Visualizing the program execution control flow of OpenMP applications

Authors:
Karl Fürlinger;Shirley Moore
Affiliations:
Innovative Computing Laboratory, EECS Department, University of Tennessee, Knoxville;Innovative Computing Laboratory, EECS Department, University of Tennessee, Knoxville
Venue:
IWOMP'08 Proceedings of the 4th international conference on OpenMP in a new era of parallelism
Year:
2008

Citing 2
Cited 0

A Portable Programming Interface for Performance Evaluation on Modern Processors

International Journal of High Performance Computing Applications
Dragon: a static and dynamic tool for OpenMP

WOMPAT'04 Proceedings of the 5th international conference on OpenMP Applications and Tools: shared Memory Parallel Programming with OpenMP

Quantified Score

Hi-index	0.00

Visualization

Abstract

One important aspect of understanding the behavior of anapplication with respect to its performance, overhead, and scalabilitycharacteristics is knowledge of its control flow. In comparison to sequentialapplications the situation is more complicated in multithreaded parallelprograms because each thread defines its own independent controlflow. On the other hand, for the most common usage models of OpenMPthe threads operate in a largely uniform way, synchronizing frequentlyat sequence points and diverging only to operate on different data itemsin worksharing constructs. This paper presents an approach to capture and visualize the controlflow of OpenMP applications in a compact way that does not require afull trace of program execution events but is instead based on a straightforwardextension to the data collected by an existing profiling tool.