1 files changed, 301 insertions, 263 deletions

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index a7bc66121330..5871ef78edba 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -33,6 +33,7 @@
 #include "delalloc-space.h"
 #include "block-group.h"
 #include "discard.h"
+#include "rcu-string.h"
 
 #undef SCRAMBLE_DELAYED_REFS
 
@@ -64,10 +65,8 @@ int btrfs_add_excluded_extent(struct btrfs_fs_info *fs_info,
 			      u64 start, u64 num_bytes)
 {
 	u64 end = start + num_bytes - 1;
-	set_extent_bits(&fs_info->freed_extents[0],
-			start, end, EXTENT_UPTODATE);
-	set_extent_bits(&fs_info->freed_extents[1],
-			start, end, EXTENT_UPTODATE);
+	set_extent_bits(&fs_info->excluded_extents, start, end,
+			EXTENT_UPTODATE);
 	return 0;
 }
 
@@ -79,10 +78,8 @@ void btrfs_free_excluded_extents(struct btrfs_block_group *cache)
 	start = cache->start;
 	end = start + cache->length - 1;
 
-	clear_extent_bits(&fs_info->freed_extents[0],
-			  start, end, EXTENT_UPTODATE);
-	clear_extent_bits(&fs_info->freed_extents[1],
-			  start, end, EXTENT_UPTODATE);
+	clear_extent_bits(&fs_info->excluded_extents, start, end,
+			  EXTENT_UPTODATE);
 }
 
 static u64 generic_ref_to_space_flags(struct btrfs_ref *ref)
@@ -1193,24 +1190,6 @@ int insert_inline_extent_backref(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
-static int insert_extent_backref(struct btrfs_trans_handle *trans,
-				 struct btrfs_path *path,
-				 u64 bytenr, u64 parent, u64 root_objectid,
-				 u64 owner, u64 offset, int refs_to_add)
-{
-	int ret;
-	if (owner < BTRFS_FIRST_FREE_OBJECTID) {
-		BUG_ON(refs_to_add != 1);
-		ret = insert_tree_block_ref(trans, path, bytenr, parent,
-					    root_objectid);
-	} else {
-		ret = insert_extent_data_ref(trans, path, bytenr, parent,
-					     root_objectid, owner, offset,
-					     refs_to_add);
-	}
-	return ret;
-}
-
 static int remove_extent_backref(struct btrfs_trans_handle *trans,
 				 struct btrfs_path *path,
 				 struct btrfs_extent_inline_ref *iref,
@@ -1469,7 +1448,6 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 	if (!path)
 		return -ENOMEM;
 
-	path->reada = READA_FORWARD;
 	path->leave_spinning = 1;
 	/* this will setup the path even if it fails to insert the back ref */
 	ret = insert_inline_extent_backref(trans, path, bytenr, num_bytes,
@@ -1494,11 +1472,17 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 	btrfs_mark_buffer_dirty(leaf);
 	btrfs_release_path(path);
 
-	path->reada = READA_FORWARD;
 	path->leave_spinning = 1;
 	/* now insert the actual backref */
-	ret = insert_extent_backref(trans, path, bytenr, parent, root_objectid,
-				    owner, offset, refs_to_add);
+	if (owner < BTRFS_FIRST_FREE_OBJECTID) {
+		BUG_ON(refs_to_add != 1);
+		ret = insert_tree_block_ref(trans, path, bytenr, parent,
+					    root_objectid);
+	} else {
+		ret = insert_extent_data_ref(trans, path, bytenr, parent,
+					     root_objectid, owner, offset,
+					     refs_to_add);
+	}
 	if (ret)
 		btrfs_abort_transaction(trans, ret);
 out:
@@ -1583,7 +1567,7 @@ static int run_delayed_extent_op(struct btrfs_trans_handle *trans,
 	int err = 0;
 	int metadata = !extent_op->is_data;
 
-	if (trans->aborted)
+	if (TRANS_ABORTED(trans))
 		return 0;
 
 	if (metadata && !btrfs_fs_incompat(fs_info, SKINNY_METADATA))
@@ -1604,7 +1588,6 @@ static int run_delayed_extent_op(struct btrfs_trans_handle *trans,
 	}
 
 again:
-	path->reada = READA_FORWARD;
 	path->leave_spinning = 1;
 	ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 1);
 	if (ret < 0) {
@@ -1703,10 +1686,9 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
 {
 	int ret = 0;
 
-	if (trans->aborted) {
+	if (TRANS_ABORTED(trans)) {
 		if (insert_reserved)
-			btrfs_pin_extent(trans->fs_info, node->bytenr,
-					 node->num_bytes, 1);
+			btrfs_pin_extent(trans, node->bytenr, node->num_bytes, 1);
 		return 0;
 	}
 
