> Isn't the call to GC_mutatorInvariant sufficient to make sure that the > heap was ok at the end of the penultimate GC? Maybe. But I'd like you to look at the heap nevertheless. I want to know when it actually gets busted.