Cilkprof : a scalability profiler for Cilk programs

This thesis describes the design and implementation of Cilkprof, a profiling tool that helps programmers to diagnose scalability problems in their Cilk programs. Cilkprof provides in-depth information about the scalability of programs, without adding excessive overhead. Cilkprof's output can be used to find scalability bottlenecks in the user's code. Cilkprof makes profiling measurements at the fork and join points of computations, which typically limits the amount of overhead incurred by the profiler. In addition, despite recording in-depth information, Cilkprof does not generate large log files that are typical of trace-based profilers. In addition, the profiling algorithm only incurs constant amortized overhead per measurement. CilkProf slows down the serial program execution by a factor of about 10 in the common case, on a well-coarsened parallel program. The slowdown is reasonable for the amount of information gained from the profiling. Finally, the approach taken by Cilkprof enables the creation of an API, which can allow users to specify their own profiling code, without having to change the Cilk runtime.

Department dc:contributor.department

Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2014

Author dc:creator

• Xing, Kerry (Kerry K.)

Advisor dc:contributor.advisor

• Charles E. Leiserson and I-Ting Angelina Lee.

• Electrical Engineering and Computer Science.

Statement dc:rights

• M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/91879

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/91879