@@ -1721,8 +1703,7 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
 	else
 		BUG();
 	if (ret && insert_reserved)
-		btrfs_pin_extent(trans->fs_info, node->bytenr,
-				 node->num_bytes, 1);
+		btrfs_pin_extent(trans, node->bytenr, node->num_bytes, 1);
 	return ret;
 }
 
@@ -1867,8 +1848,7 @@ static int cleanup_ref_head(struct btrfs_trans_handle *trans,
 	spin_unlock(&delayed_refs->lock);
 
 	if (head->must_insert_reserved) {
-		btrfs_pin_extent(fs_info, head->bytenr,
-				 head->num_bytes, 1);
+		btrfs_pin_extent(trans, head->bytenr, head->num_bytes, 1);
 		if (head->is_data) {
 			ret = btrfs_del_csums(trans, fs_info->csum_root,
 					      head->bytenr, head->num_bytes);
@@ -2135,22 +2115,6 @@ static u64 find_middle(struct rb_root *root)
 }
 #endif
 
-static inline u64 heads_to_leaves(struct btrfs_fs_info *fs_info, u64 heads)
-{
-	u64 num_bytes;
-
-	num_bytes = heads * (sizeof(struct btrfs_extent_item) +
-			     sizeof(struct btrfs_extent_inline_ref));
-	if (!btrfs_fs_incompat(fs_info, SKINNY_METADATA))
-		num_bytes += heads * sizeof(struct btrfs_tree_block_info);
-
-	/*
-	 * We don't ever fill up leaves all the way so multiply by 2 just to be
-	 * closer to what we're really going to want to use.
-	 */
-	return div_u64(num_bytes, BTRFS_LEAF_DATA_SIZE(fs_info));
-}
-
 /*
  * Takes the number of bytes to be csumm'ed and figures out how many leaves it
  * would require to store the csums for that many bytes.
@@ -2191,7 +2155,7 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
 	int run_all = count == (unsigned long)-1;
 
 	/* We'll clean this up in btrfs_cleanup_transaction */
-	if (trans->aborted)
+	if (TRANS_ABORTED(trans))
 		return 0;
 
 	if (test_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags))
@@ -2238,7 +2202,7 @@ out:
 }
 
 int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans,
-				u64 bytenr, u64 num_bytes, u64 flags,
+				struct extent_buffer *eb, u64 flags,
 				int level, int is_data)
 {
 	struct btrfs_delayed_extent_op *extent_op;
@@ -2254,7 +2218,7 @@ int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans,
 	extent_op->is_data = is_data ? true : false;
 	extent_op->level = level;
 
-	ret = btrfs_add_delayed_extent_op(trans, bytenr, num_bytes, extent_op);
+	ret = btrfs_add_delayed_extent_op(trans, eb->start, eb->len, extent_op);
 	if (ret)
 		btrfs_free_delayed_extent_op(extent_op);
 	return ret;
@@ -2342,7 +2306,8 @@ static noinline int check_delayed_ref(struct btrfs_root *root,
 
 static noinline int check_committed_ref(struct btrfs_root *root,
 					struct btrfs_path *path,
-					u64 objectid, u64 offset, u64 bytenr)
+					u64 objectid, u64 offset, u64 bytenr,
+					bool strict)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_root *extent_root = fs_info->extent_root;
@@ -2384,9 +2349,13 @@ static noinline int check_committed_ref(struct btrfs_root *root,
 	    btrfs_extent_inline_ref_size(BTRFS_EXTENT_DATA_REF_KEY))
 		goto out;
 
-	/* If extent created before last snapshot => it's definitely shared */
-	if (btrfs_extent_generation(leaf, ei) <=
-	    btrfs_root_last_snapshot(&root->root_item))
+	/*
+	 * If extent created before last snapshot => it's shared unless the
+	 * snapshot has been deleted. Use the heuristic if strict is false.
+	 */
+	if (!strict &&
+	    (btrfs_extent_generation(leaf, ei) <=
+	     btrfs_root_last_snapshot(&root->root_item)))
 		goto out;
 
 	iref = (struct btrfs_extent_inline_ref *)(ei + 1);
@@ -2411,7 +2380,7 @@ out:
 }
 
 int btrfs_cross_ref_exist(struct btrfs_root *root, u64 objectid, u64 offset,
-			  u64 bytenr)
+			  u64 bytenr, bool strict)
 {
 	struct btrfs_path *path;
 	int ret;
@@ -2422,7 +2391,7 @@ int btrfs_cross_ref_exist(struct btrfs_root *root, u64 objectid, u64 offset,
 
 	do {
 		ret = check_committed_ref(root, path, objectid,
-					  offset, bytenr);
+					  offset, bytenr, strict);
 		if (ret && ret != -ENOENT)
 			goto out;
 
@@ -2463,7 +2432,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans,
 	nritems = btrfs_header_nritems(buf);
 	level = btrfs_header_level(buf);
 
