Indeed it's not true. On-the-fly collectors coordinate with each mutator thread one at a time. The disadvantage is that a CPU is effectively dedicated to GC.
Sandro On 22/03/2010 2:40 PM, Jonathan S. Shapiro wrote: > On Mon, Mar 22, 2010 at 11:31 AM, Rodrigo Kumpera <[email protected]> wrote: >> Given that concurrent GCs have to eventually stop all threads to mark them... > > Why do you believe that this is true? It is probably true that each > thread, taken individually, must come to a state that is coordinated > with the collector. It does not seem true that there needs to be a > state in which *all* threads are simultaneously coordinated with the > collector. > _______________________________________________ > bitc-dev mailing list > [email protected] > http://www.coyotos.org/mailman/listinfo/bitc-dev _______________________________________________ bitc-dev mailing list [email protected] http://www.coyotos.org/mailman/listinfo/bitc-dev
