1 files changed, 400 insertions, 275 deletions

diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c
index e807b143b85..6d16bea94e1 100644
--- a/fs/btrfs/delayed-ref.c
+++ b/fs/btrfs/delayed-ref.c
@@ -23,6 +23,10 @@
 #include "delayed-ref.h"
 #include "transaction.h"
 
+struct kmem_cache *btrfs_delayed_ref_head_cachep;
+struct kmem_cache *btrfs_delayed_tree_ref_cachep;
+struct kmem_cache *btrfs_delayed_data_ref_cachep;
+struct kmem_cache *btrfs_delayed_extent_op_cachep;
 /*
  * delayed back reference update tracking.  For subvolume trees
  * we queue up extent allocations and backref maintenance for
@@ -36,9 +40,9 @@
  * compare two delayed tree backrefs with same bytenr and type
  */
 static int comp_tree_refs(struct btrfs_delayed_tree_ref *ref2,
-			  struct btrfs_delayed_tree_ref *ref1)
+			  struct btrfs_delayed_tree_ref *ref1, int type)
 {
-	if (ref1->node.type == BTRFS_TREE_BLOCK_REF_KEY) {
+	if (type == BTRFS_TREE_BLOCK_REF_KEY) {
 		if (ref1->root < ref2->root)
 			return -1;
 		if (ref1->root > ref2->root)
@@ -85,7 +89,8 @@ static int comp_data_refs(struct btrfs_delayed_data_ref *ref2,
  * type of the delayed backrefs and content of delayed backrefs.
  */
 static int comp_entry(struct btrfs_delayed_ref_node *ref2,
-		      struct btrfs_delayed_ref_node *ref1)
+		      struct btrfs_delayed_ref_node *ref1,
+		      bool compare_seq)
 {
 	if (ref1->bytenr < ref2->bytenr)
 		return -1;
@@ -101,10 +106,22 @@ static int comp_entry(struct btrfs_delayed_ref_node *ref2,
 		return -1;
 	if (ref1->type > ref2->type)
 		return 1;
+	if (ref1->no_quota > ref2->no_quota)
+		return 1;
+	if (ref1->no_quota < ref2->no_quota)
+		return -1;
+	/* merging of sequenced refs is not allowed */
+	if (compare_seq) {
+		if (ref1->seq < ref2->seq)
+			return -1;
+		if (ref1->seq > ref2->seq)
+			return 1;
+	}
 	if (ref1->type == BTRFS_TREE_BLOCK_REF_KEY ||
 	    ref1->type == BTRFS_SHARED_BLOCK_REF_KEY) {
 		return comp_tree_refs(btrfs_delayed_node_to_tree_ref(ref2),
-				      btrfs_delayed_node_to_tree_ref(ref1));
+				      btrfs_delayed_node_to_tree_ref(ref1),
+				      ref1->type);
 	} else if (ref1->type == BTRFS_EXTENT_DATA_REF_KEY ||
 		   ref1->type == BTRFS_SHARED_DATA_REF_KEY) {
 		return comp_data_refs(btrfs_delayed_node_to_data_ref(ref2),
@@ -134,7 +151,7 @@ static struct btrfs_delayed_ref_node *tree_insert(struct rb_root *root,
 		entry = rb_entry(parent_node, struct btrfs_delayed_ref_node,
 				 rb_node);
 
-		cmp = comp_entry(entry, ins);
+		cmp = comp_entry(entry, ins, 1);
 		if (cmp < 0)
 			p = &(*p)->rb_left;
 		else if (cmp > 0)
@@ -148,40 +165,72 @@ static struct btrfs_delayed_ref_node *tree_insert(struct rb_root *root,
 	return NULL;
 }
 