-	if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state) && level == 0)
+	if (!test_bit(BTRFS_ROOT_SHAREABLE, &root->state) && level == 0)
 		return 0;
 
 	if (full_backref)
@@ -2588,7 +2557,8 @@ static u64 first_logical_byte(struct btrfs_fs_info *fs_info, u64 search_start)
 	return bytenr;
 }
 
-static int pin_down_extent(struct btrfs_block_group *cache,
+static int pin_down_extent(struct btrfs_trans_handle *trans,
+			   struct btrfs_block_group *cache,
 			   u64 bytenr, u64 num_bytes, int reserved)
 {
 	struct btrfs_fs_info *fs_info = cache->fs_info;
@@ -2607,22 +2577,20 @@ static int pin_down_extent(struct btrfs_block_group *cache,
 
 	percpu_counter_add_batch(&cache->space_info->total_bytes_pinned,
 		    num_bytes, BTRFS_TOTAL_BYTES_PINNED_BATCH);
-	set_extent_dirty(fs_info->pinned_extents, bytenr,
+	set_extent_dirty(&trans->transaction->pinned_extents, bytenr,
 			 bytenr + num_bytes - 1, GFP_NOFS | __GFP_NOFAIL);
 	return 0;
 }
 
-int btrfs_pin_extent(struct btrfs_fs_info *fs_info,
+int btrfs_pin_extent(struct btrfs_trans_handle *trans,
 		     u64 bytenr, u64 num_bytes, int reserved)
 {
 	struct btrfs_block_group *cache;
 
-	ASSERT(fs_info->running_transaction);
-
-	cache = btrfs_lookup_block_group(fs_info, bytenr);
+	cache = btrfs_lookup_block_group(trans->fs_info, bytenr);
 	BUG_ON(!cache); /* Logic error */
 
-	pin_down_extent(cache, bytenr, num_bytes, reserved);
+	pin_down_extent(trans, cache, bytenr, num_bytes, reserved);
 
 	btrfs_put_block_group(cache);
 	return 0;
@@ -2631,13 +2599,15 @@ int btrfs_pin_extent(struct btrfs_fs_info *fs_info,
 /*
  * this function must be called within transaction
  */
-int btrfs_pin_extent_for_log_replay(struct btrfs_fs_info *fs_info,
+int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans,
 				    u64 bytenr, u64 num_bytes)
 {
 	struct btrfs_block_group *cache;
 	int ret;
 
-	cache = btrfs_lookup_block_group(fs_info, bytenr);
+	btrfs_add_excluded_extent(trans->fs_info, bytenr, num_bytes);
+
+	cache = btrfs_lookup_block_group(trans->fs_info, bytenr);
 	if (!cache)
 		return -EINVAL;
 
@@ -2649,7 +2619,7 @@ int btrfs_pin_extent_for_log_replay(struct btrfs_fs_info *fs_info,
 	 */
 	btrfs_cache_block_group(cache, 1);
 
-	pin_down_extent(cache, bytenr, num_bytes, 0);
+	pin_down_extent(trans, cache, bytenr, num_bytes, 0);
 
 	/* remove us from the free space cache (if we're there at all) */
 	ret = btrfs_remove_free_space(cache, bytenr, num_bytes);
@@ -2763,11 +2733,6 @@ void btrfs_prepare_extent_commit(struct btrfs_fs_info *fs_info)
 		}
 	}
 
-	if (fs_info->pinned_extents == &fs_info->freed_extents[0])
-		fs_info->pinned_extents = &fs_info->freed_extents[1];
-	else
-		fs_info->pinned_extents = &fs_info->freed_extents[0];
-
 	up_write(&fs_info->commit_root_sem);
 
 	btrfs_update_global_block_rsv(fs_info);
@@ -2908,12 +2873,9 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans)
 	u64 end;
 	int ret;
 
-	if (fs_info->pinned_extents == &fs_info->freed_extents[0])
-		unpin = &fs_info->freed_extents[1];
-	else
-		unpin = &fs_info->freed_extents[0];
+	unpin = &trans->transaction->pinned_extents;
 
-	while (!trans->aborted) {
+	while (!TRANS_ABORTED(trans)) {
 		struct extent_state *cached_state = NULL;
 
 		mutex_lock(&fs_info->unused_bg_unpin_mutex);
@@ -2923,6 +2885,9 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans)
 			mutex_unlock(&fs_info->unused_bg_unpin_mutex);
 			break;
 		}
+		if (test_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags))
+			clear_extent_bits(&fs_info->excluded_extents, start,
+					  end, EXTENT_UPTODATE);
 
 		if (btrfs_test_opt(fs_info, DISCARD_SYNC))
 			ret = btrfs_discard_extent(fs_info, start,
@@ -2950,14 +2915,14 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans)
 		u64 trimmed = 0;
 
 		ret = -EROFS;
-		if (!trans->aborted)
+		if (!TRANS_ABORTED(trans))
 			ret = btrfs_discard_extent(fs_info,
 						   block_group->start,
 						   block_group->length,
 						   &trimmed);
 
 		list_del_init(&block_group->bg_list);
-		btrfs_put_block_group_trimming(block_group);
+		btrfs_unfreeze_block_group(block_group);
 		btrfs_put_block_group(block_group);
 
 		if (ret) {
@@ -3000,7 +2965,6 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 	if (!path)
 		return -ENOMEM;
 
