include dirty data accounting from the tmp (fastdiff) snapshot if it exists
also, check if we're in the case where both the latest snapshots both have a zero "used" value. in this case, because zfs space accounting for "used" values for snapshots only reports blocks unique to that snapshot, we'll fail to detect changes to the filesystem in the dirty data algorithm. in this case, we need to clean up the tmp snapshot before proceeding, otherwise we risk missing filesystem changes.
include dirty data accounting from the tmp (fastdiff) snapshot if it exists
also, check if we're in the case where both the latest snapshots both have a zero "used" value. in this case, because zfs space accounting for "used" values for snapshots only reports blocks unique to that snapshot, we'll fail to detect changes to the filesystem in the dirty data algorithm. in this case, we need to clean up the tmp snapshot before proceeding, otherwise we risk missing filesystem changes.