Abstract
Today, millions of legacy programs are awaiting their parallelization. For this reason, the automatic discovery of parallelism in sequential programs is now receiving considerable attention. However, past efforts mainly concentrated on data parallelism hidden inside loops. As programming models begin to support more irregular types of parallelism, centered around the notion of tasks in various forms, methods are needed to identify code sections that could potentially represent parallel tasks. In this paper, we present a novel approach to automatically finding parallel tasks in sequential programs. We first created a dynamic dependence graph, then isolated tasks, and finally produced a task graph according to the dependences we find. With the help of a source-to-source code translator, parallel code is automatically generated. We conducted a range of experiments to cover both tasks executing the same code and tasks executing different code. Results showed that our method achieved reasonable speedups on the test cases.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
