lock);
 		spin_lock(&cache->lock);
 		cache->pinned -= len;
 		space_info->bytes_pinned -= len;
+		space_info->max_extent_size = 0;
 		percpu_counter_add(&space_info->total_bytes_pinned, -len);
 		if (cache->ro) {
 			space_info->bytes_readonly += len;
@@ -6233,7 +6423,6 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 	int extent_slot = 0;
 	int found_extent = 0;
 	int num_to_del = 1;
-	int no_quota = node->no_quota;
 	u32 item_size;
 	u64 refs;
 	u64 bytenr = node->bytenr;
@@ -6242,9 +6431,6 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 	bool skinny_metadata = btrfs_fs_incompat(root->fs_info,
 						 SKINNY_METADATA);
 
-	if (!info->quota_enabled || !is_fstree(root_objectid))
-		no_quota = 1;
-
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -6570,7 +6756,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 					buf->start, buf->len,
 					parent, root->root_key.objectid,
 					btrfs_header_level(buf),
-					BTRFS_DROP_DELAYED_REF, NULL, 0);
+					BTRFS_DROP_DELAYED_REF, NULL);
 		BUG_ON(ret); /* -ENOMEM */
 	}
 
@@ -6618,7 +6804,7 @@ out:
 /* Can return -ENOMEM */
 int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		      u64 bytenr, u64 num_bytes, u64 parent, u64 root_objectid,
-		      u64 owner, u64 offset, int no_quota)
+		      u64 owner, u64 offset)
 {
 	int ret;
 	struct btrfs_fs_info *fs_info = root->fs_info;
@@ -6641,13 +6827,13 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		ret = btrfs_add_delayed_tree_ref(fs_info, trans, bytenr,
 					num_bytes,
 					parent, root_objectid, (int)owner,
-					BTRFS_DROP_DELAYED_REF, NULL, no_quota);
+					BTRFS_DROP_DELAYED_REF, NULL);
 	} else {
 		ret = btrfs_add_delayed_data_ref(fs_info, trans, bytenr,
 						num_bytes,
 						parent, root_objectid, owner,
-						offset, BTRFS_DROP_DELAYED_REF,
-						NULL, no_quota);
+						offset, 0,
+						BTRFS_DROP_DELAYED_REF, NULL);
 	}
 	return ret;
 }
@@ -6833,7 +7019,7 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	struct btrfs_block_group_cache *block_group = NULL;
 	u64 search_start = 0;
 	u64 max_extent_size = 0;
-	int empty_cluster = 2 * 1024 * 1024;
+	u64 empty_cluster = 0;
 	struct btrfs_space_info *space_info;
 	int loop = 0;
 	int index = __get_raid_index(flags);
@@ -6843,6 +7029,8 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	bool failed_alloc = false;
 	bool use_cluster = true;
 	bool have_caching_bg = false;
+	bool orig_have_caching_bg = false;
+	bool full_search = false;
 
 	WARN_ON(num_bytes < root->sectorsize);
 	ins->type = BTRFS_EXTENT_ITEM_KEY;
@@ -6858,36 +7046,47 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	}
 
 	/*
-	 * If the space info is for both data and metadata it means we have a
-	 * small filesystem and we can't use the clustering stuff.
+	 * If our free space is heavily fragmented we may not be able to make
+	 * big contiguous allocations, so instead of doing the expensive search
+	 * for free space, simply return ENOSPC with our max_extent_size so we
+	 * can go ahead and search for a more manageable chunk.
+	 *
+	 * If our max_extent_size is large enough for our allocation simply
+	 * disable clustering since we will likely not be able to find enough
+	 * space to create a cluster and induce latency trying.
 	 */
-	if (btrfs_mixed_space_info(space_info))
-		use_cluster = false;
-
-	if (flags & BTRFS_BLOCK_GROUP_METADATA && use_cluster) {
-		last_ptr = &root->fs_info->meta_alloc_cluster;
-		if (!btrfs_test_opt(root, SSD))
-			empty_cluster = 64 * 1024;
-	}
-
-	if ((flags & BTRFS_BLOCK_GROUP_DATA) && use_cluster &&
-	    btrfs_test_opt(root, SSD)) {
-		last_ptr = &root->fs_info->data_alloc_cluster;
+	if (unlikely(space_info->max_extent_size)) {
+		spin_lock(&space_info->lock);
+		if (space_info->max_extent_size &&
+		    num_bytes > space_info->max_extent_size) {
+			ins->offset = space_info->max_extent_size;
+			spin_unlock(&space_info->lock);
+			return -ENOSPC;
+		} else if (space_info->max_extent_size) {
+			use_cluster = false;
+		}
+		spin_unlock(&space_info->lock);
 	}
 
+	last_ptr = fetch_cluster_info(orig_root, space_info, &empty_cluster);
 	if (last_ptr) {
 		spin_lock(&last_ptr->lock);
 		if (last_ptr->block_group)
 			hint_byte = last_ptr->window_start;
+		if (last_ptr->fragmented) {
+			/*
+			 * We still set window_start so we can keep track of the
+			 * last place we found an allocation to try and save
+			 * some time.
+			 */
+			hint_byte = last_ptr->window_start;
+			use_cluster = false;
+		}
 		spin_unlock(&last_ptr->lock);
 	}
 
 	search_start = max(search_start, first_logical_byte(root, 0));
 	search_start = max(search_start, hint_byte);
-
-	if (!last_ptr)
-		empty_cluster = 0;
-
 	if (search_start == hint_byte) {
 		block_group = btrfs_lookup_block_group(root->fs_info,
 						       search_start);
@@ -6922,6 +7121,8 @@ static noinline int find_free_extent(struct btrfs_root *orig_root,
 	}
 search:
 	have_caching_bg = false;
+	if (index == 0 || index == __get_raid_index(flags))
+		full_search = true;
 	down_read(&space_info->groups_sem);
 	list_for_each_entry(block_group, &space_info->block_groups[index],
 			    list) {
@@ -6955,6 +7156,7 @@ search:
 have_block_group:
 		cached = block_group_cache_done(block_group);
 		if (unlikely(!cached)) {
+			have_caching_bg = true;
 			ret = cache_block_group(block_group, 0);
 			BUG_ON(ret < 0);
 			ret = 0;
@@ -6969,7 +7171,7 @@ have_block_group:
 		 * Ok we want to try and use the cluster allocator, so
 		 * lets look there
 		 */
-		if (last_ptr) {
+		if (last_ptr && use_cluster) {
 			struct btrfs_block_group_cache *used_block_group;
 			unsigned long aligned_cluster;
 			/*
@@ -7095,6 +7297,16 @@ refill_cluster:
 		}
 
 unclustered_alloc:
+		/*
+		 * We are doing an unclustered alloc, set the fragmented flag so
+		 * we don't bother trying to setup a cluster again until we get
+		 * more space.
+		 */
+		if (unlikely(last_ptr)) {
+			spin_lock(&last_ptr->lock);
+			last_ptr->fragmented = 1;
+			spin_unlock(&last_ptr->lock);
+		}
 		spin_lock(&block_group->free_space_ctl->tree_lock);
 		if (cached &&
 		    block_group->free_space_ctl->free_space <
@@ -7127,8 +7339,6 @@ unclustered_alloc:
 			failed_alloc = true;
 			goto have_block_group;
 		} else if (!offset) {
-			if (!cached)
-				have_caching_bg = true;
 			goto loop;
 		}
 checks:
@@ -7169,6 +7379,10 @@ loop:
 	}
 	up_read(&space_info->groups_sem);
 
+	if ((loop == LOOP_CACHING_NOWAIT) && have_caching_bg
+		&& !orig_have_caching_bg)
+		orig_have_caching_bg = true;
+
 	if (!ins->objectid && loop >= LOOP_CACHING_WAIT && have_caching_bg)
 		goto search;
 
@@ -7185,7 +7399,20 @@ loop:
 	 */
 	if (!ins->objectid && loop < LOOP_NO_EMPTY_SIZE) {
 		index = 0;
-		loop++;
+		if (loop == LOOP_CACHING_NOWAIT) {
+			/*
+			 * We want to skip the LOOP_CACHING_WAIT step if we
+			 * don't have any unached bgs and we've alrelady done a
+			 * full search through.
+			 */
+			if (orig_have_caching_bg || !full_search)
+				loop = LOOP_CACHING_WAIT;
+			else
+				loop = LOOP_ALLOC_CHUNK;
+		} else {
+			loop++;
+		}
+
 		if (loop == LOOP_ALLOC_CHUNK) {
 			struct btrfs_trans_handle *trans;
 			int exist = 0;
@@ -7203,6 +7430,15 @@ loop:
 
 			ret = do_chunk_alloc(trans, root, flags,
 					     CHUNK_ALLOC_FORCE);
+
+			/*
+			 * If we can't allocate a new chunk we've already looped
+			 * through at least once, move on to the NO_EMPTY_SIZE
+			 * case.
+			 */
+			if (ret == -ENOSPC)
+				loop = LOOP_NO_EMPTY_SIZE;
+
 			/*
 			 * Do not bail out on ENOSPC since we
 			 * can do more things.
@@ -7219,6 +7455,15 @@ loop:
 		}
 
 		if (loop == LOOP_NO_EMPTY_SIZE) {
+			/*
+			 * Don't loop again if we already have no empty_size and
+			 * no empty_cluster.
+			 */
+			if (empty_size == 0 &&
+			    empty_cluster == 0) {
+				ret = -ENOSPC;
+				goto out;
+			}
 			empty_size = 0;
 			empty_cluster = 0;
 		}
@@ -7227,11 +7472,20 @@ loop:
 	} else if (!ins->objectid) {
 		ret = -ENOSPC;
 	} else if (ins->objectid) {
+		if (!use_cluster && last_ptr) {
+			spin_lock(&last_ptr->lock);
+			last_ptr->window_start = ins->objectid;
+			spin_unlock(&last_ptr->lock);
+		}
 		ret = 0;
 	}
 out:
-	if (ret == -ENOSPC)
+	if (ret == -ENOSPC) {
+		spin_lock(&space_info->lock);
+		space_info->max_extent_size = max_extent_size;
+		spin_unlock(&space_info->lock);
 		ins->offset = max_extent_size;
+	}
 	return ret;
 }
 
@@ -7280,7 +7534,7 @@ int btrfs_reserve_extent(struct btrfs_root *root,
 			 u64 empty_size, u64 hint_byte,
 			 struct btrfs_key *ins, int is_data, int delalloc)
 {
-	bool final_tried = false;
+	bool final_tried = num_bytes == min_alloc_size;
 	u64 flags;
 	int ret;
 
@@ -7429,8 +7683,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root,
 				     u64 parent, u64 root_objectid,
 				     u64 flags, struct btrfs_disk_key *key,
-				     int level, struct btrfs_key *ins,
-				     int no_quota)
+				     int level, struct btrfs_key *ins)
 {
 	int ret;
 	struct btrfs_fs_info *fs_info = root->fs_info;
@@ -7511,7 +7764,8 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans,
 int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root,
 				     u64 root_objectid, u64 owner,
-				     u64 offset, struct btrfs_key *ins)
+				     u64 offset, u64 ram_bytes,
+				     struct btrfs_key *ins)
 {
 	int ret;
 
@@ -7520,7 +7774,8 @@ int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 	ret = btrfs_add_delayed_data_ref(root->fs_info, trans, ins->objectid,
 					 ins->offset, 0,
 					 root_objectid, owner, offset,
-					 BTRFS_ADD_DELAYED_EXTENT, NULL, 0);
+					 ram_bytes, BTRFS_ADD_DELAYED_EXTENT,
+					 NULL);
 	return ret;
 }
 
@@ -7734,7 +7989,7 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 						 ins.objectid, ins.offset,
 						 parent, root_objectid, level,
 						 BTRFS_ADD_DELAYED_EXTENT,
-						 extent_op, 0);
+						 extent_op);
 		if (ret)
 			goto out_free_delayed;
 	}
@@ -8275,14 +8530,15 @@ skip:
 			ret = account_shared_subtree(trans, root, next,
 						     generation, level - 1);
 			if (ret) {
-				printk_ratelimited(KERN_ERR "BTRFS: %s Error "
+				btrfs_err_rl(root->fs_info,
+					"Error "
 					"%d accounting shared subtree. Quota "
-					"is out of sync, rescan required.\n",
-					root->fs_info->sb->s_id, ret);
+					"is out of sync, rescan required.",
+					ret);
 			}
 		}
 		ret = btrfs_free_extent(trans, root, bytenr, blocksize, parent,
-				root->root_key.objectid, level - 1, 0, 0);
+				root->root_key.objectid, level - 1, 0);
 		BUG_ON(ret); /* -ENOMEM */
 	}
 	btrfs_tree_unlock(next);
@@ -8367,10 +8623,11 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
 			BUG_ON(ret); /* -ENOMEM */
 			ret = account_leaf_items(trans, root, eb);
 			if (ret) {
-				printk_ratelimited(KERN_ERR "BTRFS: %s Error "
+				btrfs_err_rl(root->fs_info,
+					"error "
 					"%d accounting leaf items. Quota "
-					"is out of sync, rescan required.\n",
-					root->fs_info->sb->s_id, ret);
+					"is out of sync, rescan required.",
+					ret);
 			}
 		}
 		/* make block locked assertion in clean_tree_block happy */
@@ -8692,7 +8949,7 @@ out:
 	if (!for_reloc && root_dropped == false)
 		btrfs_add_dead_root(root);
 	if (err && err != -EAGAIN)
-		btrfs_std_error(root->fs_info, err);
+		btrfs_std_error(root->fs_info, err, NULL);
 	return err;
 }
 
@@ -8880,7 +9137,7 @@ again:
 	 * back off and let this transaction commit
 	 */
 	mutex_lock(&root->fs_info->ro_block_group_mutex);
-	if (trans->transaction->dirty_bg_run) {
+	if (test_bit(BTRFS_TRANS_DIRTY_BG_RUN, &trans->transaction->flags)) {
 		u64 transid = trans->transid;
 
 		mutex_unlock(&root->fs_info->ro_block_group_mutex);
@@ -9630,6 +9887,14 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans,
 
 	free_excluded_extents(root, cache);
 
+#ifdef CONFIG_BTRFS_DEBUG
+	if (btrfs_should_fragment_free_space(root, cache)) {
+		u64 new_bytes_used = size - bytes_used;
+
+		bytes_used += new_bytes_used >> 1;
+		fragment_free_space(root, cache);
+	}
+#endif
 	/*
 	 * Call to ensure the corresponding space_info object is created and
 	 * assigned to our block group, but don't update its counters just yet.
@@ -10370,8 +10635,7 @@ void btrfs_end_write_no_snapshoting(struct btrfs_root *root)
 {
 	percpu_counter_dec(&root->subv_writers->counter);
 	/*
-	 * Make sure counter is updated before we wake up
-	 * waiters.
+	 * Make sure counter is updated before we wake up waiters.
 	 */
 	smp_mb();
 	if (waitqueue_active(&root->subv_writers->wait))

fs/btrfs/extent_io.c

@@ -96,8 +96,8 @@ static inline void __btrfs_debug_check_extent_io_range(const char *caller,
 	inode = tree->mapping->host;
 	isize = i_size_read(inode);
 	if (end >= PAGE_SIZE && (end % 2) == 0 && end != isize - 1) {
-		printk_ratelimited(KERN_DEBUG
-		    "BTRFS: %s: ino %llu isize %llu odd range [%llu,%llu]\n",
+		btrfs_debug_rl(BTRFS_I(inode)->root->fs_info,
+		    "%s: ino %llu isize %llu odd range [%llu,%llu]",
 				caller, btrfs_ino(inode), isize, start, end);
 	}
 }
@@ -131,6 +131,25 @@ struct extent_page_data {
 	unsigned int sync_io:1;
 };
 
+static void add_extent_changeset(struct extent_state *state, unsigned bits,
+				 struct extent_changeset *changeset,
+				 int set)
+{
+	int ret;
+
+	if (!changeset)
+		return;
+	if (set && (state->state & bits) == bits)
+		return;
+	if (!set && (state->state & bits) == 0)
+		return;
+	changeset->bytes_changed += state->end - state->start + 1;
+	ret = ulist_add(changeset->range_changed, state->start, state->end,
+			GFP_ATOMIC);
+	/* ENOMEM */
+	BUG_ON(ret < 0);
+}
+
 static noinline void flush_write_bio(void *data);
 static inline struct btrfs_fs_info *
 tree_fs_info(struct extent_io_tree *tree)
@@ -410,7 +429,8 @@ static void clear_state_cb(struct extent_io_tree *tree,
 }
 
 static void set_state_bits(struct extent_io_tree *tree,
-			   struct extent_state *state, unsigned *bits);
+			   struct extent_state *state, unsigned *bits,
+			   struct extent_changeset *changeset);
 
 /*
  * insert an extent_state struct into the tree.  'bits' are set on the
@@ -426,7 +446,7 @@ static int insert_state(struct extent_io_tree *tree,
 			struct extent_state *state, u64 start, u64 end,
 			struct rb_node ***p,
 			struct rb_node **parent,
-			unsigned *bits)
+			unsigned *bits, struct extent_changeset *changeset)
 {
 	struct rb_node *node;
 
@@ -436,7 +456,7 @@ static int insert_state(struct extent_io_tree *tree,
 	state->start = start;
 	state->end = end;
 
-	set_state_bits(tree, state, bits);
+	set_state_bits(tree, state, bits, changeset);
 
 	node = tree_insert(&tree->state, NULL, end, &state->rb_node, p, parent);
 	if (node) {
@@ -511,7 +531,8 @@ static struct extent_state *next_state(struct extent_state *state)
  */
 static struct extent_state *clear_state_bit(struct extent_io_tree *tree,
 					    struct extent_state *state,
-					    unsigned *bits, int wake)
+					    unsigned *bits, int wake,
+					    struct extent_changeset *changeset)
 {
 	struct extent_state *next;
 	unsigned bits_to_clear = *bits & ~EXTENT_CTLBITS;
@@ -522,6 +543,7 @@ static struct extent_state *clear_state_bit(struct extent_io_tree *tree,
 		tree->dirty_bytes -= range;
 	}
 	clear_state_cb(tree, state, bits);
+	add_extent_changeset(state, bits_to_clear, changeset, 0);
 	state->state &= ~bits_to_clear;
 	if (wake)
 		wake_up(&state->wq);
@@ -569,10 +591,10 @@ static void extent_io_tree_panic(struct extent_io_tree *tree, int err)
  *
  * This takes the tree lock, and returns 0 on success and < 0 on error.
  */
-int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		     unsigned bits, int wake, int delete,
-		     struct extent_state **cached_state,
-		     gfp_t mask)
+static int __clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
+			      unsigned bits, int wake, int delete,
+			      struct extent_state **cached_state,
+			      gfp_t mask, struct extent_changeset *changeset)
 {
 	struct extent_state *state;
 	struct extent_state *cached;
@@ -671,7 +693,8 @@ hit_next:
 		if (err)
 			goto out;
 		if (state->end <= end) {
-			state = clear_state_bit(tree, state, &bits, wake);
+			state = clear_state_bit(tree, state, &bits, wake,
+						changeset);
 			goto next;
 		}
 		goto search_again;
@@ -692,13 +715,13 @@ hit_next:
 		if (wake)
 			wake_up(&state->wq);
 
-		clear_state_bit(tree, prealloc, &bits, wake);
+		clear_state_bit(tree, prealloc, &bits, wake, changeset);
 
 		prealloc = NULL;
 		goto out;
 	}
 
-	state = clear_state_bit(tree, state, &bits, wake);
+	state = clear_state_bit(tree, state, &bits, wake, changeset);
 next:
 	if (last_end == (u64)-1)
 		goto out;
@@ -789,7 +812,7 @@ out:
 
 static void set_state_bits(struct extent_io_tree *tree,
 			   struct extent_state *state,
-			   unsigned *bits)
+			   unsigned *bits, struct extent_changeset *changeset)
 {
 	unsigned bits_to_set = *bits & ~EXTENT_CTLBITS;
 
@@ -798,6 +821,7 @@ static void set_state_bits(struct extent_io_tree *tree,
 		u64 range = state->end - state->start + 1;
 		tree->dirty_bytes += range;
 	}
+	add_extent_changeset(state, bits_to_set, changeset, 1);
 	state->state |= bits_to_set;
 }
 
@@ -835,7 +859,7 @@ static int __must_check
 __set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 		 unsigned bits, unsigned exclusive_bits,
 		 u64 *failed_start, struct extent_state **cached_state,
-		 gfp_t mask)
+		 gfp_t mask, struct extent_changeset *changeset)
 {
 	struct extent_state *state;
 	struct extent_state *prealloc = NULL;
@@ -873,7 +897,7 @@ again:
 		prealloc = alloc_extent_state_atomic(prealloc);
 		BUG_ON(!prealloc);
 		err = insert_state(tree, prealloc, start, end,
-				   &p, &parent, &bits);
+				   &p, &parent, &bits, changeset);
 		if (err)
 			extent_io_tree_panic(tree, err);
 
@@ -899,7 +923,7 @@ hit_next:
 			goto out;
 		}
 
-		set_state_bits(tree, state, &bits);
+		set_state_bits(tree, state, &bits, changeset);
 		cache_state(state, cached_state);
 		merge_state(tree, state);
 		if (last_end == (u64)-1)
@@ -945,7 +969,7 @@ hit_next:
 		if (err)
 			goto out;
 		if (state->end <= end) {
-			set_state_bits(tree, state, &bits);
+			set_state_bits(tree, state, &bits, changeset);
 			cache_state(state, cached_state);
 			merge_state(tree, state);
 			if (last_end == (u64)-1)
@@ -980,7 +1004,7 @@ hit_next:
 		 * the later extent.
 		 */
 		err = insert_state(tree, prealloc, start, this_end,
-				   NULL, NULL, &bits);
+				   NULL, NULL, &bits, changeset);
 		if (err)
 			extent_io_tree_panic(tree, err);
 
@@ -1008,7 +1032,7 @@ hit_next:
 		if (err)
 			extent_io_tree_panic(tree, err);
 
-		set_state_bits(tree, prealloc, &bits);
+		set_state_bits(tree, prealloc, &bits, changeset);
 		cache_state(prealloc, cached_state);
 		merge_state(tree, prealloc);
 		prealloc = NULL;
@@ -1038,7 +1062,7 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 		   struct extent_state **cached_state, gfp_t mask)
 {
 	return __set_extent_bit(tree, start, end, bits, 0, failed_start,
-				cached_state, mask);
+				cached_state, mask, NULL);
 }
 
 
@@ -1111,7 +1135,7 @@ again:
 			goto out;
 		}
 		err = insert_state(tree, prealloc, start, end,
-				   &p, &parent, &bits);
+				   &p, &parent, &bits, NULL);
 		if (err)
 			extent_io_tree_panic(tree, err);
 		cache_state(prealloc, cached_state);
@@ -1130,9 +1154,9 @@ hit_next:
 	 * Just lock what we found and keep going
 	 */
 	if (state->start == start && state->end <= end) {
-		set_state_bits(tree, state, &bits);
+		set_state_bits(tree, state, &bits, NULL);
 		cache_state(state, cached_state);
-		state = clear_state_bit(tree, state, &clear_bits, 0);
+		state = clear_state_bit(tree, state, &clear_bits, 0, NULL);
 		if (last_end == (u64)-1)
 			goto out;
 		start = last_end + 1;
@@ -1171,9 +1195,10 @@ hit_next:
 		if (err)
 			goto out;
 		if (state->end <= end) {
-			set_state_bits(tree, state, &bits);
+			set_state_bits(tree, state, &bits, NULL);
 			cache_state(state, cached_state);
-			state = clear_state_bit(tree, state, &clear_bits, 0);
+			state = clear_state_bit(tree, state, &clear_bits, 0,
+						NULL);
 			if (last_end == (u64)-1)
 				goto out;
 			start = last_end + 1;
@@ -1208,7 +1233,7 @@ hit_next:
 		 * the later extent.
 		 */
 		err = insert_state(tree, prealloc, start, this_end,
-				   NULL, NULL, &bits);
+				   NULL, NULL, &bits, NULL);
 		if (err)
 			extent_io_tree_panic(tree, err);
 		cache_state(prealloc, cached_state);
@@ -1233,9 +1258,9 @@ hit_next:
 		if (err)
 			extent_io_tree_panic(tree, err);
 
-		set_state_bits(tree, prealloc, &bits);
+		set_state_bits(tree, prealloc, &bits, NULL);
 		cache_state(prealloc, cached_state);
-		clear_state_bit(tree, prealloc, &clear_bits, 0);
+		clear_state_bit(tree, prealloc, &clear_bits, 0, NULL);
 		prealloc = NULL;
 		goto out;
 	}
@@ -1274,6 +1299,30 @@ int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 			      NULL, mask);
 }
 
+int set_record_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
+			   unsigned bits, gfp_t mask,
+			   struct extent_changeset *changeset)
+{
+	/*
+	 * We don't support EXTENT_LOCKED yet, as current changeset will
+	 * record any bits changed, so for EXTENT_LOCKED case, it will
+	 * either fail with -EEXIST or changeset will record the whole
+	 * range.
+	 */
+	BUG_ON(bits & EXTENT_LOCKED);
+
+	return __set_extent_bit(tree, start, end, bits, 0, NULL, NULL, mask,
+				changeset);
+}
+
+int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
+		     unsigned bits, int wake, int delete,
+		     struct extent_state **cached, gfp_t mask)
+{
+	return __clear_extent_bit(tree, start, end, bits, wake, delete,
+				  cached, mask, NULL);
+}
+
 int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 		      unsigned bits, gfp_t mask)
 {
@@ -1285,6 +1334,20 @@ int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 	return clear_extent_bit(tree, start, end, bits, wake, 0, NULL, mask);
 }
 
