Load equilibration scheduling method and device

Abstract

The present invention provides a load balances scheduling method and apparatus, wherein, the method comprises: step one, a load balance monitoring module of periodic duty is started up for monitoring the distribution of task priority of each processing unit in multiprocessor system, and when the distribution unbalance of the task priority of each processing unit is monitored and the maximum difference value of the task priority distribution exceeds the preset threshold value, the load balance monitoring module can determine a source processing unit, a target processing unit and the priority of the task scheduling to be performed, and can launch a load balance scheduling message to related load balance processing module; and step two, the relating load balance processing module can execute the task scheduling according to the task priority distribution of each processing unit and relations between processing units, thereby the task load on each processing unit in a symmetrical multiprocessor system can be reached to balance.

Description

technical field [0001] The invention relates to the field of task scheduling in computer operating systems, and more specifically, to a load balancing scheduling method and device. Background technique [0002] With the development of microprocessor technology, multi-core and hyper-threading have become the mainstream of development. At present, mainstream commercial processor systems such as X86, PPC, MIPS, and ARM have been launched, or are planning to launch multi-core and hyper-threading commercial processors. The multi-core structure of these processors is based on the symmetric multi-processor system, that is, each processing unit has no priority and is equal. Compared with ordinary single-core processors, processors with multi-core and hyper-threading features have multiple execution units and can run multiple tasks at the same time, which is completely different from the situation that a single-core processor can only run one task at a time. Different, in order to s...

AI Technical Summary

Problems solved by technology

For business applications running on embedded systems in these fields, the operating system is generally required to provide features such as strong real-time performance and high efficiency. For high-priority tasks, it must be run first; thus, the existing operating system The support method for multi-core and hyper-threaded processors cannot fully meet the needs of many embedded systems, and has the following disadvantages:

[0005] 1. Its load balancing method is not based on priority balancing. In the entire symmetric multiprocessor system, the priority of the running queue on each processor is unbalanced, and the priority of running tasks on each processor in the system cannot be guaranteed. They are all higher in priority than waiting tasks, and there will be high-priority tasks waiting on one processor, while tasks with lower priority than it are running on other processors;

[0006] 2. The priority inheritance of the mutex semaphore cannot be effectively supported. Under the support mode of the existing operating system, since the priority consistency in the entire system space cannot be guaranteed, the priority inheritance of the mutex semaphore can only be performed in implemented within a processor and cannot be implemented within an entire symmetric multiprocessor system; and

[0007] 3. The timing of its balanced operation is not based on priority, but is triggered according to the operating efficiency of each processor, which cannot meet the requirement that high-priority tasks must be executed in a timely manner in the system

Images