System and method for debugging parallel program

Abstract

The invention discloses a system for debugging a parallel program, which is used in the field of information technology, and provides a system and a method for debugging the parallel program aiming at the defects in debugging the parallel program in the prior art, which can implement the cross-platform application. The system particularly comprises a front-end integrated debugging module and a background debugging management module, wherein the front-end integrated debugging module is used for providing a debugging window for a user and carrying out debugging operation on the parallel program to be debugged; and the background debugging management module is used for carrying out instrumentation by a source code according to a debugging command sent by the front-end integrated debugging module, carrying out compiling and running on the parallel program subjected to instrumentation and feeding back a running result to the front-end integrated debugging module. The method and the system for debugging the parallel program, which are disclosed by the embodiment of the invention, are convenient for the user to edit a debugging code and setting operation breakpoints of the parallel program, can be used for carrying out compiling, running and debugging on the parallel program by a remote calling mode and viewing a performance analysis result, and has good portability and expandability.

Description

technical field [0001] The invention relates to the field of information technology, in particular to a parallel program debugging system and method. Background technique [0002] At first it was thought that the increase in computer performance was mainly due to the production of faster and more efficient processors. This view is challenged by the concept of parallel processing: the fundamental meaning of parallel computing is to connect two or more computers to jointly solve some complex computing problems. For example, with the development of science and information technology, many fields such as weather forecasting, earthquake analysis, information retrieval, etc. need to process large-scale and massive data quickly and efficiently, and promote the transition from serial to parallel computing. Parallel computing is becoming more and more widely used in military, science and technology, business and other fields, and more users are beginning to use parallel technology t...

AI Technical Summary

Problems solved by technology

However, it does not use the ready-made serial debugger, and it needs to use different versions under different platforms, and does not support heterogeneous environments. Since the user interface and underlying control are all running on the Unix / Linux platform, users who are not familiar with the Unix / Linux platform can use will feel relatively difficult

Images