blob: 190018b0c64944e3463548c87e4c2d8226dcb71d [file] [log] [blame]
Paul Menageddbcc7e2007-10-18 23:39:30 -07001 CGROUPS
2 -------
3
Li Zefan45ce80f2009-01-15 13:50:59 -08004Written by Paul Menage <menage@google.com> based on
5Documentation/cgroups/cpusets.txt
Paul Menageddbcc7e2007-10-18 23:39:30 -07006
7Original copyright statements from cpusets.txt:
8Portions Copyright (C) 2004 BULL SA.
9Portions Copyright (c) 2004-2006 Silicon Graphics, Inc.
10Modified by Paul Jackson <pj@sgi.com>
11Modified by Christoph Lameter <clameter@sgi.com>
12
13CONTENTS:
14=========
15
161. Control Groups
17 1.1 What are cgroups ?
18 1.2 Why are cgroups needed ?
19 1.3 How are cgroups implemented ?
20 1.4 What does notify_on_release do ?
Daniel Lezcano97978e62010-10-27 15:33:35 -070021 1.5 What does clone_children do ?
22 1.6 How do I use cgroups ?
Paul Menageddbcc7e2007-10-18 23:39:30 -0700232. Usage Examples and Syntax
24 2.1 Basic Usage
25 2.2 Attaching processes
Kirill A. Shutemov8ca712e2010-03-10 15:22:10 -080026 2.3 Mounting hierarchies by name
Kirill A. Shutemov0dea1162010-03-10 15:22:20 -080027 2.4 Notification API
Paul Menageddbcc7e2007-10-18 23:39:30 -0700283. Kernel API
29 3.1 Overview
30 3.2 Synchronization
31 3.3 Subsystem API
324. Questions
33
341. Control Groups
Li Zefand19e0582008-02-23 15:24:08 -080035=================
Paul Menageddbcc7e2007-10-18 23:39:30 -070036
371.1 What are cgroups ?
38----------------------
39
40Control Groups provide a mechanism for aggregating/partitioning sets of
41tasks, and all their future children, into hierarchical groups with
42specialized behaviour.
43
44Definitions:
45
46A *cgroup* associates a set of tasks with a set of parameters for one
47or more subsystems.
48
49A *subsystem* is a module that makes use of the task grouping
50facilities provided by cgroups to treat groups of tasks in
51particular ways. A subsystem is typically a "resource controller" that
52schedules a resource or applies per-cgroup limits, but it may be
53anything that wants to act on a group of processes, e.g. a
54virtualization subsystem.
55
56A *hierarchy* is a set of cgroups arranged in a tree, such that
57every task in the system is in exactly one of the cgroups in the
58hierarchy, and a set of subsystems; each subsystem has system-specific
59state attached to each cgroup in the hierarchy. Each hierarchy has
60an instance of the cgroup virtual filesystem associated with it.
61
Chris Samuelcaa790b2009-01-17 00:01:18 +110062At any one time there may be multiple active hierarchies of task
Paul Menageddbcc7e2007-10-18 23:39:30 -070063cgroups. Each hierarchy is a partition of all tasks in the system.
64
65User level code may create and destroy cgroups by name in an
66instance of the cgroup virtual file system, specify and query to
67which cgroup a task is assigned, and list the task pids assigned to
68a cgroup. Those creations and assignments only affect the hierarchy
69associated with that instance of the cgroup file system.
70
71On their own, the only use for cgroups is for simple job
72tracking. The intention is that other subsystems hook into the generic
73cgroup support to provide new attributes for cgroups, such as
74accounting/limiting the resources which processes in a cgroup can
Li Zefan45ce80f2009-01-15 13:50:59 -080075access. For example, cpusets (see Documentation/cgroups/cpusets.txt) allows
Paul Menageddbcc7e2007-10-18 23:39:30 -070076you to associate a set of CPUs and a set of memory nodes with the
77tasks in each cgroup.
