btrfs: fix comment about nbytes increase at replay_one_extent()

The comment is wrong about the part where it says a prealloc extent does
not contribute to an inode's nbytes - it does. Only holes don't contribute
and that's what we are checking for, as prealloc extents always have a
disk_bytenr different from 0. So fix the comment and re-organize the code
to not set nbytes twice and set it to the extent item's number of bytes
only if it doesn't represent a hole - in case it's a hole we have already
initialized nbytes to 0 when we declared it.

Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This commit is contained in:
Filipe Manana 2025-09-18 12:58:53 +01:00 committed by David Sterba
parent 2d83ed6c6c
commit a35b3dd59b

View file

@ -740,15 +740,10 @@ static noinline int replay_one_extent(struct walk_control *wc)
if (found_type == BTRFS_FILE_EXTENT_REG ||
found_type == BTRFS_FILE_EXTENT_PREALLOC) {
nbytes = btrfs_file_extent_num_bytes(wc->log_leaf, item);
extent_end = start + nbytes;
/*
* We don't add to the inodes nbytes if we are prealloc or a
* hole.
*/
if (btrfs_file_extent_disk_bytenr(wc->log_leaf, item) == 0)
nbytes = 0;
extent_end = start + btrfs_file_extent_num_bytes(wc->log_leaf, item);
/* Holes don't take up space. */
if (btrfs_file_extent_disk_bytenr(wc->log_leaf, item) != 0)
nbytes = btrfs_file_extent_num_bytes(wc->log_leaf, item);
} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
size = btrfs_file_extent_ram_bytes(wc->log_leaf, item);
nbytes = btrfs_file_extent_ram_bytes(wc->log_leaf, item);