WCET Analysis Of Embedded System And Technology Of Cache Locking

The worst-case execution Time(Worst Case Execution Time,WCET)analysis of embedded real-time system tasks refers to the worst-case execution Time valuation of the task or program fragment before its execution.Before the program running,obtaining the WCET valuation of tasks is the premise of real-time system scheduling and schedulability analysis,also the requirements to test the performance of the real-time system.So the WCET estimation is the top priority procedure of embedded real-time system reliability analysis.Currently,the studies of WCET analysis mainly include two areas:the static and dynamic testing.Usually,the industry field often uses dynamic testing methods,but academic research field mainly uses static testing methods to analyze the WCET.Using cache in embedded system can reduce the speed difference between CPU and memory,but it has brought many problems and challenges for time estimates of tasks.So cache analysis for embedded system to analyze WECT is also a hot research.In this paper,our study includes WCET analysis method,the combination of high and low level analysis and cache analysis method for WECT valuation,specific work is as follows:This paper proposes a WCET analysis framework based on MPA algorithm of high level and Static Prediction of low level.Firstly,this framework uses MPA algorithm to analyze the program flow and obtain the Min-Tree of each code block.Then,it gets more strict constraints through the analysis of inner loop variables of function by symbolic loop bounds Computation and gets a WCET expression through the constraints of Min-Tree and the loop bounds.Finally,it uses a method of Static Prediction to solve the WCET by absolute valuation of underlying instruction cycle of each basic block and calculates the final WCET value.This paper shows the analysis process of the new framework by analyzing a small program;also it tests some benchmark programs based on five different scenario as a experiment to analyze the efficiency and accuracy of the new framework.This paper presents an optimization method of dynamic cache locking for WCET estimation.It introduces the cache locking method based on LOCK-MS,this method selects a set of instructs of each task to be cached by using the ILP model.As the ILP model is highly depend on explicit path in the tasks,if the program path is very complex,the ILP model will be increased sharply.So this paper optimizes the ILP model to avoid the path-explicit constraints,and with a top-level analysis method which is nothing to do with the specific path we get the instruction fetch numbers to replace the assumption of ILP model,which can reduce the complexity of analysis and state space of ILP model.With the compact ILP model,two sets of multitasking system tasks sets(Small and Medium)are used to do experiment and analyze of the efficiency and precision of the compact model.