78
791.2 Why are cgroups needed ?
80----------------------------
81
82There are multiple efforts to provide process aggregations in the
83Linux kernel, mainly for resource tracking purposes. Such efforts
84include cpusets, CKRM/ResGroups, UserBeanCounters, and virtual server
85namespaces. These all require the basic notion of a
86grouping/partitioning of processes, with newly forked processes ending
87in the same group (cgroup) as their parent process.
88
89The kernel cgroup patch provides the minimum essential kernel
90mechanisms required to efficiently implement such groups. It has
91minimal impact on the system fast paths, and provides hooks for
92specific subsystems such as cpusets to provide additional behaviour as
93desired.
94
95Multiple hierarchy support is provided to allow for situations where
96the division of tasks into cgroups is distinctly different for
97different subsystems - having parallel hierarchies allows each
98hierarchy to be a natural division of tasks, without having to handle
99complex combinations of tasks that would be present if several
100unrelated subsystems needed to be forced into the same tree of
101cgroups.
102
103At one extreme, each resource controller or subsystem could be in a
104separate hierarchy; at the other extreme, all subsystems
105would be attached to the same hierarchy.
106
107As an example of a scenario (originally proposed by vatsa@in.ibm.com)
108that can benefit from multiple hierarchies, consider a large
109university server with various users - students, professors, system
110tasks etc. The resource planning for this server could be along the
111following lines:
112
113 CPU : Top cpuset
114 / \
115 CPUSet1 CPUSet2
116 | |
117 (Profs) (Students)
118
119 In addition (system tasks) are attached to topcpuset (so
120 that they can run anywhere) with a limit of 20%
121
122 Memory : Professors (50%), students (30%), system (20%)
123
124 Disk : Prof (50%), students (30%), system (20%)
125
126 Network : WWW browsing (20%), Network File System (60%), others (20%)
127 / \
128 Prof (15%) students (5%)
129
Chris Samuelcaa790b2009-01-17 00:01:18 +1100130Browsers like Firefox/Lynx go into the WWW network class, while (k)nfsd go
Paul Menageddbcc7e2007-10-18 23:39:30 -0700131into NFS network class.
132
Chris Samuelcaa790b2009-01-17 00:01:18 +1100133At the same time Firefox/Lynx will share an appropriate CPU/Memory class
Paul Menageddbcc7e2007-10-18 23:39:30 -0700134depending on who launched it (prof/student).
135
136With the ability to classify tasks differently for different resources
137(by putting those resource subsystems in different hierarchies) then
138the admin can easily set up a script which receives exec notifications
139and depending on who is launching the browser he can
140
141 # echo browser_pid > /mnt/<restype>/<userclass>/tasks
142
143With only a single hierarchy, he now would potentially have to create
144a separate cgroup for every browser launched and associate it with
145approp network and other resource class. This may lead to
146proliferation of such cgroups.
147
148Also lets say that the administrator would like to give enhanced network
149access temporarily to a student's browser (since it is night and the user
Li Zefand19e0582008-02-23 15:24:08 -0800150wants to do online gaming :)) OR give one of the students simulation
Paul Menageddbcc7e2007-10-18 23:39:30 -0700151apps enhanced CPU power,
152
Li Zefand19e0582008-02-23 15:24:08 -0800153With ability to write pids directly to resource classes, it's just a
Paul Menageddbcc7e2007-10-18 23:39:30 -0700154matter of :
155
156 # echo pid > /mnt/network/<new_class>/tasks
157 (after some time)
158 # echo pid > /mnt/network/<orig_class>/tasks
159
160Without this ability, he would have to split the cgroup into
161multiple separate ones and then associate the new cgroups with the
162new resource classes.
163
164
165
1661.3 How are cgroups implemented ?
167---------------------------------
168
169Control Groups extends the kernel as follows:
170
171 - Each task in the system has a reference-counted pointer to a
172 css_set.
