A comparison of task pool variants in OpenMP and a proposal for a solution to the busy waiting problem

Authors:
Alexander Wirz;Michael Süß;Claudia Leopold
Affiliations:
University of Kassel, Research Group Programming Languages / Methodologies, Kassel, Germany;University of Kassel, Research Group Programming Languages / Methodologies, Kassel, Germany;University of Kassel, Research Group Programming Languages / Methodologies, Kassel, Germany
Venue:
IWOMP'05/IWOMP'06 Proceedings of the 2005 and 2006 international conference on OpenMP shared memory parallel programming
Year:
2005

Citing 3
Cited 1

OpenMP on networks of workstations

SC '98 Proceedings of the 1998 ACM/IEEE conference on Supercomputing
Computer Solution of Large Sparse Positive Definite

Computer Solution of Large Sparse Positive Definite
A comparison of task pools for dynamic load balancing of irregular algorithms: Research Articles

Concurrency and Computation: Practice & Experience

The GNU libstdc++ parallel mode: software engineering considerations

Proceedings of the 1st international workshop on Multicore software engineering

Quantified Score

Hi-index	0.00

Visualization

Abstract

Irregular algorithms are difficult to parallelize using existing OpenMP constructs. This paper concentrates on algorithms that deploy task pools, i.e., data structures for dynamic load balancing. We present several task pool variants that we have implemented in OpenMP, and compare their performance. Due to the lack of a mechanism in OpenMP to put a thread to sleep, we had to use busy waiting in our implementations. To eliminate this need, we suggest an extension to OpenMP that allows to put a thread to sleep on demand.