JEP 132: More-prompt finalization

Dmitry Samersoff Dmitry.Samersoff at oracle.com
Tue Dec 27 23:09:10 PST 2011


Jon,

On 2011-12-28 01:50, Jon Masamitsu wrote:
> For applications that have large heaps which don't fill up
> and so don't cause a GC (during which finalizable objects could be
> discovered), the least disruptive feature we've discussed is
> to have a concurrent phase such as CMS's concurrent
> marking phase discover finalizable objects.

Thank you for clarification. It's exactly what I'm looking for.

-Dmitry

-- 
Dmitry Samersoff
Java Hotspot development team, SPB04
* There will come soft rains ...


More information about the hotspot-dev mailing list