Early Detection of Runtime Errors and Resource Problems by Combining Model-Based Design and Static Analysis
September 24, 2014
In this webinar recording, dSPACE and AbsInt will demonstrate how to combine model-based design and automatic production code generation with static analyzers to compute worst-case execution time/stack usage and to detect run-time errors.
Instructors: Carsten Rustemeier, Product Engineer TargetLink, dSPACE GmbH; Dr. Daniel Kästner, Chief Technology Officer, AbsInt GmbH
Target Group: Function and software developers, software engineers, systems engineers, test engineers, integration engineers, quality engineers, safety managers.
This webinar is part of a series called The TargetLink Ecosystem, in which we and our TargetLink partners present tools and an integrated workflow covering TargetLink and its environment for the model-based development of safety-related software.
This webinar, which we host jointly with our TargetLink Partner AbsInt Angewandte Informatik GmbH, focuses on efficient and integrated verification of non-functional requirements for production code and object code. The static analyzers from AbsInt are used to compute worst-case execution time/stack usage and to detect run-time errors and can be invoked automatically from TargetLink models. This leads to a small analysis configuration effort and to more precise analysis results.
After a brief introduction, we will take a closer look at the following topics:
- Global and efficient data management with the TargetLink Data Dictionary
- ISO-26262 requirements for timing, storage use, and runtime errors
- Formal verification by abstract interpretation
- Verification of non-functional requirements directly from modeling level
- Analysis of worst-case execution time (aiT)
- Analysis of worst-case stack usage (StackAnalyzer)
- Proving the absence of run-time errors (Astrée)
- Tool qualification
We will use a comprehensive example to demonstrate all these different topics.