173
174 - A css_set contains a set of reference-counted pointers to
175 cgroup_subsys_state objects, one for each cgroup subsystem
176 registered in the system. There is no direct link from a task to
177 the cgroup of which it's a member in each hierarchy, but this
178 can be determined by following pointers through the
179 cgroup_subsys_state objects. This is because accessing the
180 subsystem state is something that's expected to happen frequently
181 and in performance-critical code, whereas operations that require a
182 task's actual cgroup assignments (in particular, moving between
Paul Menage817929e2007-10-18 23:39:36 -0700183 cgroups) are less common. A linked list runs through the cg_list
184 field of each task_struct using the css_set, anchored at
185 css_set->tasks.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700186
187 - A cgroup hierarchy filesystem can be mounted for browsing and
188 manipulation from user space.
189
190 - You can list all the tasks (by pid) attached to any cgroup.
191
192The implementation of cgroups requires a few, simple hooks
193into the rest of the kernel, none in performance critical paths:
194
195 - in init/main.c, to initialize the root cgroups and initial
196 css_set at system boot.
197
198 - in fork and exit, to attach and detach a task from its css_set.
199
200In addition a new file system, of type "cgroup" may be mounted, to
201enable browsing and modifying the cgroups presently known to the
202kernel. When mounting a cgroup hierarchy, you may specify a
203comma-separated list of subsystems to mount as the filesystem mount
204options. By default, mounting the cgroup filesystem attempts to
205mount a hierarchy containing all registered subsystems.
206
207If an active hierarchy with exactly the same set of subsystems already
208exists, it will be reused for the new mount. If no existing hierarchy
209matches, and any of the requested subsystems are in use in an existing
210hierarchy, the mount will fail with -EBUSY. Otherwise, a new hierarchy
211is activated, associated with the requested subsystems.
212
213It's not currently possible to bind a new subsystem to an active
214cgroup hierarchy, or to unbind a subsystem from an active cgroup
215hierarchy. This may be possible in future, but is fraught with nasty
216error-recovery issues.
217
218When a cgroup filesystem is unmounted, if there are any
219child cgroups created below the top-level cgroup, that hierarchy
220will remain active even though unmounted; if there are no
221child cgroups then the hierarchy will be deactivated.
222
223No new system calls are added for cgroups - all support for
224querying and modifying cgroups is via this cgroup file system.
225
226Each task under /proc has an added file named 'cgroup' displaying,
227for each active hierarchy, the subsystem names and the cgroup name
228as the path relative to the root of the cgroup file system.
229
230Each cgroup is represented by a directory in the cgroup file system
231containing the following files describing that cgroup:
232
Paul Menage7823da32009-10-07 16:32:26 -0700233 - tasks: list of tasks (by pid) attached to that cgroup. This list
234 is not guaranteed to be sorted. Writing a thread id into this file
235 moves the thread into this cgroup.
236 - cgroup.procs: list of tgids in the cgroup. This list is not
237 guaranteed to be sorted or free of duplicate tgids, and userspace
238 should sort/uniquify the list if this property is required.
KAMEZAWA Hiroyuki7716fa62010-04-23 13:17:51 -0400239 This is a read-only file, for now.
Li Zefand19e0582008-02-23 15:24:08 -0800240 - notify_on_release flag: run the release agent on exit?
241 - release_agent: the path to use for release notifications (this file
242 exists in the top cgroup only)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700243
244Other subsystems such as cpusets may add additional files in each
Li Zefand19e0582008-02-23 15:24:08 -0800245cgroup dir.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700246
247New cgroups are created using the mkdir system call or shell
248command. The properties of a cgroup, such as its flags, are
249modified by writing to the appropriate file in that cgroups
250directory, as listed above.
251
252The named hierarchical structure of nested cgroups allows partitioning
253a large system into nested, dynamically changeable, "soft-partitions".
254
255The attachment of each task, automatically inherited at fork by any
256children of that task, to a cgroup allows organizing the work load
257on a system into related sets of tasks. A task may be re-attached to
258any other cgroup, if allowed by the permissions on the necessary
259cgroup file system directories.
