An Empirical Study of Configuration Mismatches in Linux

This website provides additional information and links for the following publication:

Sascha El-Sharkawy, Adam Krafczyk and Klaus Schmid, An Empirical Study of Configuration Mismatches in Linux. In Proceedings of the 21st International Systems and Software Product Line Conference (SPLC 2017). Accepted.

A public repository is available, which hosts the tools described in this publication to elicit the required data and evaluate the results. This repository contains:

• The described ConfigurationMismatchDetector
• The SpaceLocator for identifying Kconfig variables in code and build files
• The results presented in the publication

Example results:

• 15.5521 Kconfig variables available in total
• 1.065 [6.86%] used in Kconfig files
• 3.507 [22.6%] used in code files
• 7.962 [51%] used in Kbuild files
• 2.987 [19.24%] used in both Kbuild and code files

A virtual machine for reproducing the data and results of the publication is available:

• Download the virtual machine
• User name and password: suh (both)
• During the log in to the virtual machine a readme-file will be shown, which further describes this machine and its usage
• A log in to the virtual machine via SSH is possible and recommended to increase usability

A technical report, which preceded this publication, is also available here.