Fix to a bug in ISO incremental tabling that showed up in the flora
testsuite. The bug what that I forgot to properly protect the heap
when the update occurred when we were not in tabling mode during the
update, but some table was created subsequent to update. In this case
hfreg had been set to the bottom of the heap, improperly.
Having said that, most of the diffs have to do with commented
debugging statements that I'm keeping around for now.