-	path->reada = READA_FORWARD;
 	path->leave_spinning = 1;
 
 	is_data = owner_objectid >= BTRFS_FIRST_FREE_OBJECTID;
@@ -3301,7 +3265,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 		cache = btrfs_lookup_block_group(fs_info, buf->start);
 
 		if (btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) {
-			pin_down_extent(cache, buf->start, buf->len, 1);
+			pin_down_extent(trans, cache, buf->start, buf->len, 1);
 			btrfs_put_block_group(cache);
 			goto out;
 		}
@@ -3345,7 +3309,7 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_ref *ref)
 	    (ref->type == BTRFS_REF_DATA &&
 	     ref->data_ref.ref_root == BTRFS_TREE_LOG_OBJECTID)) {
 		/* unlocks the pinned mutex */
-		btrfs_pin_extent(fs_info, ref->bytenr, ref->len, 1);
+		btrfs_pin_extent(trans, ref->bytenr, ref->len, 1);
 		old_ref_mod = new_ref_mod = 0;
 		ret = 0;
 	} else if (ref->type == BTRFS_REF_METADATA) {
@@ -3395,6 +3359,7 @@ static struct btrfs_block_group *btrfs_lock_cluster(
 		   struct btrfs_block_group *block_group,
 		   struct btrfs_free_cluster *cluster,
 		   int delalloc)
+	__acquires(&cluster->refill_lock)
 {
 	struct btrfs_block_group *used_bg = NULL;
 
@@ -3438,6 +3403,10 @@ btrfs_release_block_group(struct btrfs_block_group *cache,
 	btrfs_put_block_group(cache);
 }
 
+enum btrfs_extent_allocation_policy {
+	BTRFS_EXTENT_ALLOC_CLUSTERED,
+};
+
 /*
  * Structure used internally for find_free_extent() function.  Wraps needed
  * parameters.
@@ -3454,6 +3423,8 @@ struct find_free_extent_ctl {
 
 	/* For clustered allocation */
 	u64 empty_cluster;
+	struct btrfs_free_cluster *last_ptr;
+	bool use_cluster;
 
 	bool have_caching_bg;
 	bool orig_have_caching_bg;
@@ -3489,6 +3460,12 @@ struct find_free_extent_ctl {
 
 	/* Found result */
 	u64 found_offset;
+
+	/* Hint where to start looking for an empty space */
+	u64 hint_byte;
+
+	/* Allocation policy */
+	enum btrfs_extent_allocation_policy policy;
 };
 
 
@@ -3501,11 +3478,11 @@ struct find_free_extent_ctl {
  * Return 0 means we have found a location and set ffe_ctl->found_offset.
  */
 static int find_free_extent_clustered(struct btrfs_block_group *bg,
-		struct btrfs_free_cluster *last_ptr,
-		struct find_free_extent_ctl *ffe_ctl,
-		struct btrfs_block_group **cluster_bg_ret)
+				      struct find_free_extent_ctl *ffe_ctl,
+				      struct btrfs_block_group **cluster_bg_ret)
 {
 	struct btrfs_block_group *cluster_bg;
+	struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
 	u64 aligned_cluster;
 	u64 offset;
 	int ret;
@@ -3605,9 +3582,9 @@ refill_cluster:
  * Return -EAGAIN to inform caller that we need to re-search this block group
  */
 static int find_free_extent_unclustered(struct btrfs_block_group *bg,
-		struct btrfs_free_cluster *last_ptr,
-		struct find_free_extent_ctl *ffe_ctl)
+					struct find_free_extent_ctl *ffe_ctl)
 {
+	struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
 	u64 offset;
 
 	/*
@@ -3663,16 +3640,101 @@ static int find_free_extent_unclustered(struct btrfs_block_group *bg,
 	return 0;
 }
 
