Download now Free registration required
Statistical debugging uses lightweight instrumentation and statistical models to identify program behaviors that are strongly predictive of failure. However, most software is mostly correct; nearly all monitored behaviors are poor predictors of failure. The authors propose an adaptive monitoring strategy that mitigates the overhead associated with monitoring poor failure predictors. They begin by monitoring a small portion of the program, then automatically refine instrumentation over time to zero in on bugs. The authors formulate this approach as a search on the control-dependence graph of the program. They present and evaluate various heuristics that can be used for this search.
- Format: PDF
- Size: 303.5 KB