+int clear_record_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
+			     unsigned bits, gfp_t mask,
+			     struct extent_changeset *changeset)
+{
+	/*
+	 * Don't support EXTENT_LOCKED case, same reason as
+	 * set_record_extent_bits().
+	 */
+	BUG_ON(bits & EXTENT_LOCKED);
+
+	return __clear_extent_bit(tree, start, end, bits, 0, 0, NULL, mask,
+				  changeset);
+}
+
 int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end,
 			struct extent_state **cached_state, gfp_t mask)
 {
@@ -1343,7 +1406,7 @@ int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 	while (1) {
 		err = __set_extent_bit(tree, start, end, EXTENT_LOCKED | bits,
 				       EXTENT_LOCKED, &failed_start,
-				       cached_state, GFP_NOFS);
+				       cached_state, GFP_NOFS, NULL);
 		if (err == -EEXIST) {
 			wait_extent_bit(tree, failed_start, end, EXTENT_LOCKED);
 			start = failed_start;
@@ -1365,7 +1428,7 @@ int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end)
 	u64 failed_start;
 
 	err = __set_extent_bit(tree, start, end, EXTENT_LOCKED, EXTENT_LOCKED,
-			       &failed_start, NULL, GFP_NOFS);
+			       &failed_start, NULL, GFP_NOFS, NULL);
 	if (err == -EEXIST) {
 		if (failed_start > start)
 			clear_extent_bit(tree, start, failed_start - 1,
@@ -2078,8 +2141,8 @@ int repair_io_failure(struct inode *inode, u64 start, u64 length, u64 logical,
 		return -EIO;
 	}
 
-	printk_ratelimited_in_rcu(KERN_INFO
-				  "BTRFS: read error corrected: ino %llu off %llu (dev %s sector %llu)\n",
+	btrfs_info_rl_in_rcu(fs_info,
+		"read error corrected: ino %llu off %llu (dev %s sector %llu)",
 				  btrfs_ino(inode), start,
 				  rcu_str_deref(dev->name), sector);
 	bio_put(bio);
@@ -3070,8 +3133,12 @@ static int __do_readpage(struct extent_io_tree *tree,
 
 			set_extent_uptodate(tree, cur, cur + iosize - 1,
 					    &cached, GFP_NOFS);
-			unlock_extent_cached(tree, cur, cur + iosize - 1,
-			                     &cached, GFP_NOFS);
+			if (parent_locked)
+				free_extent_state(cached);
+			else
+				unlock_extent_cached(tree, cur,
+						     cur + iosize - 1,
+						     &cached, GFP_NOFS);
 			cur = cur + iosize;
 			pg_offset += iosize;
 			continue;
@@ -5566,13 +5633,15 @@ void memcpy_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 	unsigned long src_i;
 
 	if (src_offset + len > dst->len) {
-		printk(KERN_ERR "BTRFS: memmove bogus src_offset %lu move "
-		       "len %lu dst len %lu\n", src_offset, len, dst->len);
+		btrfs_err(dst->fs_info,
+			"memmove bogus src_offset %lu move "
+		       "len %lu dst len %lu", src_offset, len, dst->len);
 		BUG_ON(1);
 	}
 	if (dst_offset + len > dst->len) {
-		printk(KERN_ERR "BTRFS: memmove bogus dst_offset %lu move "
-		       "len %lu dst len %lu\n", dst_offset, len, dst->len);
+		btrfs_err(dst->fs_info,
+			"memmove bogus dst_offset %lu move "
+		       "len %lu dst len %lu", dst_offset, len, dst->len);
 		BUG_ON(1);
 	}
 
@@ -5612,13 +5681,13 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 	unsigned long src_i;
 
 	if (src_offset + len > dst->len) {
-		printk(KERN_ERR "BTRFS: memmove bogus src_offset %lu move "
-		       "len %lu len %lu\n", src_offset, len, dst->len);
+		btrfs_err(dst->fs_info, "memmove bogus src_offset %lu move "
+		       "len %lu len %lu", src_offset, len, dst->len);
 		BUG_ON(1);
 	}
 	if (dst_offset + len > dst->len) {
-		printk(KERN_ERR "BTRFS: memmove bogus dst_offset %lu move "
-		       "len %lu len %lu\n", dst_offset, len, dst->len);
+		btrfs_err(dst->fs_info, "memmove bogus dst_offset %lu move "
+		       "len %lu len %lu", dst_offset, len, dst->len);
 		BUG_ON(1);
 	}
 	if (dst_offset < src_offset) {

fs/btrfs/extent_io.h

@@ -2,6 +2,7 @@
 #define __EXTENTIO__
 
 #include <linux/rbtree.h>
+#include "ulist.h"
 
 /* bits for the extent state */
 #define EXTENT_DIRTY		(1U << 0)
@@ -18,6 +19,7 @@
 #define EXTENT_NEED_WAIT	(1U << 13)
 #define EXTENT_DAMAGED		(1U << 14)
 #define EXTENT_NORESERVE	(1U << 15)
+#define EXTENT_QGROUP_RESERVED	(1U << 16)
 #define EXTENT_IOBITS		(EXTENT_LOCKED | EXTENT_WRITEBACK)
 #define EXTENT_CTLBITS		(EXTENT_DO_ACCOUNTING | EXTENT_FIRST_DELALLOC)
 
@@ -161,6 +163,17 @@ struct extent_buffer {
 #endif
 };
 
+/*
+ * Structure to record how many bytes and which ranges are set/cleared
+ */
+struct extent_changeset {
+	/* How many bytes are set/cleared in this operation */
+	u64 bytes_changed;
+
+	/* Changed ranges */
+	struct ulist *range_changed;
+};
+
 static inline void extent_set_compress_type(unsigned long *bio_flags,
 					    int compress_type)
 {
@@ -210,11 +223,17 @@ int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end,
 		   struct extent_state *cached_state);
 int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 		      unsigned bits, gfp_t mask);
+int clear_record_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
+			     unsigned bits, gfp_t mask,
+			     struct extent_changeset *changeset);
 int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 		     unsigned bits, int wake, int delete,
 		     struct extent_state **cached, gfp_t mask);
 int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 		    unsigned bits, gfp_t mask);
+int set_record_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
+			   unsigned bits, gfp_t mask,
+			   struct extent_changeset *changeset);
 int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 		   unsigned bits, u64 *failed_start,
 		   struct extent_state **cached_state, gfp_t mask);

fs/btrfs/file.c

@@ -847,7 +847,7 @@ next_slot:
 						disk_bytenr, num_bytes, 0,
 						root->root_key.objectid,
 						new_key.objectid,
-						start - extent_offset, 1);
+						start - extent_offset);
 				BUG_ON(ret); /* -ENOMEM */
 			}
 			key.offset = start;
@@ -925,7 +925,7 @@ delete_extent_item:
 						disk_bytenr, num_bytes, 0,
 						root->root_key.objectid,
 						key.objectid, key.offset -
-						extent_offset, 0);
+						extent_offset);
 				BUG_ON(ret); /* -ENOMEM */
 				inode_sub_bytes(inode,
 						extent_end - key.offset);
@@ -1204,7 +1204,7 @@ again:
 
 		ret = btrfs_inc_extent_ref(trans, root, bytenr, num_bytes, 0,
 					   root->root_key.objectid,
-					   ino, orig_offset, 1);
+					   ino, orig_offset);
 		BUG_ON(ret); /* -ENOMEM */
 
 		if (split == start) {
@@ -1231,7 +1231,7 @@ again:
 		del_nr++;
 		ret = btrfs_free_extent(trans, root, bytenr, num_bytes,
 					0, root->root_key.objectid,
-					ino, orig_offset, 0);
+					ino, orig_offset);
 		BUG_ON(ret); /* -ENOMEM */
 	}
 	other_start = 0;
@@ -1248,7 +1248,7 @@ again:
 		del_nr++;
 		ret = btrfs_free_extent(trans, root, bytenr, num_bytes,
 					0, root->root_key.objectid,
-					ino, orig_offset, 0);
+					ino, orig_offset);
 		BUG_ON(ret); /* -ENOMEM */
 	}
 	if (del_nr == 0) {
@@ -1469,7 +1469,6 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 	u64 release_bytes = 0;
 	u64 lockstart;
 	u64 lockend;
-	unsigned long first_index;
 	size_t num_written = 0;
 	int nrptrs;
 	int ret = 0;
@@ -1485,8 +1484,6 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 	if (!pages)
 		return -ENOMEM;
 
-	first_index = pos >> PAGE_CACHE_SHIFT;
-
 	while (iov_iter_count(i) > 0) {
 		size_t offset = pos & (PAGE_CACHE_SIZE - 1);
 		size_t write_bytes = min(iov_iter_count(i),
@@ -1510,12 +1507,17 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 		}
 
 		reserve_bytes = num_pages << PAGE_CACHE_SHIFT;
-		ret = btrfs_check_data_free_space(inode, reserve_bytes, write_bytes);
-		if (ret == -ENOSPC &&
-		    (BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
-					      BTRFS_INODE_PREALLOC))) {
+
+		if (BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
+					     BTRFS_INODE_PREALLOC)) {
 			ret = check_can_nocow(inode, pos, &write_bytes);
+			if (ret < 0)
+				break;
 			if (ret > 0) {
+				/*
+				 * For nodata cow case, no need to reserve
+				 * data space.
+				 */
 				only_release_metadata = true;
 				/*
 				 * our prealloc extent may be smaller than
@@ -1524,20 +1526,19 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 				num_pages = DIV_ROUND_UP(write_bytes + offset,
 							 PAGE_CACHE_SIZE);
 				reserve_bytes = num_pages << PAGE_CACHE_SHIFT;
-				ret = 0;
-			} else {
-				ret = -ENOSPC;
+				goto reserve_metadata;
 			}
 		}
-
-		if (ret)
+		ret = btrfs_check_data_free_space(inode, pos, write_bytes);
+		if (ret < 0)
 			break;
 
+reserve_metadata:
 		ret = btrfs_delalloc_reserve_metadata(inode, reserve_bytes);
 		if (ret) {
 			if (!only_release_metadata)
-				btrfs_free_reserved_data_space(inode,
-							       reserve_bytes);
+				btrfs_free_reserved_data_space(inode, pos,
+							       write_bytes);
 			else
 				btrfs_end_write_no_snapshoting(root);
 			break;
@@ -1603,12 +1604,17 @@ again:
 				BTRFS_I(inode)->outstanding_extents++;
 				spin_unlock(&BTRFS_I(inode)->lock);
 			}
-			if (only_release_metadata)
+			if (only_release_metadata) {
 				btrfs_delalloc_release_metadata(inode,
 								release_bytes);
-			else
-				btrfs_delalloc_release_space(inode,
+			} else {
+				u64 __pos;
+
+				__pos = round_down(pos, root->sectorsize) +
+					(dirty_pages << PAGE_CACHE_SHIFT);
+				btrfs_delalloc_release_space(inode, __pos,
 							     release_bytes);
+			}
 		}
 
 		release_bytes = dirty_pages << PAGE_CACHE_SHIFT;
@@ -1660,7 +1666,7 @@ again:
 			btrfs_end_write_no_snapshoting(root);
 			btrfs_delalloc_release_metadata(inode, release_bytes);
 		} else {
-			btrfs_delalloc_release_space(inode, release_bytes);
+			btrfs_delalloc_release_space(inode, pos, release_bytes);
 		}
 	}
 
@@ -2266,7 +2272,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	u64 drop_end;
 	int ret = 0;
 	int err = 0;
-	int rsv_count;
+	unsigned int rsv_count;
 	bool same_page;
 	bool no_holes = btrfs_fs_incompat(root->fs_info, NO_HOLES);
 	u64 ino_size;
@@ -2487,6 +2493,19 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	}
 
 	trans->block_rsv = &root->fs_info->trans_block_rsv;
+	/*
+	 * If we are using the NO_HOLES feature we might have had already an
+	 * hole that overlaps a part of the region [lockstart, lockend] and
+	 * ends at (or beyond) lockend. Since we have no file extent items to
+	 * represent holes, drop_end can be less than lockend and so we must
+	 * make sure we have an extent map representing the existing hole (the
+	 * call to __btrfs_drop_extents() might have dropped the existing extent
+	 * map representing the existing hole), otherwise the fast fsync path
+	 * will not record the existence of the hole region
+	 * [existing_hole_start, lockend].
+	 */
+	if (drop_end <= lockend)
+		drop_end = lockend + 1;
 	/*
 	 * Don't insert file hole extent item if it's for a range beyond eof
 	 * (because it's useless) or if it represents a 0 bytes range (when
@@ -2541,17 +2560,61 @@ out_only_mutex:
 	return err;
 }
 
+/* Helper structure to record which range is already reserved */
+struct falloc_range {
+	struct list_head list;
+	u64 start;
+	u64 len;
+};
+
+/*
+ * Helper function to add falloc range
+ *
+ * Caller should have locked the larger range of extent containing
+ * [start, len)
+ */
+static int add_falloc_range(struct list_head *head, u64 start, u64 len)
+{
+	struct falloc_range *prev = NULL;
+	struct falloc_range *range = NULL;
+
+	if (list_empty(head))
+		goto insert;
+
+	/*
+	 * As fallocate iterate by bytenr order, we only need to check
+	 * the last range.
+	 */
+	prev = list_entry(head->prev, struct falloc_range, list);
+	if (prev->start + prev->len == start) {
+		prev->len += len;
+		return 0;
+	}
+insert:
+	range = kmalloc(sizeof(*range), GFP_NOFS);
+	if (!range)
+		return -ENOMEM;
+	range->start = start;
+	range->len = len;
+	list_add_tail(&range->list, head);
+	return 0;
+}
+
 static long btrfs_fallocate(struct file *file, int mode,
 			    loff_t offset, loff_t len)
 {
 	struct inode *inode = file_inode(file);
 	struct extent_state *cached_state = NULL;
+	struct falloc_range *range;
+	struct falloc_range *tmp;
+	struct list_head reserve_list;
 	u64 cur_offset;
 	u64 last_byte;
 	u64 alloc_start;
 	u64 alloc_end;
 	u64 alloc_hint = 0;
 	u64 locked_end;
+	u64 actual_end = 0;
 	struct extent_map *em;
 	int blocksize = BTRFS_I(inode)->root->sectorsize;
 	int ret;
@@ -2567,11 +2630,12 @@ static long btrfs_fallocate(struct file *file, int mode,
 		return btrfs_punch_hole(inode, offset, len);
 
 	/*
-	 * Make sure we have enough space before we do the
-	 * allocation.
+	 * Only trigger disk allocation, don't trigger qgroup reserve
+	 *
+	 * For qgroup space, it will be checked later.
 	 */
-	ret = btrfs_check_data_free_space(inode, alloc_end - alloc_start, alloc_end - alloc_start);
-	if (ret)
+	ret = btrfs_alloc_data_chunk_ondemand(inode, alloc_end - alloc_start);
+	if (ret < 0)
 		return ret;
 
 	mutex_lock(&inode->i_mutex);
@@ -2579,6 +2643,13 @@ static long btrfs_fallocate(struct file *file, int mode,
 	if (ret)
 		goto out;
 
+	/*
+	 * TODO: Move these two operations after we have checked
+	 * accurate reserved space, or fallocate can still fail but
+	 * with page truncated or size expanded.
+	 *
+	 * But that's a minor problem and won't do much harm BTW.
+	 */
 	if (alloc_start > inode->i_size) {
 		ret = btrfs_cont_expand(inode, i_size_read(inode),
 					alloc_start);
@@ -2637,10 +2708,10 @@ 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) {
-		u64 actual_end;
-
 		em = btrfs_get_extent(inode, NULL, 0, cur_offset,
 				      alloc_end - cur_offset, 0);
 		if (IS_ERR_OR_NULL(em)) {
@@ -2653,57 +2724,82 @@ static long btrfs_fallocate(struct file *file, int mode,
 		last_byte = min(extent_map_end(em), alloc_end);
 		actual_end = min_t(u64, extent_map_end(em), offset + len);
 		last_byte = ALIGN(last_byte, blocksize);
-
 		if (em->block_start == EXTENT_MAP_HOLE ||
 		    (cur_offset >= inode->i_size &&
 		     !test_bit(EXTENT_FLAG_PREALLOC, &em->flags))) {
-			ret = btrfs_prealloc_file_range(inode, mode, cur_offset,
-							last_byte - cur_offset,
-							1 << inode->i_blkbits,
-							offset + len,
-							&alloc_hint);
-		} else if (actual_end > inode->i_size &&
-			   !(mode & FALLOC_FL_KEEP_SIZE)) {
-			struct btrfs_trans_handle *trans;
-			struct btrfs_root *root = BTRFS_I(inode)->root;
-
-			/*
-			 * We didn't need to allocate any more space, but we
-			 * still extended the size of the file so we need to
-			 * update i_size and the inode item.
-			 */
-			trans = btrfs_start_transaction(root, 1);
-			if (IS_ERR(trans)) {
-				ret = PTR_ERR(trans);
-			} else {
-				inode->i_ctime = CURRENT_TIME;
-				i_size_write(inode, actual_end);
-				btrfs_ordered_update_i_size(inode, actual_end,
-							    NULL);
-				ret = btrfs_update_inode(trans, root, inode);
-				if (ret)
-					btrfs_end_transaction(trans, root);
-				else
-					ret = btrfs_end_transaction(trans,
-								    root);
+			ret = add_falloc_range(&reserve_list, cur_offset,
+					       last_byte - cur_offset);
+			if (ret < 0) {
+				free_extent_map(em);
+				break;
 			}
+			ret = btrfs_qgroup_reserve_data(inode, cur_offset,
+					last_byte - cur_offset);
+			if (ret < 0)
+				break;
 		}
 		free_extent_map(em);
-		if (ret < 0)
-			break;
-
 		cur_offset = last_byte;
-		if (cur_offset >= alloc_end) {
-			ret = 0;
+		if (cur_offset >= alloc_end)
 			break;
+	}
+
+	/*
+	 * If ret is still 0, means we're OK to fallocate.
+	 * Or just cleanup the list and exit.
+	 */
+	list_for_each_entry_safe(range, tmp, &reserve_list, list) {
+		if (!ret)
+			ret = btrfs_prealloc_file_range(inode, mode,
+					range->start,
+					range->len, 1 << inode->i_blkbits,
+					offset + len, &alloc_hint);
+		list_del(&range->list);
+		kfree(range);
+	}
+	if (ret < 0)
+		goto out_unlock;
+
+	if (actual_end > inode->i_size &&
+	    !(mode & FALLOC_FL_KEEP_SIZE)) {
+		struct btrfs_trans_handle *trans;
+		struct btrfs_root *root = BTRFS_I(inode)->root;
+
+		/*
+		 * We didn't need to allocate any more space, but we
+		 * still extended the size of the file so we need to
+		 * update i_size and the inode item.
+		 */
+		trans = btrfs_start_transaction(root, 1);
+		if (IS_ERR(trans)) {
+			ret = PTR_ERR(trans);
+		} else {
+			inode->i_ctime = CURRENT_TIME;
+			i_size_write(inode, actual_end);
+			btrfs_ordered_update_i_size(inode, actual_end, NULL);
+			ret = btrfs_update_inode(trans, root, inode);
+			if (ret)
+				btrfs_end_transaction(trans, root);
+			else
+				ret = btrfs_end_transaction(trans, root);
 		}
 	}
+out_unlock:
 	unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
 			     &cached_state, GFP_NOFS);
 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);
 	mutex_unlock(&inode->i_mutex);
 	/* Let go of our reservation. */
-	btrfs_free_reserved_data_space(inode, alloc_end - alloc_start);
+	btrfs_free_reserved_data_space(inode, alloc_start,
+				       alloc_end - alloc_start);
 	return ret;
 }
 

fs/btrfs/free-space-cache.c

@@ -450,9 +450,9 @@ static int io_ctl_check_generation(struct btrfs_io_ctl *io_ctl, u64 generation)
 
 	gen = io_ctl->cur;
 	if (le64_to_cpu(*gen) != generation) {
-		printk_ratelimited(KERN_ERR "BTRFS: space cache generation "
-				   "(%Lu) does not match inode (%Lu)\n", *gen,
-				   generation);
+		btrfs_err_rl(io_ctl->root->fs_info,
+			"space cache generation (%llu) does not match inode (%llu)",
+				*gen, generation);
 		io_ctl_unmap_page(io_ctl);
 		return -EIO;
 	}
@@ -506,8 +506,8 @@ static int io_ctl_check_crc(struct btrfs_io_ctl *io_ctl, int index)
 			      PAGE_CACHE_SIZE - offset);
 	btrfs_csum_final(crc, (char *)&crc);
 	if (val != crc) {
-		printk_ratelimited(KERN_ERR "BTRFS: csum mismatch on free "
-				   "space cache\n");
+		btrfs_err_rl(io_ctl->root->fs_info,
+			"csum mismatch on free space cache");
 		io_ctl_unmap_page(io_ctl);
 		return -EIO;
 	}
@@ -1215,7 +1215,7 @@ out:
  * @offset - the offset for the key we'll insert
  *
  * This function writes out a free space cache struct to disk for quick recovery
- * on mount.  This will return 0 if it was successfull in writing the cache out,
+ * on mount.  This will return 0 if it was successful in writing the cache out,
  * or an errno if it was not.
  */
 static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
@@ -1730,7 +1730,7 @@ static void bitmap_set_bits(struct btrfs_free_space_ctl *ctl,
  */
 static int search_bitmap(struct btrfs_free_space_ctl *ctl,
 			 struct btrfs_free_space *bitmap_info, u64 *offset,
-			 u64 *bytes)
+			 u64 *bytes, bool for_alloc)
 {
 	unsigned long found_bits = 0;
 	unsigned long max_bits = 0;
@@ -1738,11 +1738,26 @@ static int search_bitmap(struct btrfs_free_space_ctl *ctl,
 	unsigned long next_zero;
 	unsigned long extent_bits;
 
+	/*
+	 * Skip searching the bitmap if we don't have a contiguous section that
+	 * is large enough for this allocation.
+	 */
+	if (for_alloc &&
+	    bitmap_info->max_extent_size &&
+	    bitmap_info->max_extent_size < *bytes) {
+		*bytes = bitmap_info->max_extent_size;
+		return -1;
+	}
+
 	i = offset_to_bit(bitmap_info->offset, ctl->unit,
 			  max_t(u64, *offset, bitmap_info->offset));
 	bits = bytes_to_bits(*bytes, ctl->unit);
 
 	for_each_set_bit_from(i, bitmap_info->bitmap, BITS_PER_BITMAP) {
+		if (for_alloc && bits == 1) {
+			found_bits = 1;
+			break;
+		}
 		next_zero = find_next_zero_bit(bitmap_info->bitmap,
 					       BITS_PER_BITMAP, i);
 		extent_bits = next_zero - i;
@@ -1762,6 +1777,7 @@ static int search_bitmap(struct btrfs_free_space_ctl *ctl,
 	}
 
 	*bytes = (u64)(max_bits) * ctl->unit;
+	bitmap_info->max_extent_size = *bytes;
 	return -1;
 }
 
@@ -1813,7 +1829,7 @@ find_free_space(struct btrfs_free_space_ctl *ctl, u64 *offset, u64 *bytes,
 		if (entry->bitmap) {
 			u64 size = *bytes;
 
-			ret = search_bitmap(ctl, entry, &tmp, &size);
+			ret = search_bitmap(ctl, entry, &tmp, &size, true);
 			if (!ret) {
 				*offset = tmp;
 				*bytes = size;
@@ -1874,7 +1890,8 @@ again:
 	search_start = *offset;
 	search_bytes = ctl->unit;
 	search_bytes = min(search_bytes, end - search_start + 1);
-	ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes);
+	ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes,
+			    false);
 	if (ret < 0 || search_start != *offset)
 		return -EINVAL;
 
@@ -1919,7 +1936,7 @@ again:
 		search_start = *offset;
 		search_bytes = ctl->unit;
 		ret = search_bitmap(ctl, bitmap_info, &search_start,
-				    &search_bytes);
+				    &search_bytes, false);
 		if (ret < 0 || search_start != *offset)
 			return -EAGAIN;
 
@@ -1943,6 +1960,12 @@ static u64 add_bytes_to_bitmap(struct btrfs_free_space_ctl *ctl,
 
 	bitmap_set_bits(ctl, info, offset, bytes_to_set);
 
+	/*
+	 * We set some bytes, we have no idea what the max extent size is
+	 * anymore.
+	 */
+	info->max_extent_size = 0;
+
 	return bytes_to_set;
 
 }
@@ -1951,12 +1974,19 @@ static bool use_bitmap(struct btrfs_free_space_ctl *ctl,
 		      struct btrfs_free_space *info)
 {
 	struct btrfs_block_group_cache *block_group = ctl->private;
+	bool forced = false;
+
+#ifdef CONFIG_BTRFS_DEBUG
+	if (btrfs_should_fragment_free_space(block_group->fs_info->extent_root,
+					     block_group))
+		forced = true;
+#endif
 
