Towards Stateflow Model-Aware Debugging using Model-to-Source Tags with LLDB

Bewoayia Kebianyor and Philipp Ittershagen and Kim Grüttner
2nd International Workshop on Embedded Software for Industrial IoT (ESIIT) at DATE'19
The development of Internet-of-Things (IoT) devices are becoming more and more complex as the demand for compact devices with shrinking size of processors and microchips increase. At the same time, there is constant need to develop high quality products with a short time-to-market. Thus many companies employ model-based software development to cope with this challenge. The increase in complexity of these devices means an increase in the complexity of the software running on them, and an increase in unwanted behavior. The need for detection of errors (debugging) at all levels of the software development becomes a very important part in the software development process. We argue that there exist several solutions of debugging a software at the model level, but these may not be appropriate as some run on simulated environment or depend on proprietary trace infrastructures running on the target system.In this paper, we present a target platform independent model-based debugging approach that is neither based on simulation nor code instrumentation and makes use of existing open-source software technologies. This approach can be applied for any software model, from which production C/C++ code is generated, irrespective of the used modeling language or methodology. We further investigate our concept of customizing LLDB for debugging C/C++ code that has been generated from a Stateflow model.
03 / 2019
Cost-Efficient Smart System Software Synthesis