260
261When a task is moved from one cgroup to another, it gets a new
262css_set pointer - if there's an already existing css_set with the
263desired collection of cgroups then that group is reused, else a new
Li Zefanb851ee72009-02-18 14:48:14 -0800264css_set is allocated. The appropriate existing css_set is located by
265looking into a hash table.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700266
Paul Menage817929e2007-10-18 23:39:36 -0700267To allow access from a cgroup to the css_sets (and hence tasks)
268that comprise it, a set of cg_cgroup_link objects form a lattice;
269each cg_cgroup_link is linked into a list of cg_cgroup_links for
Li Zefand19e0582008-02-23 15:24:08 -0800270a single cgroup on its cgrp_link_list field, and a list of
Paul Menage817929e2007-10-18 23:39:36 -0700271cg_cgroup_links for a single css_set on its cg_link_list.
272
273Thus the set of tasks in a cgroup can be listed by iterating over
274each css_set that references the cgroup, and sub-iterating over
275each css_set's task set.
276
Paul Menageddbcc7e2007-10-18 23:39:30 -0700277The use of a Linux virtual file system (vfs) to represent the
278cgroup hierarchy provides for a familiar permission and name space
279for cgroups, with a minimum of additional kernel code.
280
2811.4 What does notify_on_release do ?
282------------------------------------
283
Paul Menageddbcc7e2007-10-18 23:39:30 -0700284If the notify_on_release flag is enabled (1) in a cgroup, then
285whenever the last task in the cgroup leaves (exits or attaches to
286some other cgroup) and the last child cgroup of that cgroup
287is removed, then the kernel runs the command specified by the contents
288of the "release_agent" file in that hierarchy's root directory,
289supplying the pathname (relative to the mount point of the cgroup
290file system) of the abandoned cgroup. This enables automatic
291removal of abandoned cgroups. The default value of
292notify_on_release in the root cgroup at system boot is disabled
293(0). The default value of other cgroups at creation is the current
294value of their parents notify_on_release setting. The default value of
295a cgroup hierarchy's release_agent path is empty.
296
Daniel Lezcano97978e62010-10-27 15:33:35 -07002971.5 What does clone_children do ?
298---------------------------------
299
300If the clone_children flag is enabled (1) in a cgroup, then all
301cgroups created beneath will call the post_clone callbacks for each
302subsystem of the newly created cgroup. Usually when this callback is
303implemented for a subsystem, it copies the values of the parent
304subsystem, this is the case for the cpuset.
305
3061.6 How do I use cgroups ?
Paul Menageddbcc7e2007-10-18 23:39:30 -0700307--------------------------
308
309To start a new job that is to be contained within a cgroup, using
310the "cpuset" cgroup subsystem, the steps are something like:
311
312 1) mkdir /dev/cgroup
313 2) mount -t cgroup -ocpuset cpuset /dev/cgroup
314 3) Create the new cgroup by doing mkdir's and write's (or echo's) in
315 the /dev/cgroup virtual file system.
316 4) Start a task that will be the "founding father" of the new job.
317 5) Attach that task to the new cgroup by writing its pid to the
318 /dev/cgroup tasks file for that cgroup.
319 6) fork, exec or clone the job tasks from this founding father task.
320
321For example, the following sequence of commands will setup a cgroup
322named "Charlie", containing just CPUs 2 and 3, and Memory Node 1,
323and then start a subshell 'sh' in that cgroup:
324
325 mount -t cgroup cpuset -ocpuset /dev/cgroup
326 cd /dev/cgroup
327 mkdir Charlie
328 cd Charlie
Dhaval Giani0f146a72008-05-12 14:02:31 -0700329 /bin/echo 2-3 > cpuset.cpus
330 /bin/echo 1 > cpuset.mems
Paul Menageddbcc7e2007-10-18 23:39:30 -0700331 /bin/echo $$ > tasks
332 sh
333 # The subshell 'sh' is now running in cgroup Charlie
334 # The next line should display '/Charlie'
335 cat /proc/self/cgroup
336
3372. Usage Examples and Syntax
338============================
339
3402.1 Basic Usage
341---------------
342
343Creating, modifying, using the cgroups can be done through the cgroup
344virtual filesystem.