+/* insert a new ref to head ref rbtree */
+static struct btrfs_delayed_ref_head *htree_insert(struct rb_root *root,
+						   struct rb_node *node)
+{
+	struct rb_node **p = &root->rb_node;
+	struct rb_node *parent_node = NULL;
+	struct btrfs_delayed_ref_head *entry;
+	struct btrfs_delayed_ref_head *ins;
+	u64 bytenr;
+
+	ins = rb_entry(node, struct btrfs_delayed_ref_head, href_node);
+	bytenr = ins->node.bytenr;
+	while (*p) {
+		parent_node = *p;
+		entry = rb_entry(parent_node, struct btrfs_delayed_ref_head,
+				 href_node);
+
+		if (bytenr < entry->node.bytenr)
+			p = &(*p)->rb_left;
+		else if (bytenr > entry->node.bytenr)
+			p = &(*p)->rb_right;
+		else
+			return entry;
+	}
+
+	rb_link_node(node, parent_node, p);
+	rb_insert_color(node, root);
+	return NULL;
+}
+
 /*
  * find an head entry based on bytenr. This returns the delayed ref
- * head if it was able to find one, or NULL if nothing was in that spot
+ * head if it was able to find one, or NULL if nothing was in that spot.
+ * If return_bigger is given, the next bigger entry is returned if no exact
+ * match is found.
  */
-static struct btrfs_delayed_ref_node *find_ref_head(struct rb_root *root,
-				  u64 bytenr,
-				  struct btrfs_delayed_ref_node **last)
+static struct btrfs_delayed_ref_head *
+find_ref_head(struct rb_root *root, u64 bytenr,
+	      int return_bigger)
 {
-	struct rb_node *n = root->rb_node;
-	struct btrfs_delayed_ref_node *entry;
-	int cmp;
+	struct rb_node *n;
+	struct btrfs_delayed_ref_head *entry;
 
+	n = root->rb_node;
+	entry = NULL;
 	while (n) {
-		entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node);
-		WARN_ON(!entry->in_tree);
-		if (last)
-			*last = entry;
-
-		if (bytenr < entry->bytenr)
-			cmp = -1;
-		else if (bytenr > entry->bytenr)
-			cmp = 1;
-		else if (!btrfs_delayed_ref_is_head(entry))
-			cmp = 1;
-		else
-			cmp = 0;
+		entry = rb_entry(n, struct btrfs_delayed_ref_head, href_node);
 
-		if (cmp < 0)
+		if (bytenr < entry->node.bytenr)
 			n = n->rb_left;
-		else if (cmp > 0)
+		else if (bytenr > entry->node.bytenr)
 			n = n->rb_right;
 		else
 			return entry;
 	}
+	if (entry && return_bigger) {
+		if (bytenr > entry->node.bytenr) {
+			n = rb_next(&entry->href_node);
+			if (!n)
+				n = rb_first(root);
+			entry = rb_entry(n, struct btrfs_delayed_ref_head,
+					 href_node);
+			return entry;
+		}
+		return entry;
+	}
 	return NULL;
 }
 
@@ -209,115 +258,190 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans,
-			   struct list_head *cluster, u64 start)
+static inline void drop_delayed_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)
 {
-	int count = 0;
-	struct btrfs_delayed_ref_root *delayed_refs;
-	struct rb_node *node;
-	struct btrfs_delayed_ref_node *ref;
-	struct btrfs_delayed_ref_head *head;
-
-	delayed_refs = &trans->transaction->delayed_refs;
-	if (start == 0) {
-		node = rb_first(&delayed_refs->root);
+	if (btrfs_delayed_ref_is_head(ref)) {
+		head = btrfs_delayed_node_to_head(ref);
+		rb_erase(&head->href_node, &delayed_refs->href_root);
 	} else {
-		ref = NULL;
-		find_ref_head(&delayed_refs->root, start, &ref);
-		if (ref) {
-			struct btrfs_delayed_ref_node *tmp;
-
-			node = rb_prev(&ref->rb_node);
-			while (node) {
-				tmp = rb_entry(node,
-					       struct btrfs_delayed_ref_node,
-					       rb_node);
-				if (tmp->bytenr < start)
-					break;
-				ref = tmp;
-				node = rb_prev(&ref->rb_node);
-			}
-			node = &ref->rb_node;
-		} else
-			node = rb_first(&delayed_refs->root);
+		assert_spin_locked(&head->lock);
+		rb_erase(&ref->rb_node, &head->ref_root);
 	}
-again:
-	while (node && count < 32) {
-		ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-		if (btrfs_delayed_ref_is_head(ref)) {
-			head = btrfs_delayed_node_to_head(ref);
-			if (list_empty(&head->cluster)) {
-				list_add_tail(&head->cluster, cluster);
-				delayed_refs->run_delayed_start =
-					head->node.bytenr;
-				count++;
-
-				WARN_ON(delayed_refs->num_heads_ready == 0);
-				delayed_refs->num_heads_ready--;
-			} else if (count) {
-				/* the goal of the clustering is to find extents
-				 * that are likely to end up in the same extent
-				 * leaf on disk.  So, we don't want them spread
-				 * all over the tree.  Stop now if we've hit
-				 * a head that was already in use
-				 */
-				break;
+	ref->in_tree = 0;
+	btrfs_put_delayed_ref(ref);
+	atomic_dec(&delayed_refs->num_entries);
+	if (trans->delayed_ref_updates)
+		trans->delayed_ref_updates--;
+}
+
+static int 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 rb_node *node;
+	int mod = 0;
+	int done = 0;
+
+	node = rb_next(&ref->rb_node);
+	while (!done && node) {
+		struct btrfs_delayed_ref_node *next;
+
+		next = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
+		node = rb_next(node);
+		if (seq && next->seq >= seq)
+			break;
+		if (comp_entry(ref, next, 0))
+			continue;
+
+		if (ref->action == next->action) {
+			mod = next->ref_mod;
+		} else {
+			if (ref->ref_mod < next->ref_mod) {
+				struct btrfs_delayed_ref_node *tmp;
+
+				tmp = ref;
+				ref = next;
+				next = tmp;
+				done = 1;
 			}
+			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 = 1;
+		} else {
+			/*
+			 * You 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);
 		}
-		node = rb_next(node);
-	}
-	if (count) {
-		return 0;
-	} else if (start) {
-		/*
-		 * we've gone to the end of the rbtree without finding any
-		 * clusters.  start from the beginning and try again
-		 */
-		start = 0;
-		node = rb_first(&delayed_refs->root);
-		goto again;
 	}
-	return 1;
+	return done;
 }
 
