From:                     Kendra Smith

Sent:                      Tuesday, May 09, 2000 11:29 PM

To:                         M?crosöft Research Tech Talk, Sem. Notice

Cc:                         Kendra Smith

Subject:                 UW-CSE Colloq / 5-10-2000 / Das / MSR / Toward Scalable Pointer Analysis

UW-CSE Colloq / 5-10-2000 / Das / MSR / Toward Scalable Pointer Analysis

 

This is a revised title and abstract for Manuvir Das' talk on Wednesday.

 

NOTE: THIS LECTURE WILL *NOT* BE VIDEOTAPED.

 

UNIVERSITY OF WASHINGTON

Seattle, Washington 98195

 

Department of Computer Science and Engineering

Box 352350

(206) 543-1695

 

COLLOQUIUM

 

SPEAKER:      Manuvir Das, M?crosöft Research

 

TITLE:          Towards Scalable Pointer Analysis

 

DATE:           Wednesday, May 10, 2000

 

TIME:           2:30pm

 

PLACE:                   EE1-003

 

HOST:           Susan Eggers

 

Abstract: Our research group at M?crosöft is working on developing

error detection methods for commercial applications software. Many of

these methods involve some form of static program analysis. Because

the programs being analyzed make heavy use of pointer valued

variables, the success of our error detection methods depends

critically on pointer analysis.  Unfortunately, pointer analysis

algorithms exhibit a natural trade-off between the quality of

information produced, and the efficiency of the algorithm. Fast

algorithms are too imprecise, while precise algorithms are too

slow. In this talk, I shall describe our attempts (both successes and

failures) to produce a pointer analysis that scales both in

performance, and in precision. I shall describe three new algorithms

for pointer analysis that successively improve precision, while

maintaining scalability in performance. The result is an analysis that

scales easily to large programs (it processes a 1.4MLOC program in

less than two minutes of analysis time and 200MB of memory), while

providing precision similar to that of much more expensive

analyses. Finally, I shall draw on my experiences to suggest a

methodology for developing scalable static analysis algorithms.

 

Bio: Manuvir Das is a Researcher in the Software Productivity Tools

group at M?crosöft Research. He received a B. Tech. in Computer

Science from the Indian Institute of Technology, Bombay, in 1991.  He

did his graduate work at the University of Wisconsin-Madison, from

where he received a Ph.D. in 1998. His interests lie in scalable

program analysis, its application to error detection tools, and

methods to improve the correctness of software.

 

Email: talk-info@cs.washington.edu

Info: http://www.cs.washington.edu