Kamada, T., Matsuoka, S., and Yonezawa, A. 1994. Efficient parallel garbage collection on massively parallel computers. In Proceedings of the 1994 IEEE Supercomputing Conference.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....algorithm. 6.3 Comparision to Other Distributed Collectors Unlike reference counting collection strategies [Lang et al. 1992] our asynchronous collector requires only two bits to reclaim proxies. Our approach bears greater similarity to distributed marking schemes that use bulldozing messages [Kamada et al. 1994; Venkatasubramanian et al. 1992] to confirm message arrival. Unlike these other schemes, the tick process used to provide the effect of bulldozing is itself asynchronous with respect to user processes, requires no extra acknowledgment signals, and is not centralized. The cyclic extension is ....

Kamada, T., Matsuoka, S., and Yonezawa, A. 1994. Efficient parallel garbage collection on massively parallel computers. In Proceedings of the 1994 IEEE Supercomputing Conference.

....grid architecture is considered and properties concerning message routing on the grid topology are used for detecting a consistent system state; in our proposition, timestamping of events is used for this purpose, and noassumption is madeon the underlying architecture. The algorithmproposed in [25], like the one proposed for EMERALD, considers every running object as non garbage, and thus does not detect unneeded computations. This algorithm was designed for massively parallel multiprocessors, and scales up to a 1024 nodes machine, but unlike our proposition does not consider unreliable ....

S. Matsuoka T. Kamada and A. Yonezawa. Efficient parallel garbage collection on massively parallel computers. In Proc. of Supercomputing'94, Washington D.C., November 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC