Global semaphore implementation method based on multi-core multi-processor parallel system

Abstract

The invention discloses a global semaphore implementation method based on a multi-core multi-processor parallel system. The method comprises the following steps: each processing core performs system initialization and maps a shared storage space through SRIO; each processing core creates a thread, initializes the thread, and establishes a mapping table of the thread and local semaphore; a global semaphore is created from a global semaphore buffer pool corresponding to a processing core to which the thread belongs, and the global semaphore information is filled into a global semaphore name table; and when the thread is suspended on a waiting queue of a global semaphore or is unsuspended, the thread is suspended or unsuspended through a semaphore proxy mechanism. Through a semaphore agent mechanism, the global semaphore is realized by utilizing a software method, the method can be used for occasions of thread synchronization and mutual exclusion among multi-core multiprocessors, and theproblems of thread synchronization and mutual exclusion at the same moment on a multi-core multiprocessor parallel system are solved.

Description

technical field [0001] The invention relates to a method for realizing a global semaphore, in particular to a method for realizing a global semaphore based on a multi-core multiprocessor parallel system, and belongs to the technical field of embedded computers. Background technique [0002] With the development of computer systems and the increasing demand for real-time and parallelism in its application fields, multi-core multi-processor parallel systems have become an important trend in the development of embedded computer systems, and have received close attention from academia and industry. [0003] Generally speaking, in a multi-core multi-processor parallel system, joint work between processor cores and processors is required, and multiple threads running at the same time may need to access or use the same resource; or there is a dependency between threads Relationship, the operation of a thread depends on another thread. In the case of high software complexity and st...

AI Technical Summary

Problems solved by technology

In the case of high software complexity and strict requirements on running timing, these constraints may affect the normal operation of multi-core multi-processor parallel systems. Therefore, it is necessary to solve critical resource allocation, shared resource contention, and dependencies between threads. Issues such as thread synchronization and mutual exclusion in the running order of time

Images