SARATHI: Characterization Study on Regression Bugs and Identification of Regression Bug Inducing Changes: A Case-Study on Google Chromium Project

Abstract

As a software system evolves, maintaining the system becomes increasingly difficult. A lot of times code changes or system patches cause an existing feature to misbehave or fail completely. An issue ticket reporting a defect in a feature that was working earlier, is known as a Regression Bug. Running a test suite to validate the new features getting added and faults introduced in previously working code, after every change is impractical. As a result, by the time an issue is identified and reported a lot of changes are made to the source code, which makes it very difficult for the developers to find the regression bug inducing change. Regression bugs are considered to be inevitable and truism in large and complex software systems [1]. Issue Tracking System (ITS) are applications to track and manage issue reports and to archive bug or feature enhancement requests. Version Control System (VCS) are source code control systems recording the author, timestamp, commit message and modified files. We first conduct an in-depth characterization study of regression bugs by mining issue tracking system dataset belonging to a large and complex software system i.e. Google Chromium Project. We then describe our solution approach to find the regression bug inducing change, based on mining ITS and VCS data. We build a recommendation engine Sarathi to assist a bug fixer in locating a regression bug inducing change and validate the system on real world Google Chromium project.