-/*
- * This checks to see if there are any delayed refs in the
- * btree for a given bytenr.  It returns one if it finds any
- * and zero otherwise.
- *
- * If it only finds a head node, it returns 0.
- *
- * The idea is to use this when deciding if you can safely delete an
- * extent from the extent allocation tree.  There may be a pending
- * ref in the rbtree that adds or removes references, so as long as this
- * returns one you need to leave the BTRFS_EXTENT_ITEM in the extent
- * allocation tree.
- */
-int btrfs_delayed_ref_pending(struct btrfs_trans_handle *trans, u64 bytenr)
+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;
-	struct btrfs_delayed_ref_root *delayed_refs;
-	struct rb_node *prev_node;
-	int ret = 0;
+	struct rb_node *node;
+	u64 seq = 0;
 
-	delayed_refs = &trans->transaction->delayed_refs;
-	spin_lock(&delayed_refs->lock);
+	assert_spin_locked(&head->lock);
+	/*
+	 * We don't have too much refs to merge in the case of delayed data
+	 * refs.
+	 */
+	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 = find_ref_head(&delayed_refs->root, bytenr, NULL);
-	if (ref) {
-		prev_node = rb_prev(&ref->rb_node);
-		if (!prev_node)
-			goto out;
-		ref = rb_entry(prev_node, struct btrfs_delayed_ref_node,
+	node = rb_first(&head->ref_root);
+	while (node) {
+		struct btrfs_delayed_ref_node *ref;
+
+		ref = rb_entry(node, struct btrfs_delayed_ref_node,
 			       rb_node);
-		if (ref->bytenr == bytenr)
+		/* We can't merge refs that are outside of our seq count */
+		if (seq && ref->seq >= seq)
+			break;
+		if (merge_ref(trans, delayed_refs, head, ref, seq))
+			node = rb_first(&head->ref_root);
+		else
+			node = rb_next(&ref->rb_node);
+	}
+}
+
+int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info,
+			    struct btrfs_delayed_ref_root *delayed_refs,
+			    u64 seq)
+{
+	struct seq_list *elem;
+	int ret = 0;
+
+	spin_lock(&fs_info->tree_mod_seq_lock);
+	if (!list_empty(&fs_info->tree_mod_seq_list)) {
+		elem = list_first_entry(&fs_info->tree_mod_seq_list,
+					struct seq_list, list);
+		if (seq >= elem->seq) {
+			pr_debug("holding back delayed_ref %#x.%x, lowest is %#x.%x (%p)\n",
+				 (u32)(seq >> 32), (u32)seq,
+				 (u32)(elem->seq >> 32), (u32)elem->seq,
+				 delayed_refs);
 			ret = 1;
+		}
 	}
-out:
-	spin_unlock(&delayed_refs->lock);
+
+	spin_unlock(&fs_info->tree_mod_seq_lock);
 	return ret;
 }
 
