Refactor: libpacemaker: Don't continue if cluster_status fails.

cluster_status can return FALSE for a couple error conditions, most
notably being called on a feature set that is newer than what's
supported. In that case, the caller should return its own error instead
of trying to continue on with an unpopulated scheduler object. This
prevents a cascade of error messages.

Note, however, that one caller is not handled here.
pcmkschedule_actions calls unpack_cib which calls cluster_status.
However, pcmkschedule_actions does not return any value and its
callers aren't really set up to handle a return value, either. So that
will need to be dealt with in a separate patch.

Ref T521