Digital Repository at the University of Maryland (DRUM)  >
Theses and Dissertations from UMD  >
UMD Theses and Dissertations 

Please use this identifier to cite or link to this item: http://hdl.handle.net/1903/12820

Title: AUTOMATIC CRITICAL SECTION DISCOVERY USING MEMORY USAGE PATTERNS.
Authors: Stechschulte, Lisa Marie
Advisors: Yeung, Donald
Department/Program: Electrical Engineering
Type: Thesis
Sponsors: Digital Repository at the University of Maryland
University of Maryland (College Park, Md.)
Subjects: Computer engineering
Computer science
Electrical engineering
Keywords: Atomicity
Critical Section
Memory
Issue Date: 2012
Abstract: Parallel programming introduces new types of bugs that are notoriously difficult to find. As a result researchers have put a significant amount of effort into creating tools and techniques to discover parallel bugs. One of these bugs is the violation of the assumption of <bold>atomicity</bold>-- the assumption that a region of code, called a <bold>critical section</bold>, executes without interruption from an outside operation. In this thesis, we introduce a new heuristic to infer critical sections using the temporal and spatial locality of critical sections and provide empirical results showing that the heuristic can infer critical sections in shared memory programs. Real critical sections in benchmark programs are completely covered by inferred critical sections up to 75% to 80% of the time. A programmer can use the reported critical sections to inform his addition of locks into the program.
URI: http://hdl.handle.net/1903/12820
Appears in Collections:UMD Theses and Dissertations
Electrical & Computer Engineering Theses and Dissertations

Files in This Item:

File Description SizeFormatNo. of Downloads
Stechschulte_umd_0117N_13240.pdf2.63 MBAdobe PDF286View/Open

All items in DRUM are protected by copyright, with all rights reserved.

 

DRUM is brought to you by the University of Maryland Libraries
University of Maryland, College Park, MD 20742-7011 (301)314-1328.
Please send us your comments