+static int do_allocation_clustered(struct btrfs_block_group *block_group,
+				   struct find_free_extent_ctl *ffe_ctl,
+				   struct btrfs_block_group **bg_ret)
+{
+	int ret;
+
+	/* We want to try and use the cluster allocator, so lets look there */
+	if (ffe_ctl->last_ptr && ffe_ctl->use_cluster) {
+		ret = find_free_extent_clustered(block_group, ffe_ctl, bg_ret);
+		if (ret >= 0 || ret == -EAGAIN)
+			return ret;
+		/* ret == -ENOENT case falls through */
+	}
+
+	return find_free_extent_unclustered(block_group, ffe_ctl);
+}
+
+static int do_allocation(struct btrfs_block_group *block_group,
+			 struct find_free_extent_ctl *ffe_ctl,
+			 struct btrfs_block_group **bg_ret)
+{
+	switch (ffe_ctl->policy) {
+	case BTRFS_EXTENT_ALLOC_CLUSTERED:
+		return do_allocation_clustered(block_group, ffe_ctl, bg_ret);
+	default:
+		BUG();
+	}
+}
+
+static void release_block_group(struct btrfs_block_group *block_group,
+				struct find_free_extent_ctl *ffe_ctl,
+				int delalloc)
+{
+	switch (ffe_ctl->policy) {
+	case BTRFS_EXTENT_ALLOC_CLUSTERED:
+		ffe_ctl->retry_clustered = false;
+		ffe_ctl->retry_unclustered = false;
+		break;
+	default:
+		BUG();
+	}
+
+	BUG_ON(btrfs_bg_flags_to_raid_index(block_group->flags) !=
+	       ffe_ctl->index);
+	btrfs_release_block_group(block_group, delalloc);
+}
+
+static void found_extent_clustered(struct find_free_extent_ctl *ffe_ctl,
+				   struct btrfs_key *ins)
+{
+	struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
+
+	if (!ffe_ctl->use_cluster && last_ptr) {
+		spin_lock(&last_ptr->lock);
+		last_ptr->window_start = ins->objectid;
+		spin_unlock(&last_ptr->lock);
+	}
+}
+
+static void found_extent(struct find_free_extent_ctl *ffe_ctl,
+			 struct btrfs_key *ins)
+{
+	switch (ffe_ctl->policy) {
+	case BTRFS_EXTENT_ALLOC_CLUSTERED:
+		found_extent_clustered(ffe_ctl, ins);
+		break;
+	default:
+		BUG();
+	}
+}
+
+static int chunk_allocation_failed(struct find_free_extent_ctl *ffe_ctl)
+{
+	switch (ffe_ctl->policy) {
+	case BTRFS_EXTENT_ALLOC_CLUSTERED:
+		/*
+		 * If we can't allocate a new chunk we've already looped through
+		 * at least once, move on to the NO_EMPTY_SIZE case.
+		 */
+		ffe_ctl->loop = LOOP_NO_EMPTY_SIZE;
+		return 0;
+	default:
+		BUG();
+	}
+}
+
 /*
  * Return >0 means caller needs to re-search for free extent
  * Return 0 means we have the needed free extent.
  * Return <0 means we failed to locate any free extent.
  */
 static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
-					struct btrfs_free_cluster *last_ptr,
 					struct btrfs_key *ins,
 					struct find_free_extent_ctl *ffe_ctl,
-					int full_search, bool use_cluster)
+					bool full_search)
 {
 	struct btrfs_root *root = fs_info->extent_root;
 	int ret;
@@ -3689,11 +3751,7 @@ static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
 		return 1;
 
 	if (ins->objectid) {
-		if (!use_cluster && last_ptr) {
-			spin_lock(&last_ptr->lock);
-			last_ptr->window_start = ins->objectid;
-			spin_unlock(&last_ptr->lock);
-		}
+		found_extent(ffe_ctl, ins);
 		return 0;
 	}
 
@@ -3739,16 +3797,10 @@ static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
 			ret = btrfs_chunk_alloc(trans, ffe_ctl->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)
-				ffe_ctl->loop = LOOP_NO_EMPTY_SIZE;
-
 			/* Do not bail out on ENOSPC since we can do more. */
-			if (ret < 0 && ret != -ENOSPC)
+			if (ret == -ENOSPC)
+				ret = chunk_allocation_failed(ffe_ctl);
+			else if (ret < 0)
 				btrfs_abort_transaction(trans, ret);
 			else
 				ret = 0;
@@ -3759,6 +3811,9 @@ static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
 		}
 
 		if (ffe_ctl->loop == LOOP_NO_EMPTY_SIZE) {
+			if (ffe_ctl->policy != BTRFS_EXTENT_ALLOC_CLUSTERED)
+				return -ENOSPC;
+
 			/*
 			 * Don't loop again if we already have no empty_size and
 			 * no empty_cluster.
@@ -3774,6 +3829,71 @@ static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
 	return -ENOSPC;
 }
 
