Performance enhancements for a dynamic invariant detector
Name
227813645-MIT.pdf
Description
Full printable version
Size
4.66 MB
Format
Adobe PDF
Checksum (MD5)
a31897d1cf6e26dcad7c5430342a16cc
Author(s)
Xiao, Chen, M. Eng. Massachusetts Institute of Technology
Advisor(s)
Michael D. Ernst and Jeff H. Perkins.
Date Issued
2007
Publisher
Massachusetts Institute of Technology
Abstract
Dynamic invariant detection is the identification of the likely properties about a program based on observed variable values during program execution. While other dynamic invariant detectors use a brute force algorithm, Daikon adds powerful optimizations to provide more scalable invariant detection without sacrificing the richness of the reported invariants. Daikon improves scalability by eliminating redundant invariants. For example, the suppression optimization allows Daikon to delay the creation of invariants that are logically implied by other true invariants. Although conceptually simple, the implementation of this optimization in Daikon has a, large fixed cost and scales polynomially with the number of program variables. I investigated performance problems in two implementations of the suppression optimization in Daikon and evaluated several methods for improving the algorithm for the suppression optimization: optimizing existing algorithms, using a hybrid, context-sensitive approach to maximize the effectiveness of the two algorithms, and batching applications of the algorithm to lower costs. Experimental results showed a 10% runtime improvement in Daikon runtime. In addition, I implemented an oracle to verify the implementation of these improvements and the other optimizations in Daikon.
Description
Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2007.
Includes bibliographical references (p. 93-95).
Subjects
Electrical Engineering and Computer Science.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
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.
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.
Persistent DSpace Link