+struct btrfs_delayed_ref_head *
+btrfs_select_ref_head(struct btrfs_trans_handle *trans)
+{
+	struct btrfs_delayed_ref_root *delayed_refs;
+	struct btrfs_delayed_ref_head *head;
+	u64 start;
+	bool loop = false;
+
+	delayed_refs = &trans->transaction->delayed_refs;
+
+again:
+	start = delayed_refs->run_delayed_start;
+	head = find_ref_head(&delayed_refs->href_root, start, 1);
+	if (!head && !loop) {
+		delayed_refs->run_delayed_start = 0;
+		start = 0;
+		loop = true;
+		head = find_ref_head(&delayed_refs->href_root, start, 1);
+		if (!head)
+			return NULL;
+	} else if (!head && loop) {
+		return NULL;
+	}
+
+	while (head->processing) {
+		struct rb_node *node;
+
+		node = rb_next(&head->href_node);
+		if (!node) {
+			if (loop)
+				return NULL;
+			delayed_refs->run_delayed_start = 0;
+			start = 0;
+			loop = true;
+			goto again;
+		}
+		head = rb_entry(node, struct btrfs_delayed_ref_head,
+				href_node);
+	}
+
+	head->processing = 1;
+	WARN_ON(delayed_refs->num_heads_ready == 0);
+	delayed_refs->num_heads_ready--;
+	delayed_refs->run_delayed_start = head->node.bytenr +
+		head->node.num_bytes;
+	return head;
+}
+
 /*
  * helper function to update an extent delayed ref in the
  * rbtree.  existing and update must both have the same
@@ -329,6 +453,7 @@ out:
 static noinline void
 update_existing_ref(struct btrfs_trans_handle *trans,
 		    struct btrfs_delayed_ref_root *delayed_refs,
+		    struct btrfs_delayed_ref_head *head,
 		    struct btrfs_delayed_ref_node *existing,
 		    struct btrfs_delayed_ref_node *update)
 {
@@ -340,18 +465,11 @@ update_existing_ref(struct btrfs_trans_handle *trans,
 		 * every changing the extent allocation tree.
 		 */
 		existing->ref_mod--;
-		if (existing->ref_mod == 0) {
-			rb_erase(&existing->rb_node,
-				 &delayed_refs->root);
-			existing->in_tree = 0;
-			btrfs_put_delayed_ref(existing);
-			delayed_refs->num_entries--;
-			if (trans->delayed_ref_updates)
-				trans->delayed_ref_updates--;
-		} else {
+		if (existing->ref_mod == 0)
+			drop_delayed_ref(trans, delayed_refs, head, existing);
+		else
 			WARN_ON(existing->type == BTRFS_TREE_BLOCK_REF_KEY ||
 				existing->type == BTRFS_SHARED_BLOCK_REF_KEY);
-		}
 	} else {
 		WARN_ON(existing->type == BTRFS_TREE_BLOCK_REF_KEY ||
 			existing->type == BTRFS_SHARED_BLOCK_REF_KEY);
@@ -381,6 +499,7 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing,
 	ref = btrfs_delayed_node_to_head(update);
 	BUG_ON(existing_ref->is_data != ref->is_data);
 
+	spin_lock(&existing_ref->lock);
 	if (ref->must_insert_reserved) {
 		/* if the extent was freed and then
 		 * reallocated before the delayed ref
@@ -414,13 +533,16 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing,
 					ref->extent_op->flags_to_set;
 				existing_ref->extent_op->update_flags = 1;
 			}
-			kfree(ref->extent_op);
+			btrfs_free_delayed_extent_op(ref->extent_op);
 		}
 	}
 	/*
-	 * update the reference mod on the head to reflect this new operation
+	 * update the reference mod on the head to reflect this new operation,
+	 * only need the lock for this case cause we could be processing it
+	 * currently, for refs we just added we know we're a-ok.
 	 */
 	existing->ref_mod += update->ref_mod;
+	spin_unlock(&existing_ref->lock);
 }
 
