No abstract

Automated diagnosis of errors detected during software testing can improve the efficiency of the debugging process, and can thus help to make software more reliable. In this paper we discuss the application of a specific automated debugging technique, namely software fault localization through the analysis of program spectra, in the area of embedded software in high-volume consumer electronics products. We discuss why the technique is particularly well suited for this application domain, and through experiments on an industrial test case we demonstrate that it can lead to highly accurate diagnoses of realistic errors.

No abstract

All aspect orientation languages provide a onesize-fits-all methodology for reflection on join points. However, the amount of resources necessary for this approach is too high to be applicable in the context of consumer products. In this industrial research paper, we describe a solution to this problem and prove via an experiment that it is suitable for our context. In particular, we advocate that in the context of consumer products the reflective information should be passed explicitly using dedicated reflection parameters. Furthermore, since reflective information should efficiently encode the relevant domain knowledge, the user must be in control of the type of the dedicated reflection parameters. We describe how we implemented user-controlled reflection on join points in our aspect-oriented framework AspectKoala on top of the component model Koala [13]. We compare the resource consumption of different approaches to add reflective information on join points using this implementation. The difference in resource consumption clearly demonstrates the benefits of our solution for consumer products.