Research Statement

My research areas of interest are programming languages and software engineering, with a focus on program analysis and software testing. For my doctoral dissertation, I developed and applied static program analyses to make error handling in large software systems more reliable [3, 5, 7, 6, 8]. My analyses aim to help developers understand how run-time errors propagate through software and find error-propagation bugs that could lead to serious problems including system crashes, security vulnerabilities, silent data loss, and data corruption. In addition to finding hundreds of confirmed bugs in real-world systems such as the Linux kernel and Mozilla Firefox, my error-propagation analysis also found a critical bug in heavily tested code used for space missions by the NASA/JPL Laboratory. Most recently, I have been exploring a different application domain: numerical programs. Testing and de-bugging numerical programs pose unique challenges due to the use of floating-point arithmetic. In general, it is difficult to reason about floating-point programs, and developers often resort to using the highest available floating-point precision whenever possible. This can severely impact program performance. To alleviate this problem, I developed Precimonious, a dynamic analysis tool to assist developers in tuning the precision of their floating-point programs to improve performance [9]. I am currently designing and implementing a framework for developing dynamic program analyses to improve the reliability of numerical programs