 /*
@@ -428,12 +550,13 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing,
  * this does all the dirty work in terms of maintaining the correct
  * overall modification count.
  */
-static noinline int add_delayed_ref_head(struct btrfs_trans_handle *trans,
-					struct btrfs_delayed_ref_node *ref,
-					u64 bytenr, u64 num_bytes,
-					int action, int is_data)
+static noinline struct btrfs_delayed_ref_head *
+add_delayed_ref_head(struct btrfs_fs_info *fs_info,
+		     struct btrfs_trans_handle *trans,
+		     struct btrfs_delayed_ref_node *ref, u64 bytenr,
+		     u64 num_bytes, int action, int is_data)
 {
-	struct btrfs_delayed_ref_node *existing;
+	struct btrfs_delayed_ref_head *existing;
 	struct btrfs_delayed_ref_head *head_ref = NULL;
 	struct btrfs_delayed_ref_root *delayed_refs;
 	int count_mod = 1;
@@ -475,47 +598,59 @@ static noinline int add_delayed_ref_head(struct btrfs_trans_handle *trans,
 	ref->action  = 0;
 	ref->is_head = 1;
 	ref->in_tree = 1;
+	ref->seq = 0;
 
 	head_ref = btrfs_delayed_node_to_head(ref);
 	head_ref->must_insert_reserved = must_insert_reserved;
 	head_ref->is_data = is_data;
+	head_ref->ref_root = RB_ROOT;
+	head_ref->processing = 0;
 
-	INIT_LIST_HEAD(&head_ref->cluster);
+	spin_lock_init(&head_ref->lock);
 	mutex_init(&head_ref->mutex);
 
-	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
+	trace_add_delayed_ref_head(ref, head_ref, action);
 
+	existing = htree_insert(&delayed_refs->href_root,
+				&head_ref->href_node);
 	if (existing) {
-		update_existing_head_ref(existing, ref);
+		update_existing_head_ref(&existing->node, ref);
 		/*
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
-		kfree(ref);
+		kmem_cache_free(btrfs_delayed_ref_head_cachep, head_ref);
+		head_ref = existing;
 	} else {
 		delayed_refs->num_heads++;
 		delayed_refs->num_heads_ready++;
-		delayed_refs->num_entries++;
+		atomic_inc(&delayed_refs->num_entries);
 		trans->delayed_ref_updates++;
 	}
-	return 0;
+	return head_ref;
 }
 
 /*
  * helper to insert a delayed tree ref into the rbtree.
  */
-static noinline int add_delayed_tree_ref(struct btrfs_trans_handle *trans,
-					 struct btrfs_delayed_ref_node *ref,
-					 u64 bytenr, u64 num_bytes, u64 parent,
-					 u64 ref_root, int level, int action)
+static noinline void
+add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
+		     struct btrfs_trans_handle *trans,
+		     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)
 {
 	struct btrfs_delayed_ref_node *existing;
 	struct btrfs_delayed_tree_ref *full_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
+	u64 seq = 0;
 
 	if (action == BTRFS_ADD_DELAYED_EXTENT)
 		action = BTRFS_ADD_DELAYED_REF;
 
+	if (is_fstree(ref_root))
+		seq = atomic64_read(&fs_info->tree_mod_seq);
 	delayed_refs = &trans->transaction->delayed_refs;
 
 	/* first set the basic ref node struct up */
@@ -526,51 +661,61 @@ static noinline int add_delayed_tree_ref(struct btrfs_trans_handle *trans,
 	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);
-	if (parent) {
-		full_ref->parent = parent;
+	full_ref->parent = parent;
+	full_ref->root = ref_root;
+	if (parent)
 		ref->type = BTRFS_SHARED_BLOCK_REF_KEY;
-	} else {
-		full_ref->root = ref_root;
+	else
 		ref->type = BTRFS_TREE_BLOCK_REF_KEY;
-	}
 	full_ref->level = level;
 
-	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
+	trace_add_delayed_tree_ref(ref, full_ref, action);
 