345
Chris Samuelcaa790b2009-01-17 00:01:18 +1100346To mount a cgroup hierarchy with all available subsystems, type:
Paul Menageddbcc7e2007-10-18 23:39:30 -0700347# mount -t cgroup xxx /dev/cgroup
348
349The "xxx" is not interpreted by the cgroup code, but will appear in
350/proc/mounts so may be any useful identifying string that you like.
351
Trevor Woerner595f4b62010-05-26 14:42:35 -0700352To mount a cgroup hierarchy with just the cpuset and memory
Paul Menageddbcc7e2007-10-18 23:39:30 -0700353subsystems, type:
Li Zefanb6719ec2009-04-02 16:57:28 -0700354# mount -t cgroup -o cpuset,memory hier1 /dev/cgroup
Paul Menageddbcc7e2007-10-18 23:39:30 -0700355
356To change the set of subsystems bound to a mounted hierarchy, just
357remount with different options:
Li Zefanb6719ec2009-04-02 16:57:28 -0700358# mount -o remount,cpuset,ns hier1 /dev/cgroup
Paul Menageddbcc7e2007-10-18 23:39:30 -0700359
Li Zefanb6719ec2009-04-02 16:57:28 -0700360Now memory is removed from the hierarchy and ns is added.
361
362Note this will add ns to the hierarchy but won't remove memory or
363cpuset, because the new options are appended to the old ones:
364# mount -o remount,ns /dev/cgroup
365
366To Specify a hierarchy's release_agent:
367# mount -t cgroup -o cpuset,release_agent="/sbin/cpuset_release_agent" \
368 xxx /dev/cgroup
369
370Note that specifying 'release_agent' more than once will return failure.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700371
372Note that changing the set of subsystems is currently only supported
373when the hierarchy consists of a single (root) cgroup. Supporting
374the ability to arbitrarily bind/unbind subsystems from an existing
375cgroup hierarchy is intended to be implemented in the future.
376
377Then under /dev/cgroup you can find a tree that corresponds to the
378tree of the cgroups in the system. For instance, /dev/cgroup
379is the cgroup that holds the whole system.
380
Li Zefanb6719ec2009-04-02 16:57:28 -0700381If you want to change the value of release_agent:
382# echo "/sbin/new_release_agent" > /dev/cgroup/release_agent
383
384It can also be changed via remount.
385
Paul Menageddbcc7e2007-10-18 23:39:30 -0700386If you want to create a new cgroup under /dev/cgroup:
387# cd /dev/cgroup
388# mkdir my_cgroup
389
390Now you want to do something with this cgroup.
391# cd my_cgroup
392
393In this directory you can find several files:
394# ls
Paul Menage7823da32009-10-07 16:32:26 -0700395cgroup.procs notify_on_release tasks
Li Zefand19e0582008-02-23 15:24:08 -0800396(plus whatever files added by the attached subsystems)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700397
398Now attach your shell to this cgroup:
399# /bin/echo $$ > tasks
400
401You can also create cgroups inside your cgroup by using mkdir in this
402directory.
403# mkdir my_sub_cs
404
405To remove a cgroup, just use rmdir:
406# rmdir my_sub_cs
407
408This will fail if the cgroup is in use (has cgroups inside, or
409has processes attached, or is held alive by other subsystem-specific
410reference).
411
4122.2 Attaching processes
413-----------------------
414
415# /bin/echo PID > tasks
416
417Note that it is PID, not PIDs. You can only attach ONE task at a time.
418If you have several tasks to attach, you have to do it one after another:
419
420# /bin/echo PID1 > tasks
421# /bin/echo PID2 > tasks
422 ...
