Commit ae1238b7 authored by Mark Harmstone's avatar Mark Harmstone Committed by David Sterba
Browse files

btrfs: read key again after incrementing slot in move_existing_remaps() 

Fix move_existing_remaps() so that if we increment the slot because the
key we encounter isn't a REMAP_BACKREF, we don't reuse the objectid and
offset of the old item.

Link: https://lore.kernel.org/linux-btrfs/20260125123908.2096548-1-clm@meta.com/


Reported-by: default avatarChris Mason <clm@fb.com>
Fixes: bbea42df ("btrfs: move existing remaps before relocating block group")
Reviewed-by: default avatarJohannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: default avatarMark Harmstone <mark@harmstone.com>
Reviewed-by: default avatarDavid Sterba <dsterba@suse.com>
Signed-off-by: default avatarDavid Sterba <dsterba@suse.com>
parent b2840e33

fs/btrfs/relocation.c

+2 −0
Original line number Diff line number Diff line
@@ -4399,6 +4399,8 @@ static int move_existing_remaps(struct btrfs_fs_info *fs_info, 


				leaf = path->nodes[0];

			}



			btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);

		}



		remap = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_remap_item);