 	/*
 	 * If we are below the extents threshold then we can add this as an
 	 * extent, and don't have to deal with the bitmap
 	 */
-	if (ctl->free_extents < ctl->extents_thresh) {
+	if (!forced && ctl->free_extents < ctl->extents_thresh) {
 		/*
 		 * If this block group has some small extents we don't want to
 		 * use up all of our free slots in the cache with them, we want
@@ -2661,7 +2691,7 @@ static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group,
 	search_start = min_start;
 	search_bytes = bytes;
 
-	err = search_bitmap(ctl, entry, &search_start, &search_bytes);
+	err = search_bitmap(ctl, entry, &search_start, &search_bytes, true);
 	if (err) {
 		if (search_bytes > *max_extent_size)
 			*max_extent_size = search_bytes;
@@ -2775,6 +2805,7 @@ static int btrfs_bitmap_cluster(struct btrfs_block_group_cache *block_group,
 	unsigned long want_bits;
 	unsigned long min_bits;
 	unsigned long found_bits;
+	unsigned long max_bits = 0;
 	unsigned long start = 0;
 	unsigned long total_found = 0;
 	int ret;
@@ -2784,6 +2815,13 @@ static int btrfs_bitmap_cluster(struct btrfs_block_group_cache *block_group,
 	want_bits = bytes_to_bits(bytes, ctl->unit);
 	min_bits = bytes_to_bits(min_bytes, ctl->unit);
 
+	/*
+	 * Don't bother looking for a cluster in this bitmap if it's heavily
+	 * fragmented.
+	 */
+	if (entry->max_extent_size &&
+	    entry->max_extent_size < cont1_bytes)
+		return -ENOSPC;
 again:
 	found_bits = 0;
 	for_each_set_bit_from(i, entry->bitmap, BITS_PER_BITMAP) {
@@ -2791,13 +2829,19 @@ again:
 					       BITS_PER_BITMAP, i);
 		if (next_zero - i >= min_bits) {
 			found_bits = next_zero - i;
+			if (found_bits > max_bits)
+				max_bits = found_bits;
 			break;
 		}
+		if (next_zero - i > max_bits)
+			max_bits = next_zero - i;
 		i = next_zero;
 	}
 
-	if (!found_bits)
+	if (!found_bits) {
+		entry->max_extent_size = (u64)max_bits * ctl->unit;
 		return -ENOSPC;
+	}
 
 	if (!total_found) {
 		start = i;
@@ -3056,6 +3100,7 @@ void btrfs_init_free_cluster(struct btrfs_free_cluster *cluster)
 	spin_lock_init(&cluster->refill_lock);
 	cluster->root = RB_ROOT;
 	cluster->max_size = 0;
+	cluster->fragmented = false;
 	INIT_LIST_HEAD(&cluster->block_group_list);
 	cluster->block_group = NULL;
 }
@@ -3223,7 +3268,7 @@ static int trim_bitmaps(struct btrfs_block_group_cache *block_group,
 		}
 
 		bytes = minlen;
-		ret2 = search_bitmap(ctl, entry, &start, &bytes);
+		ret2 = search_bitmap(ctl, entry, &start, &bytes, false);
 		if (ret2 || start >= end) {
 			spin_unlock(&ctl->tree_lock);
 			mutex_unlock(&ctl->cache_writeout_mutex);
@@ -3376,7 +3421,7 @@ u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root)
 		u64 count = 1;
 		int ret;
 
-		ret = search_bitmap(ctl, entry, &offset, &count);
+		ret = search_bitmap(ctl, entry, &offset, &count, true);
 		/* Logic error; Should be empty if it can't find anything */
 		ASSERT(!ret);
 
@@ -3532,6 +3577,7 @@ again:
 		spin_lock(&ctl->tree_lock);
 		info->offset = offset;
 		info->bytes = bytes;
+		info->max_extent_size = 0;
 		ret = link_free_space(ctl, info);
 		spin_unlock(&ctl->tree_lock);
 		if (ret)
@@ -3559,6 +3605,7 @@ again:
 	}
 
 	bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes);
+
 	bytes -= bytes_added;
 	offset += bytes_added;
 	spin_unlock(&ctl->tree_lock);
@@ -3602,7 +3649,7 @@ have_info:
 
 		bit_off = offset;
 		bit_bytes = ctl->unit;
-		ret = search_bitmap(ctl, info, &bit_off, &bit_bytes);
+		ret = search_bitmap(ctl, info, &bit_off, &bit_bytes, false);
 		if (!ret) {
 			if (bit_off == offset) {
 				ret = 1;

fs/btrfs/free-space-cache.h

@@ -23,6 +23,7 @@ struct btrfs_free_space {
 	struct rb_node offset_index;
 	u64 offset;
 	u64 bytes;
+	u64 max_extent_size;
 	unsigned long *bitmap;
 	struct list_head list;
 };

fs/btrfs/inode-item.c

@@ -157,7 +157,7 @@ static int btrfs_del_inode_extref(struct btrfs_trans_handle *trans,
 	 */
 	if (!btrfs_find_name_in_ext_backref(path, ref_objectid,
 					    name, name_len, &extref)) {
-		btrfs_std_error(root->fs_info, -ENOENT);
+		btrfs_std_error(root->fs_info, -ENOENT, NULL);
 		ret = -EROFS;
 		goto out;
 	}

fs/btrfs/inode-map.c

@@ -488,17 +488,17 @@ again:
 	/* Just to make sure we have enough space */
 	prealloc += 8 * PAGE_CACHE_SIZE;
 
-	ret = btrfs_delalloc_reserve_space(inode, prealloc);
+	ret = btrfs_delalloc_reserve_space(inode, 0, prealloc);
 	if (ret)
 		goto out_put;
 
 	ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, prealloc,
 					      prealloc, prealloc, &alloc_hint);
 	if (ret) {
-		btrfs_delalloc_release_space(inode, prealloc);
+		btrfs_delalloc_release_space(inode, 0, prealloc);
 		goto out_put;
 	}
-	btrfs_free_reserved_data_space(inode, prealloc);
+	btrfs_free_reserved_data_space(inode, 0, prealloc);
 
 	ret = btrfs_write_out_ino_cache(root, trans, path, inode);
 out_put:

fs/btrfs/inode.c

@@ -310,6 +310,13 @@ static noinline int cow_file_range_inline(struct btrfs_root *root,
 	btrfs_delalloc_release_metadata(inode, end + 1 - start);
 	btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
 out:
+	/*
+	 * Don't forget to free the reserved space, as for inlined extent
+	 * it won't count as data extent, free them directly here.
+	 * And at reserve time, it's always aligned to page size, so
+	 * just free one page here.
+	 */
+	btrfs_qgroup_free_data(inode, 0, PAGE_CACHE_SIZE);
 	btrfs_free_path(path);
 	btrfs_end_transaction(trans, root);
 	return ret;
@@ -1096,6 +1103,9 @@ static noinline void async_cow_submit(struct btrfs_work *work)
 	nr_pages = (async_cow->end - async_cow->start + PAGE_CACHE_SIZE) >>
 		PAGE_CACHE_SHIFT;
 
+	/*
+	 * atomic_sub_return implies a barrier for waitqueue_active
+	 */
 	if (atomic_sub_return(nr_pages, &root->fs_info->async_delalloc_pages) <
 	    5 * 1024 * 1024 &&
 	    waitqueue_active(&root->fs_info->async_submit_wait))
@@ -1766,7 +1776,8 @@ static void btrfs_clear_bit_hook(struct inode *inode,
 
 		if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID
 		    && do_list && !(state->state & EXTENT_NORESERVE))
-			btrfs_free_reserved_data_space(inode, len);
+			btrfs_free_reserved_data_space_noquota(inode,
+					state->start, len);
 
 		__percpu_counter_add(&root->fs_info->delalloc_bytes, -len,
 				     root->fs_info->delalloc_batch);
@@ -1861,15 +1872,15 @@ static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
 			  u64 bio_offset)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
+	enum btrfs_wq_endio_type metadata = BTRFS_WQ_ENDIO_DATA;
 	int ret = 0;
 	int skip_sum;
-	int metadata = 0;
 	int async = !atomic_read(&BTRFS_I(inode)->sync_writers);
 
 	skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
 
 	if (btrfs_is_free_space_inode(inode))
-		metadata = 2;
+		metadata = BTRFS_WQ_ENDIO_FREE_SPACE;
 
 	if (!(rw & REQ_WRITE)) {
 		ret = btrfs_bio_wq_end_io(root->fs_info, bio, metadata);
@@ -1989,7 +2000,8 @@ again:
 		goto again;
 	}
 
-	ret = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE);
+	ret = btrfs_delalloc_reserve_space(inode, page_start,
+					   PAGE_CACHE_SIZE);
 	if (ret) {
 		mapping_set_error(page->mapping, ret);
 		end_extent_writepage(page, ret, page_start, page_end);
@@ -2115,7 +2127,13 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
 	ins.type = BTRFS_EXTENT_ITEM_KEY;
 	ret = btrfs_alloc_reserved_file_extent(trans, root,
 					root->root_key.objectid,
-					btrfs_ino(inode), file_pos, &ins);
+					btrfs_ino(inode), file_pos,
+					ram_bytes, &ins);
+	/*
+	 * Release the reserved range from inode dirty range map, as it is
+	 * already moved into delayed_ref_head
+	 */
+	btrfs_qgroup_release_data(inode, file_pos, ram_bytes);
 out:
 	btrfs_free_path(path);
 
@@ -2573,7 +2591,7 @@ again:
 	ret = btrfs_inc_extent_ref(trans, root, new->bytenr,
 			new->disk_len, 0,
 			backref->root_id, backref->inum,
-			new->file_pos, 0);	/* start - extent_offset */
+			new->file_pos);	/* start - extent_offset */
 	if (ret) {
 		btrfs_abort_transaction(trans, root, ret);
 		goto out_free_path;
@@ -2599,7 +2617,6 @@ static void free_sa_defrag_extent(struct new_sa_defrag_extent *new)
 		return;
 
 	list_for_each_entry_safe(old, tmp, &new->head, list) {
-		list_del(&old->list);
 		kfree(old);
 	}
 	kfree(new);
@@ -2824,6 +2841,14 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 
 	if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) {
 		BUG_ON(!list_empty(&ordered_extent->list)); /* Logic error */
+
+		/*
+		 * For mwrite(mmap + memset to write) case, we still reserve
+		 * space for NOCOW range.
+		 * As NOCOW won't cause a new delayed ref, just free the space
+		 */
+		btrfs_qgroup_free_data(inode, ordered_extent->file_offset,
+				       ordered_extent->len);
 		btrfs_ordered_update_i_size(inode, 0, ordered_extent);
 		if (nolock)
 			trans = btrfs_join_transaction_nolock(root);
@@ -3018,8 +3043,6 @@ static int __readpage_endio_check(struct inode *inode,
 	char *kaddr;
 	u32 csum_expected;
 	u32 csum = ~(u32)0;
-	static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
-				      DEFAULT_RATELIMIT_BURST);
 
 	csum_expected = *(((u32 *)io_bio->csum) + icsum);
 
@@ -3032,9 +3055,8 @@ static int __readpage_endio_check(struct inode *inode,
 	kunmap_atomic(kaddr);
 	return 0;
 zeroit:
-	if (__ratelimit(&_rs))
-		btrfs_warn(BTRFS_I(inode)->root->fs_info,
-			   "csum failed ino %llu off %llu csum %u expected csum %u",
+	btrfs_warn_rl(BTRFS_I(inode)->root->fs_info,
+		"csum failed ino %llu off %llu csum %u expected csum %u",
 			   btrfs_ino(inode), start, csum, csum_expected);
 	memset(kaddr + pgoff, 1, len);
 	flush_dcache_page(page);
@@ -4217,6 +4239,47 @@ static int truncate_space_check(struct btrfs_trans_handle *trans,
 
 }
 
+static int truncate_inline_extent(struct inode *inode,
+				  struct btrfs_path *path,
+				  struct btrfs_key *found_key,
+				  const u64 item_end,
+				  const u64 new_size)
+{
+	struct extent_buffer *leaf = path->nodes[0];
+	int slot = path->slots[0];
+	struct btrfs_file_extent_item *fi;
+	u32 size = (u32)(new_size - found_key->offset);
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+
+	fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
+
+	if (btrfs_file_extent_compression(leaf, fi) != BTRFS_COMPRESS_NONE) {
+		loff_t offset = new_size;
+		loff_t page_end = ALIGN(offset, PAGE_CACHE_SIZE);
+
+		/*
+		 * Zero out the remaining of the last page of our inline extent,
+		 * instead of directly truncating our inline extent here - that
+		 * would be much more complex (decompressing all the data, then
+		 * compressing the truncated data, which might be bigger than
+		 * the size of the inline extent, resize the extent, etc).
+		 * We release the path because to get the page we might need to
+		 * read the extent item from disk (data not in the page cache).
+		 */
+		btrfs_release_path(path);
+		return btrfs_truncate_page(inode, offset, page_end - offset, 0);
+	}
+
+	btrfs_set_file_extent_ram_bytes(leaf, fi, size);
+	size = btrfs_file_extent_calc_inline_size(size);
+	btrfs_truncate_item(root, path, size, 1);
+
+	if (test_bit(BTRFS_ROOT_REF_COWS, &root->state))
+		inode_sub_bytes(inode, item_end + 1 - new_size);
+
+	return 0;
+}
+
 /*
  * this can truncate away extent items, csum items and directory items.
  * It starts at a high offset and removes keys until it can't find
@@ -4411,27 +4474,40 @@ search_again:
 			 * special encodings
 			 */
 			if (!del_item &&
-			    btrfs_file_extent_compression(leaf, fi) == 0 &&
 			    btrfs_file_extent_encryption(leaf, fi) == 0 &&
 			    btrfs_file_extent_other_encoding(leaf, fi) == 0) {
-				u32 size = new_size - found_key.offset;
-
-				if (test_bit(BTRFS_ROOT_REF_COWS, &root->state))
-					inode_sub_bytes(inode, item_end + 1 -
-							new_size);
 
 				/*
-				 * update the ram bytes to properly reflect
-				 * the new size of our item
+				 * Need to release path in order to truncate a
+				 * compressed extent. So delete any accumulated
+				 * extent items so far.
 				 */
-				btrfs_set_file_extent_ram_bytes(leaf, fi, size);
-				size =
-				    btrfs_file_extent_calc_inline_size(size);
-				btrfs_truncate_item(root, path, size, 1);
+				if (btrfs_file_extent_compression(leaf, fi) !=
+				    BTRFS_COMPRESS_NONE && pending_del_nr) {
+					err = btrfs_del_items(trans, root, path,
+							      pending_del_slot,
+							      pending_del_nr);
+					if (err) {
+						btrfs_abort_transaction(trans,
+									root,
+									err);
+						goto error;
+					}
+					pending_del_nr = 0;
+				}
+
+				err = truncate_inline_extent(inode, path,
+							     &found_key,
+							     item_end,
+							     new_size);
+				if (err) {
+					btrfs_abort_transaction(trans,
+								root, err);
+					goto error;
+				}
 			} else if (test_bit(BTRFS_ROOT_REF_COWS,
 					    &root->state)) {
-				inode_sub_bytes(inode, item_end + 1 -
-						found_key.offset);
+				inode_sub_bytes(inode, item_end + 1 - new_size);
 			}
 		}
 delete:
@@ -4461,7 +4537,7 @@ delete:
 			ret = btrfs_free_extent(trans, root, extent_start,
 						extent_num_bytes, 0,
 						btrfs_header_owner(leaf),
-						ino, extent_offset, 0);
+						ino, extent_offset);
 			BUG_ON(ret);
 			if (btrfs_should_throttle_delayed_refs(trans, root))
 				btrfs_async_run_delayed_refs(root,
@@ -4575,14 +4651,17 @@ int btrfs_truncate_page(struct inode *inode, loff_t from, loff_t len,
 	if ((offset & (blocksize - 1)) == 0 &&
 	    (!len || ((len & (blocksize - 1)) == 0)))
 		goto out;
-	ret = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE);
+	ret = btrfs_delalloc_reserve_space(inode,
+			round_down(from, PAGE_CACHE_SIZE), PAGE_CACHE_SIZE);
 	if (ret)
 		goto out;
 
 again:
 	page = find_or_create_page(mapping, index, mask);
 	if (!page) {
-		btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE);
+		btrfs_delalloc_release_space(inode,
+				round_down(from, PAGE_CACHE_SIZE),
+				PAGE_CACHE_SIZE);
 		ret = -ENOMEM;
 		goto out;
 	}
@@ -4650,7 +4729,8 @@ again:
 
 out_unlock:
 	if (ret)
-		btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE);
+		btrfs_delalloc_release_space(inode, page_start,
+					     PAGE_CACHE_SIZE);
 	unlock_page(page);
 	page_cache_release(page);
 out:
@@ -5048,6 +5128,18 @@ static void evict_inode_truncate_pages(struct inode *inode)
 		spin_unlock(&io_tree->lock);
 
 		lock_extent_bits(io_tree, start, end, 0, &cached_state);
+
+		/*
+		 * If still has DELALLOC flag, the extent didn't reach disk,
+		 * and its reserved space won't be freed by delayed_ref.
+		 * So we need to free its reserved space here.
+		 * (Refer to comment in btrfs_invalidatepage, case 2)
+		 *
+		 * Note, end is the bytenr of last byte, so we need + 1 here.
+		 */
+		if (state->state & EXTENT_DELALLOC)
+			btrfs_qgroup_free_data(inode, start, end - start + 1);
+
 		clear_extent_bit(io_tree, start, end,
 				 EXTENT_LOCKED | EXTENT_DIRTY |
 				 EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING |
@@ -7581,7 +7673,7 @@ unlock:
 			spin_unlock(&BTRFS_I(inode)->lock);
 		}
 
-		btrfs_free_reserved_data_space(inode, len);
+		btrfs_free_reserved_data_space(inode, start, len);
 		WARN_ON(dio_data->reserve < len);
 		dio_data->reserve -= len;
 		current->journal_info = dio_data;
@@ -8371,7 +8463,7 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter,
 			mutex_unlock(&inode->i_mutex);
 			relock = true;
 		}
-		ret = btrfs_delalloc_reserve_space(inode, count);
+		ret = btrfs_delalloc_reserve_space(inode, offset, count);
 		if (ret)
 			goto out;
 		dio_data.outstanding_extents = div64_u64(count +
@@ -8400,10 +8492,10 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter,
 		current->journal_info = NULL;
 		if (ret < 0 && ret != -EIOCBQUEUED) {
 			if (dio_data.reserve)
-				btrfs_delalloc_release_space(inode,
-							dio_data.reserve);
+				btrfs_delalloc_release_space(inode, offset,
+							     dio_data.reserve);
 		} else if (ret >= 0 && (size_t)ret < count)
-			btrfs_delalloc_release_space(inode,
+			btrfs_delalloc_release_space(inode, offset,
 						     count - (size_t)ret);
 	}
 out:
@@ -8562,6 +8654,18 @@ static void btrfs_invalidatepage(struct page *page, unsigned int offset,
 		}
 	}
 
+	/*
+	 * Qgroup reserved space handler
+	 * Page here will be either
+	 * 1) Already written to disk
+	 *    In this case, its reserved space is released from data rsv map
+	 *    and will be freed by delayed_ref handler finally.
+	 *    So even we call qgroup_free_data(), it won't decrease reserved
+	 *    space.
+	 * 2) Not written to disk
+	 *    This means the reserved space should be freed here.
+	 */
+	btrfs_qgroup_free_data(inode, page_start, PAGE_CACHE_SIZE);
 	if (!inode_evicting) {
 		clear_extent_bit(tree, page_start, page_end,
 				 EXTENT_LOCKED | EXTENT_DIRTY |
@@ -8612,7 +8716,11 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	u64 page_end;
 
 	sb_start_pagefault(inode->i_sb);
-	ret  = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE);
+	page_start = page_offset(page);
+	page_end = page_start + PAGE_CACHE_SIZE - 1;
+
+	ret = btrfs_delalloc_reserve_space(inode, page_start,
+					   PAGE_CACHE_SIZE);
 	if (!ret) {
 		ret = file_update_time(vma->vm_file);
 		reserved = 1;
@@ -8631,8 +8739,6 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 again:
 	lock_page(page);
 	size = i_size_read(inode);
-	page_start = page_offset(page);
-	page_end = page_start + PAGE_CACHE_SIZE - 1;
 
 	if ((page->mapping != inode->i_mapping) ||
 	    (page_start >= size)) {
@@ -8709,7 +8815,7 @@ out_unlock:
 	}
 	unlock_page(page);
 out:
-	btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE);
+	btrfs_delalloc_release_space(inode, page_start, PAGE_CACHE_SIZE);
 out_noreserve:
 	sb_end_pagefault(inode->i_sb);
 	return ret;
@@ -8998,6 +9104,7 @@ void btrfs_destroy_inode(struct inode *inode)
 			btrfs_put_ordered_extent(ordered);
 		}
 	}
+	btrfs_qgroup_check_reserved_leak(inode);
 	inode_tree_del(inode);
 	btrfs_drop_extent_cache(inode, 0, (u64)-1, 0);
 free:
@@ -9634,6 +9741,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 	u64 cur_offset = start;
 	u64 i_size;
 	u64 cur_bytes;
+	u64 last_alloc = (u64)-1;
 	int ret = 0;
 	bool own_trans = true;
 
@@ -9650,6 +9758,13 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 
 		cur_bytes = min(num_bytes, 256ULL * 1024 * 1024);
 		cur_bytes = max(cur_bytes, min_size);
+		/*
+		 * If we are severely fragmented we could end up with really
+		 * small allocations, so if the allocator is returning small
+		 * chunks lets make its job easier by only searching for those
+		 * sized chunks.
+		 */
+		cur_bytes = min(cur_bytes, last_alloc);
 		ret = btrfs_reserve_extent(root, cur_bytes, min_size, 0,
 					   *alloc_hint, &ins, 1, 0);
 		if (ret) {
@@ -9658,6 +9773,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
 			break;
 		}
 
+		last_alloc = ins.offset;
 		ret = insert_reserved_file_extent(trans, inode,
 						  cur_offset, ins.objectid,
 						  ins.offset, ins.offset,

fs/btrfs/ioctl.c

@@ -1120,7 +1120,8 @@ static int cluster_pages_for_defrag(struct inode *inode,
 	page_cnt = min_t(u64, (u64)num_pages, (u64)file_end - start_index + 1);
 
 	ret = btrfs_delalloc_reserve_space(inode,
-					   page_cnt << PAGE_CACHE_SHIFT);
+			start_index << PAGE_CACHE_SHIFT,
+			page_cnt << PAGE_CACHE_SHIFT);
 	if (ret)
 		return ret;
 	i_done = 0;
@@ -1210,7 +1211,8 @@ again:
 		BTRFS_I(inode)->outstanding_extents++;
 		spin_unlock(&BTRFS_I(inode)->lock);
 		btrfs_delalloc_release_space(inode,
-				     (page_cnt - i_done) << PAGE_CACHE_SHIFT);
+				start_index << PAGE_CACHE_SHIFT,
+				(page_cnt - i_done) << PAGE_CACHE_SHIFT);
 	}
 
 
@@ -1235,7 +1237,9 @@ out:
 		unlock_page(pages[i]);
 		page_cache_release(pages[i]);
 	}
-	btrfs_delalloc_release_space(inode, page_cnt << PAGE_CACHE_SHIFT);
+	btrfs_delalloc_release_space(inode,
+			start_index << PAGE_CACHE_SHIFT,
+			page_cnt << PAGE_CACHE_SHIFT);
 	return ret;
 
 }
@@ -1342,7 +1346,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			break;
 
 		if (btrfs_defrag_cancelled(root->fs_info)) {
-			printk(KERN_DEBUG "BTRFS: defrag_file cancelled\n");
+			btrfs_debug(root->fs_info, "defrag_file cancelled");
 			ret = -EAGAIN;
 			break;
 		}
@@ -1579,7 +1583,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 	new_size = div_u64(new_size, root->sectorsize);
 	new_size *= root->sectorsize;
 
-	printk_in_rcu(KERN_INFO "BTRFS: new size for %s is %llu\n",
+	btrfs_info_in_rcu(root->fs_info, "new size for %s is %llu",
 		      rcu_str_deref(device->name), new_size);
 
 	if (new_size > old_size) {
@@ -2081,7 +2085,7 @@ static noinline int search_ioctl(struct inode *inode,
 		key.offset = (u64)-1;
 		root = btrfs_read_fs_root_no_name(info, &key);
 		if (IS_ERR(root)) {
-			printk(KERN_ERR "BTRFS: could not find root %llu\n",
+			btrfs_err(info, "could not find root %llu",
 			       sk->tree_id);
 			btrfs_free_path(path);
 			return -ENOENT;
@@ -2221,7 +2225,7 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
 	key.offset = (u64)-1;
 	root = btrfs_read_fs_root_no_name(info, &key);
 	if (IS_ERR(root)) {
-		printk(KERN_ERR "BTRFS: could not find root %llu\n", tree_id);
+		btrfs_err(info, "could not find root %llu", tree_id);
 		ret = -ENOENT;
 		goto out;
 	}
@@ -2699,7 +2703,6 @@ static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
 {
 	struct btrfs_ioctl_fs_info_args *fi_args;
 	struct btrfs_device *device;
-	struct btrfs_device *next;
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 	int ret = 0;
 
