ABSTRACT
The schedulability analysis techniques presented are based on the hypothesis that task instances never block for any reason, unless they have been executed completely. It analyzes the effects of task interaction on their worst-case response time and schedulability. The self-suspension of a task has an impact on how it interacts with other tasks, and on the properties of the interaction, too. The chapter shows that the self-suspension of a task has not only a local effect on the response time of that task—this is quite obvious—but it may also give rise to an extra interference affecting lower-priority tasks, and further increase the blocking times. It looks at all these effects, calculate their worst-case contribution to the blocking time of each task, and further extend the RTA method to take them into account.
