Mapping performance data for high-level and data views of parallel program performance
R. Bruce Irvin, Barton P. Miller · 1996
Programs written inhigh-level parallel languages need profiling tools that provide performance data in terms of the semantics of the high-level language.But high-level performance data can be incomplete when the cause of a performance problem camot be explained in terms of the semantics of the language.We also need the ability to view the performance of the underlying mechanisms used by the language and correlate the underlying activity to the language source code.The key techniques for providing these performance views is the ability to map low-level performance data up to the language abstractions.We describe how we use this information to produce performance data at the higher levels, and how we present this data in terms of both the code and parallel data structures.We have developed an implementation of these mapping techniques for the data parallel CM Fortran language running on the TMC CM-5.We have augmented the Paradyn Parallel Performance TOOIS with these mapping and high-level language facilities and used them to study several real data parallel Fortran (CM Fortran) applications.1 :[NTRODUCTION High-level parallel languages offer portable, concise notations for specifying parallel programs, and their compilers automatically map programs onto complex parallel machines.These languages can free programmers from the difficult, error-prone, and sometimes ineffective task of specifying parallel computations explicitly.We describe a tool for profiling the performance of parallel prc,grams written with a high-level parallel language or library.Thn paper concentrates on the mechanisms for mapping low-level performance data to the high-level language control and data struc-Tlrk work is supported in pan by Wright Laboratory Avionics Dkectorate, Air Force Matenaf Command, USAF, under grarrt F33615-94-1-1525 (ARPA order no.B550), arrd CDA-90246 18, arrd Department of Energy Grant DE-FG02-93ER25 176.