Raoul - Thanks for posting that reference to Azul. While we've heard of Azul before (and already mentioned them in this thread), I was not aware that they had already done exactly what I had wondered about. They mated their Pauseless GC algorithm (now called C4) to an x86 MMU supported barrier. Their paper claims sub 0.02s worst-case pause times with a 30GB+ heap, where Sun's concurrent collector has a worst-case pause of over 10 seconds.
http://www.azulsystems.com/technology/c4-garbage-collector http://dl.acm.org/citation.cfm?id=1993491&dl=ACM&coll=DL&CFID=96709212&CFTOKEN=86276521 Very interesting read for anyone who hasn't seen it.
_______________________________________________ bitc-dev mailing list [email protected] http://www.coyotos.org/mailman/listinfo/bitc-dev