+	spin_lock(&head_ref->lock);
+	existing = tree_insert(&head_ref->ref_root, &ref->rb_node);
 	if (existing) {
-		update_existing_ref(trans, delayed_refs, existing, ref);
+		update_existing_ref(trans, delayed_refs, head_ref, existing,
+				    ref);
 		/*
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
-		kfree(ref);
+		kmem_cache_free(btrfs_delayed_tree_ref_cachep, full_ref);
 	} else {
-		delayed_refs->num_entries++;
+		atomic_inc(&delayed_refs->num_entries);
 		trans->delayed_ref_updates++;
 	}
-	return 0;
+	spin_unlock(&head_ref->lock);
 }
 
 /*
  * helper to insert a delayed data ref into the rbtree.
  */
-static noinline int add_delayed_data_ref(struct btrfs_trans_handle *trans,
-					 struct btrfs_delayed_ref_node *ref,
-					 u64 bytenr, u64 num_bytes, u64 parent,
-					 u64 ref_root, u64 owner, u64 offset,
-					 int action)
+static noinline void
+add_delayed_data_ref(struct btrfs_fs_info *fs_info,
+		     struct btrfs_trans_handle *trans,
+		     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)
 {
 	struct btrfs_delayed_ref_node *existing;
 	struct btrfs_delayed_data_ref *full_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
+	u64 seq = 0;
 
 	if (action == BTRFS_ADD_DELAYED_EXTENT)
 		action = BTRFS_ADD_DELAYED_REF;
 
 	delayed_refs = &trans->transaction->delayed_refs;
 
+	if (is_fstree(ref_root))
+		seq = atomic64_read(&fs_info->tree_mod_seq);
+
 	/* first set the basic ref node struct up */
 	atomic_set(&ref->refs, 1);
 	ref->bytenr = bytenr;
@@ -579,32 +724,37 @@ static noinline int add_delayed_data_ref(struct btrfs_trans_handle *trans,
 	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);
-	if (parent) {
-		full_ref->parent = parent;
+	full_ref->parent = parent;
+	full_ref->root = ref_root;
+	if (parent)
 		ref->type = BTRFS_SHARED_DATA_REF_KEY;
-	} else {
-		full_ref->root = ref_root;
+	else
 		ref->type = BTRFS_EXTENT_DATA_REF_KEY;
-	}
+
 	full_ref->objectid = owner;
 	full_ref->offset = offset;
 
-	existing = tree_insert(&delayed_refs->root, &ref->rb_node);
+	trace_add_delayed_data_ref(ref, full_ref, action);
 
+	spin_lock(&head_ref->lock);
+	existing = tree_insert(&head_ref->ref_root, &ref->rb_node);
 	if (existing) {
-		update_existing_ref(trans, delayed_refs, existing, ref);
+		update_existing_ref(trans, delayed_refs, head_ref, existing,
+				    ref);
 		/*
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
-		kfree(ref);
+		kmem_cache_free(btrfs_delayed_data_ref_cachep, full_ref);
 	} else {
-		delayed_refs->num_entries++;
+		atomic_inc(&delayed_refs->num_entries);
 		trans->delayed_ref_updates++;
 	}
-	return 0;
+	spin_unlock(&head_ref->lock);
 }
 
 /*
@@ -612,24 +762,28 @@ static noinline int add_delayed_data_ref(struct btrfs_trans_handle *trans,
  * to make sure the delayed ref is eventually processed before this
  * transaction commits.
  */
-int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans,
+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)
+			       struct btrfs_delayed_extent_op *extent_op,
+			       int no_quota)
 {
 	struct btrfs_delayed_tree_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	int ret;
+
+	if (!is_fstree(ref_root) || !fs_info->quota_enabled)
+		no_quota = 0;
 
 	BUG_ON(extent_op && extent_op->is_data);
-	ref = kmalloc(sizeof(*ref), GFP_NOFS);
+	ref = kmem_cache_alloc(btrfs_delayed_tree_ref_cachep, GFP_NOFS);
 	if (!ref)
 		return -ENOMEM;
 
-	head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS);
+	head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS);
 	if (!head_ref) {
-		kfree(ref);
+		kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref);
 		return -ENOMEM;
 	}
 
@@ -642,39 +796,43 @@ int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans,
 	 * insert both the head node and the new ref without dropping
 	 * the spin lock
 	 */
-	ret = add_delayed_ref_head(trans, &head_ref->node, bytenr, num_bytes,
-				   action, 0);
-	BUG_ON(ret);
+	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node,
+					bytenr, num_bytes, action, 0);
 
