1 .\" Copyright (C) 2011, Eric Biederman <ebiederm@xmission.com>
2 .\" and Copyright (C) 2011, 2012, Michael Kerrisk <mtk.manpages@gmail.com>
4 .\" %%%LICENSE_START(GPLv2_ONELINE)
5 .\" Licensed under the GPLv2
8 .TH SETNS 2 2016-03-15 "Linux" "Linux Programmer's Manual"
10 setns \- reassociate thread with a namespace
13 .BR "#define _GNU_SOURCE" " /* See feature_test_macros(7) */"
16 .BI "int setns(int " fd ", int " nstype );
19 Given a file descriptor referring to a namespace,
20 reassociate the calling thread with that namespace.
24 argument is a file descriptor referring to one of the namespace entries in a
28 for further information on
30 The calling thread will be reassociated with the corresponding namespace,
31 subject to any constraints imposed by the
37 argument specifies which type of namespace
38 the calling thread may be reassociated with.
39 This argument can have one of the following values:
42 Allow any type of namespace to be joined.
44 .BR CLONE_NEWCGROUP " (since Linux 4.6)"
46 must refer to a cgroup namespace.
48 .BR CLONE_NEWIPC " (since Linux 3.0)"
50 must refer to an IPC namespace.
52 .BR CLONE_NEWNET " (since Linux 3.0)"
54 must refer to a network namespace.
56 .BR CLONE_NEWNS " (since Linux 3.8)"
58 must refer to a mount namespace.
60 .BR CLONE_NEWPID " (since Linux 3.8)"
62 must refer to a descendant PID namespace.
64 .BR CLONE_NEWUSER " (since Linux 3.8)"
66 must refer to a user namespace.
68 .BR CLONE_NEWUTS " (since Linux 3.0)"
70 must refer to a UTS namespace.
74 as 0 suffices if the caller knows (or does not care)
75 what type of namespace is referred to by
77 Specifying a nonzero value for
79 is useful if the caller does not know what type of namespace is referred to by
81 and wants to ensure that the namespace is of a particular type.
82 (The caller might not know the type of the namespace referred to by
84 if the file descriptor was opened by another process and, for example,
85 passed to the caller via a UNIX domain socket.)
88 behaves somewhat differently from the other
91 reassociating the calling thread with a PID namespace changes only
92 the PID namespace that child processes of the caller will be created in;
93 it does not change the PID namespace of the caller itself.
94 Reassociating with a PID namespace is allowed only if the
95 PID namespace specified by
97 is a descendant (child, grandchild, etc.)
98 of the PID namespace of the caller.
99 For further details on PID namespaces, see
100 .BR pid_namespaces (7).
102 A process reassociating itself with a user namespace must have the
104 .\" See kernel/user_namespace.c:userns_install() [3.8 source]
105 capability in the target user namespace.
106 Upon successfully joining a user namespace,
107 a process is granted all capabilities in that namespace,
108 regardless of its user and group IDs.
109 A multithreaded process may not change user namespace with
111 It is not permitted to use
113 to reenter the caller's current user namespace.
114 This prevents a caller that has dropped capabilities from regaining
115 those capabilities via a call to
117 For security reasons,
118 .\" commit e66eded8309ebf679d3d3c1f5820d1f2ca332c71
119 .\" https://lwn.net/Articles/543273/
120 a process can't join a new user namespace if it is sharing
121 filesystem-related attributes
122 (the attributes whose sharing is controlled by the
125 flag) with another process.
126 For further details on user namespaces, see
127 .BR user_namespaces (7).
129 A process may not be reassociated with a new mount namespace if it is
131 .\" Above check is in fs/namespace.c:mntns_install() [3.8 source]
132 Changing the mount namespace requires that the caller possess both
136 capabilities in its own user namespace and
138 in the target mount namespace.
140 .BR user_namespaces (7)
141 for details on the interaction of user namespaces and mount namespaces.
145 to change the caller's cgroup namespace does not change
146 the caller's cgroup memberships.
151 On failure, \-1 is returned and
153 is set to indicate the error.
158 is not a valid file descriptor.
162 refers to a namespace whose type does not match that specified in
166 There is problem with reassociating
167 the thread with the specified namespace.
169 .\" See kernel/pid_namespace.c::pidns_install() [kernel 3.18 sources]
171 The caller tried to join an ancestor (parent, grandparent, and so on)
175 The caller attempted to join the user namespace
176 in which it is already a member.
179 .\" commit e66eded8309ebf679d3d3c1f5820d1f2ca332c71
180 The caller shares filesystem
182 state (in particular, the root directory)
183 with other processes and tried to join a new user namespace.
186 .\" See kernel/user_namespace.c::userns_install() [kernel 3.15 sources]
187 The caller is multithreaded and tried to join a new user namespace.
190 Cannot allocate sufficient memory to change the specified namespace.
193 The calling thread did not have the required capability
198 system call first appeared in Linux in kernel 3.0;
199 library support was added to glibc in version 2.14.
203 system call is Linux-specific.
205 Not all of the attributes that can be shared when
206 a new thread is created using
211 The program below takes two or more arguments.
212 The first argument specifies the pathname of a namespace file in an existing
215 The remaining arguments specify a command and its arguments.
216 The program opens the namespace file, joins that namespace using
218 and executes the specified command inside that namespace.
220 The following shell session demonstrates the use of this program
221 (compiled as a binary named
223 in conjunction with the
225 example program in the
227 man page (complied as a binary named
230 We begin by executing the example program in
233 That program creates a child in a separate UTS namespace.
234 The child changes the hostname in its namespace,
235 and then both processes display the hostnames in their UTS namespaces,
236 so that we can see that they are different.
240 $ \fBsu\fP # Need privilege for namespace operations
242 # \fB./newuts bizarro &\fP
244 clone() returned 3550
245 uts.nodename in child: bizarro
246 uts.nodename in parent: antero
247 # \fBuname \-n\fP # Verify hostname in the shell
252 We then run the program shown below,
253 using it to execute a shell.
254 Inside that shell, we verify that the hostname is the one
255 set by the child created by the first program:
259 # \fB./ns_exec /proc/3550/ns/uts /bin/bash\fP
260 # \fBuname \-n\fP # Executed in shell started by ns_exec
273 #define errExit(msg) do { perror(msg); exit(EXIT_FAILURE); \\
277 main(int argc, char *argv[])
282 fprintf(stderr, "%s /proc/PID/ns/FILE cmd args...\\n", argv[0]);
286 fd = open(argv[1], O_RDONLY); /* Get file descriptor for namespace */
290 if (setns(fd, 0) == \-1) /* Join that namespace */
293 execvp(argv[2], &argv[2]); /* Execute a command in namespace */