Btrfs: fix early enospc during balancing (7ce618db) · Commits · 戴 / test

We now do extra checks before a balance to make sure there is room for the balance to take place. One of the checks was testing to see if we were trying to balance away the last block group of a given type. If there is no space available for new chunks, we should not try and balance away the last block group of a give type. But, the code wasn't checking for available chunk space, and so it was exiting too soon. The fix here is to combine some of the checks and make sure we try to allocate new chunks when we're balancing the last block group. Signed-off-by:

Chris Mason <chris.mason@oracle.com>

fs/btrfs/extent-tree.c

+7 −13

Original line number	Diff line number	Diff line
		@@ -6756,22 +6756,16 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr)

		/*
		* if this is the last block group we have in this space, we can't
		* relocate it.
		*/
		if (space_info->total_bytes == block_group->key.offset) {
		ret = -1;
		spin_unlock(&space_info->lock);
		goto out;
		}

		/*
		* need to make sure we have room in the space to handle all of the
		* extents from this block group. If we can, we're good
		* relocate it unless we're able to allocate a new chunk below.
		*
		* Otherwise, we need to make sure we have room in the space to handle
		* all of the extents from this block group. If we can, we're good
		*/
		if (space_info->bytes_used + space_info->bytes_reserved +
		if ((space_info->total_bytes != block_group->key.offset) &&
		(space_info->bytes_used + space_info->bytes_reserved +
		space_info->bytes_pinned + space_info->bytes_readonly +
		btrfs_block_group_used(&block_group->item) <
		space_info->total_bytes) {
		space_info->total_bytes)) {
		spin_unlock(&space_info->lock);
		goto out;
		}

Admin message