btrfs: zoned: don't skip block groups with 100% zone unusable
authorJohannes Thumshirn <johannes.thumshirn@wdc.com>
Wed, 21 Feb 2024 15:35:52 +0000 (07:35 -0800)
committerDavid Sterba <dsterba@suse.com>
Tue, 26 Mar 2024 15:42:39 +0000 (16:42 +0100)
Commit f4a9f219411f ("btrfs: do not delete unused block group if it may be
used soon") changed the behaviour of deleting unused block-groups on zoned
filesystems. Starting with this commit, we're using
btrfs_space_info_used() to calculate the number of used bytes in a
space_info. But btrfs_space_info_used() also accounts
btrfs_space_info::bytes_zone_unusable as used bytes.

So if a block group is 100% zone_unusable it is skipped from the deletion
step.

In order not to skip fully zone_unusable block-groups, also check if the
block-group has bytes left that can be used on a zoned filesystem.

Fixes: f4a9f219411f ("btrfs: do not delete unused block group if it may be used soon")
CC: stable@vger.kernel.org # 6.1+
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
fs/btrfs/block-group.c

index 5f7587ca1ca7720d26a7cd192fdcd7ccd97c55bd..1e09aeea69c22e011b5a8f305421b342d04aa8b4 100644 (file)
@@ -1559,7 +1559,8 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info)
                 * needing to allocate extents from the block group.
                 */
                used = btrfs_space_info_used(space_info, true);
-               if (space_info->total_bytes - block_group->length < used) {
+               if (space_info->total_bytes - block_group->length < used &&
+                   block_group->zone_unusable < block_group->length) {
                        /*
                         * Add a reference for the list, compensate for the ref
                         * drop under the "next" label for the