Abstract

Modular design is an important requirement in modern embedded system design flows because of the widespread acceptance of new paradigms such as IP core reuse and platform-based design. Cosimulation frameworks must thus support modular design, since programmable devices, ad-hoc HW components, and the interconnect infrastructure must be easily interchangeable in order to allow design exploration while keeping the SW portion unchanged or only marginally changed. The proposed co-simulation framework implements such a modular approach to co-simulation by means of a novel paradigm in which HW models can be modified on the fly by keeping the SW parts unchanged. This is achieved through an ISS-centric co-simulation strategy in which modularity is provided in terms of (i) the replacement of HW components thanks to the use of a common interface based on the device address space, or (ii) the use of different ISS's, thanks to a re-configurable simulator. We demonstrate our approach onto an industrial-strength embedded application, showing that the proposed co-simulation strategy provides both high speed and accuracy.

No abstract

The structure and the functionalities of a device driver are strongly influenced by the target platform architecture, as well as by the device communication protocol. This makes the generation of device drivers designed for complex embedded platforms a very time consuming and error prone activity. Validation becomes then a nodal point in the design flow. The aim of this paper is to present a co-simulation framework that allows validation of device drivers. The proposed framework supports all mechanisms used by device drivers to communicate with HW devices so that both modeled and actual components can be included in the simulated embedded platform. In this way, the generated code can be tested and validated even if the final platform is not ready yet. The framework has been applied to some examples to highlight the performance and effectiveness of this approach.

In the development of application software for networked embedded systems a significant step to improve interoperability and reuse consisted in relying on a service layer providing an abstract interface through a well-defined programming paradigm (e.g., object-oriented, tuplespace or database). During application development this layer can provide a simulation model of the actual platform while during implementation it can be mapped onto an actual middleware. During application development the choice of the programming paradigm is driven by productivity while during deployment the choice of the middleware is driven by efficiency and availability. Keeping the same programming paradigm in these two phases limits design-space exploration while changing paradigm requires to re-write the application. In this work we describe 1) a SystemC-based service layer providing different paradigms, i.e., objectoriented and tuple space, 2) design reasons for translating the application between these paradigms, and 3) a methodology for this translation.