两种基于符号的最坏执行时间分析方法的比较与分析

通常的最坏执行时间分析方法只是得到一个常数,然后将这个常数应用在所有的地方,这将造成所得的值过于悲观（over pessimistic）.如果能以一个代数表达式代替,其中影响程序执行时间的一些未知值（如调用参数,运行平台参数等）都用参数来表示,然后在具体的环境中再根据具体的情况进行实例化,那么所获得的最坏执行时间将更加接近于实际的值.文章讨论了基于符号的最坏执行时间分析方法,分析了其特点,对两种基于符号的最坏执行时间分析方法进行了比较和研究,并给出了将来的研究方向.

Comparison and Analysis of two Symbolic Worst-Case Execution Time Analysis Methods

The result of a traditional real-time WCET(worst-case execution time) analysis method is usually a constant, and is applied everywhere. This causes over-pessimism. If we can replace the constant with an algebraic expression parameterized with some of the subprogram parameters or (and) operation configuration parameters, and later evaluated it using specific parameters in specific environment. This will result in a tighter value. This paper discuses symbolic worst case execution time analysis methods, analyzes the characteristics of symbolic worst case execution time analysis methods, compares and studies two existing symbolic worst case execution time analysis methods, gives future research directions.