| | |
Summary: Banshee: A Scalable Constraint-Based Analysis
Toolkit
John Kodumal1
and Alex Aiken2
1
EECS Department, University of California, Berkeley
2
Computer Science Department, Stanford University
Abstract. We introduce Banshee, a toolkit for constructing constraint-
based analyses. Banshee's novel features include a code generator for
creating customized constraint resolution engines, incremental analysis
based on backtracking, and fast persistence. These features make Ban-
shee useful as a foundation for production program analyses.
1 Introduction
Program analyses that are simultaneously scalable, accurate, and efficient re-
main expensive to develop. One approach to lowering implementation cost is to
express the analysis using constraints. Constraints separate analysis specification
(constraint generation) from analysis implementation (constraint resolution). By
exploiting this separation, designers can benefit from existing algorithms for
constraint resolution. This separation helps, but leaves several problems unad-
|
