Analysis method of mpi code communication process based on text analysis

Abstract

The invention provides a text analysis-based MPI code communication process analysis method, which is applied to parallel application programs developed based on MPI. The text analysis-based MPI code communication process analysis method can simulate MPI tasks by analyzing the source code of the MPI parallel application. During the execution process, the peer-to-peer communication information can be quickly and accurately mined, reducing the resource overhead in the analysis process. It includes the following steps: (a) analyze the MPI source code line by line, and capture the key communication variables; (b) perform a corresponding number of cycle analysis on the source code according to the number of processes that the user expects to start, and simulate an MPI for each cycle analysis Execution of the task; (c) during each cycle, execute the corresponding analysis process according to the different keyword statements analyzed, and search whether the MPI communication keyword is included; (d) according to the different MPI communication keywords searched, the The communication process is converted into point-to-point communication information; (e) integrating point-to-point communication information in multiple processes.

Description

Technical field: [0001] The present invention relates to a code analysis method, more specifically, to a text analysis-based MPI code communication process analysis method. Background technique: [0002] In computer science, High Performance Computing (High Performance Computing, HPC) is an important branch, which is a technology to research and develop high-performance computers from many aspects, such as system structure, parallel algorithm and software development. High-performance computing has gradually become a new pillar of scientific research beyond traditional theoretical research and laboratory experiments. [0003] In the process of using high-performance computing, some applications cannot effectively utilize the computing resources of the high-performance computing platform they occupy, resulting in a waste of many computing resources. In order to utilize the high-performance computing platform more efficiently, it needs to be able to guide it to match the actu...

AI Technical Summary

Problems solved by technology

In the face of large-scale applications, this method will consume redundant computing and time resources due to the large number of processes, long computing time, and complicated communication processes.

Images