Michael Martin

Ph.D. Candidate

Stanford University

My primary research interests involve custom runtimes to efficiently track program executions for safety or program understanding purposes, instead of traditional applications such as profiling. Most of this work is being done on the Java Virtual Machine, because the JVM's bytecodes have so much metadata attached to them that you don't even need source to be able to do nifty analyses.

The Program Query Language runtime matching system is the most direct fruit of this. It also ties into many of the advanced static analyses done my by group. You can see the full suite of papers on the SUIF Program Analysis page.

Papers

Automatic Extraction of Object-Oriented Component Interfaces. by John Whaley, Michael C. Martin and Monica S. Lam. In Proceedings of the International Symposium on Software Testing and Analysis, pages 218-228, July 2002. This won the ACM SIGSOFT Distinguished Paper Award in 2002. We built a system that would automatically infer state machine models based on the usage of a software component -- the paper discusses our results. The system we built for this never really got past the research prototype phase, and developed into other things. The problem itself wound up being picked up and developed to good effect by the University of Pennsylvania's JIST project.
Design, implementation and testing of extended and mixed precision BLAS, by Xiaoye S. Li, James W. Demmel, David H. Bailey, Greg Henry, Yozo Hida, Jimmy Iskandar, William Kahan, Suh Y. Kang, Anil Kapur, Michael C. Martin, Brandon J. Thompson, Teresa Tung, and Daniel J. Yoo. In ACM Transactions on Mathematical Software, 28(2):152-205, 2002. No link here because I didn't actually contribute to the paper at all, but this paper appears to be the fruit of the research team I worked with as an undergraduate at UC Berkeley back in 1999. Section 5 (pp. 187-189) describes the part we undergrads were working on - using M4 to automatically generate templated C code for the many Extended BLAS functions.
Context-Sensitive Program Analysis as Database Queries, by Monica S. Lam, John Whaley, V. Benjamin Livshits, Michael C. Martin, Dzintars Avots, Michael Carbin and Christopher Unkel. In Proceedings of the 24th SIGMOD-SIGACT-SIGART Symposium on Principles of Database Systems, pages 1-12, June 2005. This was an invited talk by my advisor, and as such is a collaborative paper by the entire group.
Finding Application Errors and Security Flaws Using PQL: a Program Query Language, by Michael C. Martin, V. Benjamin Livshits, and Monica S. Lam. In Proceedings of the Conference on Object-Oriented Programming, Systems, Languages and Applications (OOPSLA '05), pages 365-383, October 2005. This paper describes the PQL system, which turns program queries (which are, essentially, snippets of Java code) into matchers. The base system finds matches in a program as the program runs, and can trigger events happpening on the match. It can also be used to drive static analyses. The paper discusses both. PQL itself is hosted on SourceForge and includes demonstration code. It is updated as bits of it become less "research prototype" and more "actually usable." If you read one paper, read this one.
Securing Web Applications with Static and Dynamic Information Flow Tracking, by Monica S. Lam, Michael C. Martin, Benjamin Livshits, and John Whaley. Invited paper at the ACM SIGPLAN 2008 Workshop on Partial Evaluation and Program Manipulation (PEPM '08). Provides a nice unified view of the Web-based analysis our group has done over the years.

Michael Martin

Ph.D. Candidate

Stanford University

Papers

Other Material