@@ -2711,7 +2714,7 @@ static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
 	fi_args->num_devices = fs_devices->num_devices;
 	memcpy(&fi_args->fsid, root->fs_info->fsid, sizeof(fi_args->fsid));
 
-	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
+	list_for_each_entry(device, &fs_devices->devices, dev_list) {
 		if (device->devid > fi_args->max_id)
 			fi_args->max_id = device->devid;
 	}
@@ -3203,41 +3206,6 @@ out:
 	return ret;
 }
 
-/* Helper to check and see if this root currently has a ref on the given disk
- * bytenr.  If it does then we need to update the quota for this root.  This
- * doesn't do anything if quotas aren't enabled.
- */
-static int check_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
-		     u64 disko)
-{
-	struct seq_list tree_mod_seq_elem = SEQ_LIST_INIT(tree_mod_seq_elem);
-	struct ulist *roots;
-	struct ulist_iterator uiter;
-	struct ulist_node *root_node = NULL;
-	int ret;
-
-	if (!root->fs_info->quota_enabled)
-		return 1;
-
-	btrfs_get_tree_mod_seq(root->fs_info, &tree_mod_seq_elem);
-	ret = btrfs_find_all_roots(trans, root->fs_info, disko,
-				   tree_mod_seq_elem.seq, &roots);
-	if (ret < 0)
-		goto out;
-	ret = 0;
-	ULIST_ITER_INIT(&uiter);
-	while ((root_node = ulist_next(roots, &uiter))) {
-		if (root_node->val == root->objectid) {
-			ret = 1;
-			break;
-		}
-	}
-	ulist_free(roots);
-out:
-	btrfs_put_tree_mod_seq(root->fs_info, &tree_mod_seq_elem);
-	return ret;
-}
-
 static int clone_finish_inode_update(struct btrfs_trans_handle *trans,
 				     struct inode *inode,
 				     u64 endoff,
@@ -3328,6 +3296,150 @@ static void clone_update_extent_map(struct inode *inode,
 			&BTRFS_I(inode)->runtime_flags);
 }
 
+/*
+ * Make sure we do not end up inserting an inline extent into a file that has
+ * already other (non-inline) extents. If a file has an inline extent it can
+ * not have any other extents and the (single) inline extent must start at the
+ * file offset 0. Failing to respect these rules will lead to file corruption,
+ * resulting in EIO errors on read/write operations, hitting BUG_ON's in mm, etc
+ *
+ * We can have extents that have been already written to disk or we can have
+ * dirty ranges still in delalloc, in which case the extent maps and items are
+ * created only when we run delalloc, and the delalloc ranges might fall outside
+ * the range we are currently locking in the inode's io tree. So we check the
+ * inode's i_size because of that (i_size updates are done while holding the
+ * i_mutex, which we are holding here).
+ * We also check to see if the inode has a size not greater than "datal" but has
+ * extents beyond it, due to an fallocate with FALLOC_FL_KEEP_SIZE (and we are
+ * protected against such concurrent fallocate calls by the i_mutex).
+ *
+ * If the file has no extents but a size greater than datal, do not allow the
+ * copy because we would need turn the inline extent into a non-inline one (even
+ * with NO_HOLES enabled). If we find our destination inode only has one inline
+ * extent, just overwrite it with the source inline extent if its size is less
+ * than the source extent's size, or we could copy the source inline extent's
+ * data into the destination inode's inline extent if the later is greater then
+ * the former.
+ */
+static int clone_copy_inline_extent(struct inode *src,
+				    struct inode *dst,
+				    struct btrfs_trans_handle *trans,
+				    struct btrfs_path *path,
+				    struct btrfs_key *new_key,
+				    const u64 drop_start,
+				    const u64 datal,
+				    const u64 skip,
+				    const u64 size,
+				    char *inline_data)
+{
+	struct btrfs_root *root = BTRFS_I(dst)->root;
+	const u64 aligned_end = ALIGN(new_key->offset + datal,
+				      root->sectorsize);
+	int ret;
+	struct btrfs_key key;
+
+	if (new_key->offset > 0)
+		return -EOPNOTSUPP;
+
+	key.objectid = btrfs_ino(dst);
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = 0;
+	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	if (ret < 0) {
+		return ret;
+	} else if (ret > 0) {
+		if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
+			ret = btrfs_next_leaf(root, path);
+			if (ret < 0)
+				return ret;
+			else if (ret > 0)
+				goto copy_inline_extent;
+		}
+		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
+		if (key.objectid == btrfs_ino(dst) &&
+		    key.type == BTRFS_EXTENT_DATA_KEY) {
+			ASSERT(key.offset > 0);
+			return -EOPNOTSUPP;
+		}
+	} else if (i_size_read(dst) <= datal) {
+		struct btrfs_file_extent_item *ei;
+		u64 ext_len;
+
+		/*
+		 * If the file size is <= datal, make sure there are no other
+		 * extents following (can happen do to an fallocate call with
+		 * the flag FALLOC_FL_KEEP_SIZE).
+		 */
+		ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
+				    struct btrfs_file_extent_item);
+		/*
+		 * If it's an inline extent, it can not have other extents
+		 * following it.
+		 */
+		if (btrfs_file_extent_type(path->nodes[0], ei) ==
+		    BTRFS_FILE_EXTENT_INLINE)
+			goto copy_inline_extent;
+
+		ext_len = btrfs_file_extent_num_bytes(path->nodes[0], ei);
+		if (ext_len > aligned_end)
+			return -EOPNOTSUPP;
+
+		ret = btrfs_next_item(root, path);
+		if (ret < 0) {
+			return ret;
+		} else if (ret == 0) {
+			btrfs_item_key_to_cpu(path->nodes[0], &key,
+					      path->slots[0]);
+			if (key.objectid == btrfs_ino(dst) &&
+			    key.type == BTRFS_EXTENT_DATA_KEY)
+				return -EOPNOTSUPP;
+		}
+	}
+
+copy_inline_extent:
+	/*
+	 * We have no extent items, or we have an extent at offset 0 which may
+	 * or may not be inlined. All these cases are dealt the same way.
+	 */
+	if (i_size_read(dst) > datal) {
+		/*
+		 * If the destination inode has an inline extent...
+		 * This would require copying the data from the source inline
+		 * extent into the beginning of the destination's inline extent.
+		 * But this is really complex, both extents can be compressed
+		 * or just one of them, which would require decompressing and
+		 * re-compressing data (which could increase the new compressed
+		 * size, not allowing the compressed data to fit anymore in an
+		 * inline extent).
+		 * So just don't support this case for now (it should be rare,
+		 * we are not really saving space when cloning inline extents).
+		 */
+		return -EOPNOTSUPP;
+	}
+
+	btrfs_release_path(path);
+	ret = btrfs_drop_extents(trans, root, dst, drop_start, aligned_end, 1);
+	if (ret)
+		return ret;
+	ret = btrfs_insert_empty_item(trans, root, path, new_key, size);
+	if (ret)
+		return ret;
+
+	if (skip) {
+		const u32 start = btrfs_file_extent_calc_inline_size(0);
+
+		memmove(inline_data + start, inline_data + start + skip, datal);
+	}
+
+	write_extent_buffer(path->nodes[0], inline_data,
+			    btrfs_item_ptr_offset(path->nodes[0],
+						  path->slots[0]),
+			    size);
+	inode_add_bytes(dst, datal);
+
+	return 0;
+}
+
 /**
  * btrfs_clone() - clone a range from inode file to another
  *
@@ -3352,9 +3464,7 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 	u32 nritems;
 	int slot;
 	int ret;
-	int no_quota;
 	const u64 len = olen_aligned;
-	u64 last_disko = 0;
 	u64 last_dest_end = destoff;
 
 	ret = -ENOMEM;
@@ -3400,7 +3510,6 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 
 		nritems = btrfs_header_nritems(path->nodes[0]);
 process_slot:
-		no_quota = 1;
 		if (path->slots[0] >= nritems) {
 			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
 			if (ret < 0)
@@ -3552,35 +3661,13 @@ process_slot:
 				btrfs_set_file_extent_num_bytes(leaf, extent,
 								datal);
 
-				/*
-				 * We need to look up the roots that point at
-				 * this bytenr and see if the new root does.  If
-				 * it does not we need to make sure we update
-				 * quotas appropriately.
-				 */
-				if (disko && root != BTRFS_I(src)->root &&
-				    disko != last_disko) {
-					no_quota = check_ref(trans, root,
-							     disko);
-					if (no_quota < 0) {
-						btrfs_abort_transaction(trans,
-									root,
-									ret);
-						btrfs_end_transaction(trans,
-								      root);
-						ret = no_quota;
-						goto out;
-					}
-				}
-
 				if (disko) {
 					inode_add_bytes(inode, datal);
 					ret = btrfs_inc_extent_ref(trans, root,
 							disko, diskl, 0,
 							root->root_key.objectid,
 							btrfs_ino(inode),
-							new_key.offset - datao,
-							no_quota);
+							new_key.offset - datao);
 					if (ret) {
 						btrfs_abort_transaction(trans,
 									root,
@@ -3594,21 +3681,6 @@ process_slot:
 			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
 				u64 skip = 0;
 				u64 trim = 0;
-				u64 aligned_end = 0;
-
-				/*
-				 * Don't copy an inline extent into an offset
-				 * greater than zero. Having an inline extent
-				 * at such an offset results in chaos as btrfs
-				 * isn't prepared for such cases. Just skip
-				 * this case for the same reasons as commented
-				 * at btrfs_ioctl_clone().
-				 */
-				if (last_dest_end > 0) {
-					ret = -EOPNOTSUPP;
-					btrfs_end_transaction(trans, root);
-					goto out;
-				}
 
 				if (off > key.offset) {
 					skip = off - key.offset;
@@ -3626,42 +3698,22 @@ process_slot:
 				size -= skip + trim;
 				datal -= skip + trim;
 
-				aligned_end = ALIGN(new_key.offset + datal,
-						    root->sectorsize);
-				ret = btrfs_drop_extents(trans, root, inode,
-							 drop_start,
-							 aligned_end,
-							 1);
+				ret = clone_copy_inline_extent(src, inode,
+							       trans, path,
+							       &new_key,
+							       drop_start,
+							       datal,
+							       skip, size, buf);
 				if (ret) {
 					if (ret != -EOPNOTSUPP)
 						btrfs_abort_transaction(trans,
-							root, ret);
-					btrfs_end_transaction(trans, root);
-					goto out;
-				}
-
-				ret = btrfs_insert_empty_item(trans, root, path,
-							      &new_key, size);
-				if (ret) {
-					btrfs_abort_transaction(trans, root,
-								ret);
+									root,
+									ret);
 					btrfs_end_transaction(trans, root);
 					goto out;
 				}
-
-				if (skip) {
-					u32 start =
-					  btrfs_file_extent_calc_inline_size(0);
-					memmove(buf+start, buf+start+skip,
-						datal);
-				}
-
 				leaf = path->nodes[0];
 				slot = path->slots[0];
-				write_extent_buffer(leaf, buf,
-					    btrfs_item_ptr_offset(leaf, slot),
-					    size);
-				inode_add_bytes(inode, datal);
 			}
 
 			/* If we have an implicit hole (NO_HOLES feature). */
@@ -4814,7 +4866,7 @@ static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
 	/* update qgroup status and info */
 	err = btrfs_run_qgroups(trans, root->fs_info);
 	if (err < 0)
-		btrfs_error(root->fs_info, ret,
+		btrfs_std_error(root->fs_info, ret,
 			    "failed to update qgroup status and info\n");
 	err = btrfs_end_transaction(trans, root);
 	if (err && !ret)

fs/btrfs/locking.c

@@ -79,6 +79,9 @@ void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw)
 		write_lock(&eb->lock);
 		WARN_ON(atomic_read(&eb->spinning_writers));
 		atomic_inc(&eb->spinning_writers);
+		/*
+		 * atomic_dec_and_test implies a barrier for waitqueue_active
+		 */
 		if (atomic_dec_and_test(&eb->blocking_writers) &&
 		    waitqueue_active(&eb->write_lock_wq))
 			wake_up(&eb->write_lock_wq);
@@ -86,6 +89,9 @@ void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw)
 		BUG_ON(atomic_read(&eb->blocking_readers) == 0);
 		read_lock(&eb->lock);
 		atomic_inc(&eb->spinning_readers);
+		/*
+		 * atomic_dec_and_test implies a barrier for waitqueue_active
+		 */
 		if (atomic_dec_and_test(&eb->blocking_readers) &&
 		    waitqueue_active(&eb->read_lock_wq))
 			wake_up(&eb->read_lock_wq);
@@ -229,6 +235,9 @@ void btrfs_tree_read_unlock_blocking(struct extent_buffer *eb)
 	}
 	btrfs_assert_tree_read_locked(eb);
 	WARN_ON(atomic_read(&eb->blocking_readers) == 0);
+	/*
+	 * atomic_dec_and_test implies a barrier for waitqueue_active
+	 */
 	if (atomic_dec_and_test(&eb->blocking_readers) &&
 	    waitqueue_active(&eb->read_lock_wq))
 		wake_up(&eb->read_lock_wq);
@@ -280,6 +289,9 @@ void btrfs_tree_unlock(struct extent_buffer *eb)
 	if (blockers) {
 		WARN_ON(atomic_read(&eb->spinning_writers));
 		atomic_dec(&eb->blocking_writers);
+		/*
+		 * Make sure counter is updated before we wake up waiters.
+		 */
 		smp_mb();
 		if (waitqueue_active(&eb->write_lock_wq))
 			wake_up(&eb->write_lock_wq);

fs/btrfs/ordered-data.c

@@ -345,6 +345,9 @@ int btrfs_dec_test_first_ordered_pending(struct inode *inode,
 
 	if (entry->bytes_left == 0) {
 		ret = test_and_set_bit(BTRFS_ORDERED_IO_DONE, &entry->flags);
+		/*
+		 * Implicit memory barrier after test_and_set_bit
+		 */
 		if (waitqueue_active(&entry->wait))
 			wake_up(&entry->wait);
 	} else {
@@ -409,6 +412,9 @@ have_entry:
 
 	if (entry->bytes_left == 0) {
 		ret = test_and_set_bit(BTRFS_ORDERED_IO_DONE, &entry->flags);
+		/*
+		 * Implicit memory barrier after test_and_set_bit
+		 */
 		if (waitqueue_active(&entry->wait))
 			wake_up(&entry->wait);
 	} else {
@@ -484,15 +490,16 @@ void btrfs_wait_logged_extents(struct btrfs_trans_handle *trans,
 
 	spin_lock_irq(&log->log_extents_lock[index]);
 	while (!list_empty(&log->logged_list[index])) {
+		struct inode *inode;
 		ordered = list_first_entry(&log->logged_list[index],
 					   struct btrfs_ordered_extent,
 					   log_list);
 		list_del_init(&ordered->log_list);
+		inode = ordered->inode;
 		spin_unlock_irq(&log->log_extents_lock[index]);
 
 		if (!test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags) &&
 		    !test_bit(BTRFS_ORDERED_DIRECT, &ordered->flags)) {
-			struct inode *inode = ordered->inode;
 			u64 start = ordered->file_offset;
 			u64 end = ordered->file_offset + ordered->len - 1;
 
@@ -503,20 +510,25 @@ void btrfs_wait_logged_extents(struct btrfs_trans_handle *trans,
 						   &ordered->flags));
 
 		/*
-		 * If our ordered extent completed it means it updated the
-		 * fs/subvol and csum trees already, so no need to make the
-		 * current transaction's commit wait for it, as we end up
-		 * holding memory unnecessarily and delaying the inode's iput
-		 * until the transaction commit (we schedule an iput for the
-		 * inode when the ordered extent's refcount drops to 0), which
-		 * prevents it from being evictable until the transaction
-		 * commits.
+		 * In order to keep us from losing our ordered extent
+		 * information when committing the transaction we have to make
+		 * sure that any logged extents are completed when we go to
+		 * commit the transaction.  To do this we simply increase the
+		 * current transactions pending_ordered counter and decrement it
+		 * when the ordered extent completes.
 		 */
-		if (test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags))
-			btrfs_put_ordered_extent(ordered);
-		else
-			list_add_tail(&ordered->trans_list, &trans->ordered);
-
+		if (!test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags)) {
+			struct btrfs_ordered_inode_tree *tree;
+
+			tree = &BTRFS_I(inode)->ordered_tree;
+			spin_lock_irq(&tree->lock);
+			if (!test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags)) {
+				set_bit(BTRFS_ORDERED_PENDING, &ordered->flags);
+				atomic_inc(&trans->transaction->pending_ordered);
+			}
+			spin_unlock_irq(&tree->lock);
+		}
+		btrfs_put_ordered_extent(ordered);
 		spin_lock_irq(&log->log_extents_lock[index]);
 	}
 	spin_unlock_irq(&log->log_extents_lock[index]);
@@ -578,6 +590,7 @@ void btrfs_remove_ordered_extent(struct inode *inode,
 	struct btrfs_ordered_inode_tree *tree;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct rb_node *node;
+	bool dec_pending_ordered = false;
 
 	tree = &BTRFS_I(inode)->ordered_tree;
 	spin_lock_irq(&tree->lock);
@@ -587,8 +600,37 @@ void btrfs_remove_ordered_extent(struct inode *inode,
 	if (tree->last == node)
 		tree->last = NULL;
 	set_bit(BTRFS_ORDERED_COMPLETE, &entry->flags);
+	if (test_and_clear_bit(BTRFS_ORDERED_PENDING, &entry->flags))
+		dec_pending_ordered = true;
 	spin_unlock_irq(&tree->lock);
 
+	/*
+	 * The current running transaction is waiting on us, we need to let it
+	 * know that we're complete and wake it up.
+	 */
+	if (dec_pending_ordered) {
+		struct btrfs_transaction *trans;
+
+		/*
+		 * The checks for trans are just a formality, it should be set,
+		 * but if it isn't we don't want to deref/assert under the spin
+		 * lock, so be nice and check if trans is set, but ASSERT() so
+		 * if it isn't set a developer will notice.
+		 */
+		spin_lock(&root->fs_info->trans_lock);
+		trans = root->fs_info->running_transaction;
+		if (trans)
+			atomic_inc(&trans->use_count);
+		spin_unlock(&root->fs_info->trans_lock);
+
+		ASSERT(trans);
+		if (trans) {
+			if (atomic_dec_and_test(&trans->pending_ordered))
+				wake_up(&trans->pending_wait);
+			btrfs_put_transaction(trans);
+		}
+	}
+
 	spin_lock(&root->ordered_extent_lock);
 	list_del_init(&entry->root_extent_list);
 	root->nr_ordered_extents--;

fs/btrfs/ordered-data.h

@@ -73,6 +73,8 @@ struct btrfs_ordered_sum {
 
 #define BTRFS_ORDERED_LOGGED 10 /* Set when we've waited on this ordered extent
 				 * in the logging code. */
+#define BTRFS_ORDERED_PENDING 11 /* We are waiting for this ordered extent to
+				  * complete in the current transaction. */
 struct btrfs_ordered_extent {
 	/* logical offset in the file */
 	u64 file_offset;

fs/btrfs/props.c

@@ -49,18 +49,16 @@ static struct prop_handler prop_handlers[] = {
 		.extract = prop_compression_extract,
 		.inheritable = 1
 	},
-	{
-		.xattr_name = NULL
-	}
 };
 
 void __init btrfs_props_init(void)
 {
-	struct prop_handler *p;
+	int i;
 
 	hash_init(prop_handlers_ht);
 
-	for (p = &prop_handlers[0]; p->xattr_name; p++) {
+	for (i = 0; i < ARRAY_SIZE(prop_handlers); i++) {
+		struct prop_handler *p = &prop_handlers[i];
 		u64 h = btrfs_name_hash(p->xattr_name, strlen(p->xattr_name));
 
 		hash_add(prop_handlers_ht, &p->node, h);
@@ -301,15 +299,16 @@ static int inherit_props(struct btrfs_trans_handle *trans,
 			 struct inode *inode,
 			 struct inode *parent)
 {
-	const struct prop_handler *h;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	int ret;
+	int i;
 
 	if (!test_bit(BTRFS_INODE_HAS_PROPS,
 		      &BTRFS_I(parent)->runtime_flags))
 		return 0;
 
-	for (h = &prop_handlers[0]; h->xattr_name; h++) {
+	for (i = 0; i < ARRAY_SIZE(prop_handlers); i++) {
+		const struct prop_handler *h = &prop_handlers[i];
 		const char *value;
 		u64 num_bytes;
 

fs/btrfs/qgroup.c

@@ -1652,10 +1652,6 @@ static int qgroup_update_counters(struct btrfs_fs_info *fs_info,
 			}
 		}
 
-		/* For exclusive extent, free its reserved bytes too */
-		if (nr_old_roots == 0 && nr_new_roots == 1 &&
-		    cur_new_count == nr_new_roots)
-			qg->reserved -= num_bytes;
 		if (dirty)
 			qgroup_dirty(fs_info, qg);
 	}
@@ -2035,7 +2031,7 @@ out:
 	return ret;
 }
 
-int btrfs_qgroup_reserve(struct btrfs_root *root, u64 num_bytes)
+static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes)
 {
 	struct btrfs_root *quota_root;
 	struct btrfs_qgroup *qgroup;
@@ -2116,14 +2112,13 @@ out:
 	return ret;
 }
 
-void btrfs_qgroup_free(struct btrfs_root *root, u64 num_bytes)
+void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info,
+			       u64 ref_root, u64 num_bytes)
 {
 	struct btrfs_root *quota_root;
 	struct btrfs_qgroup *qgroup;
-	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct ulist_node *unode;
 	struct ulist_iterator uiter;
-	u64 ref_root = root->root_key.objectid;
 	int ret = 0;
 
 	if (!is_fstree(ref_root))
@@ -2169,6 +2164,11 @@ out:
 	spin_unlock(&fs_info->qgroup_lock);
 }
 
