Search for: All records

Researchers have reported that static analysis tools rarely achieve a false-positive rate that would make them attractive to developers. We overcome this problem by a technique that leads to reporting fewer bugs but also much fewer false positives. Our technique prunes the static call graph that sits at the core of many static analyses. Specifically, static call-graph construction proceeds as usual, after which a call-graph pruner removes many false-positive edges but few true edges. The challenge is to strike a balance between being aggressive in removing false-positive edges but not so aggressive that no true edges remain. We achieve this goal by automatically producing a call-graph pruner through an automatic, ahead-of-time learning process. We added such a call-graph pruner to a software tool for null-pointer analysis and found that the false-positive rate decreased from 73% to 23%. This improvement makes the tool more useful to developers. 

Reducing a failure-inducing input to a smaller one is challenging for input with internal dependencies because most sub-inputs are invalid. Kalhauge and Palsberg made progress on this problem by mapping the task to a reduction problem for dependency graphs that avoids invalid inputs entirely. Their tool J-Reduce efficiently reduces Java bytecode to 24 percent of its original size, which made it the most effective tool until now. However, the output from their tool is often too large to be helpful in a bug report. In this paper, we show that more fine-grained modeling of dependencies leads to much more reduction. Specifically, we use propositional logic for specifying dependencies and we show how this works for Java bytecode. Once we have a propositional formula that specifies all valid sub-inputs, we run an algorithm that finds a small, valid, failure-inducing input. Our algorithm interleaves runs of the buggy program and calls to a procedure that finds a minimal satisfying assignment. Our experiments show that we can reduce Java bytecode to 4.6 percent of its original size, which is 5.3 times better than the 24.3 percent achieved by J-Reduce. The much smaller output is more suitable for bug reports.