[gclist] Finalizers & Reference counting.
Sat, 31 Aug 2002 18:19:39 +0100
At 2002-08-31 16:53:33+0000, Andrew Shi-hwa Chen writes:
> I think "don't promote" is overly specific.
Absolutely; I was responding to the question of whether a generational
collector has particular problems with compilers (or other strongly
phased mutators). A generational collector might well have an
interface controlling promotion; a generic memory management layer on
top of such a collector might provide some sort of liveness interface
which translates into promotion control.