robustness vs. conservative GC

[Jeff Sturm]

On Fri, 10 Aug 2001, Tom Lord wrote:
>        Supposing you do have exact type information for the stack, how can GC 
>        be certain a stack slot is initialized?
> 
> If the information you have doesn't tell you that, then it isn't
> exact.

Fair enough.  But now instead of one stack frame description per function,
you may need several.  And observe correct sequential ordering on a
multiprocessor.

I'm still finding it hard to imagine how that can be done for reasonable
cost.

Jeff