Process real-time scheduling method

Abstract

The embodiment of the invention discloses a process real-time scheduling method. The method comprises the following steps of: A, presetting a value of a process task, wherein the value of a task is selected from a priority value of a process and the maximal priority values of all current processes according to the value of the task per se; and B, when the priority levels of process tasks are close or equal to each other, comparing the values of the process tasks, and executing the task with a maximal value preferentially; or comparing the values of all current tasks when a system is overloaded, and executing the process task with the maximal value so as to ensure overall value maximization of the system and maintain the system at an acceptable level. In the process real-time scheduling method of the embodiment of the invention, by introducing a value-based preemptive threshold value, timely completion of a key task of the system can be ensured and the system performance is maintained and degraded gracefully without a system failure or even collapse when the system is overloaded, so as to achieve the maximization of the overall system finally; and resource waste caused by a trashing phenomenon can be reduced.

Description

technical field [0001] The invention relates to the field of computer process scheduling, in particular to a process real-time scheduling method. Background technique [0002] As the Linux system is widely used in embedded systems, real-time control and other fields, it is particularly important to enhance the real-time performance of the Linux kernel. The core of the Linux kernel is the scheduling strategy. The optimization of the scheduling strategy and the improvement of real-time performance are based on the scheduling algorithm. The study of real-time scheduling algorithms is an important research field in the real-time field. The priority-driven method is the most common method of real-time system scheduling. Each task is given a priority according to the specific information of the task. When the system needs to be scheduled, the next task to be run is selected according to the priority to ensure that the system is fair and effective. The response completes the task....

AI Technical Summary

Problems solved by technology

Although the EDF algorithm and the LSF algorithm show their optimality under normal system load, in the case of overload, the system cannot guarantee that all tasks can meet the deadline, and the performance of the EDF or LSF algorithm will drop sharply. , which can lead to instability due to missing important task deadlines

Furthermore, deadlines or idle time tasks are not necessarily the most critical. In the case of overload, the system cannot guarantee the timely completion of critical tasks, so system performance cannot be gracefully degraded, resulting in system failure or even crash.

In addition, in the LSF algorithm, when the system has more than one task whose minimum margin is similar or equal and has the minimum margin in the system, the system will undergo frequent task switching, that is, thrashing, which will cause waste of system resources.

Images