When presented with the problem of scheduling a real-time task set many timing impediments may manifest themselves. This is especially evident when developing tasks within a prototyping environment. For example, the Computer Aided Prototyping System (CAPS) software [Luqi] development environment provides limited monitoring techniques for real-time task execution timing. These techniques are restricted to typical error messages which are present upon execution failure. This approach necessitates that the software system designer presumes a task set execution time.
This paper focuses upon the examination and feedback of fine grain execution timing data useful for task scheduling. This work was accomplished through the development of a program to perform true dynamic run time data collection. The collected data is representative of a typical task set execution exhibited within a real-time environment. The results of developing an accurate and efficient real-time task execution monitoring program which assists in overcoming the problem of task run time prediction are presented in this paper.