+static int prepare_allocation_clustered(struct btrfs_fs_info *fs_info,
+					struct find_free_extent_ctl *ffe_ctl,
+					struct btrfs_space_info *space_info,
+					struct btrfs_key *ins)
+{
+	/*
+	 * 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 (space_info->max_extent_size) {
+		spin_lock(&space_info->lock);
+		if (space_info->max_extent_size &&
+		    ffe_ctl->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) {
+			ffe_ctl->use_cluster = false;
+		}
+		spin_unlock(&space_info->lock);
+	}
+
+	ffe_ctl->last_ptr = fetch_cluster_info(fs_info, space_info,
+					       &ffe_ctl->empty_cluster);
+	if (ffe_ctl->last_ptr) {
+		struct btrfs_free_cluster *last_ptr = ffe_ctl->last_ptr;
+
+		spin_lock(&last_ptr->lock);
+		if (last_ptr->block_group)
+			ffe_ctl->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.
+			 */
+			ffe_ctl->hint_byte = last_ptr->window_start;
+			ffe_ctl->use_cluster = false;
+		}
+		spin_unlock(&last_ptr->lock);
+	}
+
+	return 0;
+}
+
+static int prepare_allocation(struct btrfs_fs_info *fs_info,
+			      struct find_free_extent_ctl *ffe_ctl,
+			      struct btrfs_space_info *space_info,
+			      struct btrfs_key *ins)
+{
+	switch (ffe_ctl->policy) {
+	case BTRFS_EXTENT_ALLOC_CLUSTERED:
+		return prepare_allocation_clustered(fs_info, ffe_ctl,
+						    space_info, ins);
+	default:
+		BUG();
+	}
+}
+
 /*
  * walks the btree of allocated extents and find a hole of a given size.
  * The key ins is changed to record the hole:
@@ -3801,16 +3921,14 @@ static int find_free_extent_update_loop(struct btrfs_fs_info *fs_info,
  */
 static noinline int find_free_extent(struct btrfs_fs_info *fs_info,
 				u64 ram_bytes, u64 num_bytes, u64 empty_size,
-				u64 hint_byte, struct btrfs_key *ins,
+				u64 hint_byte_orig, struct btrfs_key *ins,
 				u64 flags, int delalloc)
 {
 	int ret = 0;
 	int cache_block_group_error = 0;
-	struct btrfs_free_cluster *last_ptr = NULL;
 	struct btrfs_block_group *block_group = NULL;
 	struct find_free_extent_ctl ffe_ctl = {0};
 	struct btrfs_space_info *space_info;
-	bool use_cluster = true;
 	bool full_search = false;
 
 	WARN_ON(num_bytes < fs_info->sectorsize);
@@ -3819,13 +3937,19 @@ static noinline int find_free_extent(struct btrfs_fs_info *fs_info,
 	ffe_ctl.empty_size = empty_size;
 	ffe_ctl.flags = flags;
 	ffe_ctl.search_start = 0;
-	ffe_ctl.retry_clustered = false;
-	ffe_ctl.retry_unclustered = false;
 	ffe_ctl.delalloc = delalloc;
 	ffe_ctl.index = btrfs_bg_flags_to_raid_index(flags);
 	ffe_ctl.have_caching_bg = false;
 	ffe_ctl.orig_have_caching_bg = false;
 	ffe_ctl.found_offset = 0;
+	ffe_ctl.hint_byte = hint_byte_orig;
+	ffe_ctl.policy = BTRFS_EXTENT_ALLOC_CLUSTERED;
+
+	/* For clustered allocation */
+	ffe_ctl.retry_clustered = false;
+	ffe_ctl.retry_unclustered = false;
+	ffe_ctl.last_ptr = NULL;
+	ffe_ctl.use_cluster = true;
 
 	ins->type = BTRFS_EXTENT_ITEM_KEY;
 	ins->objectid = 0;
@@ -3839,51 +3963,14 @@ static noinline int find_free_extent(struct btrfs_fs_info *fs_info,
 		return -ENOSPC;
 	}
 
-	/*
-	 * 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 (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(fs_info, space_info,
-				      &ffe_ctl.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);
-	}
+	ret = prepare_allocation(fs_info, &ffe_ctl, space_info, ins);
+	if (ret < 0)
+		return ret;
 
 	ffe_ctl.search_start = max(ffe_ctl.search_start,
 				   first_logical_byte(fs_info, 0));
-	ffe_ctl.search_start = max(ffe_ctl.search_start, hint_byte);
-	if (ffe_ctl.search_start == hint_byte) {
+	ffe_ctl.search_start = max(ffe_ctl.search_start, ffe_ctl.hint_byte);
+	if (ffe_ctl.search_start == ffe_ctl.hint_byte) {
 		block_group = btrfs_lookup_block_group(fs_info,
 						       ffe_ctl.search_start);
 		/*
@@ -3924,6 +4011,8 @@ search:
 	down_read(&space_info->groups_sem);
 	list_for_each_entry(block_group,
 			    &space_info->block_groups[ffe_ctl.index], list) {
+		struct btrfs_block_group *bg_ret;
+
 		/* If the block group is read-only, we can skip it entirely. */
 		if (unlikely(block_group->ro))
 			continue;
@@ -3984,39 +4073,20 @@ have_block_group:
 		if (unlikely(block_group->cached == BTRFS_CACHE_ERROR))
 			goto loop;
 