423# /bin/echo PIDn > tasks
424
Li Zefanbef67c52008-07-04 09:59:55 -0700425You can attach the current shell task by echoing 0:
426
427# echo 0 > tasks
428
Paul Menagec6d57f32009-09-23 15:56:19 -07004292.3 Mounting hierarchies by name
430--------------------------------
431
432Passing the name=<x> option when mounting a cgroups hierarchy
433associates the given name with the hierarchy. This can be used when
434mounting a pre-existing hierarchy, in order to refer to it by name
435rather than by its set of active subsystems. Each hierarchy is either
436nameless, or has a unique name.
437
438The name should match [\w.-]+
439
440When passing a name=<x> option for a new hierarchy, you need to
441specify subsystems manually; the legacy behaviour of mounting all
442subsystems when none are explicitly specified is not supported when
443you give a subsystem a name.
444
445The name of the subsystem appears as part of the hierarchy description
446in /proc/mounts and /proc/<pid>/cgroups.
447
Kirill A. Shutemov0dea1162010-03-10 15:22:20 -08004482.4 Notification API
449--------------------
450
451There is mechanism which allows to get notifications about changing
452status of a cgroup.
453
454To register new notification handler you need:
455 - create a file descriptor for event notification using eventfd(2);
456 - open a control file to be monitored (e.g. memory.usage_in_bytes);
457 - write "<event_fd> <control_fd> <args>" to cgroup.event_control.
458 Interpretation of args is defined by control file implementation;
459
460eventfd will be woken up by control file implementation or when the
461cgroup is removed.
462
463To unregister notification handler just close eventfd.
464
465NOTE: Support of notifications should be implemented for the control
466file. See documentation for the subsystem.
Paul Menagec6d57f32009-09-23 15:56:19 -0700467
Paul Menageddbcc7e2007-10-18 23:39:30 -07004683. Kernel API
469=============
470
4713.1 Overview
472------------
473
474Each kernel subsystem that wants to hook into the generic cgroup
475system needs to create a cgroup_subsys object. This contains
476various methods, which are callbacks from the cgroup system, along
477with a subsystem id which will be assigned by the cgroup system.
478
479Other fields in the cgroup_subsys object include:
480
481- subsys_id: a unique array index for the subsystem, indicating which
Li Zefand19e0582008-02-23 15:24:08 -0800482 entry in cgroup->subsys[] this subsystem should be managing.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700483
Li Zefand19e0582008-02-23 15:24:08 -0800484- name: should be initialized to a unique subsystem name. Should be
485 no longer than MAX_CGROUP_TYPE_NAMELEN.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700486
Li Zefand19e0582008-02-23 15:24:08 -0800487- early_init: indicate if the subsystem needs early initialization
488 at system boot.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700489
490Each cgroup object created by the system has an array of pointers,
491indexed by subsystem id; this pointer is entirely managed by the
492subsystem; the generic cgroup code will never touch this pointer.
493
4943.2 Synchronization
495-------------------
496
497There is a global mutex, cgroup_mutex, used by the cgroup
498system. This should be taken by anything that wants to modify a
499cgroup. It may also be taken to prevent cgroups from being
500modified, but more specific locks may be more appropriate in that
501situation.
502
503See kernel/cgroup.c for more details.
504
505Subsystems can take/release the cgroup_mutex via the functions
Paul Menageddbcc7e2007-10-18 23:39:30 -0700506cgroup_lock()/cgroup_unlock().
507
508Accessing a task's cgroup pointer may be done in the following ways:
509- while holding cgroup_mutex
510- while holding the task's alloc_lock (via task_lock())
511- inside an rcu_read_lock() section via rcu_dereference()
512
5133.3 Subsystem API
Li Zefand19e0582008-02-23 15:24:08 -0800514-----------------
Paul Menageddbcc7e2007-10-18 23:39:30 -0700515
516Each subsystem should:
517
518- add an entry in linux/cgroup_subsys.h
519- define a cgroup_subsys object called <name>_subsys
520
Ben Blume6a11052010-03-10 15:22:09 -0800521If a subsystem can be compiled as a module, it should also have in its
Ben Blumcf5d5942010-03-10 15:22:09 -0800522module initcall a call to cgroup_load_subsys(), and in its exitcall a
523call to cgroup_unload_subsys(). It should also set its_subsys.module =
524THIS_MODULE in its .c file.
