Abstract
As a result of the industrial deployment of real-time systems, there is an increasing demand for methods to perform safe and tight calculation of the worst case execution time (WCET) of programs. The WCET is a necessary prerequisite to be able to guarantee correct timing behaviour of the real-time system. WCET calculation means to find the path, often among a huge number of paths, that takes the longest time to execute. The calculation requires path information for the program, such as the maximum number of iterations in loops and identification of paths that are never executed. In most existing WCET analysis methods, this information is given as manual annotations by the programmer.

In this thesis we present a method which automatically calculates path information for object-oriented real-time programs by static analysis of the program. Thus, the method can be used in automating the WCET analysis, by relieving the programmer from the tedious and error-prone manual annotation work.

The method, which is based on abstract interpretation, generates safe but not necessarily exact path information. A trade-off between quality and calculation cost has to be made, since finding the exact information is a very complex problem. We propose time budgets to guarantee termination of the analysis for all programs.

We show how the general abstract interpretation theory can be used, in a structured way, to approximate the semantics of an imperative or object-oriented programming language.

We have chosen to analyze RealTimeTalk (RTT), an object-oriented language based on Smalltalk. We have developed a prototype tool which implements our analysis for a subset of the language. We show that the tool is capable of analyzing programs with a complexity which would make manual annotation of the program all but trivial.

We also discuss the role of the method in a WCET analysis tool framework.

Disertation: May 8, 2000