-		/*
-		 * Ok we want to try and use the cluster allocator, so
-		 * lets look there
-		 */
-		if (last_ptr && use_cluster) {
-			struct btrfs_block_group *cluster_bg = NULL;
-
-			ret = find_free_extent_clustered(block_group, last_ptr,
-							 &ffe_ctl, &cluster_bg);
-
-			if (ret == 0) {
-				if (cluster_bg && cluster_bg != block_group) {
-					btrfs_release_block_group(block_group,
-								  delalloc);
-					block_group = cluster_bg;
-				}
-				goto checks;
-			} else if (ret == -EAGAIN) {
-				goto have_block_group;
-			} else if (ret > 0) {
-				goto loop;
+		bg_ret = NULL;
+		ret = do_allocation(block_group, &ffe_ctl, &bg_ret);
+		if (ret == 0) {
+			if (bg_ret && bg_ret != block_group) {
+				btrfs_release_block_group(block_group, delalloc);
+				block_group = bg_ret;
 			}
-			/* ret == -ENOENT case falls through */
-		}
-
-		ret = find_free_extent_unclustered(block_group, last_ptr,
-						   &ffe_ctl);
-		if (ret == -EAGAIN)
+		} else if (ret == -EAGAIN) {
 			goto have_block_group;
-		else if (ret > 0)
+		} else if (ret > 0) {
 			goto loop;
-		/* ret == 0 case falls through */
-checks:
+		}
+
+		/* Checks */
 		ffe_ctl.search_start = round_up(ffe_ctl.found_offset,
 					     fs_info->stripesize);
 
@@ -4050,17 +4120,12 @@ checks:
 		btrfs_release_block_group(block_group, delalloc);
 		break;
 loop:
-		ffe_ctl.retry_clustered = false;
-		ffe_ctl.retry_unclustered = false;
-		BUG_ON(btrfs_bg_flags_to_raid_index(block_group->flags) !=
-		       ffe_ctl.index);
-		btrfs_release_block_group(block_group, delalloc);
+		release_block_group(block_group, &ffe_ctl, delalloc);
 		cond_resched();
 	}
 	up_read(&space_info->groups_sem);
 
-	ret = find_free_extent_update_loop(fs_info, last_ptr, ins, &ffe_ctl,
-					   full_search, use_cluster);
+	ret = find_free_extent_update_loop(fs_info, ins, &ffe_ctl, full_search);
 	if (ret > 0)
 		goto search;
 
@@ -4189,18 +4254,20 @@ int btrfs_free_reserved_extent(struct btrfs_fs_info *fs_info,
 	return 0;
 }
 
-int btrfs_pin_reserved_extent(struct btrfs_fs_info *fs_info, u64 start, u64 len)
+int btrfs_pin_reserved_extent(struct btrfs_trans_handle *trans, u64 start,
+			      u64 len)
 {
 	struct btrfs_block_group *cache;
 	int ret = 0;
 
-	cache = btrfs_lookup_block_group(fs_info, start);
+	cache = btrfs_lookup_block_group(trans->fs_info, start);
 	if (!cache) {
-		btrfs_err(fs_info, "unable to find block group for %llu", start);
+		btrfs_err(trans->fs_info, "unable to find block group for %llu",
+			  start);
 		return -ENOSPC;
 	}
 
-	ret = pin_down_extent(cache, start, len, 1);
+	ret = pin_down_extent(trans, cache, start, len, 1);
 	btrfs_put_block_group(cache);
 	return ret;
 }
@@ -4431,7 +4498,7 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
 	ret = alloc_reserved_file_extent(trans, 0, root_objectid, 0, owner,
 					 offset, ins, 1);
 	if (ret)
-		btrfs_pin_extent(fs_info, ins->objectid, ins->offset, 1);
+		btrfs_pin_extent(trans, ins->objectid, ins->offset, 1);
 	btrfs_put_block_group(block_group);
 	return ret;
 }
@@ -4750,8 +4817,7 @@ static noinline int walk_down_proc(struct btrfs_trans_handle *trans,
 		BUG_ON(ret); /* -ENOMEM */
 		ret = btrfs_dec_ref(trans, root, eb, 0);
 		BUG_ON(ret); /* -ENOMEM */
-		ret = btrfs_set_disk_extent_flags(trans, eb->start,
-						  eb->len, flag,
+		ret = btrfs_set_disk_extent_flags(trans, eb, flag,
 						  btrfs_header_level(eb), 0);
 		BUG_ON(ret); /* -ENOMEM */
 		wc->flags[level] |= flag;
@@ -5209,9 +5275,7 @@ static noinline int walk_up_tree(struct btrfs_trans_handle *trans,
  *
  * If called with for_reloc == 0, may exit early with -EAGAIN
  */
-int btrfs_drop_snapshot(struct btrfs_root *root,
-			 struct btrfs_block_rsv *block_rsv, int update_ref,
-			 int for_reloc)
+int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref, int for_reloc)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_path *path;
@@ -5240,7 +5304,14 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 		goto out;
 	}
 
