[j_seigh]

Only using way more bits. The original IBM lock-free stack algorithm assumed 32 bits was safe because it would take 100 years for a 32 bit counter to wrap at the time. Now it's less than 1 second.

There's some Bugblatter Beast logic here. They assume that if they can't imagine a counter/pointer value reuse ever reoccurring then malicious actors won't be able to imagine it either.

Here's an idea. How about fixing the buffer overflows and lack of data sanitation that allow code injection to occur in the first place.