Ben Blume6a11052010-03-10 15:22:09 -0800525
Paul Menageddbcc7e2007-10-18 23:39:30 -0700526Each subsystem may export the following methods. The only mandatory
527methods are create/destroy. Any others that are null are presumed to
528be successful no-ops.
529
Li Zefand19e0582008-02-23 15:24:08 -0800530struct cgroup_subsys_state *create(struct cgroup_subsys *ss,
531 struct cgroup *cgrp)
Paul Menage8dc4f3e2008-02-07 00:13:45 -0800532(cgroup_mutex held by caller)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700533
534Called to create a subsystem state object for a cgroup. The
535subsystem should allocate its subsystem state object for the passed
536cgroup, returning a pointer to the new object on success or a
537negative error code. On success, the subsystem pointer should point to
538a structure of type cgroup_subsys_state (typically embedded in a
539larger subsystem-specific object), which will be initialized by the
540cgroup system. Note that this will be called at initialization to
541create the root subsystem state for this subsystem; this case can be
542identified by the passed cgroup object having a NULL parent (since
543it's the root of the hierarchy) and may be an appropriate place for
544initialization code.
545
Li Zefand19e0582008-02-23 15:24:08 -0800546void destroy(struct cgroup_subsys *ss, struct cgroup *cgrp)
Paul Menage8dc4f3e2008-02-07 00:13:45 -0800547(cgroup_mutex held by caller)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700548
Paul Menage8dc4f3e2008-02-07 00:13:45 -0800549The cgroup system is about to destroy the passed cgroup; the subsystem
550should do any necessary cleanup and free its subsystem state
551object. By the time this method is called, the cgroup has already been
552unlinked from the file system and from the child list of its parent;
553cgroup->parent is still valid. (Note - can also be called for a
554newly-created cgroup if an error occurs after this subsystem's
555create() method has been called for the new cgroup).
Paul Menageddbcc7e2007-10-18 23:39:30 -0700556
KAMEZAWA Hiroyukiec64f512009-04-02 16:57:26 -0700557int pre_destroy(struct cgroup_subsys *ss, struct cgroup *cgrp);
Li Zefand19e0582008-02-23 15:24:08 -0800558
559Called before checking the reference count on each subsystem. This may
560be useful for subsystems which have some extra references even if
KAMEZAWA Hiroyukiec64f512009-04-02 16:57:26 -0700561there are not tasks in the cgroup. If pre_destroy() returns error code,
562rmdir() will fail with it. From this behavior, pre_destroy() can be
563called multiple times against a cgroup.
Li Zefand19e0582008-02-23 15:24:08 -0800564
565int can_attach(struct cgroup_subsys *ss, struct cgroup *cgrp,
Ben Blumbe367d02009-09-23 15:56:31 -0700566 struct task_struct *task, bool threadgroup)
Paul Menage8dc4f3e2008-02-07 00:13:45 -0800567(cgroup_mutex held by caller)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700568
569Called prior to moving a task into a cgroup; if the subsystem
570returns an error, this will abort the attach operation. If a NULL
571task is passed, then a successful result indicates that *any*
572unspecified task can be moved into the cgroup. Note that this isn't
573called on a fork. If this method returns 0 (success) then this should
Daisuke Nishimura2468c722010-03-10 15:22:03 -0800574remain valid while the caller holds cgroup_mutex and it is ensured that either
575attach() or cancel_attach() will be called in future. If threadgroup is
Ben Blumbe367d02009-09-23 15:56:31 -0700576true, then a successful result indicates that all threads in the given
577thread's threadgroup can be moved together.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700578
Daisuke Nishimura2468c722010-03-10 15:22:03 -0800579void cancel_attach(struct cgroup_subsys *ss, struct cgroup *cgrp,
580 struct task_struct *task, bool threadgroup)
581(cgroup_mutex held by caller)
582
583Called when a task attach operation has failed after can_attach() has succeeded.
