Btrfs: unlock nodes earlier when inserting items in a btree

In ctree.c:setup_items_for_insert(), we can unlock all nodes in our
path before we process the leaf (shift items and data, adjust data
offsets, etc). This allows for better btree concurrency, as we're
often holding a write lock on at least the node at level 1.

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
This commit is contained in:
Filipe Manana 2014-07-28 19:34:35 +01:00 committed by Chris Mason
parent d1b00a4711
commit 24cdc847d9

View File

@ -4738,6 +4738,12 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
int slot;
struct btrfs_map_token token;
if (path->slots[0] == 0) {
btrfs_cpu_key_to_disk(&disk_key, cpu_key);
fixup_low_keys(root, path, &disk_key, 1);
}
btrfs_unlock_up_safe(path, 1);
btrfs_init_map_token(&token);
leaf = path->nodes[0];
@ -4798,12 +4804,6 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
}
btrfs_set_header_nritems(leaf, nritems + nr);
if (slot == 0) {
btrfs_cpu_key_to_disk(&disk_key, cpu_key);
fixup_low_keys(root, path, &disk_key, 1);
}
btrfs_unlock_up_safe(path, 1);
btrfs_mark_buffer_dirty(leaf);
if (btrfs_leaf_free_space(root, leaf) < 0) {