This research was conducted by Guoliang Jin, Linhai Song, Wei Zhang, Shan Lu, and Ben Liblit. The paper appeared in the ACM SIGPLAN 2011 Conference on Programming Language Design and Implementation (PLDI 2011).
This paper has been nominated as a SIGPLAN CACM Research Highlight. The nomination reads, in part,
This paper was well motivated (the example in the introduction was very nicely chosen to illustrate the pertinent issues), well written, and explained a set of intuitive techniques clearly. In addition, the results seemed very promising. The strategy presented by the paper is fairly simple, but surprisingly effective, at least for fixing bugs that fall under its failure model. While it’s easy to see many sources of potential inefficiency and incompleteness in the proposed algorithm, the empirical results make the case that the technique can fix real bugs with negligible overhead. This is one of the first papers to attack the problem of automated bug fixing, so it should be of wide interest.
Fixing software bugs has always been an important and time-consuming process in software development. Fixing concurrency bugs has become especially critical in the multicore era. However, fixing concurrency bugs is challenging, in part due to non-deterministic failures and tricky parallel reasoning. Beyond correctly fixing the original problem in the software, a good patch should also avoid introducing new bugs, degrading performance unnecessarily, or damaging software readability. Existing tools cannot automate the whole fixing process and provide good-quality patches.
We present AFix, a tool that automates the whole process of fixing one common type of concurrency bug: single-variable atomicity violations. AFix starts from the bug reports of existing bug-detection tools. It augments these with static analysis to construct a suitable patch for each bug report. It further tries to combine the patches of multiple bugs for better performance and code readability. Finally, AFix’s run-time component provides testing customized for each patch. Our evaluation shows that patches automatically generated by AFix correctly eliminate six out of eight real-world bugs and significantly decrease the failure probability in the other two cases. AFix patches never introduce new bugs and usually have similar performance to manually-designed patches.
The full paper is available as a single PDF document. A suggested BibTeX citation record is also available.