Linux Kernel Architecture

Chapter 2: Process Management and Scheduling

Implementation

The implementation of namespaces requires two components: per-subsystem namespace structures that

wrap all formerly global components on a per-namespace basis, and a mechanism that associates a given

process with the individual namespaces to which it belongs. Figure 2-4 illustrates the situation.

struct task_struct

struct nsproxy

struct nsproxy

uts_namespc

user_namespc

mnt_namespc

uts_namespc

user_namespc

mnt_namespc

UTS Namespace

User Namespace

0

0

1

Tasks

struct task_struct

struct task_struct

Figure 2-4: Connection between processes and namespaces.

Formerly global properties of subsystems are wrapped up in namespaces, and each process is associated

with a particular selection of namespaces. Each kernel subsystem that is aware of namespaces must

provide a data structure that collects all objects that must be available on a per-namespace basis.struct

nsproxyis used to collect pointers to the subsystem-specific namespace wrappers:

<nsproxy.h>

struct nsproxy {

atomic_t count;

struct uts_namespace *uts_ns;

struct ipc_namespace *ipc_ns;

struct mnt_namespace *mnt_ns;

struct pid_namespace *pid_ns;

struct user_namespace *user_ns;

struct net *net_ns;

};

Currently the following areas of the kernel are aware of namespaces:

❑ The UTS namespace contains the name of the running kernel, and its version, the underlying

architecture type, and so on. UTS is a shorthand forUnixTimesharing System.

❑ All information related to inter-process communication (IPC) is stored instruct

ipc_namespace.