Zach Saw <zach.saw at ...> writes: > The user's finalizer routine then registers this object for a second finalization > via GC_register_finalizer_no_order. However, this does not ever get triggered. > It appears that the problem isn't with the GC at all. Sorry for the confusion.