[gclist] Finalization and object orientation.

Hans Boehm boehm@hoh.mti.sgi.com
Fri, 28 Mar 1997 11:15:44 -0800

On Mar 28, 10:39am, Eliot & Linda wrote:

> So can we agree that in languages which reify execution state and
> don't allow pointer forging that reachability is decidable?

I don't think so.  Whether or not the context will be accessed is certainly
undecidable in general.  In special cases, an agressive collector/compiler
might sometimes determine that it will not be accessed, and collect dead parts
of the context.

I think you really have a different notion of reachability in mind.  I don't
know how easy it is to formally define your notion for Smalltalk.  Do contexts
include compiler-introduced temporaries for subexpression results?  Is argument
evaluation order defined?

I think that pointer-forging is really not relevant here.  Even in environments
in which pointer-forging is an issue, I think we could all agree that legal
programs shouldn't access objects through forged pointers, and hence the
possibility shouldn't influence the notion of reachability.


Hans-Juergen Boehm