Porting Of Mobile IPv6Open Source From User Space To Kernel

With the rapid development of mobile Internet, mobile IPv6will be widely used in the future. But, currently, mobile IPv6protocol implementations mostly run in the user space of Linux operating system. Not only the efficiency of running is lower, but also they cannot run in the Android or other Linux kernel based operating systems. To resolve these two problems, this paper ports the mobile IPv6implementation from user space to kernel. So that most Linux kernel based operating systems can use the implementation more effective and directly, and avoid duplication of development.According to the general steps of software porting, this paper firstly analyses the feasibility of porting considering of the feature of Linux network protocol stack and the differences between Linux user space and Linux kernel. The distinction of user space programs and kernel programs includes several aspects, such as boot mode, system interface, support functions and some others. And In the kernel there are some similar mechanisms or functions to replace the ones in the user space. And so we can know, that the porting is feasible.Secondly, this paper analyses the mobile IPv6open source(UMIP)’s software architecture and implementation. Then divide the code which needs to be ported into four types. Different types use different porting strategies.Then, according to the porting strategies, this paper describes the specific methods and processes from four aspects, namely boot mode, GLIB library, Linux system interface and some others. Porting of GLIB library includes the porting of file manipulation functions, multi-threads functions, mutex, etc; porting of Linux system interface includes the porting of sysctl file system, ioctl interface, Netlink socketand some other interfaces; the porting of rest styles includes the porting of reading ways of configuration file, etc.At the completion of the porting, this paper introduces how to add the code which has been ported into the Linux kernel source. Then compileand debug the code. Because of the debugging workload is much bigger, the debugging work is not over yet. The next step is to complete the debugging and testing, and then release the new implementation.