How to reconcile event-based performance analysis with tasking in OpenMP

Authors:
Daniel Lorenz;Bernd Mohr;Christian Rössel;Dirk Schmidl;Felix Wolf
Affiliations:
Jülich Supercomputing Centre, Forschungszentrum Jülich, Germany;Jülich Supercomputing Centre, Forschungszentrum Jülich, Germany;Jülich Supercomputing Centre, Forschungszentrum Jülich, Germany;Dept. of Computer Science, RWTH Aachen University, Germany;Jülich Supercomputing Centre, Forschungszentrum Jülich, Germany
Venue:
IWOMP'10 Proceedings of the 6th international conference on Beyond Loop Level Parallelism in OpenMP: accelerators, Tasking and more
Year:
2010

Citing 5
Cited 3

Design and Prototype of a Performance Tool Interface for OpenMP

The Journal of Supercomputing
The Tau Parallel Performance System

International Journal of High Performance Computing Applications
Providing Observability for OpenMP 3.0 Applications

IWOMP '09 Proceedings of the 5th International Workshop on OpenMP: Evolving OpenMP in an Age of Extreme Parallelism
Performance Profiling for OpenMP Tasks

IWOMP '09 Proceedings of the 5th International Workshop on OpenMP: Evolving OpenMP in an Age of Extreme Parallelism
The Scalasca performance toolset architecture

Concurrency and Computation: Practice & Experience - Scalable Tools for High-End Computing

Performance analysis techniques for task-based OpenMP applications

IWOMP'12 Proceedings of the 8th international conference on OpenMP in a Heterogeneous World
On the instrumentation of OpenMP and ompss tasking constructs

Euro-Par'12 Proceedings of the 18th international conference on Parallel processing workshops
Extending the scope of the controlled logical clock

Cluster Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

With version 3.0, the OpenMP specification introduced a task construct and with it an additional dimension of concurrency. While offering a convenient means to express task parallelism, the new construct presents a serious challenge to event-based performance analysis. Since tasking may disrupt the classic sequence of region entry and exit events, essential analysis procedures such as reconstructing dynamic call paths or correctly attributing performance metrics to individual task region instances may become impossible. To overcome this limitation, we describe a portable method to distinguish individual task instances and to track their suspension and resumption with event-based instrumentation. Implemented as an extension of the OPARI source-code instrumenter, our portable solution supports C/C++ programs with tied tasks and with untied tasks that are suspended only at implied scheduling points, while introducing only negligible measurement overhead. Finally, we discuss possible extensions of the OpenMP specification to provide general support for task identifiers with untied tasks.