The Linux kernel as a case study in software evolution

Abstract

We use 810 versions of the Linux kernel, released over a period of 14 years, to characterize the system's evolution, using Lehman's laws of software evolution as a basis. We investigate different possible interpretations of these laws, as reflected by different metrics that can be used to quantify them. For example, system growth has traditionally been quantified using lines of code or number of functions, but functional growth of an operating system like Linux can also be quantified using the number of system calls. In addition we use the availability of the source code to track metrics, such as McCabe's cyclomatic complexity, that have not been tracked across so many versions previously. We find that the data supports several of Lehman's laws, mainly those concerned with growth and with the stability of the process. We also make some novel observations, e.g. that the average complexity of functions is decreasing with time, but this is mainly due to the addition of many small functions.

High-MCC Functions in the Linux Kernel

“…In the process of studying the evolution of the Linux kernel, and in particular how various code metrics change with time, we found that some Linux kernel functions have MCC values in the hundreds [10]. Specifically, we found that the distribution of MCC values has a heavy tail, the absolute number of high-MCC functions is growing, but their fraction out of all functions in Linux is shrinking.…”

“…Nevertheless, we have found functions with MCC values in the hundreds in the Linux kernel [10]. This chance discovery immediately led to a set of research questions:…”

“…Linux is an evolving system [10]. It has shown phenomenal growth during the 17 years till the time the kernel we studied was released in 2011: version 1.0 had 122,442 lines of actual code, and version 2.6.37.5 had 9,185,179 lines, an average annual growth rate of 29%.…”

“…At an earlier time, in version 2.3.46, it had reached an MCC value of 563, but then in version 2.3.47 this dropped to 8. The reason was a design change, where a large switch was replaced by a table lookup [10]. A similar change occurred in function usb stor show sense, where a large switch statement was replaced by a call to a new function implementing a lookup table.…”

Evolution of the Linux Kernel Variability Model

“…We have found that 3578 patches that modify Kconfig files are driver related. Similarly, Israeli [12] collected several software metrics of the Linux kernel source code from 1994 and 2008, and also observed the functional growth by counting features. Adams studies the evolution of the Linux kernel build system [13] and finds that considerable maintenance to the system is performed to reduce the build complexity, that grows, partly due to the increase in number of features.…”