Btrfs: faster file extent item replace operations

When writing to a file we drop existing file extent items that cover the
write range and then add a new file extent item that represents that write
range.

Before this change we were doing a tree lookup to remove the file extent
items, and then after we did another tree lookup to insert the new file
extent item.
Most of the time all the file extent items we need to drop are located
within a single leaf - this is the leaf where our new file extent item ends
up at. Therefore, in this common case just combine these 2 operations into
a single one.

By avoiding the second btree navigation for insertion of the new file extent
item, we reduce btree node/leaf lock acquisitions/releases, btree block/leaf
COW operations, CPU time on btree node/leaf key binary searches, etc.

Besides for file writes, this is an operation that happens for file fsync's
as well. However log btrees are much less likely to big as big as regular
fs btrees, therefore the impact of this change is smaller.

The following benchmark was performed against an SSD drive and a
HDD drive, both for random and sequential writes:

  sysbench --test=fileio --file-num=4096 --file-total-size=8G \
     --file-test-mode=[rndwr|seqwr] --num-threads=512 \
     --file-block-size=8192 \ --max-requests=1000000 \
     --file-fsync-freq=0 --file-io-mode=sync [prepare|run]

All results below are averages of 10 runs of the respective test.

** SSD sequential writes

Before this change: 225.88 Mb/sec
After this change:  277.26 Mb/sec

** SSD random writes

Before this change: 49.91 Mb/sec
After this change:  56.39 Mb/sec

** HDD sequential writes

Before this change: 68.53 Mb/sec
After this change:  69.87 Mb/sec

** HDD random writes

Before this change: 13.04 Mb/sec
After this change:  14.39 Mb/sec

Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>

1 files changed, 40 insertions, 4 deletions

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 35bf83876f3f..030012e1710c 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -692,7 +692,10 @@ next:
 int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
 			 struct btrfs_root *root, struct inode *inode,
 			 struct btrfs_path *path, u64 start, u64 end,
-			 u64 *drop_end, int drop_cache)
+			 u64 *drop_end, int drop_cache,
+			 int replace_extent,
+			 u32 extent_item_size,
+			 int *key_inserted)
 {
 	struct extent_buffer *leaf;
 	struct btrfs_file_extent_item *fi;
@@ -712,6 +715,7 @@ int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
 	int modify_tree = -1;
 	int update_refs = (root->ref_cows || root == root->fs_info->tree_root);
 	int found = 0;
+	int leafs_visited = 0;
 
 	if (drop_cache)
 		btrfs_drop_extent_cache(inode, start, end - 1, 0);
@@ -733,6 +737,7 @@ int __btrfs_drop_extents(struct btrfs_trans_handle *trans,
 				path->slots[0]--;
 		}
 		ret = 0;
+		leafs_visited++;
 next_slot:
 		leaf = path->nodes[0];
 		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
@@ -744,6 +749,7 @@ next_slot:
 				ret = 0;
 				break;
 			}
+			leafs_visited++;
 			leaf = path->nodes[0];
 			recow = 1;
 		}
@@ -927,14 +933,44 @@ next_slot:
 	}
 
 	if (!ret && del_nr > 0) {
+		/*
+		 * Set path->slots[0] to first slot, so that after the delete
+		 * if items are move off from our leaf to its immediate left or
+		 * right neighbor leafs, we end up with a correct and adjusted
+		 * path->slots[0] for our insertion.
+		 */
+		path->slots[0] = del_slot;
 		ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
 		if (ret)
 			btrfs_abort_transaction(trans, root, ret);
+
+		leaf = path->nodes[0];
+		/*
+		 * leaf eb has flag EXTENT_BUFFER_STALE if it was deleted (that
+		 * is, its contents got pushed to its neighbors), in which case
+		 * it means path->locks[0] == 0
+		 */
+		if (!ret && replace_extent && leafs_visited == 1 &&
+		    path->locks[0] &&
+		    btrfs_leaf_free_space(root, leaf) >=
+		    sizeof(struct btrfs_item) + extent_item_size) {
+
+			key.objectid = ino;
+			key.type = BTRFS_EXTENT_DATA_KEY;
+			key.offset = start;
+			setup_items_for_insert(root, path, &key,
+					       &extent_item_size,
+					       extent_item_size,
+					       sizeof(struct btrfs_item) +
+					       extent_item_size, 1);
+			*key_inserted = 1;
+		}
 	}
 
+	if (!replace_extent || !(*key_inserted))
+		btrfs_release_path(path);
 	if (drop_end)
 		*drop_end = found ? min(end, extent_end) : end;
-	btrfs_release_path(path);
 	return ret;
 }
 
@@ -949,7 +985,7 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 	if (!path)
 		return -ENOMEM;
 	ret = __btrfs_drop_extents(trans, root, inode, path, start, end, NULL,
-				   drop_cache);
+				   drop_cache, 0, 0, NULL);
 	btrfs_free_path(path);
 	return ret;
 }
@@ -2219,7 +2255,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	while (cur_offset < lockend) {
 		ret = __btrfs_drop_extents(trans, root, inode, path,
 					   cur_offset, lockend + 1,
-					   &drop_end, 1);
+					   &drop_end, 1, 0, 0, NULL);
 		if (ret != -ENOSPC)
 			break;