-	trans = btrfs_start_transaction(tree_root, 0);
+	/*
+	 * Use join to avoid potential EINTR from transaction start. See
+	 * wait_reserve_ticket and the whole reservation callchain.
+	 */
+	if (for_reloc)
+		trans = btrfs_join_transaction(tree_root);
+	else
+		trans = btrfs_start_transaction(tree_root, 0);
 	if (IS_ERR(trans)) {
 		err = PTR_ERR(trans);
 		goto out_free;
@@ -5250,9 +5321,6 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 	if (err)
 		goto out_end_trans;
 
-	if (block_rsv)
-		trans->block_rsv = block_rsv;
-
 	/*
 	 * This will help us catch people modifying the fs tree while we're
 	 * dropping it.  It is unsafe to mess with the fs tree while it's being
@@ -5380,8 +5448,6 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 				err = PTR_ERR(trans);
 				goto out_free;
 			}
-			if (block_rsv)
-				trans->block_rsv = block_rsv;
 		}
 	}
 	btrfs_release_path(path);
@@ -5413,13 +5479,18 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 		}
 	}
 
-	if (test_bit(BTRFS_ROOT_IN_RADIX, &root->state)) {
+	/*
+	 * This subvolume is going to be completely dropped, and won't be
+	 * recorded as dirty roots, thus pertrans meta rsv will not be freed at
+	 * commit transaction time.  So free it here manually.
+	 */
+	btrfs_qgroup_convert_reserved_meta(root, INT_MAX);
+	btrfs_qgroup_free_meta_all_pertrans(root);
+
+	if (test_bit(BTRFS_ROOT_IN_RADIX, &root->state))
 		btrfs_add_dropped_root(trans, root);
-	} else {
-		free_extent_buffer(root->node);
-		free_extent_buffer(root->commit_root);
-		btrfs_put_fs_root(root);
-	}
+	else
+		btrfs_put_root(root);
 	root_dropped = true;
 out_end_trans:
 	btrfs_end_transaction_throttle(trans);
@@ -5436,8 +5507,6 @@ out:
 	 */
 	if (!for_reloc && !root_dropped)
 		btrfs_add_dead_root(root);
-	if (err && err != -EAGAIN)
-		btrfs_handle_fs_error(fs_info, err, NULL);
 	return err;
 }
 
@@ -5605,6 +5674,19 @@ static int btrfs_trim_free_extents(struct btrfs_device *device, u64 *trimmed)
 					    &start, &end,
 					    CHUNK_TRIMMED | CHUNK_ALLOCATED);
 
+		/* Check if there are any CHUNK_* bits left */
+		if (start > device->total_bytes) {
+			WARN_ON(IS_ENABLED(CONFIG_BTRFS_DEBUG));
+			btrfs_warn_in_rcu(fs_info,
+"ignoring attempt to trim beyond device size: offset %llu length %llu device %s device size %llu",
+					  start, end - start + 1,
+					  rcu_str_deref(device->name),
+					  device->total_bytes);
+			mutex_unlock(&fs_info->chunk_mutex);
+			ret = 0;
+			break;
+		}
+
 		/* Ensure we skip the reserved area in the first 1M */
 		start = max_t(u64, start, SZ_1M);
 
@@ -5749,47 +5831,3 @@ int btrfs_trim_fs(struct btrfs_fs_info *fs_info, struct fstrim_range *range)
 		return bg_ret;
 	return dev_ret;
 }
-
-/*
- * btrfs_{start,end}_write_no_snapshotting() are similar to
- * mnt_{want,drop}_write(), they are used to prevent some tasks from writing
- * data into the page cache through nocow before the subvolume is snapshoted,
- * but flush the data into disk after the snapshot creation, or to prevent
- * operations while snapshotting is ongoing and that cause the snapshot to be
- * inconsistent (writes followed by expanding truncates for example).
- */
-void btrfs_end_write_no_snapshotting(struct btrfs_root *root)
-{
-	percpu_counter_dec(&root->subv_writers->counter);
-	cond_wake_up(&root->subv_writers->wait);
-}
-
-int btrfs_start_write_no_snapshotting(struct btrfs_root *root)
-{
-	if (atomic_read(&root->will_be_snapshotted))
-		return 0;
-
-	percpu_counter_inc(&root->subv_writers->counter);
-	/*
-	 * Make sure counter is updated before we check for snapshot creation.
-	 */
-	smp_mb();
-	if (atomic_read(&root->will_be_snapshotted)) {
-		btrfs_end_write_no_snapshotting(root);
-		return 0;
-	}
-	return 1;
-}
-
-void btrfs_wait_for_snapshot_creation(struct btrfs_root *root)
-{
-	while (true) {
-		int ret;
-
-		ret = btrfs_start_write_no_snapshotting(root);
-		if (ret)
-			break;
-		wait_var_event(&root->will_be_snapshotted,
-			       !atomic_read(&root->will_be_snapshotted));
-	}
-}