584A subsystem whose can_attach() has some side-effects should provide this
Thomas Weber88393162010-03-16 11:47:56 +0100585function, so that the subsystem can implement a rollback. If not, not necessary.
Daisuke Nishimura2468c722010-03-10 15:22:03 -0800586This will be called only about subsystems whose can_attach() operation have
587succeeded.
588
Li Zefand19e0582008-02-23 15:24:08 -0800589void attach(struct cgroup_subsys *ss, struct cgroup *cgrp,
Ben Blumbe367d02009-09-23 15:56:31 -0700590 struct cgroup *old_cgrp, struct task_struct *task,
591 bool threadgroup)
Li Zefan18e7f1f2009-01-07 18:07:32 -0800592(cgroup_mutex held by caller)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700593
594Called after the task has been attached to the cgroup, to allow any
595post-attachment activity that requires memory allocations or blocking.
Ben Blumbe367d02009-09-23 15:56:31 -0700596If threadgroup is true, the subsystem should take care of all threads
597in the specified thread's threadgroup. Currently does not support any
598subsystem that might need the old_cgrp for every thread in the group.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700599
600void fork(struct cgroup_subsy *ss, struct task_struct *task)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700601
Li Zefane8d55fd2008-04-29 01:00:13 -0700602Called when a task is forked into a cgroup.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700603
604void exit(struct cgroup_subsys *ss, struct task_struct *task)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700605
Li Zefand19e0582008-02-23 15:24:08 -0800606Called during task exit.
Paul Menageddbcc7e2007-10-18 23:39:30 -0700607
Li Zefand19e0582008-02-23 15:24:08 -0800608int populate(struct cgroup_subsys *ss, struct cgroup *cgrp)
Li Zefan18e7f1f2009-01-07 18:07:32 -0800609(cgroup_mutex held by caller)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700610
611Called after creation of a cgroup to allow a subsystem to populate
612the cgroup directory with file entries. The subsystem should make
613calls to cgroup_add_file() with objects of type cftype (see
614include/linux/cgroup.h for details). Note that although this
615method can return an error code, the error code is currently not
616always handled well.
617
Li Zefand19e0582008-02-23 15:24:08 -0800618void post_clone(struct cgroup_subsys *ss, struct cgroup *cgrp)
Li Zefan18e7f1f2009-01-07 18:07:32 -0800619(cgroup_mutex held by caller)
Paul Menage697f4162007-10-18 23:39:34 -0700620
Chris Samuelcaa790b2009-01-17 00:01:18 +1100621Called at the end of cgroup_clone() to do any parameter
Paul Menage697f4162007-10-18 23:39:34 -0700622initialization which might be required before a task could attach. For
623example in cpusets, no task may attach before 'cpus' and 'mems' are set
624up.
625
Paul Menageddbcc7e2007-10-18 23:39:30 -0700626void bind(struct cgroup_subsys *ss, struct cgroup *root)
Paul Menage999cd8a2009-01-07 18:08:36 -0800627(cgroup_mutex and ss->hierarchy_mutex held by caller)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700628
629Called when a cgroup subsystem is rebound to a different hierarchy
630and root cgroup. Currently this will only involve movement between
631the default hierarchy (which never has sub-cgroups) and a hierarchy
632that is being created/destroyed (and hence has no sub-cgroups).
633
6344. Questions
635============
636
637Q: what's up with this '/bin/echo' ?
638A: bash's builtin 'echo' command does not check calls to write() against
639 errors. If you use it in the cgroup file system, you won't be
640 able to tell whether a command succeeded or failed.
641
642Q: When I attach processes, only the first of the line gets really attached !
643A: We can only return one error code per call to write(). So you should also
644 put only ONE pid.
645