[Back]

P. Puschner, R. Kirner:
"Timing Analysis for Embedded Systems and Time-Predictable Computing";
Poster: Siemens PSE Technology Day, Vienna, Austria; 2005-11-25.

This presentation consists of two parts: In the first part we introduce a new measurement-based worst-case execution time (WCET) analysis method; in the second part we present a strategy for generating code that is highly temporally predictable, which simplifies timing analysis.
Exhaustive end-to-end measurements are in most cases computationally intractable for determining the WCET. Therefore, we propose to measure execution times of subparts of the application and combine those partial results into a final result. We use heuristic methods and model checking to generate test data, forcing the execution of selected paths to perform runtime measurements. The measured times are used to calculate the WCET in a final computation step. As we operate on source code level our approach is platform independent except for the run-time measurements performed on the target host. We show the feasibility of the required steps and explain our approach by means of a case study.
The second part presents a strategy for generating embedded real-time systems code that is highly temporally predictable, which, in turn, makes timing analysis almost trivial. The core of this strategy is the so-called single-path software/hardware architecture. Single-path programs run on processors that support a conditional move instruction (and possibly other predicated instructions) with a constant execution time, and, as the name says, these programs have only a single execution path. We show how single-path code can be generated out of any code whose worst-case execution time (WCET) can be bounded. Thus, converting WCET-analyzable code into single-path code provides a general solution to simplify WCET analysis and to produce time-predictable code. The talk also introduces the WCET-oriented programming strategy that, if used together with the single-path approach, yields not only deterministic but also very competitive task execution times.