Has the bug really been fixed?

• DOI: 10.1145/1806799.1806812
• Publication Date: 2010-06-08T12:37:34Z
• Authors: Zhongxian Gu; Earl T. Barr; David J. Hamilton; Zhendong Su
• Keywords: AspectJ; Software bug; Software regression

Software has bugs, and fixing those bugs pervades the software engineering process. It is folklore that bug fixes are often buggy themselves, resulting in bad fixes, either failing to fix a or creating new bugs. To confirm this folklore, we explored databases of Ant, AspectJ, Rhino projects, found comprise as much 9% all Thus, detecting correcting important for improving quality reliability software. However, no prior work systematically considered problem, which paper introduces formalizes. In particular, formalizes two criteria determine whether resolves bug: coverage disruption. The measures extent correctly handles inputs may trigger bug, while disruption deviations from program's intended behavior after application fix. This also novel notion distance-bounded weakest precondition basis developed practical techniques compute