Variants of embedded software are created, for example, to allow adaptation to different hardware or customer requirements. During testing, there is a risk that some variants will remain untested because conventional code coverage measurements may not be sufficient.
Code access analysis is used to identify untested variants. It compares the executable lines of the source file and the preprocessed code of all variants to identify untested lines. Code access analysis is a new feature of version 5.1 of the TESSY unit testing tool. TESSY's code access analysis determines the code access value as a percentage. A value of 100% means that there are no untested variants.
In safety-critical projects with many variants, it is important to prove that all variants have been tested and to have a clear hint to untested source code lines.
You can find the full article (written in German) in the e-paper of Elektronik 12/2024 on page 82: Detecting blind spots in code.