Porting CHAOS library to MPI
CHAOS is a very flexible library for runtime support of irregular problems, yet unfortunately, its original implementation is not applicable to parallel programs which are running under MPI. This paper presents details of porting CHAOS to MPI. The most important feature of the new improved version consists in separate MPI communicators: for internal CHAOS use and for an application program. The code of the library was carefully revised and now it is ANSI C compliant. Porting CHAOS to MPI allows for a much wider usability of this library as now it may be applied on virtually every parallel computer.
KeywordsDirect Monte Carlo Simulation Runtime Library Communication Schedule Runtime Support Distribute Memory Machine
Unable to display preview. Download preview PDF.
- 1.Mukherjee, S., Sharma, S., Hill, M., Larus, J., Rogers, A., and Saltz, J.: Efficient Support for Irregular Applications on Distributed-Memory Machines, Proceedings of the Fifth ACM SIGPLAN Symposium on Principles & Practice of Parallel Programming '95, Santa Barbara, California, July 19–21 (1995) 68–79Google Scholar
- 2.Hwang, Y.-S., Moon, B., Sharma, S., Ponnusamy, R., Das, R., and Saltz, J.: Runtime and Language Support for Compiling Adaptive Irregular Programs on Distributed Memory Machines, Software: Practice & Experience 25 (6) (1995) 597–621Google Scholar
- 3.Saltz, J., Ponnusammy, R., Sharma, S., Moon, B., Hwang,Y.-S., Uysal, M., and Das, R.: A Manual for the CHAOS Runtime Library, UMIACS Technical Reports CS-TR-3437 and UMIACS-TR-95-34, University of Maryland: Department of Computer Science, March 1995; ftp://hpsl.cs.umd.edu/pub/chaos_distributionGoogle Scholar
- 5.Edjlali, G., Sussman, A., and Saltz, J.: Interoperability of Data Parallel Runtime Libraries, International Parallel Processing Symposium 1997, April 1997Google Scholar
- 6.Brezany, P., Choudhary, A.: Techniques and Optimization for Developing Irregular Out-of-Core Applications on Distributed-memory Systems, Institute for Software technology and Parallel Systems, University of Vienna, TR 96-4, December 1996Google Scholar
- 7.Brezany, P.: Input/Output Intensively Parallel Computing, Lecture Notes in Computer Science, Vol. 1220, Springer-Verlag, Berlin Heildelberg New York (1997)Google Scholar
- 8.Message Passing Interface Forum: MPI-2: Extensions to the Message-Passing Interface, July 18, 1997; http://www.mpi-forum.org/docs/mpi-20.psGoogle Scholar
- 9.Berryman, H., Saltz, J., and Scroggs, J.: Execution time support for adaptive scientific algorithms on distributed memory machines, Concurency: Practice and Experience 3 (1991) 159–178Google Scholar