+static inline void qgroup_free(struct btrfs_root *root, u64 num_bytes)
+{
+	return btrfs_qgroup_free_refroot(root->fs_info, root->objectid,
+					 num_bytes);
+}
 void assert_qgroups_uptodate(struct btrfs_trans_handle *trans)
 {
 	if (list_empty(&trans->qgroup_ref_list) && !trans->delayed_ref_elem.seq)
@@ -2188,10 +2188,10 @@ void assert_qgroups_uptodate(struct btrfs_trans_handle *trans)
  */
 static int
 qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
-		   struct btrfs_trans_handle *trans,
-		   struct extent_buffer *scratch_leaf)
+		   struct btrfs_trans_handle *trans)
 {
 	struct btrfs_key found;
+	struct extent_buffer *scratch_leaf = NULL;
 	struct ulist *roots = NULL;
 	struct seq_list tree_mod_seq_elem = SEQ_LIST_INIT(tree_mod_seq_elem);
 	u64 num_bytes;
@@ -2229,7 +2229,15 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 	fs_info->qgroup_rescan_progress.objectid = found.objectid + 1;
 
 	btrfs_get_tree_mod_seq(fs_info, &tree_mod_seq_elem);
-	memcpy(scratch_leaf, path->nodes[0], sizeof(*scratch_leaf));
+	scratch_leaf = btrfs_clone_extent_buffer(path->nodes[0]);
+	if (!scratch_leaf) {
+		ret = -ENOMEM;
+		mutex_unlock(&fs_info->qgroup_rescan_lock);
+		goto out;
+	}
+	extent_buffer_get(scratch_leaf);
+	btrfs_tree_read_lock(scratch_leaf);
+	btrfs_set_lock_blocking_rw(scratch_leaf, BTRFS_READ_LOCK);
 	slot = path->slots[0];
 	btrfs_release_path(path);
 	mutex_unlock(&fs_info->qgroup_rescan_lock);
@@ -2255,6 +2263,10 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 			goto out;
 	}
 out:
+	if (scratch_leaf) {
+		btrfs_tree_read_unlock_blocking(scratch_leaf);
+		free_extent_buffer(scratch_leaf);
+	}
 	btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem);
 
 	return ret;
@@ -2266,16 +2278,12 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
 						     qgroup_rescan_work);
 	struct btrfs_path *path;
 	struct btrfs_trans_handle *trans = NULL;
-	struct extent_buffer *scratch_leaf = NULL;
 	int err = -ENOMEM;
 	int ret = 0;
 
 	path = btrfs_alloc_path();
 	if (!path)
 		goto out;
-	scratch_leaf = kmalloc(sizeof(*scratch_leaf), GFP_NOFS);
-	if (!scratch_leaf)
-		goto out;
 
 	err = 0;
 	while (!err) {
@@ -2287,8 +2295,7 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
 		if (!fs_info->quota_enabled) {
 			err = -EINTR;
 		} else {
-			err = qgroup_rescan_leaf(fs_info, path, trans,
-						 scratch_leaf);
+			err = qgroup_rescan_leaf(fs_info, path, trans);
 		}
 		if (err > 0)
 			btrfs_commit_transaction(trans, fs_info->fs_root);
@@ -2297,7 +2304,6 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
 	}
 
 out:
-	kfree(scratch_leaf);
 	btrfs_free_path(path);
 
 	mutex_lock(&fs_info->qgroup_rescan_lock);
@@ -2486,3 +2492,190 @@ btrfs_qgroup_rescan_resume(struct btrfs_fs_info *fs_info)
 		btrfs_queue_work(fs_info->qgroup_rescan_workers,
 				 &fs_info->qgroup_rescan_work);
 }
+
+/*
+ * Reserve qgroup space for range [start, start + len).
+ *
+ * This function will either reserve space from related qgroups or doing
+ * nothing if the range is already reserved.
+ *
+ * Return 0 for successful reserve
+ * Return <0 for error (including -EQUOT)
+ *
+ * NOTE: this function may sleep for memory allocation.
+ */
+int btrfs_qgroup_reserve_data(struct inode *inode, u64 start, u64 len)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct extent_changeset changeset;
+	struct ulist_node *unode;
+	struct ulist_iterator uiter;
+	int ret;
+
+	if (!root->fs_info->quota_enabled || !is_fstree(root->objectid) ||
+	    len == 0)
+		return 0;
+
+	changeset.bytes_changed = 0;
+	changeset.range_changed = ulist_alloc(GFP_NOFS);
+	ret = set_record_extent_bits(&BTRFS_I(inode)->io_tree, start,
+			start + len -1, EXTENT_QGROUP_RESERVED, GFP_NOFS,
+			&changeset);
+	trace_btrfs_qgroup_reserve_data(inode, start, len,
+					changeset.bytes_changed,
+					QGROUP_RESERVE);
+	if (ret < 0)
+		goto cleanup;
+	ret = qgroup_reserve(root, changeset.bytes_changed);
+	if (ret < 0)
+		goto cleanup;
+
+	ulist_free(changeset.range_changed);
+	return ret;
+
+cleanup:
+	/* cleanup already reserved ranges */
+	ULIST_ITER_INIT(&uiter);
+	while ((unode = ulist_next(changeset.range_changed, &uiter)))
+		clear_extent_bit(&BTRFS_I(inode)->io_tree, unode->val,
+				 unode->aux, EXTENT_QGROUP_RESERVED, 0, 0, NULL,
+				 GFP_NOFS);
+	ulist_free(changeset.range_changed);
+	return ret;
+}
+
+static int __btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len,
+				       int free)
+{
+	struct extent_changeset changeset;
+	int trace_op = QGROUP_RELEASE;
+	int ret;
+
+	changeset.bytes_changed = 0;
+	changeset.range_changed = ulist_alloc(GFP_NOFS);
+	if (!changeset.range_changed)
+		return -ENOMEM;
+
+	ret = clear_record_extent_bits(&BTRFS_I(inode)->io_tree, start, 
+			start + len -1, EXTENT_QGROUP_RESERVED, GFP_NOFS,
+			&changeset);
+	if (ret < 0)
+		goto out;
+
+	if (free) {
+		qgroup_free(BTRFS_I(inode)->root, changeset.bytes_changed);
+		trace_op = QGROUP_FREE;
+	}
+	trace_btrfs_qgroup_release_data(inode, start, len,
+					changeset.bytes_changed, trace_op);
+out:
+	ulist_free(changeset.range_changed);
+	return ret;
+}
+
+/*
+ * Free a reserved space range from io_tree and related qgroups
+ *
+ * Should be called when a range of pages get invalidated before reaching disk.
+ * Or for error cleanup case.
+ *
+ * For data written to disk, use btrfs_qgroup_release_data().
+ *
+ * NOTE: This function may sleep for memory allocation.
+ */
+int btrfs_qgroup_free_data(struct inode *inode, u64 start, u64 len)
+{
+	return __btrfs_qgroup_release_data(inode, start, len, 1);
+}
+
+/*
+ * Release a reserved space range from io_tree only.
+ *
+ * Should be called when a range of pages get written to disk and corresponding
+ * FILE_EXTENT is inserted into corresponding root.
+ *
+ * Since new qgroup accounting framework will only update qgroup numbers at
+ * commit_transaction() time, its reserved space shouldn't be freed from
+ * related qgroups.
+ *
+ * But we should release the range from io_tree, to allow further write to be
+ * COWed.
+ *
+ * NOTE: This function may sleep for memory allocation.
+ */
+int btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len)
+{
+	return __btrfs_qgroup_release_data(inode, start, len, 0);
+}
+
+int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes)
+{
+	int ret;
+
+	if (!root->fs_info->quota_enabled || !is_fstree(root->objectid) ||
+	    num_bytes == 0)
+		return 0;
+
+	BUG_ON(num_bytes != round_down(num_bytes, root->nodesize));
+	ret = qgroup_reserve(root, num_bytes);
+	if (ret < 0)
+		return ret;
+	atomic_add(num_bytes, &root->qgroup_meta_rsv);
+	return ret;
+}
+
+void btrfs_qgroup_free_meta_all(struct btrfs_root *root)
+{
+	int reserved;
+
+	if (!root->fs_info->quota_enabled || !is_fstree(root->objectid))
+		return;
+
+	reserved = atomic_xchg(&root->qgroup_meta_rsv, 0);
+	if (reserved == 0)
+		return;
+	qgroup_free(root, reserved);
+}
+
+void btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes)
+{
+	if (!root->fs_info->quota_enabled || !is_fstree(root->objectid))
+		return;
+
+	BUG_ON(num_bytes != round_down(num_bytes, root->nodesize));
+	WARN_ON(atomic_read(&root->qgroup_meta_rsv) < num_bytes);
+	atomic_sub(num_bytes, &root->qgroup_meta_rsv);
+	qgroup_free(root, num_bytes);
+}
+
+/*
+ * Check qgroup reserved space leaking, normally at destory inode
+ * time
+ */
+void btrfs_qgroup_check_reserved_leak(struct inode *inode)
+{
+	struct extent_changeset changeset;
+	struct ulist_node *unode;
+	struct ulist_iterator iter;
+	int ret;
+
+	changeset.bytes_changed = 0;
+	changeset.range_changed = ulist_alloc(GFP_NOFS);
+	if (WARN_ON(!changeset.range_changed))
+		return;
+
+	ret = clear_record_extent_bits(&BTRFS_I(inode)->io_tree, 0, (u64)-1,
+			EXTENT_QGROUP_RESERVED, GFP_NOFS, &changeset);
+
+	WARN_ON(ret < 0);
+	if (WARN_ON(changeset.bytes_changed)) {
+		ULIST_ITER_INIT(&iter);
+		while ((unode = ulist_next(changeset.range_changed, &iter))) {
+			btrfs_warn(BTRFS_I(inode)->root->fs_info,
+				"leaking qgroup reserved space, ino: %lu, start: %llu, end: %llu",
+				inode->i_ino, unode->val, unode->aux);
+		}
+		qgroup_free(BTRFS_I(inode)->root, changeset.bytes_changed);
+	}
+	ulist_free(changeset.range_changed);
+}

fs/btrfs/qgroup.h

@@ -33,6 +33,13 @@ struct btrfs_qgroup_extent_record {
 	struct ulist *old_roots;
 };
 
+/*
+ * For qgroup event trace points only
+ */
+#define QGROUP_RESERVE		(1<<0)
+#define QGROUP_RELEASE		(1<<1)
+#define QGROUP_FREE		(1<<2)
+
 int btrfs_quota_enable(struct btrfs_trans_handle *trans,
 		       struct btrfs_fs_info *fs_info);
 int btrfs_quota_disable(struct btrfs_trans_handle *trans,
@@ -71,9 +78,18 @@ int btrfs_run_qgroups(struct btrfs_trans_handle *trans,
 int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans,
 			 struct btrfs_fs_info *fs_info, u64 srcid, u64 objectid,
 			 struct btrfs_qgroup_inherit *inherit);
-int btrfs_qgroup_reserve(struct btrfs_root *root, u64 num_bytes);
-void btrfs_qgroup_free(struct btrfs_root *root, u64 num_bytes);
-
+void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info,
+			       u64 ref_root, u64 num_bytes);
+/*
+ * TODO: Add proper trace point for it, as btrfs_qgroup_free() is
+ * called by everywhere, can't provide good trace for delayed ref case.
+ */
+static inline void btrfs_qgroup_free_delayed_ref(struct btrfs_fs_info *fs_info,
+						 u64 ref_root, u64 num_bytes)
+{
+	btrfs_qgroup_free_refroot(fs_info, ref_root, num_bytes);
+	trace_btrfs_qgroup_free_delayed_ref(ref_root, num_bytes);
+}
 void assert_qgroups_uptodate(struct btrfs_trans_handle *trans);
 
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
@@ -81,4 +97,13 @@ int btrfs_verify_qgroup_counts(struct btrfs_fs_info *fs_info, u64 qgroupid,
 			       u64 rfer, u64 excl);
 #endif
 
+/* New io_tree based accurate qgroup reserve API */
+int btrfs_qgroup_reserve_data(struct inode *inode, u64 start, u64 len);
+int btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len);
+int btrfs_qgroup_free_data(struct inode *inode, u64 start, u64 len);
+
+int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes);
+void btrfs_qgroup_free_meta_all(struct btrfs_root *root);
+void btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes);
+void btrfs_qgroup_check_reserved_leak(struct inode *inode);
 #endif /* __BTRFS_QGROUP__ */

fs/btrfs/raid56.c

@@ -810,7 +810,11 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio)
 			}
 
 			goto done_nolock;
-		} else  if (waitqueue_active(&h->wait)) {
+			/*
+			 * The barrier for this waitqueue_active is not needed,
+			 * we're protected by h->lock and can't miss a wakeup.
+			 */
+		} else if (waitqueue_active(&h->wait)) {
 			spin_unlock(&rbio->bio_list_lock);
 			spin_unlock_irqrestore(&h->lock, flags);
 			wake_up(&h->wait);

fs/btrfs/reada.c

@@ -569,7 +569,7 @@ static int reada_add_block(struct reada_control *rc, u64 logical,
 	rec = kzalloc(sizeof(*rec), GFP_NOFS);
 	if (!rec) {
 		reada_extent_put(root->fs_info, re);
-		return -1;
+		return -ENOMEM;
 	}
 
 	rec->rc = rc;
@@ -918,6 +918,7 @@ struct reada_control *btrfs_reada_add(struct btrfs_root *root,
 	u64 start;
 	u64 generation;
 	int level;
+	int ret;
 	struct extent_buffer *node;
 	static struct btrfs_key max_key = {
 		.objectid = (u64)-1,
@@ -943,9 +944,10 @@ struct reada_control *btrfs_reada_add(struct btrfs_root *root,
 	generation = btrfs_header_generation(node);
 	free_extent_buffer(node);
 
-	if (reada_add_block(rc, start, &max_key, level, generation)) {
+	ret = reada_add_block(rc, start, &max_key, level, generation);
+	if (ret) {
 		kfree(rc);
-		return ERR_PTR(-ENOMEM);
+		return ERR_PTR(ret);
 	}
 
 	reada_start_machine(root->fs_info);

fs/btrfs/relocation.c

@@ -1716,7 +1716,7 @@ int replace_file_extents(struct btrfs_trans_handle *trans,
 		ret = btrfs_inc_extent_ref(trans, root, new_bytenr,
 					   num_bytes, parent,
 					   btrfs_header_owner(leaf),
-					   key.objectid, key.offset, 1);
+					   key.objectid, key.offset);
 		if (ret) {
 			btrfs_abort_transaction(trans, root, ret);
 			break;
@@ -1724,7 +1724,7 @@ int replace_file_extents(struct btrfs_trans_handle *trans,
 
 		ret = btrfs_free_extent(trans, root, bytenr, num_bytes,
 					parent, btrfs_header_owner(leaf),
-					key.objectid, key.offset, 1);
+					key.objectid, key.offset);
 		if (ret) {
 			btrfs_abort_transaction(trans, root, ret);
 			break;
@@ -1900,23 +1900,21 @@ again:
 
 		ret = btrfs_inc_extent_ref(trans, src, old_bytenr, blocksize,
 					path->nodes[level]->start,
-					src->root_key.objectid, level - 1, 0,
-					1);
+					src->root_key.objectid, level - 1, 0);
 		BUG_ON(ret);
 		ret = btrfs_inc_extent_ref(trans, dest, new_bytenr, blocksize,
 					0, dest->root_key.objectid, level - 1,
-					0, 1);
+					0);
 		BUG_ON(ret);
 
 		ret = btrfs_free_extent(trans, src, new_bytenr, blocksize,
 					path->nodes[level]->start,
-					src->root_key.objectid, level - 1, 0,
-					1);
+					src->root_key.objectid, level - 1, 0);
 		BUG_ON(ret);
 
 		ret = btrfs_free_extent(trans, dest, old_bytenr, blocksize,
 					0, dest->root_key.objectid, level - 1,
-					0, 1);
+					0);
 		BUG_ON(ret);
 
 		btrfs_unlock_up_safe(path, 0);
@@ -2418,7 +2416,7 @@ again:
 	}
 out:
 	if (ret) {
-		btrfs_std_error(root->fs_info, ret);
+		btrfs_std_error(root->fs_info, ret, NULL);
 		if (!list_empty(&reloc_roots))
 			free_reloc_roots(&reloc_roots);
 
@@ -2745,7 +2743,7 @@ static int do_relocation(struct btrfs_trans_handle *trans,
 						node->eb->start, blocksize,
 						upper->eb->start,
 						btrfs_header_owner(upper->eb),
-						node->level, 0, 1);
+						node->level, 0);
 			BUG_ON(ret);
 
 			ret = btrfs_drop_subtree(trans, root, eb, upper->eb);
@@ -3034,8 +3032,8 @@ int prealloc_file_extent_cluster(struct inode *inode,
 	BUG_ON(cluster->start != cluster->boundary[0]);
 	mutex_lock(&inode->i_mutex);
 
-	ret = btrfs_check_data_free_space(inode, cluster->end +
-					  1 - cluster->start, 0);
+	ret = btrfs_check_data_free_space(inode, cluster->start,
+					  cluster->end + 1 - cluster->start);
 	if (ret)
 		goto out;
 
@@ -3056,8 +3054,8 @@ int prealloc_file_extent_cluster(struct inode *inode,
 			break;
 		nr++;
 	}
-	btrfs_free_reserved_data_space(inode, cluster->end +
-				       1 - cluster->start);
+	btrfs_free_reserved_data_space(inode, cluster->start,
+				       cluster->end + 1 - cluster->start);
 out:
 	mutex_unlock(&inode->i_mutex);
 	return ret;

fs/btrfs/root-tree.c

@@ -45,12 +45,13 @@ static void btrfs_read_root_item(struct extent_buffer *eb, int slot,
 	if (!need_reset && btrfs_root_generation(item)
 		!= btrfs_root_generation_v2(item)) {
 		if (btrfs_root_generation_v2(item) != 0) {
-			printk(KERN_WARNING "BTRFS: mismatching "
+			btrfs_warn(eb->fs_info,
+					"mismatching "
 					"generation and generation_v2 "
 					"found in root item. This root "
 					"was probably mounted with an "
 					"older kernel. Resetting all "
-					"new fields.\n");
+					"new fields.");
 		}
 		need_reset = 1;
 	}
@@ -141,7 +142,7 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root
 	int ret;
 	int slot;
 	unsigned long ptr;
-	int old_len;
+	u32 old_len;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -283,7 +284,7 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root)
 			trans = btrfs_join_transaction(tree_root);
 			if (IS_ERR(trans)) {
 				err = PTR_ERR(trans);
-				btrfs_error(tree_root->fs_info, err,
+				btrfs_std_error(tree_root->fs_info, err,
 					    "Failed to start trans to delete "
 					    "orphan item");
 				break;
@@ -292,7 +293,7 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root)
 						    root_key.objectid);
 			btrfs_end_transaction(trans, tree_root);
 			if (err) {
-				btrfs_error(tree_root->fs_info, err,
+				btrfs_std_error(tree_root->fs_info, err,
 					    "Failed to delete root orphan "
 					    "item");
 				break;

fs/btrfs/scrub.c

@@ -580,9 +580,9 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root,
 	 * hold all of the paths here
 	 */
 	for (i = 0; i < ipath->fspath->elem_cnt; ++i)
-		printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev "
+		btrfs_warn_in_rcu(fs_info, "%s at logical %llu on dev "
 			"%s, sector %llu, root %llu, inode %llu, offset %llu, "
-			"length %llu, links %u (path: %s)\n", swarn->errstr,
+			"length %llu, links %u (path: %s)", swarn->errstr,
 			swarn->logical, rcu_str_deref(swarn->dev->name),
 			(unsigned long long)swarn->sector, root, inum, offset,
 			min(isize - offset, (u64)PAGE_SIZE), nlink,
@@ -592,9 +592,9 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root,
 	return 0;
 
 err:
-	printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev "
+	btrfs_warn_in_rcu(fs_info, "%s at logical %llu on dev "
 		"%s, sector %llu, root %llu, inode %llu, offset %llu: path "
-		"resolving failed with ret=%d\n", swarn->errstr,
+		"resolving failed with ret=%d", swarn->errstr,
 		swarn->logical, rcu_str_deref(swarn->dev->name),
 		(unsigned long long)swarn->sector, root, inum, offset, ret);
 
@@ -649,10 +649,10 @@ static void scrub_print_warning(const char *errstr, struct scrub_block *sblock)
 			ret = tree_backref_for_extent(&ptr, eb, &found_key, ei,
 						      item_size, &ref_root,
 						      &ref_level);
-			printk_in_rcu(KERN_WARNING
-				"BTRFS: %s at logical %llu on dev %s, "
+			btrfs_warn_in_rcu(fs_info,
+				"%s at logical %llu on dev %s, "
 				"sector %llu: metadata %s (level %d) in tree "
-				"%llu\n", errstr, swarn.logical,
+				"%llu", errstr, swarn.logical,
 				rcu_str_deref(dev->name),
 				(unsigned long long)swarn.sector,
 				ref_level ? "node" : "leaf",
@@ -850,8 +850,8 @@ out:
 		btrfs_dev_replace_stats_inc(
 			&sctx->dev_root->fs_info->dev_replace.
 			num_uncorrectable_read_errors);
-		printk_ratelimited_in_rcu(KERN_ERR "BTRFS: "
-		    "unable to fixup (nodatasum) error at logical %llu on dev %s\n",
+		btrfs_err_rl_in_rcu(sctx->dev_root->fs_info,
+		    "unable to fixup (nodatasum) error at logical %llu on dev %s",
 			fixup->logical, rcu_str_deref(fixup->dev->name));
 	}
 
@@ -1230,8 +1230,8 @@ corrected_error:
 			sctx->stat.corrected_errors++;
 			sblock_to_check->data_corrected = 1;
 			spin_unlock(&sctx->stat_lock);
-			printk_ratelimited_in_rcu(KERN_ERR
-				"BTRFS: fixed up error at logical %llu on dev %s\n",
+			btrfs_err_rl_in_rcu(fs_info,
+				"fixed up error at logical %llu on dev %s",
 				logical, rcu_str_deref(dev->name));
 		}
 	} else {
@@ -1239,8 +1239,8 @@ did_not_correct_error:
 		spin_lock(&sctx->stat_lock);
 		sctx->stat.uncorrectable_errors++;
 		spin_unlock(&sctx->stat_lock);
-		printk_ratelimited_in_rcu(KERN_ERR
-			"BTRFS: unable to fixup (regular) error at logical %llu on dev %s\n",
+		btrfs_err_rl_in_rcu(fs_info,
+			"unable to fixup (regular) error at logical %llu on dev %s",
 			logical, rcu_str_deref(dev->name));
 	}
 
@@ -1626,9 +1626,9 @@ static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad,
 		int ret;
 
 		if (!page_bad->dev->bdev) {
-			printk_ratelimited(KERN_WARNING "BTRFS: "
+			btrfs_warn_rl(sblock_bad->sctx->dev_root->fs_info,
 				"scrub_repair_page_from_good_copy(bdev == NULL) "
-				"is unexpected!\n");
+				"is unexpected");
 			return -EIO;
 		}
 
@@ -2201,15 +2201,15 @@ static void scrub_missing_raid56_worker(struct btrfs_work *work)
 		spin_lock(&sctx->stat_lock);
 		sctx->stat.read_errors++;
 		spin_unlock(&sctx->stat_lock);
-		printk_ratelimited_in_rcu(KERN_ERR
-			"BTRFS: I/O error rebulding logical %llu for dev %s\n",
+		btrfs_err_rl_in_rcu(fs_info,
+			"IO error rebuilding logical %llu for dev %s",
 			logical, rcu_str_deref(dev->name));
 	} else if (sblock->header_error || sblock->checksum_error) {
 		spin_lock(&sctx->stat_lock);
 		sctx->stat.uncorrectable_errors++;
 		spin_unlock(&sctx->stat_lock);
-		printk_ratelimited_in_rcu(KERN_ERR
-			"BTRFS: failed to rebuild valid logical %llu for dev %s\n",
+		btrfs_err_rl_in_rcu(fs_info,
+			"failed to rebuild valid logical %llu for dev %s",
 			logical, rcu_str_deref(dev->name));
 	} else {
 		scrub_write_block_to_dev_replace(sblock);
@@ -4375,8 +4375,8 @@ static int write_page_nocow(struct scrub_ctx *sctx,
 	if (!dev)
 		return -EIO;
 	if (!dev->bdev) {
-		printk_ratelimited(KERN_WARNING
-			"BTRFS: scrub write_page_nocow(bdev == NULL) is unexpected!\n");
+		btrfs_warn_rl(dev->dev_root->fs_info,
+			"scrub write_page_nocow(bdev == NULL) is unexpected");
 		return -EIO;
 	}
 	bio = btrfs_io_bio_alloc(GFP_NOFS, 1);

fs/btrfs/send.c

@@ -1434,16 +1434,6 @@ verbose_printk(KERN_DEBUG "btrfs: find_extent_clone: data_offset=%llu, "
 	}
 
 	if (cur_clone_root) {
-		if (compressed != BTRFS_COMPRESS_NONE) {
-			/*
-			 * Offsets given by iterate_extent_inodes() are relative
-			 * to the start of the extent, we need to add logical
-			 * offset from the file extent item.
-			 * (See why at backref.c:check_extent_in_eb())
-			 */
-			cur_clone_root->offset += btrfs_file_extent_offset(eb,
-									   fi);
-		}
 		*found = cur_clone_root;
 		ret = 0;
 	} else {
@@ -2353,8 +2343,14 @@ static int send_subvol_begin(struct send_ctx *sctx)
 	}
 
 	TLV_PUT_STRING(sctx, BTRFS_SEND_A_PATH, name, namelen);
-	TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID,
-			sctx->send_root->root_item.uuid);
+
+	if (!btrfs_is_empty_uuid(sctx->send_root->root_item.received_uuid))
+		TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID,
+			    sctx->send_root->root_item.received_uuid);
+	else
+		TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID,
+			    sctx->send_root->root_item.uuid);
+
 	TLV_PUT_U64(sctx, BTRFS_SEND_A_CTRANSID,
 		    le64_to_cpu(sctx->send_root->root_item.ctransid));
 	if (parent_root) {
@@ -2564,7 +2560,7 @@ verbose_printk("btrfs: send_create_inode %llu\n", ino);
 	} else if (S_ISSOCK(mode)) {
 		cmd = BTRFS_SEND_C_MKSOCK;
 	} else {
-		printk(KERN_WARNING "btrfs: unexpected inode type %o",
+		btrfs_warn(sctx->send_root->fs_info, "unexpected inode type %o",
 				(int)(mode & S_IFMT));
 		ret = -ENOTSUPP;
 		goto out;
@@ -4687,6 +4683,171 @@ tlv_put_failure:
 	return ret;
 }
 
+static int send_extent_data(struct send_ctx *sctx,
+			    const u64 offset,
+			    const u64 len)
+{
+	u64 sent = 0;
+
+	if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA)
+		return send_update_extent(sctx, offset, len);
+
+	while (sent < len) {
+		u64 size = len - sent;
+		int ret;
+
+		if (size > BTRFS_SEND_READ_SIZE)
+			size = BTRFS_SEND_READ_SIZE;
+		ret = send_write(sctx, offset + sent, size);
+		if (ret < 0)
+			return ret;
+		if (!ret)
+			break;
+		sent += ret;
+	}
+	return 0;
+}
+
+static int clone_range(struct send_ctx *sctx,
+		       struct clone_root *clone_root,
+		       const u64 disk_byte,
+		       u64 data_offset,
+		       u64 offset,
+		       u64 len)
+{
+	struct btrfs_path *path;
+	struct btrfs_key key;
+	int ret;
+
+	path = alloc_path_for_send();
+	if (!path)
+		return -ENOMEM;
+
+	/*
+	 * We can't send a clone operation for the entire range if we find
+	 * extent items in the respective range in the source file that
+	 * refer to different extents or if we find holes.
+	 * So check for that and do a mix of clone and regular write/copy
+	 * operations if needed.
+	 *
+	 * Example:
+	 *
+	 * mkfs.btrfs -f /dev/sda
+	 * mount /dev/sda /mnt
+	 * xfs_io -f -c "pwrite -S 0xaa 0K 100K" /mnt/foo
+	 * cp --reflink=always /mnt/foo /mnt/bar
+	 * xfs_io -c "pwrite -S 0xbb 50K 50K" /mnt/foo
+	 * btrfs subvolume snapshot -r /mnt /mnt/snap
+	 *
+	 * If when we send the snapshot and we are processing file bar (which
+	 * has a higher inode number than foo) we blindly send a clone operation
+	 * for the [0, 100K[ range from foo to bar, the receiver ends up getting
+	 * a file bar that matches the content of file foo - iow, doesn't match
+	 * the content from bar in the original filesystem.
+	 */
+	key.objectid = clone_root->ino;
+	key.type = BTRFS_EXTENT_DATA_KEY;
+	key.offset = clone_root->offset;
+	ret = btrfs_search_slot(NULL, clone_root->root, &key, path, 0, 0);
+	if (ret < 0)
+		goto out;
+	if (ret > 0 && path->slots[0] > 0) {
+		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0] - 1);
+		if (key.objectid == clone_root->ino &&
+		    key.type == BTRFS_EXTENT_DATA_KEY)
+			path->slots[0]--;
+	}
+
+	while (true) {
+		struct extent_buffer *leaf = path->nodes[0];
+		int slot = path->slots[0];
+		struct btrfs_file_extent_item *ei;
+		u8 type;
+		u64 ext_len;
+		u64 clone_len;
+
+		if (slot >= btrfs_header_nritems(leaf)) {
+			ret = btrfs_next_leaf(clone_root->root, path);
+			if (ret < 0)
+				goto out;
+			else if (ret > 0)
+				break;
+			continue;
+		}
+
+		btrfs_item_key_to_cpu(leaf, &key, slot);
+
+		/*
+		 * We might have an implicit trailing hole (NO_HOLES feature
+		 * enabled). We deal with it after leaving this loop.
+		 */
+		if (key.objectid != clone_root->ino ||
+		    key.type != BTRFS_EXTENT_DATA_KEY)
+			break;
+
+		ei = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
+		type = btrfs_file_extent_type(leaf, ei);
+		if (type == BTRFS_FILE_EXTENT_INLINE) {
+			ext_len = btrfs_file_extent_inline_len(leaf, slot, ei);
+			ext_len = PAGE_CACHE_ALIGN(ext_len);
+		} else {
+			ext_len = btrfs_file_extent_num_bytes(leaf, ei);
+		}
+
+		if (key.offset + ext_len <= clone_root->offset)
+			goto next;
+
+		if (key.offset > clone_root->offset) {
+			/* Implicit hole, NO_HOLES feature enabled. */
+			u64 hole_len = key.offset - clone_root->offset;
+
+			if (hole_len > len)
+				hole_len = len;
+			ret = send_extent_data(sctx, offset, hole_len);
+			if (ret < 0)
+				goto out;
+
+			len -= hole_len;
+			if (len == 0)
+				break;
+			offset += hole_len;
+			clone_root->offset += hole_len;
+			data_offset += hole_len;
+		}
+
+		if (key.offset >= clone_root->offset + len)
+			break;
+
+		clone_len = min_t(u64, ext_len, len);
+
+		if (btrfs_file_extent_disk_bytenr(leaf, ei) == disk_byte &&
+		    btrfs_file_extent_offset(leaf, ei) == data_offset)
+			ret = send_clone(sctx, offset, clone_len, clone_root);
+		else
+			ret = send_extent_data(sctx, offset, clone_len);
+
+		if (ret < 0)
+			goto out;
+
+		len -= clone_len;
+		if (len == 0)
+			break;
+		offset += clone_len;
+		clone_root->offset += clone_len;
+		data_offset += clone_len;
+next:
+		path->slots[0]++;
+	}
+
+	if (len > 0)
+		ret = send_extent_data(sctx, offset, len);
+	else
+		ret = 0;
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
 static int send_write_or_clone(struct send_ctx *sctx,
 			       struct btrfs_path *path,
 			       struct btrfs_key *key,
@@ -4695,9 +4856,7 @@ static int send_write_or_clone(struct send_ctx *sctx,
 	int ret = 0;
 	struct btrfs_file_extent_item *ei;
 	u64 offset = key->offset;
-	u64 pos = 0;
 	u64 len;
-	u32 l;
 	u8 type;
 	u64 bs = sctx->send_root->fs_info->sb->s_blocksize;
 
@@ -4725,22 +4884,15 @@ static int send_write_or_clone(struct send_ctx *sctx,
 	}
 
 	if (clone_root && IS_ALIGNED(offset + len, bs)) {
-		ret = send_clone(sctx, offset, len, clone_root);
-	} else if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA) {
-		ret = send_update_extent(sctx, offset, len);
+		u64 disk_byte;
+		u64 data_offset;
+
+		disk_byte = btrfs_file_extent_disk_bytenr(path->nodes[0], ei);
+		data_offset = btrfs_file_extent_offset(path->nodes[0], ei);
+		ret = clone_range(sctx, clone_root, disk_byte, data_offset,
+				  offset, len);
 	} else {
-		while (pos < len) {
-			l = len - pos;
-			if (l > BTRFS_SEND_READ_SIZE)
-				l = BTRFS_SEND_READ_SIZE;
-			ret = send_write(sctx, pos + offset, l);
-			if (ret < 0)
-				goto out;
-			if (!ret)
-				break;
-			pos += ret;
-		}
-		ret = 0;
+		ret = send_extent_data(sctx, offset, len);
 	}
 out:
 	return ret;

fs/btrfs/super.c

@@ -130,7 +130,6 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info)
 	}
 }
 
-#ifdef CONFIG_PRINTK
 /*
  * __btrfs_std_error decodes expected errors from the caller and
  * invokes the approciate error response.
@@ -140,7 +139,9 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		       unsigned int line, int errno, const char *fmt, ...)
 {
 	struct super_block *sb = fs_info->sb;
+#ifdef CONFIG_PRINTK
 	const char *errstr;
+#endif
 
 	/*
 	 * Special case: if the error is EROFS, and we're already
@@ -149,6 +150,7 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 	if (errno == -EROFS && (sb->s_flags & MS_RDONLY))
   		return;
 
+#ifdef CONFIG_PRINTK
 	errstr = btrfs_decode_error(errno);
 	if (fmt) {
 		struct va_format vaf;
@@ -166,6 +168,7 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		printk(KERN_CRIT "BTRFS: error (device %s) in %s:%d: errno=%d %s\n",
 			sb->s_id, function, line, errno, errstr);
 	}
+#endif
 
 	/* Don't go through full error handling during mount */
 	save_error_info(fs_info);
@@ -173,6 +176,7 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		btrfs_handle_error(fs_info);
 }
 
+#ifdef CONFIG_PRINTK
 static const char * const logtypes[] = {
 	"emergency",
 	"alert",
@@ -212,27 +216,6 @@ void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...)
 
 	va_end(args);
 }
-
-#else
-
-void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
-		       unsigned int line, int errno, const char *fmt, ...)
-{
-	struct super_block *sb = fs_info->sb;
-
-	/*
-	 * Special case: if the error is EROFS, and we're already
-	 * under MS_RDONLY, then it is safe here.
-	 */
-	if (errno == -EROFS && (sb->s_flags & MS_RDONLY))
-		return;
-
-	/* Don't go through full error handling during mount */
-	if (sb->s_flags & MS_BORN) {
-		save_error_info(fs_info);
-		btrfs_handle_error(fs_info);
-	}
-}
 #endif
 
 /*
@@ -320,6 +303,9 @@ enum {
 	Opt_commit_interval, Opt_barrier, Opt_nodefrag, Opt_nodiscard,
 	Opt_noenospc_debug, Opt_noflushoncommit, Opt_acl, Opt_datacow,
 	Opt_datasum, Opt_treelog, Opt_noinode_cache,
+#ifdef CONFIG_BTRFS_DEBUG
+	Opt_fragment_data, Opt_fragment_metadata, Opt_fragment_all,
+#endif
 	Opt_err,
 };
 
@@ -372,6 +358,11 @@ static match_table_t tokens = {
 	{Opt_rescan_uuid_tree, "rescan_uuid_tree"},
 	{Opt_fatal_errors, "fatal_errors=%s"},
 	{Opt_commit_interval, "commit=%d"},
+#ifdef CONFIG_BTRFS_DEBUG
+	{Opt_fragment_data, "fragment=data"},
+	{Opt_fragment_metadata, "fragment=metadata"},
+	{Opt_fragment_all, "fragment=all"},
+#endif
 	{Opt_err, NULL},
 };
 
@@ -738,6 +729,22 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
 			}
 			break;
+#ifdef CONFIG_BTRFS_DEBUG
+		case Opt_fragment_all:
+			btrfs_info(root->fs_info, "fragmenting all space");
+			btrfs_set_opt(info->mount_opt, FRAGMENT_DATA);
+			btrfs_set_opt(info->mount_opt, FRAGMENT_METADATA);
+			break;
+		case Opt_fragment_metadata:
+			btrfs_info(root->fs_info, "fragmenting metadata");
+			btrfs_set_opt(info->mount_opt,
+				      FRAGMENT_METADATA);
+			break;
+		case Opt_fragment_data:
+			btrfs_info(root->fs_info, "fragmenting data");
+			btrfs_set_opt(info->mount_opt, FRAGMENT_DATA);
+			break;
+#endif
 		case Opt_err:
 			btrfs_info(root->fs_info, "unrecognized mount option '%s'", p);
 			ret = -EINVAL;
@@ -1189,6 +1196,12 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",fatal_errors=panic");
 	if (info->commit_interval != BTRFS_DEFAULT_COMMIT_INTERVAL)
 		seq_printf(seq, ",commit=%d", info->commit_interval);
+#ifdef CONFIG_BTRFS_DEBUG
+	if (btrfs_test_opt(root, FRAGMENT_DATA))
+		seq_puts(seq, ",fragment=data");
+	if (btrfs_test_opt(root, FRAGMENT_METADATA))
+		seq_puts(seq, ",fragment=metadata");
+#endif
 	seq_printf(seq, ",subvolid=%llu",
 		  BTRFS_I(d_inode(dentry))->root->root_key.objectid);
 	seq_puts(seq, ",subvol=");

fs/btrfs/sysfs.c

@@ -437,24 +437,24 @@ static const struct attribute *btrfs_attrs[] = {
 	NULL,
 };
 
-static void btrfs_release_super_kobj(struct kobject *kobj)
+static void btrfs_release_fsid_kobj(struct kobject *kobj)
 {
 	struct btrfs_fs_devices *fs_devs = to_fs_devs(kobj);
 
-	memset(&fs_devs->super_kobj, 0, sizeof(struct kobject));
+	memset(&fs_devs->fsid_kobj, 0, sizeof(struct kobject));
 	complete(&fs_devs->kobj_unregister);
 }
 
 static struct kobj_type btrfs_ktype = {
 	.sysfs_ops	= &kobj_sysfs_ops,
-	.release	= btrfs_release_super_kobj,
+	.release	= btrfs_release_fsid_kobj,
 };
 
 static inline struct btrfs_fs_devices *to_fs_devs(struct kobject *kobj)
 {
 	if (kobj->ktype != &btrfs_ktype)
 		return NULL;
-	return container_of(kobj, struct btrfs_fs_devices, super_kobj);
+	return container_of(kobj, struct btrfs_fs_devices, fsid_kobj);
 }
 
 static inline struct btrfs_fs_info *to_fs_info(struct kobject *kobj)
@@ -502,12 +502,12 @@ static int addrm_unknown_feature_attrs(struct btrfs_fs_info *fs_info, bool add)
 			attrs[0] = &fa->kobj_attr.attr;
 			if (add) {
 				int ret;
-				ret = sysfs_merge_group(&fs_info->fs_devices->super_kobj,
+				ret = sysfs_merge_group(&fs_info->fs_devices->fsid_kobj,
 							&agroup);
 				if (ret)
 					return ret;
 			} else
-				sysfs_unmerge_group(&fs_info->fs_devices->super_kobj,
+				sysfs_unmerge_group(&fs_info->fs_devices->fsid_kobj,
 						    &agroup);
 		}
 
@@ -523,9 +523,9 @@ static void __btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs)
 		fs_devs->device_dir_kobj = NULL;
 	}
 
-	if (fs_devs->super_kobj.state_initialized) {
-		kobject_del(&fs_devs->super_kobj);
-		kobject_put(&fs_devs->super_kobj);
+	if (fs_devs->fsid_kobj.state_initialized) {
+		kobject_del(&fs_devs->fsid_kobj);
+		kobject_put(&fs_devs->fsid_kobj);
 		wait_for_completion(&fs_devs->kobj_unregister);
 	}
 }
@@ -545,7 +545,7 @@ void btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs)
 	}
 }
 
-void btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info)
+void btrfs_sysfs_remove_mounted(struct btrfs_fs_info *fs_info)
 {
 	btrfs_reset_fs_info_ptr(fs_info);
 
@@ -555,9 +555,9 @@ void btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info)
 		kobject_put(fs_info->space_info_kobj);
 	}
 	addrm_unknown_feature_attrs(fs_info, false);
-	sysfs_remove_group(&fs_info->fs_devices->super_kobj, &btrfs_feature_attr_group);
-	sysfs_remove_files(&fs_info->fs_devices->super_kobj, btrfs_attrs);
-	btrfs_kobj_rm_device(fs_info->fs_devices, NULL);
+	sysfs_remove_group(&fs_info->fs_devices->fsid_kobj, &btrfs_feature_attr_group);
+	sysfs_remove_files(&fs_info->fs_devices->fsid_kobj, btrfs_attrs);
+	btrfs_sysfs_rm_device_link(fs_info->fs_devices, NULL);
 }
 
 const char * const btrfs_feature_set_names[3] = {
@@ -637,7 +637,7 @@ static void init_feature_attrs(void)
 
 /* when one_device is NULL, it removes all device links */
 
-int btrfs_kobj_rm_device(struct btrfs_fs_devices *fs_devices,
+int btrfs_sysfs_rm_device_link(struct btrfs_fs_devices *fs_devices,
 		struct btrfs_device *one_device)
 {
 	struct hd_struct *disk;
@@ -675,7 +675,7 @@ int btrfs_sysfs_add_device(struct btrfs_fs_devices *fs_devs)
 {
 	if (!fs_devs->device_dir_kobj)
 		fs_devs->device_dir_kobj = kobject_create_and_add("devices",
-						&fs_devs->super_kobj);
+						&fs_devs->fsid_kobj);
 
 	if (!fs_devs->device_dir_kobj)
 		return -ENOMEM;
@@ -683,7 +683,7 @@ int btrfs_sysfs_add_device(struct btrfs_fs_devices *fs_devs)
 	return 0;
 }
 
-int btrfs_kobj_add_device(struct btrfs_fs_devices *fs_devices,
+int btrfs_sysfs_add_device_link(struct btrfs_fs_devices *fs_devices,
 				struct btrfs_device *one_device)
 {
 	int error = 0;
@@ -730,31 +730,31 @@ int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs,
 	int error;
 
 	init_completion(&fs_devs->kobj_unregister);
-	fs_devs->super_kobj.kset = btrfs_kset;
-	error = kobject_init_and_add(&fs_devs->super_kobj,
+	fs_devs->fsid_kobj.kset = btrfs_kset;
+	error = kobject_init_and_add(&fs_devs->fsid_kobj,
 				&btrfs_ktype, parent, "%pU", fs_devs->fsid);
 	return error;
 }
 
-int btrfs_sysfs_add_one(struct btrfs_fs_info *fs_info)
+int btrfs_sysfs_add_mounted(struct btrfs_fs_info *fs_info)
 {
 	int error;
 	struct btrfs_fs_devices *fs_devs = fs_info->fs_devices;
-	struct kobject *super_kobj = &fs_devs->super_kobj;
+	struct kobject *fsid_kobj = &fs_devs->fsid_kobj;
 
 	btrfs_set_fs_info_ptr(fs_info);
 
-	error = btrfs_kobj_add_device(fs_devs, NULL);
+	error = btrfs_sysfs_add_device_link(fs_devs, NULL);
 	if (error)
 		return error;
 
-	error = sysfs_create_files(super_kobj, btrfs_attrs);
+	error = sysfs_create_files(fsid_kobj, btrfs_attrs);
 	if (error) {
-		btrfs_kobj_rm_device(fs_devs, NULL);
+		btrfs_sysfs_rm_device_link(fs_devs, NULL);
 		return error;
 	}
 
-	error = sysfs_create_group(super_kobj,
+	error = sysfs_create_group(fsid_kobj,
 				   &btrfs_feature_attr_group);
 	if (error)
 		goto failure;
@@ -764,7 +764,7 @@ int btrfs_sysfs_add_one(struct btrfs_fs_info *fs_info)
 		goto failure;
 
 	fs_info->space_info_kobj = kobject_create_and_add("allocation",
-						  super_kobj);
+						  fsid_kobj);
 	if (!fs_info->space_info_kobj) {
 		error = -ENOMEM;
 		goto failure;
@@ -776,7 +776,7 @@ int btrfs_sysfs_add_one(struct btrfs_fs_info *fs_info)
 
 	return 0;
 failure:
-	btrfs_sysfs_remove_one(fs_info);
+	btrfs_sysfs_remove_mounted(fs_info);
 	return error;
 }
 

fs/btrfs/sysfs.h

@@ -82,9 +82,9 @@ char *btrfs_printable_features(enum btrfs_feature_set set, u64 flags);
 extern const char * const btrfs_feature_set_names[3];
 extern struct kobj_type space_info_ktype;
 extern struct kobj_type btrfs_raid_ktype;
-int btrfs_kobj_add_device(struct btrfs_fs_devices *fs_devices,
+int btrfs_sysfs_add_device_link(struct btrfs_fs_devices *fs_devices,
 		struct btrfs_device *one_device);
-int btrfs_kobj_rm_device(struct btrfs_fs_devices *fs_devices,
+int btrfs_sysfs_rm_device_link(struct btrfs_fs_devices *fs_devices,
                 struct btrfs_device *one_device);
 int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs,
 				struct kobject *parent);

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

@@ -19,6 +19,7 @@
 #include <linux/slab.h>
 #include "btrfs-tests.h"
 #include "../ctree.h"
+#include "../disk-io.h"
 #include "../free-space-cache.h"
 
 #define BITS_PER_BITMAP		(PAGE_CACHE_SIZE * 8)
@@ -35,6 +36,12 @@ static struct btrfs_block_group_cache *init_test_block_group(void)
 		kfree(cache);
 		return NULL;
 	}
+	cache->fs_info = btrfs_alloc_dummy_fs_info();
+	if (!cache->fs_info) {
+		kfree(cache->free_space_ctl);
+		kfree(cache);
+		return NULL;
+	}
 
 	cache->key.objectid = 0;
 	cache->key.offset = 1024 * 1024 * 1024;
@@ -879,7 +886,8 @@ test_steal_space_from_bitmap_to_extent(struct btrfs_block_group_cache *cache)
 int btrfs_test_free_space_cache(void)
 {
 	struct btrfs_block_group_cache *cache;
-	int ret;
+	struct btrfs_root *root = NULL;
+	int ret = -ENOMEM;
 
 	test_msg("Running btrfs free space cache tests\n");
 
@@ -889,6 +897,17 @@ int btrfs_test_free_space_cache(void)
 		return 0;
 	}
 
+	root = btrfs_alloc_dummy_root();
+	if (!root)
+		goto out;
+
+	root->fs_info = btrfs_alloc_dummy_fs_info();
+	if (!root->fs_info)
+		goto out;
+
+	root->fs_info->extent_root = root;
+	cache->fs_info = root->fs_info;
+
 	ret = test_extents(cache);
 	if (ret)
 		goto out;
@@ -904,6 +923,7 @@ out:
 	__btrfs_remove_free_space_cache(cache->free_space_ctl);
 	kfree(cache->free_space_ctl);
 	kfree(cache);
+	btrfs_free_dummy_root(root);
 	test_msg("Free space cache tests finished\n");
 	return ret;
 }

fs/btrfs/transaction.c

@@ -82,6 +82,12 @@ void btrfs_put_transaction(struct btrfs_transaction *transaction)
 static void clear_btree_io_tree(struct extent_io_tree *tree)
 {
 	spin_lock(&tree->lock);
+	/*
+	 * Do a single barrier for the waitqueue_active check here, the state
+	 * of the waitqueue should not change once clear_btree_io_tree is
+	 * called.
+	 */
+	smp_mb();
 	while (!RB_EMPTY_ROOT(&tree->state)) {
 		struct rb_node *node;
 		struct extent_state *state;
@@ -226,25 +232,22 @@ loop:
 	extwriter_counter_init(cur_trans, type);
 	init_waitqueue_head(&cur_trans->writer_wait);
 	init_waitqueue_head(&cur_trans->commit_wait);
+	init_waitqueue_head(&cur_trans->pending_wait);
 	cur_trans->state = TRANS_STATE_RUNNING;
 	/*
 	 * One for this trans handle, one so it will live on until we
 	 * commit the transaction.
 	 */
 	atomic_set(&cur_trans->use_count, 2);
-	cur_trans->have_free_bgs = 0;
+	atomic_set(&cur_trans->pending_ordered, 0);
+	cur_trans->flags = 0;
 	cur_trans->start_time = get_seconds();
-	cur_trans->dirty_bg_run = 0;
+
+	memset(&cur_trans->delayed_refs, 0, sizeof(cur_trans->delayed_refs));
 
 	cur_trans->delayed_refs.href_root = RB_ROOT;
 	cur_trans->delayed_refs.dirty_extent_root = RB_ROOT;
 	atomic_set(&cur_trans->delayed_refs.num_entries, 0);
-	cur_trans->delayed_refs.num_heads_ready = 0;
-	cur_trans->delayed_refs.pending_csums = 0;
-	cur_trans->delayed_refs.num_heads = 0;
-	cur_trans->delayed_refs.flushing = 0;
-	cur_trans->delayed_refs.run_delayed_start = 0;
-	cur_trans->delayed_refs.qgroup_to_skip = 0;
 
 	/*
 	 * although the tree mod log is per file system and not per transaction,
@@ -264,7 +267,6 @@ loop:
 	INIT_LIST_HEAD(&cur_trans->pending_snapshots);
 	INIT_LIST_HEAD(&cur_trans->pending_chunks);
 	INIT_LIST_HEAD(&cur_trans->switch_commits);
-	INIT_LIST_HEAD(&cur_trans->pending_ordered);
 	INIT_LIST_HEAD(&cur_trans->dirty_bgs);
 	INIT_LIST_HEAD(&cur_trans->io_bgs);
 	INIT_LIST_HEAD(&cur_trans->dropped_roots);
@@ -447,8 +449,8 @@ static inline bool need_reserve_reloc_root(struct btrfs_root *root)
 }
 
 static struct btrfs_trans_handle *
-start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
-		  enum btrfs_reserve_flush_enum flush)
+start_transaction(struct btrfs_root *root, unsigned int num_items,
+		  unsigned int type, enum btrfs_reserve_flush_enum flush)
 {
 	struct btrfs_trans_handle *h;
 	struct btrfs_transaction *cur_trans;
@@ -478,13 +480,10 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 	 * the appropriate flushing if need be.
 	 */
 	if (num_items > 0 && root != root->fs_info->chunk_root) {
-		if (root->fs_info->quota_enabled &&
-		    is_fstree(root->root_key.objectid)) {
-			qgroup_reserved = num_items * root->nodesize;
-			ret = btrfs_qgroup_reserve(root, qgroup_reserved);
-			if (ret)
-				return ERR_PTR(ret);
-		}
+		qgroup_reserved = num_items * root->nodesize;
+		ret = btrfs_qgroup_reserve_meta(root, qgroup_reserved);
+		if (ret)
+			return ERR_PTR(ret);
 
 		num_bytes = btrfs_calc_trans_metadata_size(root, num_items);
 		/*
@@ -502,7 +501,7 @@ start_transaction(struct btrfs_root *root, u64 num_items, unsigned int type,
 			goto reserve_fail;
 	}
 again:
-	h = kmem_cache_alloc(btrfs_trans_handle_cachep, GFP_NOFS);
+	h = kmem_cache_zalloc(btrfs_trans_handle_cachep, GFP_NOFS);
 	if (!h) {
 		ret = -ENOMEM;
 		goto alloc_fail;
@@ -543,26 +542,13 @@ again:
 
 	h->transid = cur_trans->transid;
 	h->transaction = cur_trans;
-	h->blocks_used = 0;
-	h->bytes_reserved = 0;
-	h->chunk_bytes_reserved = 0;
 	h->root = root;
-	h->delayed_ref_updates = 0;
 	h->use_count = 1;
-	h->adding_csums = 0;
-	h->block_rsv = NULL;
-	h->orig_rsv = NULL;
-	h->aborted = 0;
-	h->qgroup_reserved = 0;
-	h->delayed_ref_elem.seq = 0;
+
 	h->type = type;
-	h->allocating_chunk = false;
 	h->can_flush_pending_bgs = true;
-	h->reloc_reserved = false;
-	h->sync = false;
 	INIT_LIST_HEAD(&h->qgroup_ref_list);
 	INIT_LIST_HEAD(&h->new_bgs);
-	INIT_LIST_HEAD(&h->ordered);
 
 	smp_mb();
 	if (cur_trans->state >= TRANS_STATE_BLOCKED &&
@@ -579,7 +565,6 @@ again:
 		h->bytes_reserved = num_bytes;
 		h->reloc_reserved = reloc_reserved;
 	}
-	h->qgroup_reserved = qgroup_reserved;
 
 got_it:
 	btrfs_record_root_in_trans(h, root);
@@ -597,20 +582,20 @@ alloc_fail:
 		btrfs_block_rsv_release(root, &root->fs_info->trans_block_rsv,
 					num_bytes);
 reserve_fail:
-	if (qgroup_reserved)
-		btrfs_qgroup_free(root, qgroup_reserved);
+	btrfs_qgroup_free_meta(root, qgroup_reserved);
 	return ERR_PTR(ret);
 }
 
 struct btrfs_trans_handle *btrfs_start_transaction(struct btrfs_root *root,
-						   int num_items)
+						   unsigned int num_items)
 {
 	return start_transaction(root, num_items, TRANS_START,
 				 BTRFS_RESERVE_FLUSH_ALL);
 }
 
 struct btrfs_trans_handle *btrfs_start_transaction_lflush(
-					struct btrfs_root *root, int num_items)
+					struct btrfs_root *root,
+					unsigned int num_items)
 {
 	return start_transaction(root, num_items, TRANS_START,
 				 BTRFS_RESERVE_FLUSH_LIMIT);
@@ -794,12 +779,6 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 	if (!list_empty(&trans->new_bgs))
 		btrfs_create_pending_block_groups(trans, root);
 