-	ret = add_delayed_tree_ref(trans, &ref->node, bytenr, num_bytes,
-				   parent, ref_root, level, action);
-	BUG_ON(ret);
+	add_delayed_tree_ref(fs_info, trans, head_ref, &ref->node, bytenr,
+				   num_bytes, parent, ref_root, level, action,
+				   no_quota);
 	spin_unlock(&delayed_refs->lock);
+
 	return 0;
 }
 
 /*
  * add a delayed data ref. it's similar to btrfs_add_delayed_tree_ref.
  */
-int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans,
+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)
+			       struct btrfs_delayed_extent_op *extent_op,
+			       int no_quota)
 {
 	struct btrfs_delayed_data_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	int ret;
+
+	if (!is_fstree(ref_root) || !fs_info->quota_enabled)
+		no_quota = 0;
 
 	BUG_ON(extent_op && !extent_op->is_data);
-	ref = kmalloc(sizeof(*ref), GFP_NOFS);
+	ref = kmem_cache_alloc(btrfs_delayed_data_ref_cachep, GFP_NOFS);
 	if (!ref)
 		return -ENOMEM;
 
-	head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS);
+	head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS);
 	if (!head_ref) {
-		kfree(ref);
+		kmem_cache_free(btrfs_delayed_data_ref_cachep, ref);
 		return -ENOMEM;
 	}
 
@@ -687,26 +845,26 @@ int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans,
 	 * insert both the head node and the new ref without dropping
 	 * the spin lock
 	 */
-	ret = add_delayed_ref_head(trans, &head_ref->node, bytenr, num_bytes,
-				   action, 1);
-	BUG_ON(ret);
+	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node,
+					bytenr, num_bytes, action, 1);
 
-	ret = add_delayed_data_ref(trans, &ref->node, bytenr, num_bytes,
-				   parent, ref_root, owner, offset, action);
-	BUG_ON(ret);
+	add_delayed_data_ref(fs_info, trans, head_ref, &ref->node, bytenr,
+				   num_bytes, parent, ref_root, owner, offset,
+				   action, no_quota);
 	spin_unlock(&delayed_refs->lock);
+
 	return 0;
 }
 
-int btrfs_add_delayed_extent_op(struct btrfs_trans_handle *trans,
+int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
+				struct btrfs_trans_handle *trans,
 				u64 bytenr, u64 num_bytes,
 				struct btrfs_delayed_extent_op *extent_op)
 {
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	int ret;
 
-	head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS);
+	head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS);
 	if (!head_ref)
 		return -ENOMEM;
 
@@ -715,10 +873,9 @@ int btrfs_add_delayed_extent_op(struct btrfs_trans_handle *trans,
 	delayed_refs = &trans->transaction->delayed_refs;
 	spin_lock(&delayed_refs->lock);
 
-	ret = add_delayed_ref_head(trans, &head_ref->node, bytenr,
+	add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
 				   num_bytes, BTRFS_UPDATE_DELAYED_HEAD,
 				   extent_op->is_data);
-	BUG_ON(ret);
 
 	spin_unlock(&delayed_refs->lock);
 	return 0;
@@ -732,88 +889,56 @@ int btrfs_add_delayed_extent_op(struct btrfs_trans_handle *trans,
 struct btrfs_delayed_ref_head *
 btrfs_find_delayed_ref_head(struct btrfs_trans_handle *trans, u64 bytenr)
 {
-	struct btrfs_delayed_ref_node *ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
 
 	delayed_refs = &trans->transaction->delayed_refs;
-	ref = find_ref_head(&delayed_refs->root, bytenr, NULL);
-	if (ref)
-		return btrfs_delayed_node_to_head(ref);
-	return NULL;
+	return find_ref_head(&delayed_refs->href_root, bytenr, 0);
 }
 
