man7/pid_namespaces.7

   1 .\" Copyright (c) 2013 by Michael Kerrisk <mtk.manpages@gmail.com>
   2 .\" and Copyright (c) 2012 by Eric W. Biederman <ebiederm@xmission.com>
   3 .\"
   4 .\" %%%LICENSE_START(VERBATIM)
   5 .\" Permission is granted to make and distribute verbatim copies of this
   6 .\" manual provided the copyright notice and this permission notice are
   7 .\" preserved on all copies.
   8 .\"
   9 .\" Permission is granted to copy and distribute modified versions of this
  10 .\" manual under the conditions for verbatim copying, provided that the
  11 .\" entire resulting derived work is distributed under the terms of a
  12 .\" permission notice identical to this one.
  13 .\"
  14 .\" Since the Linux kernel and libraries are constantly changing, this
  15 .\" manual page may be incorrect or out-of-date.  The author(s) assume no
  16 .\" responsibility for errors or omissions, or for damages resulting from
  17 .\" the use of the information contained herein.  The author(s) may not
  18 .\" have taken the same level of care in the production of this manual,
  19 .\" which is licensed free of charge, as they might when working
  20 .\" professionally.
  21 .\"
  22 .\" Formatted or processed versions of this manual, if unaccompanied by
  23 .\" the source, must acknowledge the copyright and authors of this work.
  24 .\" %%%LICENSE_END
  25 .\"
  26 .\"
  27 .TH PID_NAMESPACES 7 2017-05-03 "Linux" "Linux Programmer's Manual"
  28 .SH NAME
  29 pid_namespaces \- overview of Linux PID namespaces
  30 .SH DESCRIPTION
  31 For an overview of namespaces, see
  32 .BR namespaces (7).
  33
  34 PID namespaces isolate the process ID number space,
  35 meaning that processes in different PID namespaces can have the same PID.
  36 PID namespaces allow containers to provide functionality
  37 such as suspending/resuming the set of processes in the container and
  38 migrating the container to a new host
  39 while the processes inside the container maintain the same PIDs.
  40
  41 PIDs in a new PID namespace start at 1,
  42 somewhat like a standalone system, and calls to
  43 .BR fork (2),
  44 .BR vfork (2),
  45 or
  46 .BR clone (2)
  47 will produce processes with PIDs that are unique within the namespace.
  48
  49 Use of PID namespaces requires a kernel that is configured with the
  50 .B CONFIG_PID_NS
  51 option.
  52 .\"
  53 .\" ============================================================
  54 .\"
  55 .SS The namespace "init" process
  56 The first process created in a new namespace
  57 (i.e., the process created using
  58 .BR clone (2)
  59 with the
  60 .BR CLONE_NEWPID
  61 flag, or the first child created by a process after a call to
  62 .BR unshare (2)
  63 using the
  64 .BR CLONE_NEWPID
  65 flag) has the PID 1, and is the "init" process for the namespace (see
  66 .BR init (1)).
  67 A child process that is orphaned within the namespace will be reparented
  68 to this process rather than
  69 .BR init (1)
  70 (unless one of the ancestors of the child
  71 in the same PID namespace employed the
  72 .BR prctl (2)
  73 .B PR_SET_CHILD_SUBREAPER
  74 command to mark itself as the reaper of orphaned descendant processes).
  75
  76 If the "init" process of a PID namespace terminates,
  77 the kernel terminates all of the processes in the namespace via a
  78 .BR SIGKILL
  79 signal.
  80 This behavior reflects the fact that the "init" process
  81 is essential for the correct operation of a PID namespace.
  82 In this case, a subsequent
  83 .BR fork (2)
  84 into this PID namespace will fail with the error
  85 .BR ENOMEM ;
  86 it is not possible to create a new processes in a PID namespace whose "init"
  87 process has terminated.
  88 Such scenarios can occur when, for example,
  89 a process uses an open file descriptor for a
  90 .I /proc/[pid]/ns/pid
  91 file corresponding to a process that was in a namespace to
  92 .BR setns (2)
  93 into that namespace after the "init" process has terminated.
  94 Another possible scenario can occur after a call to
  95 .BR unshare (2):
  96 if the first child subsequently created by a
  97 .BR fork (2)
  98 terminates, then subsequent calls to
  99 .BR fork (2)
 100 will fail with
 101 .BR ENOMEM .
 102
 103 Only signals for which the "init" process has established a signal handler
 104 can be sent to the "init" process by other members of the PID namespace.