-	if (!list_empty(&trans->ordered)) {
-		spin_lock(&info->trans_lock);
-		list_splice_init(&trans->ordered, &cur_trans->pending_ordered);
-		spin_unlock(&info->trans_lock);
-	}
-
 	trans->delayed_ref_updates = 0;
 	if (!trans->sync) {
 		must_run_delayed_refs =
@@ -815,15 +794,6 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 			must_run_delayed_refs = 2;
 	}
 
-	if (trans->qgroup_reserved) {
-		/*
-		 * the same root has to be passed here between start_transaction
-		 * and end_transaction. Subvolume quota depends on this.
-		 */
-		btrfs_qgroup_free(trans->root, trans->qgroup_reserved);
-		trans->qgroup_reserved = 0;
-	}
-
 	btrfs_trans_release_metadata(trans, root);
 	trans->block_rsv = NULL;
 
@@ -856,6 +826,9 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 	atomic_dec(&cur_trans->num_writers);
 	extwriter_counter_dec(cur_trans, trans->type);
 
+	/*
+	 * Make sure counter is updated before we wake up waiters.
+	 */
 	smp_mb();
 	if (waitqueue_active(&cur_trans->writer_wait))
 		wake_up(&cur_trans->writer_wait);
@@ -1238,6 +1211,7 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans,
 			spin_lock(&fs_info->fs_roots_radix_lock);
 			if (err)
 				break;
+			btrfs_qgroup_free_meta_all(root);
 		}
 	}
 	spin_unlock(&fs_info->fs_roots_radix_lock);
@@ -1795,25 +1769,10 @@ static inline void btrfs_wait_delalloc_flush(struct btrfs_fs_info *fs_info)
 }
 
 static inline void
-btrfs_wait_pending_ordered(struct btrfs_transaction *cur_trans,
-			   struct btrfs_fs_info *fs_info)
+btrfs_wait_pending_ordered(struct btrfs_transaction *cur_trans)
 {
-	struct btrfs_ordered_extent *ordered;
-
-	spin_lock(&fs_info->trans_lock);
-	while (!list_empty(&cur_trans->pending_ordered)) {
-		ordered = list_first_entry(&cur_trans->pending_ordered,
-					   struct btrfs_ordered_extent,
-					   trans_list);
-		list_del_init(&ordered->trans_list);
-		spin_unlock(&fs_info->trans_lock);
-
-		wait_event(ordered->wait, test_bit(BTRFS_ORDERED_COMPLETE,
-						   &ordered->flags));
-		btrfs_put_ordered_extent(ordered);
-		spin_lock(&fs_info->trans_lock);
-	}
-	spin_unlock(&fs_info->trans_lock);
+	wait_event(cur_trans->pending_wait,
+		   atomic_read(&cur_trans->pending_ordered) == 0);
 }
 
 int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
@@ -1842,10 +1801,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	btrfs_trans_release_metadata(trans, root);
 	trans->block_rsv = NULL;
-	if (trans->qgroup_reserved) {
-		btrfs_qgroup_free(root, trans->qgroup_reserved);
-		trans->qgroup_reserved = 0;
-	}
 
 	cur_trans = trans->transaction;
 
@@ -1865,7 +1820,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 		return ret;
 	}
 
-	if (!cur_trans->dirty_bg_run) {
+	if (!test_bit(BTRFS_TRANS_DIRTY_BG_RUN, &cur_trans->flags)) {
 		int run_it = 0;
 
 		/* this mutex is also taken before trying to set
@@ -1874,18 +1829,17 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 		 * after a extents from that block group have been
 		 * allocated for cache files.  btrfs_set_block_group_ro
 		 * will wait for the transaction to commit if it
-		 * finds dirty_bg_run = 1
+		 * finds BTRFS_TRANS_DIRTY_BG_RUN set.
 		 *
-		 * The dirty_bg_run flag is also used to make sure only
-		 * one process starts all the block group IO.  It wouldn't
+		 * The BTRFS_TRANS_DIRTY_BG_RUN flag is also used to make sure
+		 * only one process starts all the block group IO.  It wouldn't
 		 * hurt to have more than one go through, but there's no
 		 * real advantage to it either.
 		 */
 		mutex_lock(&root->fs_info->ro_block_group_mutex);
-		if (!cur_trans->dirty_bg_run) {
+		if (!test_and_set_bit(BTRFS_TRANS_DIRTY_BG_RUN,
+				      &cur_trans->flags))
 			run_it = 1;
-			cur_trans->dirty_bg_run = 1;
-		}
 		mutex_unlock(&root->fs_info->ro_block_group_mutex);
 
 		if (run_it)
@@ -1897,7 +1851,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	}
 
 	spin_lock(&root->fs_info->trans_lock);
-	list_splice_init(&trans->ordered, &cur_trans->pending_ordered);
 	if (cur_trans->state >= TRANS_STATE_COMMIT_START) {
 		spin_unlock(&root->fs_info->trans_lock);
 		atomic_inc(&cur_trans->use_count);
@@ -1956,7 +1909,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	btrfs_wait_delalloc_flush(root->fs_info);
 
-	btrfs_wait_pending_ordered(cur_trans, root->fs_info);
+	btrfs_wait_pending_ordered(cur_trans);
 
 	btrfs_scrub_pause(root);
 	/*
@@ -2136,7 +2089,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	ret = btrfs_write_and_wait_transaction(trans, root);
 	if (ret) {
-		btrfs_error(root->fs_info, ret,
+		btrfs_std_error(root->fs_info, ret,
 			    "Error while writing out transaction");
 		mutex_unlock(&root->fs_info->tree_log_mutex);
 		goto scrub_continue;
@@ -2156,7 +2109,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	btrfs_finish_extent_commit(trans, root);
 
-	if (cur_trans->have_free_bgs)
+	if (test_bit(BTRFS_TRANS_HAVE_FREE_BGS, &cur_trans->flags))
 		btrfs_clear_space_info_full(root->fs_info);
 
 	root->fs_info->last_trans_committed = cur_trans->transid;
@@ -2198,10 +2151,6 @@ cleanup_transaction:
 	btrfs_trans_release_metadata(trans, root);
 	btrfs_trans_release_chunk_metadata(trans);
 	trans->block_rsv = NULL;
-	if (trans->qgroup_reserved) {
-		btrfs_qgroup_free(root, trans->qgroup_reserved);
-		trans->qgroup_reserved = 0;
-	}
 	btrfs_warn(root->fs_info, "Skipping commit of aborted transaction.");
 	if (current->journal_info == trans)
 		current->journal_info = NULL;

fs/btrfs/transaction.h

@@ -32,6 +32,10 @@ enum btrfs_trans_state {
 	TRANS_STATE_MAX			= 6,
 };
 
+#define BTRFS_TRANS_HAVE_FREE_BGS	0
+#define BTRFS_TRANS_DIRTY_BG_RUN	1
+#define BTRFS_TRANS_CACHE_ENOSPC	2
+
 struct btrfs_transaction {
 	u64 transid;
 	/*
@@ -46,11 +50,9 @@ struct btrfs_transaction {
 	 */
 	atomic_t num_writers;
 	atomic_t use_count;
+	atomic_t pending_ordered;
 
-	/*
-	 * true if there is free bgs operations in this transaction
-	 */
-	int have_free_bgs;
+	unsigned long flags;
 
 	/* Be protected by fs_info->trans_lock when we want to change it. */
 	enum btrfs_trans_state state;
@@ -59,9 +61,9 @@ struct btrfs_transaction {
 	unsigned long start_time;
 	wait_queue_head_t writer_wait;
 	wait_queue_head_t commit_wait;
+	wait_queue_head_t pending_wait;
 	struct list_head pending_snapshots;
 	struct list_head pending_chunks;
-	struct list_head pending_ordered;
 	struct list_head switch_commits;
 	struct list_head dirty_bgs;
 	struct list_head io_bgs;
@@ -80,7 +82,6 @@ struct btrfs_transaction {
 	spinlock_t dropped_roots_lock;
 	struct btrfs_delayed_ref_root delayed_refs;
 	int aborted;
-	int dirty_bg_run;
 };
 
 #define __TRANS_FREEZABLE	(1U << 0)
@@ -107,7 +108,6 @@ struct btrfs_trans_handle {
 	u64 transid;
 	u64 bytes_reserved;
 	u64 chunk_bytes_reserved;
-	u64 qgroup_reserved;
 	unsigned long use_count;
 	unsigned long blocks_reserved;
 	unsigned long blocks_used;
@@ -129,7 +129,6 @@ struct btrfs_trans_handle {
 	 */
 	struct btrfs_root *root;
 	struct seq_list delayed_ref_elem;
-	struct list_head ordered;
 	struct list_head qgroup_ref_list;
 	struct list_head new_bgs;
 };
@@ -185,9 +184,10 @@ static inline void btrfs_clear_skip_qgroup(struct btrfs_trans_handle *trans)
 int btrfs_end_transaction(struct btrfs_trans_handle *trans,
 			  struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_start_transaction(struct btrfs_root *root,
-						   int num_items);
+						   unsigned int num_items);
 struct btrfs_trans_handle *btrfs_start_transaction_lflush(
-					struct btrfs_root *root, int num_items);
+					struct btrfs_root *root,
+					unsigned int num_items);
 struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_attach_transaction(struct btrfs_root *root);

fs/btrfs/tree-log.c

@@ -229,7 +229,9 @@ int btrfs_pin_log_trans(struct btrfs_root *root)
 void btrfs_end_log_trans(struct btrfs_root *root)
 {
 	if (atomic_dec_and_test(&root->log_writers)) {
-		smp_mb();
+		/*
+		 * Implicit memory barrier after atomic_dec_and_test
+		 */
 		if (waitqueue_active(&root->log_writer_wait))
 			wake_up(&root->log_writer_wait);
 	}
@@ -691,7 +693,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 				ret = btrfs_inc_extent_ref(trans, root,
 						ins.objectid, ins.offset,
 						0, root->root_key.objectid,
-						key->objectid, offset, 0);
+						key->objectid, offset);
 				if (ret)
 					goto out;
 			} else {
@@ -2820,7 +2822,9 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 
 	mutex_lock(&log_root_tree->log_mutex);
 	if (atomic_dec_and_test(&log_root_tree->log_writers)) {
-		smp_mb();
+		/*
+		 * Implicit memory barrier after atomic_dec_and_test
+		 */
 		if (waitqueue_active(&log_root_tree->log_writer_wait))
 			wake_up(&log_root_tree->log_writer_wait);
 	}
@@ -2950,6 +2954,9 @@ out_wake_log_root:
 	atomic_set(&log_root_tree->log_commit[index2], 0);
 	mutex_unlock(&log_root_tree->log_mutex);
 
+	/*
+	 * The barrier before waitqueue_active is implied by mutex_unlock
+	 */
 	if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
 		wake_up(&log_root_tree->log_commit_wait[index2]);
 out:
@@ -2961,6 +2968,9 @@ out:
 	atomic_set(&root->log_commit[index1], 0);
 	mutex_unlock(&root->log_mutex);
 
+	/*
+	 * The barrier before waitqueue_active is implied by mutex_unlock
+	 */
 	if (waitqueue_active(&root->log_commit_wait[index1]))
 		wake_up(&root->log_commit_wait[index1]);
 	return ret;
@@ -5314,7 +5324,7 @@ int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
 
 	ret = walk_log_tree(trans, log_root_tree, &wc);
 	if (ret) {
-		btrfs_error(fs_info, ret, "Failed to pin buffers while "
+		btrfs_std_error(fs_info, ret, "Failed to pin buffers while "
 			    "recovering log root tree.");
 		goto error;
 	}
@@ -5328,7 +5338,7 @@ again:
 		ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0);
 
 		if (ret < 0) {
-			btrfs_error(fs_info, ret,
+			btrfs_std_error(fs_info, ret,
 				    "Couldn't find tree log root.");
 			goto error;
 		}
@@ -5346,7 +5356,7 @@ again:
 		log = btrfs_read_fs_root(log_root_tree, &found_key);
 		if (IS_ERR(log)) {
 			ret = PTR_ERR(log);
-			btrfs_error(fs_info, ret,
+			btrfs_std_error(fs_info, ret,
 				    "Couldn't read tree log root.");
 			goto error;
 		}
@@ -5361,7 +5371,7 @@ again:
 			free_extent_buffer(log->node);
 			free_extent_buffer(log->commit_root);
 			kfree(log);
-			btrfs_error(fs_info, ret, "Couldn't read target root "
+			btrfs_std_error(fs_info, ret, "Couldn't read target root "
 				    "for tree log recovery.");
 			goto error;
 		}

fs/btrfs/volumes.c

@@ -42,6 +42,82 @@
 #include "dev-replace.h"
 #include "sysfs.h"
 
+const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
+	[BTRFS_RAID_RAID10] = {
+		.sub_stripes	= 2,
+		.dev_stripes	= 1,
+		.devs_max	= 0,	/* 0 == as many as possible */
+		.devs_min	= 4,
+		.tolerated_failures = 1,
+		.devs_increment	= 2,
+		.ncopies	= 2,
+	},
+	[BTRFS_RAID_RAID1] = {
+		.sub_stripes	= 1,
+		.dev_stripes	= 1,
+		.devs_max	= 2,
+		.devs_min	= 2,
+		.tolerated_failures = 1,
+		.devs_increment	= 2,
+		.ncopies	= 2,
+	},
+	[BTRFS_RAID_DUP] = {
+		.sub_stripes	= 1,
+		.dev_stripes	= 2,
+		.devs_max	= 1,
+		.devs_min	= 1,
+		.tolerated_failures = 0,
+		.devs_increment	= 1,
+		.ncopies	= 2,
+	},
+	[BTRFS_RAID_RAID0] = {
+		.sub_stripes	= 1,
+		.dev_stripes	= 1,
+		.devs_max	= 0,
+		.devs_min	= 2,
+		.tolerated_failures = 0,
+		.devs_increment	= 1,
+		.ncopies	= 1,
+	},
+	[BTRFS_RAID_SINGLE] = {
+		.sub_stripes	= 1,
+		.dev_stripes	= 1,
+		.devs_max	= 1,
+		.devs_min	= 1,
+		.tolerated_failures = 0,
+		.devs_increment	= 1,
+		.ncopies	= 1,
+	},
+	[BTRFS_RAID_RAID5] = {
+		.sub_stripes	= 1,
+		.dev_stripes	= 1,
+		.devs_max	= 0,
+		.devs_min	= 2,
+		.tolerated_failures = 1,
+		.devs_increment	= 1,
+		.ncopies	= 2,
+	},
+	[BTRFS_RAID_RAID6] = {
+		.sub_stripes	= 1,
+		.dev_stripes	= 1,
+		.devs_max	= 0,
+		.devs_min	= 3,
+		.tolerated_failures = 2,
+		.devs_increment	= 1,
+		.ncopies	= 3,
+	},
+};
+
+const u64 const btrfs_raid_group[BTRFS_NR_RAID_TYPES] = {
+	[BTRFS_RAID_RAID10] = BTRFS_BLOCK_GROUP_RAID10,
+	[BTRFS_RAID_RAID1]  = BTRFS_BLOCK_GROUP_RAID1,
+	[BTRFS_RAID_DUP]    = BTRFS_BLOCK_GROUP_DUP,
+	[BTRFS_RAID_RAID0]  = BTRFS_BLOCK_GROUP_RAID0,
+	[BTRFS_RAID_SINGLE] = 0,
+	[BTRFS_RAID_RAID5]  = BTRFS_BLOCK_GROUP_RAID5,
+	[BTRFS_RAID_RAID6]  = BTRFS_BLOCK_GROUP_RAID6,
+};
+
 static int init_first_rw_device(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root,
 				struct btrfs_device *device);
@@ -198,7 +274,6 @@ btrfs_get_bdev_and_sb(const char *device_path, fmode_t flags, void *holder,
 
 	if (IS_ERR(*bdev)) {
 		ret = PTR_ERR(*bdev);
-		printk(KERN_INFO "BTRFS: open %s failed\n", device_path);
 		goto error;
 	}
 
@@ -211,8 +286,8 @@ btrfs_get_bdev_and_sb(const char *device_path, fmode_t flags, void *holder,
 	}
 	invalidate_bdev(*bdev);
 	*bh = btrfs_read_dev_super(*bdev);
-	if (!*bh) {
-		ret = -EINVAL;
+	if (IS_ERR(*bh)) {
+		ret = PTR_ERR(*bh);
 		blkdev_put(*bdev, flags);
 		goto error;
 	}
@@ -345,6 +420,9 @@ loop_lock:
 		pending = pending->bi_next;
 		cur->bi_next = NULL;
 
+		/*
+		 * atomic_dec_return implies a barrier for waitqueue_active
+		 */
 		if (atomic_dec_return(&fs_info->nr_async_bios) < limit &&
 		    waitqueue_active(&fs_info->async_submit_wait))
 			wake_up(&fs_info->async_submit_wait);
@@ -765,36 +843,7 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 
 	mutex_lock(&fs_devices->device_list_mutex);
 	list_for_each_entry_safe(device, tmp, &fs_devices->devices, dev_list) {
-		struct btrfs_device *new_device;
-		struct rcu_string *name;
-
-		if (device->bdev)
-			fs_devices->open_devices--;
-
-		if (device->writeable &&
-		    device->devid != BTRFS_DEV_REPLACE_DEVID) {
-			list_del_init(&device->dev_alloc_list);
-			fs_devices->rw_devices--;
-		}
-
-		if (device->missing)
-			fs_devices->missing_devices--;
-
-		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(!name); /* -ENOMEM */
-			rcu_assign_pointer(new_device->name, name);
-		}
-
-		list_replace_rcu(&device->dev_list, &new_device->dev_list);
-		new_device->fs_devices = device->fs_devices;
-
-		call_rcu(&device->rcu, free_device);
+		btrfs_close_one_device(device);
 	}
 	mutex_unlock(&fs_devices->device_list_mutex);
 
@@ -1402,7 +1451,7 @@ again:
 		extent = btrfs_item_ptr(leaf, path->slots[0],
 					struct btrfs_dev_extent);
 	} else {
-		btrfs_error(root->fs_info, ret, "Slot search failed");
+		btrfs_std_error(root->fs_info, ret, "Slot search failed");
 		goto out;
 	}
 
@@ -1410,10 +1459,10 @@ again:
 
 	ret = btrfs_del_item(trans, root, path);
 	if (ret) {
-		btrfs_error(root->fs_info, ret,
+		btrfs_std_error(root->fs_info, ret,
 			    "Failed to remove dev extent item");
 	} else {
-		trans->transaction->have_free_bgs = 1;
+		set_bit(BTRFS_TRANS_HAVE_FREE_BGS, &trans->transaction->flags);
 	}
 out:
 	btrfs_free_path(path);
@@ -1801,7 +1850,7 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	if (device->bdev) {
 		device->fs_devices->open_devices--;
 		/* remove sysfs entry */
-		btrfs_kobj_rm_device(root->fs_info->fs_devices, device);
+		btrfs_sysfs_rm_device_link(root->fs_info->fs_devices, device);
 	}
 
 	call_rcu(&device->rcu, free_device);
@@ -1924,7 +1973,8 @@ void btrfs_rm_dev_replace_remove_srcdev(struct btrfs_fs_info *fs_info,
 	if (srcdev->writeable) {
 		fs_devices->rw_devices--;
 		/* zero out the old super if it is writable */
-		btrfs_scratch_superblock(srcdev);
+		btrfs_scratch_superblocks(srcdev->bdev,
+					rcu_str_deref(srcdev->name));
 	}
 
 	if (srcdev->bdev)
@@ -1971,10 +2021,11 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	WARN_ON(!tgtdev);
 	mutex_lock(&fs_info->fs_devices->device_list_mutex);
 
-	btrfs_kobj_rm_device(fs_info->fs_devices, tgtdev);
+	btrfs_sysfs_rm_device_link(fs_info->fs_devices, tgtdev);
 
 	if (tgtdev->bdev) {
-		btrfs_scratch_superblock(tgtdev);
+		btrfs_scratch_superblocks(tgtdev->bdev,
+					rcu_str_deref(tgtdev->name));
 		fs_info->fs_devices->open_devices--;
 	}
 	fs_info->fs_devices->num_devices--;
@@ -2041,10 +2092,8 @@ int btrfs_find_device_missing_or_by_path(struct btrfs_root *root,
 			}
 		}
 
-		if (!*device) {
-			btrfs_err(root->fs_info, "no missing device found");
-			return -ENOENT;
-		}
+		if (!*device)
+			return BTRFS_ERROR_DEV_MISSING_NOT_FOUND;
 
 		return 0;
 	} else {
@@ -2309,7 +2358,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 				    tmp + 1);
 
 	/* add sysfs device entry */
-	btrfs_kobj_add_device(root->fs_info->fs_devices, device);
+	btrfs_sysfs_add_device_link(root->fs_info->fs_devices, device);
 
 	/*
 	 * we've got more storage, clear any full flags on the space
@@ -2350,9 +2399,10 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 		 */
 		snprintf(fsid_buf, BTRFS_UUID_UNPARSED_SIZE, "%pU",
 						root->fs_info->fsid);
-		if (kobject_rename(&root->fs_info->fs_devices->super_kobj,
+		if (kobject_rename(&root->fs_info->fs_devices->fsid_kobj,
 								fsid_buf))
-			pr_warn("BTRFS: sysfs: failed to create fsid for sprout\n");
+			btrfs_warn(root->fs_info,
+				"sysfs: failed to create fsid for sprout");
 	}
 
 	root->fs_info->num_tolerated_disk_barrier_failures =
@@ -2368,7 +2418,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 
 		ret = btrfs_relocate_sys_chunks(root);
 		if (ret < 0)
-			btrfs_error(root->fs_info, ret,
+			btrfs_std_error(root->fs_info, ret,
 				    "Failed to relocate sys chunks after "
 				    "device initialization. This can be fixed "
 				    "using the \"btrfs balance\" command.");
@@ -2388,7 +2438,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 error_trans:
 	btrfs_end_transaction(trans, root);
 	rcu_string_free(device->name);
-	btrfs_kobj_rm_device(root->fs_info->fs_devices, device);
+	btrfs_sysfs_rm_device_link(root->fs_info->fs_devices, device);
 	kfree(device);
 error:
 	blkdev_put(bdev, FMODE_EXCL);
@@ -2613,7 +2663,7 @@ static int btrfs_free_chunk(struct btrfs_trans_handle *trans,
 	if (ret < 0)
 		goto out;
 	else if (ret > 0) { /* Logic error or corruption */
-		btrfs_error(root->fs_info, -ENOENT,
+		btrfs_std_error(root->fs_info, -ENOENT,
 			    "Failed lookup while freeing chunk.");
 		ret = -ENOENT;
 		goto out;
@@ -2621,7 +2671,7 @@ static int btrfs_free_chunk(struct btrfs_trans_handle *trans,
 
 	ret = btrfs_del_item(trans, root, path);
 	if (ret < 0)
-		btrfs_error(root->fs_info, ret,
+		btrfs_std_error(root->fs_info, ret,
 			    "Failed to delete chunk item.");
 out:
 	btrfs_free_path(path);
@@ -2806,7 +2856,7 @@ static int btrfs_relocate_chunk(struct btrfs_root *root, u64 chunk_offset)
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
-		btrfs_std_error(root->fs_info, ret);
+		btrfs_std_error(root->fs_info, ret, NULL);
 		return ret;
 	}
 