-/*
- * add a delayed ref to the tree.  This does all of the accounting required
- * to make sure the delayed ref is eventually processed before this
- * transaction commits.
- *
- * The main point of this call is to add and remove a backreference in a single
- * shot, taking the lock only once, and only searching for the head node once.
- *
- * It is the same as doing a ref add and delete in two separate calls.
- */
-#if 0
-int btrfs_update_delayed_ref(struct btrfs_trans_handle *trans,
-			  u64 bytenr, u64 num_bytes, u64 orig_parent,
-			  u64 parent, u64 orig_ref_root, u64 ref_root,
-			  u64 orig_ref_generation, u64 ref_generation,
-			  u64 owner_objectid, int pin)
+void btrfs_delayed_ref_exit(void)
 {
-	struct btrfs_delayed_ref *ref;
-	struct btrfs_delayed_ref *old_ref;
-	struct btrfs_delayed_ref_head *head_ref;
-	struct btrfs_delayed_ref_root *delayed_refs;
-	int ret;
-
-	ref = kmalloc(sizeof(*ref), GFP_NOFS);
-	if (!ref)
-		return -ENOMEM;
-
-	old_ref = kmalloc(sizeof(*old_ref), GFP_NOFS);
-	if (!old_ref) {
-		kfree(ref);
-		return -ENOMEM;
-	}
-
-	/*
-	 * the parent = 0 case comes from cases where we don't actually
-	 * know the parent yet.  It will get updated later via a add/drop
-	 * pair.
-	 */
-	if (parent == 0)
-		parent = bytenr;
-	if (orig_parent == 0)
-		orig_parent = bytenr;
+	if (btrfs_delayed_ref_head_cachep)
+		kmem_cache_destroy(btrfs_delayed_ref_head_cachep);
+	if (btrfs_delayed_tree_ref_cachep)
+		kmem_cache_destroy(btrfs_delayed_tree_ref_cachep);
+	if (btrfs_delayed_data_ref_cachep)
+		kmem_cache_destroy(btrfs_delayed_data_ref_cachep);
+	if (btrfs_delayed_extent_op_cachep)
+		kmem_cache_destroy(btrfs_delayed_extent_op_cachep);
+}
 
-	head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS);
-	if (!head_ref) {
-		kfree(ref);
-		kfree(old_ref);
-		return -ENOMEM;
-	}
-	delayed_refs = &trans->transaction->delayed_refs;
-	spin_lock(&delayed_refs->lock);
+int btrfs_delayed_ref_init(void)
+{
+	btrfs_delayed_ref_head_cachep = kmem_cache_create(
+				"btrfs_delayed_ref_head",
+				sizeof(struct btrfs_delayed_ref_head), 0,
+				SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL);
+	if (!btrfs_delayed_ref_head_cachep)
+		goto fail;
+
+	btrfs_delayed_tree_ref_cachep = kmem_cache_create(
+				"btrfs_delayed_tree_ref",
+				sizeof(struct btrfs_delayed_tree_ref), 0,
+				SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL);
+	if (!btrfs_delayed_tree_ref_cachep)
+		goto fail;
+
+	btrfs_delayed_data_ref_cachep = kmem_cache_create(
+				"btrfs_delayed_data_ref",
+				sizeof(struct btrfs_delayed_data_ref), 0,
+				SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL);
+	if (!btrfs_delayed_data_ref_cachep)
+		goto fail;
+
+	btrfs_delayed_extent_op_cachep = kmem_cache_create(
+				"btrfs_delayed_extent_op",
+				sizeof(struct btrfs_delayed_extent_op), 0,
+				SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL);
+	if (!btrfs_delayed_extent_op_cachep)
+		goto fail;
 
-	/*
-	 * insert both the head node and the new ref without dropping
-	 * the spin lock
-	 */
-	ret = __btrfs_add_delayed_ref(trans, &head_ref->node, bytenr, num_bytes,
-				      (u64)-1, 0, 0, 0,
-				      BTRFS_UPDATE_DELAYED_HEAD, 0);
-	BUG_ON(ret);
-
-	ret = __btrfs_add_delayed_ref(trans, &ref->node, bytenr, num_bytes,
-				      parent, ref_root, ref_generation,
-				      owner_objectid, BTRFS_ADD_DELAYED_REF, 0);
-	BUG_ON(ret);
-
-	ret = __btrfs_add_delayed_ref(trans, &old_ref->node, bytenr, num_bytes,
-				      orig_parent, orig_ref_root,
-				      orig_ref_generation, owner_objectid,
-				      BTRFS_DROP_DELAYED_REF, pin);
-	BUG_ON(ret);
-	spin_unlock(&delayed_refs->lock);
 	return 0;
+fail:
+	btrfs_delayed_ref_exit();
+	return -ENOMEM;
 }
-#endif