Description And Data Collection Of Key States In Running Linux Process Visualization

As a typical representative of the OpenSource operating system, with its free, open nature and advanced technology, Linux entered the field of educational research and has been playing as an outstanding and practical operating system platform in research and teaching purpose. More and more people realize the necessity to understand the Linux kernel, but source code's complexity and huge scale make understanding of its internal structure difficulty, especially new kernel with more functions is introduced continuously. But in recent years visualization technology develops fast and shows more obvious advantages in improving the human cognitive abilities. It goes deep into many vocations, such as geography, health care software development, education and so on, it combines with Internet information search, software visualization , data mining and shows more extensive research and application value. The main purpose of this paper is to explore the feasibility to combine visualization with the study on key executing processes of Linux kernel.Presently, a lot work have done to introduce visualization into operating system teaching field or combined with Linux kernel research, but there're some deficiencies. Some algorithm visualization tools or software have excellent visual effect but pay more attention to demonstrate algorithm itself, they're separated from actual running system, and so the data source is unreal. Meanwhile their main animation form is program or algorithm visualization so they are lack of interactive with user; Some visualization work use running operating system as data source, they pre-defined several process or activities by static binding, then simulate process visually, This binding limits its scope of application; Some visualizations aim at analyze or monitor system's performance, they pay more attention to the information about the overall state and the utilization rate of resources, and they don't care kernel data structures and functions which can't meet the need to thorough understand kernel. To improve the current work, this paper based on actual data collected from running kernel, mainly visualize data structures and functions of source code, aims at visualize various key operations of a relatively complete executing process during a certain period of time.This paper first gives out the workflow of key process visualization, focus on the contents following: 1.Set running Linux operating system as data sources, extract key states of kernel during fixed periods, define kinds of data collection points, classify them then specify their data recording format, in this way get one method to describe kernel states; 2. Analyze the basic principles should follow and fundamental issues should solve during data acquisition, discuss the usability of current kernel data collection methods under these principles, settle corresponding problems and give out a feasible method.Then, demonstrate how to insert data collection points in source code by tracing a specific process example. Analyze several data records collected from running kernel and deal with them visual to verify the usability of the description method defined before.Finally, from the results of visual pictures implemented by program, we can see combined with actual kernel data and mainly focus on program and data structure visualization, key process visualization can help to understand the function on the basis of source code and the visualization result is more real and acceptable, so it must have the significance for further research and development.