@@ -3009,16 +3059,19 @@ static void update_balance_args(struct btrfs_balance_control *bctl)
 	 * (albeit full) chunks.
 	 */
 	if (!(bctl->data.flags & BTRFS_BALANCE_ARGS_USAGE) &&
+	    !(bctl->data.flags & BTRFS_BALANCE_ARGS_USAGE_RANGE) &&
 	    !(bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT)) {
 		bctl->data.flags |= BTRFS_BALANCE_ARGS_USAGE;
 		bctl->data.usage = 90;
 	}
 	if (!(bctl->sys.flags & BTRFS_BALANCE_ARGS_USAGE) &&
+	    !(bctl->sys.flags & BTRFS_BALANCE_ARGS_USAGE_RANGE) &&
 	    !(bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT)) {
 		bctl->sys.flags |= BTRFS_BALANCE_ARGS_USAGE;
 		bctl->sys.usage = 90;
 	}
 	if (!(bctl->meta.flags & BTRFS_BALANCE_ARGS_USAGE) &&
+	    !(bctl->meta.flags & BTRFS_BALANCE_ARGS_USAGE_RANGE) &&
 	    !(bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT)) {
 		bctl->meta.flags |= BTRFS_BALANCE_ARGS_USAGE;
 		bctl->meta.usage = 90;
@@ -3072,6 +3125,39 @@ static int chunk_profiles_filter(u64 chunk_type,
 
 static int chunk_usage_filter(struct btrfs_fs_info *fs_info, u64 chunk_offset,
 			      struct btrfs_balance_args *bargs)
+{
+	struct btrfs_block_group_cache *cache;
+	u64 chunk_used;
+	u64 user_thresh_min;
+	u64 user_thresh_max;
+	int ret = 1;
+
+	cache = btrfs_lookup_block_group(fs_info, chunk_offset);
+	chunk_used = btrfs_block_group_used(&cache->item);
+
+	if (bargs->usage_min == 0)
+		user_thresh_min = 0;
+	else
+		user_thresh_min = div_factor_fine(cache->key.offset,
+					bargs->usage_min);
+
+	if (bargs->usage_max == 0)
+		user_thresh_max = 1;
+	else if (bargs->usage_max > 100)
+		user_thresh_max = cache->key.offset;
+	else
+		user_thresh_max = div_factor_fine(cache->key.offset,
+					bargs->usage_max);
+
+	if (user_thresh_min <= chunk_used && chunk_used < user_thresh_max)
+		ret = 0;
+
+	btrfs_put_block_group(cache);
+	return ret;
+}
+
+static int chunk_usage_range_filter(struct btrfs_fs_info *fs_info,
+		u64 chunk_offset, struct btrfs_balance_args *bargs)
 {
 	struct btrfs_block_group_cache *cache;
 	u64 chunk_used, user_thresh;
@@ -3080,7 +3166,7 @@ static int chunk_usage_filter(struct btrfs_fs_info *fs_info, u64 chunk_offset,
 	cache = btrfs_lookup_block_group(fs_info, chunk_offset);
 	chunk_used = btrfs_block_group_used(&cache->item);
 
-	if (bargs->usage == 0)
+	if (bargs->usage_min == 0)
 		user_thresh = 1;
 	else if (bargs->usage > 100)
 		user_thresh = cache->key.offset;
@@ -3170,6 +3256,19 @@ static int chunk_vrange_filter(struct extent_buffer *leaf,
 	return 1;
 }
 
+static int chunk_stripes_range_filter(struct extent_buffer *leaf,
+			       struct btrfs_chunk *chunk,
+			       struct btrfs_balance_args *bargs)
+{
+	int num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
+
+	if (bargs->stripes_min <= num_stripes
+			&& num_stripes <= bargs->stripes_max)
+		return 0;
+
+	return 1;
+}
+
 static int chunk_soft_convert_filter(u64 chunk_type,
 				     struct btrfs_balance_args *bargs)
 {
@@ -3216,6 +3315,9 @@ static int should_balance_chunk(struct btrfs_root *root,
 	if ((bargs->flags & BTRFS_BALANCE_ARGS_USAGE) &&
 	    chunk_usage_filter(bctl->fs_info, chunk_offset, bargs)) {
 		return 0;
+	} else if ((bargs->flags & BTRFS_BALANCE_ARGS_USAGE_RANGE) &&
+	    chunk_usage_range_filter(bctl->fs_info, chunk_offset, bargs)) {
+		return 0;
 	}
 
 	/* devid filter */
@@ -3236,6 +3338,12 @@ static int should_balance_chunk(struct btrfs_root *root,
 		return 0;
 	}
 
+	/* stripes filter */
+	if ((bargs->flags & BTRFS_BALANCE_ARGS_STRIPES_RANGE) &&
+	    chunk_stripes_range_filter(leaf, chunk, bargs)) {
+		return 0;
+	}
+
 	/* soft profile changing mode */
 	if ((bargs->flags & BTRFS_BALANCE_ARGS_SOFT) &&
 	    chunk_soft_convert_filter(chunk_type, bargs)) {
@@ -3250,6 +3358,16 @@ static int should_balance_chunk(struct btrfs_root *root,
 			return 0;
 		else
 			bargs->limit--;
+	} else if ((bargs->flags & BTRFS_BALANCE_ARGS_LIMIT_RANGE)) {
+		/*
+		 * Same logic as the 'limit' filter; the minimum cannot be
+		 * determined here because we do not have the global informatoin
+		 * about the count of all chunks that satisfy the filters.
+		 */
+		if (bargs->limit_max == 0)
+			return 0;
+		else
+			bargs->limit_max--;
 	}
 
 	return 1;
@@ -3264,6 +3382,7 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info)
 	struct btrfs_device *device;
 	u64 old_size;
 	u64 size_to_free;
+	u64 chunk_type;
 	struct btrfs_chunk *chunk;
 	struct btrfs_path *path;
 	struct btrfs_key key;
@@ -3274,9 +3393,13 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info)
 	int ret;
 	int enospc_errors = 0;
 	bool counting = true;
+	/* The single value limit and min/max limits use the same bytes in the */
 	u64 limit_data = bctl->data.limit;
 	u64 limit_meta = bctl->meta.limit;
 	u64 limit_sys = bctl->sys.limit;
+	u32 count_data = 0;
+	u32 count_meta = 0;
+	u32 count_sys = 0;
 
 	/* step one make some room on all the devices */
 	devices = &fs_info->fs_devices->devices;
@@ -3317,6 +3440,10 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info)
 	spin_unlock(&fs_info->balance_lock);
 again:
 	if (!counting) {
+		/*
+		 * The single value limit and min/max limits use the same bytes
+		 * in the
+		 */
 		bctl->data.limit = limit_data;
 		bctl->meta.limit = limit_meta;
 		bctl->sys.limit = limit_sys;
@@ -3364,6 +3491,7 @@ again:
 		}
 
 		chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
+		chunk_type = btrfs_chunk_type(leaf, chunk);
 
 		if (!counting) {
 			spin_lock(&fs_info->balance_lock);
@@ -3384,6 +3512,28 @@ again:
 			spin_lock(&fs_info->balance_lock);
 			bctl->stat.expected++;
 			spin_unlock(&fs_info->balance_lock);
+
+			if (chunk_type & BTRFS_BLOCK_GROUP_DATA)
+				count_data++;
+			else if (chunk_type & BTRFS_BLOCK_GROUP_SYSTEM)
+				count_sys++;
+			else if (chunk_type & BTRFS_BLOCK_GROUP_METADATA)
+				count_meta++;
+
+			goto loop;
+		}
+
+		/*
+		 * Apply limit_min filter, no need to check if the LIMITS
+		 * filter is used, limit_min is 0 by default
+		 */
+		if (((chunk_type & BTRFS_BLOCK_GROUP_DATA) &&
+					count_data < bctl->data.limit_min)
+				|| ((chunk_type & BTRFS_BLOCK_GROUP_METADATA) &&
+					count_meta < bctl->meta.limit_min)
+				|| ((chunk_type & BTRFS_BLOCK_GROUP_SYSTEM) &&
+					count_sys < bctl->sys.limit_min)) {
+			mutex_unlock(&fs_info->delete_unused_bgs_mutex);
 			goto loop;
 		}
 
@@ -3461,11 +3611,20 @@ static void __cancel_balance(struct btrfs_fs_info *fs_info)
 	unset_balance_control(fs_info);
 	ret = del_balance_item(fs_info->tree_root);
 	if (ret)
-		btrfs_std_error(fs_info, ret);
+		btrfs_std_error(fs_info, ret, NULL);
 
 	atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
 }
 
+/* Non-zero return value signifies invalidity */
+static inline int validate_convert_profile(struct btrfs_balance_args *bctl_arg,
+		u64 allowed)
+{
+	return ((bctl_arg->flags & BTRFS_BALANCE_ARGS_CONVERT) &&
+		(!alloc_profile_is_valid(bctl_arg->target, 1) ||
+		 (bctl_arg->target & ~allowed)));
+}
+
 /*
  * Should be called with both balance and volume mutexes held
  */
@@ -3523,27 +3682,21 @@ int btrfs_balance(struct btrfs_balance_control *bctl,
 	if (num_devices > 3)
 		allowed |= (BTRFS_BLOCK_GROUP_RAID10 |
 			    BTRFS_BLOCK_GROUP_RAID6);
-	if ((bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
-	    (!alloc_profile_is_valid(bctl->data.target, 1) ||
-	     (bctl->data.target & ~allowed))) {
+	if (validate_convert_profile(&bctl->data, allowed)) {
 		btrfs_err(fs_info, "unable to start balance with target "
 			   "data profile %llu",
 		       bctl->data.target);
 		ret = -EINVAL;
 		goto out;
 	}
-	if ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
-	    (!alloc_profile_is_valid(bctl->meta.target, 1) ||
-	     (bctl->meta.target & ~allowed))) {
+	if (validate_convert_profile(&bctl->meta, allowed)) {
 		btrfs_err(fs_info,
 			   "unable to start balance with target metadata profile %llu",
 		       bctl->meta.target);
 		ret = -EINVAL;
 		goto out;
 	}
-	if ((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
-	    (!alloc_profile_is_valid(bctl->sys.target, 1) ||
-	     (bctl->sys.target & ~allowed))) {
+	if (validate_convert_profile(&bctl->sys, allowed)) {
 		btrfs_err(fs_info,
 			   "unable to start balance with target system profile %llu",
 		       bctl->sys.target);
@@ -4285,65 +4438,6 @@ static int btrfs_cmp_device_info(const void *a, const void *b)
 	return 0;
 }
 
-static const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
-	[BTRFS_RAID_RAID10] = {
-		.sub_stripes	= 2,
-		.dev_stripes	= 1,
-		.devs_max	= 0,	/* 0 == as many as possible */
-		.devs_min	= 4,
-		.devs_increment	= 2,
-		.ncopies	= 2,
-	},
-	[BTRFS_RAID_RAID1] = {
-		.sub_stripes	= 1,
-		.dev_stripes	= 1,
-		.devs_max	= 2,
-		.devs_min	= 2,
-		.devs_increment	= 2,
-		.ncopies	= 2,
-	},
-	[BTRFS_RAID_DUP] = {
-		.sub_stripes	= 1,
-		.dev_stripes	= 2,
-		.devs_max	= 1,
-		.devs_min	= 1,
-		.devs_increment	= 1,
-		.ncopies	= 2,
-	},
-	[BTRFS_RAID_RAID0] = {
-		.sub_stripes	= 1,
-		.dev_stripes	= 1,
-		.devs_max	= 0,
-		.devs_min	= 2,
-		.devs_increment	= 1,
-		.ncopies	= 1,
-	},
-	[BTRFS_RAID_SINGLE] = {
-		.sub_stripes	= 1,
-		.dev_stripes	= 1,
-		.devs_max	= 1,
-		.devs_min	= 1,
-		.devs_increment	= 1,
-		.ncopies	= 1,
-	},
-	[BTRFS_RAID_RAID5] = {
-		.sub_stripes	= 1,
-		.dev_stripes	= 1,
-		.devs_max	= 0,
-		.devs_min	= 2,
-		.devs_increment	= 1,
-		.ncopies	= 2,
-	},
-	[BTRFS_RAID_RAID6] = {
-		.sub_stripes	= 1,
-		.dev_stripes	= 1,
-		.devs_max	= 0,
-		.devs_min	= 3,
-		.devs_increment	= 1,
-		.ncopies	= 3,
-	},
-};
-
 static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target)
 {
 	/* TODO allow them to set a preferred stripe size */
@@ -6594,8 +6688,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 	BUG_ON(!path);
 	ret = btrfs_search_slot(trans, dev_root, &key, path, -1, 1);
 	if (ret < 0) {
-		printk_in_rcu(KERN_WARNING "BTRFS: "
-			"error %d while searching for dev_stats item for device %s!\n",
+		btrfs_warn_in_rcu(dev_root->fs_info,
+			"error %d while searching for dev_stats item for device %s",
 			      ret, rcu_str_deref(device->name));
 		goto out;
 	}
@@ -6605,8 +6699,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 		/* need to delete old one and insert a new one */
 		ret = btrfs_del_item(trans, dev_root, path);
 		if (ret != 0) {
-			printk_in_rcu(KERN_WARNING "BTRFS: "
-				"delete too small dev_stats item for device %s failed %d!\n",
+			btrfs_warn_in_rcu(dev_root->fs_info,
+				"delete too small dev_stats item for device %s failed %d",
 				      rcu_str_deref(device->name), ret);
 			goto out;
 		}
@@ -6619,9 +6713,9 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 		ret = btrfs_insert_empty_item(trans, dev_root, path,
 					      &key, sizeof(*ptr));
 		if (ret < 0) {
-			printk_in_rcu(KERN_WARNING "BTRFS: "
-					  "insert dev_stats item for device %s failed %d!\n",
-				      rcu_str_deref(device->name), ret);
+			btrfs_warn_in_rcu(dev_root->fs_info,
+				"insert dev_stats item for device %s failed %d",
+				rcu_str_deref(device->name), ret);
 			goto out;
 		}
 	}
@@ -6675,8 +6769,8 @@ static void btrfs_dev_stat_print_on_error(struct btrfs_device *dev)
 {
 	if (!dev->dev_stats_valid)
 		return;
-	printk_ratelimited_in_rcu(KERN_ERR "BTRFS: "
-			   "bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
+	btrfs_err_rl_in_rcu(dev->dev_root->fs_info,
+		"bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u",
 			   rcu_str_deref(dev->name),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
@@ -6695,8 +6789,8 @@ static void btrfs_dev_stat_print_on_load(struct btrfs_device *dev)
 	if (i == BTRFS_DEV_STAT_VALUES_MAX)
 		return; /* all values == 0, suppress message */
 
-	printk_in_rcu(KERN_INFO "BTRFS: "
-		   "bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
+	btrfs_info_in_rcu(dev->dev_root->fs_info,
+		"bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u",
 	       rcu_str_deref(dev->name),
 	       btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
 	       btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
@@ -6740,22 +6834,34 @@ int btrfs_get_dev_stats(struct btrfs_root *root,
 	return 0;
 }
 
-int btrfs_scratch_superblock(struct btrfs_device *device)
+void btrfs_scratch_superblocks(struct block_device *bdev, char *device_path)
 {
 	struct buffer_head *bh;
 	struct btrfs_super_block *disk_super;
+	int copy_num;
 
-	bh = btrfs_read_dev_super(device->bdev);
-	if (!bh)
-		return -EINVAL;
-	disk_super = (struct btrfs_super_block *)bh->b_data;
+	if (!bdev)
+		return;
 
-	memset(&disk_super->magic, 0, sizeof(disk_super->magic));
-	set_buffer_dirty(bh);
-	sync_dirty_buffer(bh);
-	brelse(bh);
+	for (copy_num = 0; copy_num < BTRFS_SUPER_MIRROR_MAX;
+		copy_num++) {
 
-	return 0;
+		if (btrfs_read_dev_one_super(bdev, copy_num, &bh))
+			continue;
+
+		disk_super = (struct btrfs_super_block *)bh->b_data;
+
+		memset(&disk_super->magic, 0, sizeof(disk_super->magic));
+		set_buffer_dirty(bh);
+		sync_dirty_buffer(bh);
+		brelse(bh);
+	}
+
+	/* Notify udev that device has changed */
+	btrfs_kobject_uevent(bdev, KOBJ_CHANGE);
+
+	/* Update ctime/mtime for device path for libblkid */
+	update_dev_time(device_path);
 }
 
 /*
@@ -6823,3 +6929,38 @@ void btrfs_reset_fs_info_ptr(struct btrfs_fs_info *fs_info)
 		fs_devices = fs_devices->seed;
 	}
 }
+
+void btrfs_close_one_device(struct btrfs_device *device)
+{
+	struct btrfs_fs_devices *fs_devices = device->fs_devices;
+	struct btrfs_device *new_device;
+	struct rcu_string *name;
+
+	if (device->bdev)
+		fs_devices->open_devices--;
+
+	if (device->writeable &&
+	    device->devid != BTRFS_DEV_REPLACE_DEVID) {
+		list_del_init(&device->dev_alloc_list);
+		fs_devices->rw_devices--;
+	}
+
+	if (device->missing)
+		fs_devices->missing_devices--;
+
+	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(!name); /* -ENOMEM */
+		rcu_assign_pointer(new_device->name, name);
+	}
+
+	list_replace_rcu(&device->dev_list, &new_device->dev_list);
+	new_device->fs_devices = device->fs_devices;
+
+	call_rcu(&device->rcu, free_device);
+}

fs/btrfs/volumes.h

@@ -256,7 +256,7 @@ struct btrfs_fs_devices {
 
 	struct btrfs_fs_info *fs_info;
 	/* sysfs kobjects */
-	struct kobject super_kobj;
+	struct kobject fsid_kobj;
 	struct kobject *device_dir_kobj;
 	struct completion kobj_unregister;
 };
@@ -334,10 +334,15 @@ struct btrfs_raid_attr {
 	int dev_stripes;	/* stripes per dev */
 	int devs_max;		/* max devs to use */
 	int devs_min;		/* min devs needed */
+	int tolerated_failures; /* max tolerated fail devs */
 	int devs_increment;	/* ndevs has to be a multiple of this */
 	int ncopies;		/* how many copies to data has */
 };
 
+extern const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES];
+
+extern const u64 btrfs_raid_group[BTRFS_NR_RAID_TYPES];
+
 struct map_lookup {
 	u64 type;
 	int io_align;
@@ -375,6 +380,9 @@ struct map_lookup {
 #define BTRFS_BALANCE_ARGS_DRANGE	(1ULL << 3)
 #define BTRFS_BALANCE_ARGS_VRANGE	(1ULL << 4)
 #define BTRFS_BALANCE_ARGS_LIMIT	(1ULL << 5)
+#define BTRFS_BALANCE_ARGS_LIMIT_RANGE	(1ULL << 6)
+#define BTRFS_BALANCE_ARGS_STRIPES_RANGE (1ULL << 7)
+#define BTRFS_BALANCE_ARGS_USAGE_RANGE	(1ULL << 8)
 
 #define BTRFS_BALANCE_ARGS_MASK			\
 	(BTRFS_BALANCE_ARGS_PROFILES |		\
@@ -382,7 +390,10 @@ struct map_lookup {
 	 BTRFS_BALANCE_ARGS_DEVID | 		\
 	 BTRFS_BALANCE_ARGS_DRANGE |		\
 	 BTRFS_BALANCE_ARGS_VRANGE |		\
-	 BTRFS_BALANCE_ARGS_LIMIT)
+	 BTRFS_BALANCE_ARGS_LIMIT |		\
+	 BTRFS_BALANCE_ARGS_LIMIT_RANGE |	\
+	 BTRFS_BALANCE_ARGS_STRIPES_RANGE |	\
+	 BTRFS_BALANCE_ARGS_USAGE_RANGE)
 
 /*
  * Profile changing flags.  When SOFT is set we won't relocate chunk if
@@ -482,7 +493,7 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 				      struct btrfs_device *tgtdev);
 void btrfs_init_dev_replace_tgtdev_for_resume(struct btrfs_fs_info *fs_info,
 					      struct btrfs_device *tgtdev);
-int btrfs_scratch_superblock(struct btrfs_device *device);
+void btrfs_scratch_superblocks(struct block_device *bdev, char *device_path);
 int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree,
 			   u64 logical, u64 len, int mirror_num);
 unsigned long btrfs_full_stripe_len(struct btrfs_root *root,
@@ -555,5 +566,6 @@ static inline void unlock_chunks(struct btrfs_root *root)
 struct list_head *btrfs_get_fs_uuids(void);
 void btrfs_set_fs_info_ptr(struct btrfs_fs_info *fs_info);
 void btrfs_reset_fs_info_ptr(struct btrfs_fs_info *fs_info);
+void btrfs_close_one_device(struct btrfs_device *device);
 
 #endif

include/trace/events/btrfs.h

@@ -1117,6 +1117,119 @@ DEFINE_EVENT(btrfs__workqueue_done, btrfs_workqueue_destroy,
 	TP_ARGS(wq)
 );
 
+DECLARE_EVENT_CLASS(btrfs__qgroup_data_map,
+
+	TP_PROTO(struct inode *inode, u64 free_reserved),
+
+	TP_ARGS(inode, free_reserved),
+
+	TP_STRUCT__entry(
+		__field(	u64,		rootid		)
+		__field(	unsigned long,	ino		)
+		__field(	u64,		free_reserved	)
+	),
+
+	TP_fast_assign(
+		__entry->rootid		=	BTRFS_I(inode)->root->objectid;
+		__entry->ino		=	inode->i_ino;
+		__entry->free_reserved	=	free_reserved;
+	),
+
+	TP_printk("rootid=%llu, ino=%lu, free_reserved=%llu",
+		  __entry->rootid, __entry->ino, __entry->free_reserved)
+);
+
+DEFINE_EVENT(btrfs__qgroup_data_map, btrfs_qgroup_init_data_rsv_map,
+
+	TP_PROTO(struct inode *inode, u64 free_reserved),
+
+	TP_ARGS(inode, free_reserved)
+);
+
+DEFINE_EVENT(btrfs__qgroup_data_map, btrfs_qgroup_free_data_rsv_map,
+
+	TP_PROTO(struct inode *inode, u64 free_reserved),
+
+	TP_ARGS(inode, free_reserved)
+);
+
+#define BTRFS_QGROUP_OPERATIONS				\
+	{ QGROUP_RESERVE,	"reserve"	},	\
+	{ QGROUP_RELEASE,	"release"	},	\
+	{ QGROUP_FREE,		"free"		}
+
+DECLARE_EVENT_CLASS(btrfs__qgroup_rsv_data,
+
+	TP_PROTO(struct inode *inode, u64 start, u64 len, u64 reserved, int op),
+
+	TP_ARGS(inode, start, len, reserved, op),
+
+	TP_STRUCT__entry(
+		__field(	u64,		rootid		)
+		__field(	unsigned long,	ino		)
+		__field(	u64,		start		)
+		__field(	u64,		len		)
+		__field(	u64,		reserved	)
+		__field(	int,		op		)
+	),
+
+	TP_fast_assign(
+		__entry->rootid		= BTRFS_I(inode)->root->objectid;
+		__entry->ino		= inode->i_ino;
+		__entry->start		= start;
+		__entry->len		= len;
+		__entry->reserved	= reserved;
+		__entry->op		= op;
+	),
+
+	TP_printk("root=%llu, ino=%lu, start=%llu, len=%llu, reserved=%llu, op=%s",
+		  __entry->rootid, __entry->ino, __entry->start, __entry->len,
+		  __entry->reserved,
+		  __print_flags((unsigned long)__entry->op, "",
+				BTRFS_QGROUP_OPERATIONS)
+	)
+);
+
+DEFINE_EVENT(btrfs__qgroup_rsv_data, btrfs_qgroup_reserve_data,
+
+	TP_PROTO(struct inode *inode, u64 start, u64 len, u64 reserved, int op),
+
+	TP_ARGS(inode, start, len, reserved, op)
+);
+
+DEFINE_EVENT(btrfs__qgroup_rsv_data, btrfs_qgroup_release_data,
+
+	TP_PROTO(struct inode *inode, u64 start, u64 len, u64 reserved, int op),
+
+	TP_ARGS(inode, start, len, reserved, op)
+);
+
+DECLARE_EVENT_CLASS(btrfs__qgroup_delayed_ref,
+
+	TP_PROTO(u64 ref_root, u64 reserved),
+
+	TP_ARGS(ref_root, reserved),
+
+	TP_STRUCT__entry(
+		__field(	u64,		ref_root	)
+		__field(	u64,		reserved	)
+	),
+
+	TP_fast_assign(
+		__entry->ref_root	= ref_root;
+		__entry->reserved	= reserved;
+	),
+
+	TP_printk("root=%llu, reserved=%llu, op=free",
+		  __entry->ref_root, __entry->reserved)
+);
+
+DEFINE_EVENT(btrfs__qgroup_delayed_ref, btrfs_qgroup_free_delayed_ref,
+
+	TP_PROTO(u64 ref_root, u64 reserved),
+
+	TP_ARGS(ref_root, reserved)
+);
 #endif /* _TRACE_BTRFS_H */
 
 /* This part must be outside protection */

include/uapi/linux/btrfs.h

@@ -206,7 +206,13 @@ struct btrfs_ioctl_feature_flags {
  */
 struct btrfs_balance_args {
 	__u64 profiles;
-	__u64 usage;
+	union {
+		__le64 usage;
+		struct {
+			__le32 usage_min;
+			__le32 usage_max;
+		};
+	};
 	__u64 devid;
 	__u64 pstart;
 	__u64 pend;
@@ -217,8 +223,27 @@ struct btrfs_balance_args {
 
 	__u64 flags;
 
-	__u64 limit;		/* limit number of processed chunks */
-	__u64 unused[7];
+	/*
+	 * BTRFS_BALANCE_ARGS_LIMIT with value 'limit'
+	 * BTRFS_BALANCE_ARGS_LIMIT_RANGE - the extend version can use minimum
+	 * and maximum
+	 */
+	union {
+		__u64 limit;		/* limit number of processed chunks */
+		struct {
+			__u32 limit_min;
+			__u32 limit_max;
+		};
+	};
+
+	/*
+	 * Process chunks that cross stripes_min..stripes_max devices,
+	 * BTRFS_BALANCE_ARGS_STRIPES_RANGE
+	 */
+	__le32 stripes_min;
+	__le32 stripes_max;
+
+	__u64 unused[6];
 } __attribute__ ((__packed__));
 
 /* report balance progress to userspace */