Enabling Data Structure Oriented Performance Analysis with Hardware Performance Counter Support
An interesting and as of yet under-represented aspect of program development and optimization are data structures. Instead of analyzing data with respect to code regions, the objective here is to see how performance metrics are related to data structures. With the advanced performance monitoring unit of Intel’s Itanium processor series such an analysis becomes possible. This paper describes how the hardware features of the Itanium 2 processor are exploited by the perfmon and PAPI performance monitoring APIs and how PAPI’s support for address range restrictions has been integrated into an existing profiling tool to achieve the goal of data structure oriented profiling in the context of OpenMP applications.
KeywordsCritical Section Data Address Virtual Address Address Range Hardware Capability
Unable to display preview. Download preview PDF.
- 3.Gaugler, T.: Ein Werkzeug zur Untersuchung des Cacheverhaltens von Datenstrukturen mittels Ereigniszählern. Diplomarbeit, Universität Karlsruhe (2005)Google Scholar
- 5.Itzkowitz, M., Mazurov, O., Copty, N., Lin, Y.: An OpenMP runtime API for profiling. The OpenMP ARB as an official ARB White Paper (accepted), http://www.compunity.org/futures/omp-api.html
- 7.Levon, J.: OProfile, A system-wide profiler for Linux systems, http://oprofile.sourceforge.net
- 8.Mohr, B., Malony, A.D., Shende, S.S., Wolf, F.: Towards a performance tool interface for OpenMP: An approach based on directive rewriting. In: Proceedings of the Third Workshop on OpenMP, EWOMP 2001 (September 2001)Google Scholar
- 9.Perfmon2, the hardware-based performance monitoring interface for linux, http://perfmon2.sourceforge.net/