 105 This restriction applies even to privileged processes,
 106 and prevents other members of the PID namespace from
 107 accidentally killing the "init" process.
 108
 109 Likewise, a process in an ancestor namespace
 110 can\(emsubject to the usual permission checks described in
 111 .BR kill (2)\(emsend
 112 signals to the "init" process of a child PID namespace only
 113 if the "init" process has established a handler for that signal.
 114 (Within the handler, the
 115 .I siginfo_t
 116 .I si_pid
 117 field described in
 118 .BR sigaction (2)
 119 will be zero.)
 120 .B SIGKILL
 121 or
 122 .B SIGSTOP
 123 are treated exceptionally:
 124 these signals are forcibly delivered when sent from an ancestor PID namespace.
 125 Neither of these signals can be caught by the "init" process,
 126 and so will result in the usual actions associated with those signals
 127 (respectively, terminating and stopping the process).
 128
 129 Starting with Linux 3.4, the
 130 .BR reboot (2)
 131 system call causes a signal to be sent to the namespace "init" process.
 132 See
 133 .BR reboot (2)
 134 for more details.
 135 .\"
 136 .\" ============================================================
 137 .\"
 138 .SS Nesting PID namespaces
 139 PID namespaces can be nested:
 140 each PID namespace has a parent,
 141 except for the initial ("root") PID namespace.
 142 The parent of a PID namespace is the PID namespace of the process that
 143 created the namespace using
 144 .BR clone (2)
 145 or
 146 .BR unshare (2).
 147 PID namespaces thus form a tree,
 148 with all namespaces ultimately tracing their ancestry to the root namespace.
 149 Since Linux 3.7,
 150 .\" commit f2302505775fd13ba93f034206f1e2a587017929
 151 .\" The kernel constant MAX_PID_NS_LEVEL
 152 the kernel limits the maximum nesting depth for PID namespaces to 32.
 153
 154 A process is visible to other processes in its PID namespace,
 155 and to the processes in each direct ancestor PID namespace
 156 going back to the root PID namespace.
 157 In this context, "visible" means that one process
 158 can be the target of operations by another process using
 159 system calls that specify a process ID.
 160 Conversely, the processes in a child PID namespace can't see
 161 processes in the parent and further removed ancestor namespaces.
 162 More succinctly: a process can see (e.g., send signals with
 163 .BR kill (2),
 164 set nice values with
 165 .BR setpriority (2),
 166 etc.) only processes contained in its own PID namespace
 167 and in descendants of that namespace.
 168
 169 A process has one process ID in each of the layers of the PID
 170 namespace hierarchy in which is visible,
 171 and walking back though each direct ancestor namespace
 172 through to the root PID namespace.
 173 System calls that operate on process IDs always
 174 operate using the process ID that is visible in the
 175 PID namespace of the caller.
 176 A call to
 177 .BR getpid (2)
 178 always returns the PID associated with the namespace in which
 179 the process was created.
 180
 181 Some processes in a PID namespace may have parents
 182 that are outside of the namespace.
 183 For example, the parent of the initial process in the namespace
 184 (i.e., the
 185 .BR init (1)
 186 process with PID 1) is necessarily in another namespace.
 187 Likewise, the direct children of a process that uses
 188 .BR setns (2)
 189 to cause its children to join a PID namespace are in a different
 190 PID namespace from the caller of
 191 .BR setns (2).
 192 Calls to
 193 .BR getppid (2)
 194 for such processes return 0.
 195
 196 While processes may freely descend into child PID namespaces
 197 (e.g., using
 198 .BR setns (2)
 199 with
 200 .BR CLONE_NEWPID ),
 201 they may not move in the other direction.
 202 That is to say, processes may not enter any ancestor namespaces
 203 (parent, grandparent, etc.).
 204 Changing PID namespaces is a one-way operation.
 205
 206 The
 207 .BR NS_GET_PARENT
 208 .BR ioctl (2)
 209 operation can be used to discover the parental relationship
 210 between PID namespaces; see
 211 .BR ioctl_ns (2).
 212 .\"
 213 .\" ============================================================
 214 .\"
 215 .SS setns(2) and unshare(2) semantics
 216 Calls to
 217 .BR setns (2)
 218 that specify a PID namespace file descriptor
 219 and calls to
 220 .BR unshare (2)
 221 with the
 222 .BR CLONE_NEWPID
 223 flag cause children subsequently created
 224 by the caller to be placed in a different PID namespace from the caller.
 225 These calls do not, however,
 226 change the PID namespace of the calling process,
 227 because doing so would change the caller's idea of its own PID
 228 (as reported by
 229 .BR getpid ()),
 230 which would break many applications and libraries.
 231
 232 To put things another way:
 233 a process's PID namespace membership is determined when the process is created
 234 and cannot be changed thereafter.
 235 Among other things, this means that the parental relationship
 236 between processes mirrors the parental relationship between PID namespaces:
 237 the parent of a process is either in the same namespace
 238 or resides in the immediate parent PID namespace.
 239 .SS Compatibility of CLONE_NEWPID with other CLONE_* flags
 240 In current versions of Linux,
 241 .BR CLONE_NEWPID
 242 can't be combined with
 243 .BR CLONE_THREAD .
 244 Threads are required to be in the same PID namespace such that
 245 the threads in a process can send signals to each other.
 246 Similarly, it must be possible to see all of the threads
 247 of a processes in the
 248 .BR proc (5)
 249 filesystem.
 250 Additionally, if two threads were in different PID
 251 namespaces, the process ID of the process sending a signal
 252 could not be meaningfully encoded when a signal is sent
 253 (see the description of the
 254 .I siginfo_t
 255 type in
 256 .BR sigaction (2)).
 257 Since this is computed when a signal is enqueued,
 258 a signal queue shared by processes in multiple PID namespaces
 259 would defeat that.
 260
 261 .\" Note these restrictions were all introduced in
 262 .\" 8382fcac1b813ad0a4e68a838fc7ae93fa39eda0
 263 .\" when CLONE_NEWPID|CLONE_VM was disallowed
 264 In earlier versions of Linux,
 265 .BR CLONE_NEWPID
 266 was additionally disallowed (failing with the error
 267 .BR EINVAL )
 268 in combination with
 269 .BR CLONE_SIGHAND
 270 .\" (restriction lifted in faf00da544045fdc1454f3b9e6d7f65c841de302)
 271 (before Linux 4.3) as well as
 272 .\" (restriction lifted in e79f525e99b04390ca4d2366309545a836c03bf1)
 273 .BR CLONE_VM
 274 (before Linux 3.12).
 275 The changes that lifted these restrictions have also been ported to
 276 earlier stable kernels.
 277 .\"
 278 .\" ============================================================
 279 .\"
 280 .SS /proc and PID namespaces
 281 A
 282 .I /proc
 283 filesystem shows (in the
 284 .I /proc/[pid]
 285 directories) only processes visible in the PID namespace
 286 of the process that performed the mount, even if the
 287 .I /proc
 288 filesystem is viewed from processes in other namespaces.
 289
 290 After creating a new PID namespace,
 291 it is useful for the child to change its root directory
 292 and mount a new procfs instance at
 293 .I /proc
 294 so that tools such as
 295 .BR ps (1)
 296 work correctly.
 297 If a new mount namespace is simultaneously created by including
 298 .BR CLONE_NEWNS
 299 in the
 300 .IR flags
 301 argument of
 302 .BR clone (2)
 303 or
 304 .BR unshare (2),
 305 then it isn't necessary to change the root directory:
 306 a new procfs instance can be mounted directly over
 307 .IR /proc .
 308
 309 From a shell, the command to mount
 310 .I /proc
 311 is:
 312
 313     $ mount -t proc proc /proc
 314
 315 Calling
 316 .BR readlink (2)
 317 on the path
 318 .I /proc/self
 319 yields the process ID of the caller in the PID namespace of the procfs mount
 320 (i.e., the PID namespace of the process that mounted the procfs).
 321 This can be useful for introspection purposes,
 322 when a process wants to discover its PID in other namespaces.
 323 .\"
 324 .\" ============================================================
 325 .\"
 326 .SS Miscellaneous
 327 When a process ID is passed over a UNIX domain socket to a
 328 process in a different PID namespace (see the description of
 329 .B SCM_CREDENTIALS
 330 in
 331 .BR unix (7)),
 332 it is translated into the corresponding PID value in
 333 the receiving process's PID namespace.
 334 .SH CONFORMING TO
 335 Namespaces are a Linux-specific feature.
 336 .SH EXAMPLE
 337 See
 338 .BR user_namespaces (7).
 339 .SH SEE ALSO
 340 .BR clone (2),
 341 .BR setns (2),
 342 .BR unshare (2),
 343 .BR proc (5),
 344 .BR capabilities (7),
 345 .BR credentials (7),
 346 .BR namespaces (7),
 347 .BR user_namespaces (7),
 348 .BR switch_root (8)