Blame view
kernel/cgroup.c
93.7 KB
ddbcc7e8e Task Control Grou... |
1 |
/* |
ddbcc7e8e Task Control Grou... |
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 |
* Generic process-grouping system. * * Based originally on the cpuset system, extracted by Paul Menage * Copyright (C) 2006 Google, Inc * * Copyright notices from the original cpuset code: * -------------------------------------------------- * Copyright (C) 2003 BULL SA. * Copyright (C) 2004-2006 Silicon Graphics, Inc. * * Portions derived from Patrick Mochel's sysfs code. * sysfs is Copyright (c) 2001-3 Patrick Mochel * * 2003-10-10 Written by Simon Derr. * 2003-10-22 Updates by Stephen Hemminger. * 2004 May-July Rework by Paul Jackson. * --------------------------------------------------- * * This file is subject to the terms and conditions of the GNU General Public * License. See the file COPYING in the main directory of the Linux * distribution for more details. */ #include <linux/cgroup.h> #include <linux/errno.h> #include <linux/fs.h> #include <linux/kernel.h> #include <linux/list.h> #include <linux/mm.h> #include <linux/mutex.h> #include <linux/mount.h> #include <linux/pagemap.h> |
a424316ca Task Control Grou... |
34 |
#include <linux/proc_fs.h> |
ddbcc7e8e Task Control Grou... |
35 36 |
#include <linux/rcupdate.h> #include <linux/sched.h> |
817929ec2 Task Control Grou... |
37 |
#include <linux/backing-dev.h> |
ddbcc7e8e Task Control Grou... |
38 39 40 41 42 |
#include <linux/seq_file.h> #include <linux/slab.h> #include <linux/magic.h> #include <linux/spinlock.h> #include <linux/string.h> |
bbcb81d09 Task Control Grou... |
43 |
#include <linux/sort.h> |
81a6a5cdd Task Control Grou... |
44 |
#include <linux/kmod.h> |
846c7bb05 Add cgroupstats |
45 46 |
#include <linux/delayacct.h> #include <linux/cgroupstats.h> |
472b1053f cgroups: use a ha... |
47 |
#include <linux/hash.h> |
3f8206d49 [PATCH] get rid o... |
48 |
#include <linux/namei.h> |
337eb00a2 Push BKL down int... |
49 |
#include <linux/smp_lock.h> |
846c7bb05 Add cgroupstats |
50 |
|
ddbcc7e8e Task Control Grou... |
51 |
#include <asm/atomic.h> |
81a6a5cdd Task Control Grou... |
52 |
static DEFINE_MUTEX(cgroup_mutex); |
ddbcc7e8e Task Control Grou... |
53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 |
/* Generate an array of cgroup subsystem pointers */ #define SUBSYS(_x) &_x ## _subsys, static struct cgroup_subsys *subsys[] = { #include <linux/cgroup_subsys.h> }; /* * A cgroupfs_root represents the root of a cgroup hierarchy, * and may be associated with a superblock to form an active * hierarchy */ struct cgroupfs_root { struct super_block *sb; /* * The bitmask of subsystems intended to be attached to this * hierarchy */ unsigned long subsys_bits; /* The bitmask of subsystems currently attached to this hierarchy */ unsigned long actual_subsys_bits; /* A list running through the attached subsystems */ struct list_head subsys_list; /* The root cgroup for this hierarchy */ struct cgroup top_cgroup; /* Tracks how many cgroups are currently defined in hierarchy.*/ int number_of_cgroups; |
e5f6a8609 cgroups: make roo... |
85 |
/* A list running through the active hierarchies */ |
ddbcc7e8e Task Control Grou... |
86 87 88 89 |
struct list_head root_list; /* Hierarchy-specific flags */ unsigned long flags; |
81a6a5cdd Task Control Grou... |
90 |
|
e788e066c cgroup files: mov... |
91 |
/* The path to use for release notifications. */ |
81a6a5cdd Task Control Grou... |
92 |
char release_agent_path[PATH_MAX]; |
ddbcc7e8e Task Control Grou... |
93 |
}; |
ddbcc7e8e Task Control Grou... |
94 95 96 97 98 99 |
/* * The "rootnode" hierarchy is the "dummy hierarchy", reserved for the * subsystems that are otherwise unattached - it never has more than a * single cgroup, and all tasks are part of that cgroup. */ static struct cgroupfs_root rootnode; |
38460b48d cgroup: CSS ID su... |
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 |
/* * CSS ID -- ID per subsys's Cgroup Subsys State(CSS). used only when * cgroup_subsys->use_id != 0. */ #define CSS_ID_MAX (65535) struct css_id { /* * The css to which this ID points. This pointer is set to valid value * after cgroup is populated. If cgroup is removed, this will be NULL. * This pointer is expected to be RCU-safe because destroy() * is called after synchronize_rcu(). But for safe use, css_is_removed() * css_tryget() should be used for avoiding race. */ struct cgroup_subsys_state *css; /* * ID of this css. */ unsigned short id; /* * Depth in hierarchy which this ID belongs to. */ unsigned short depth; /* * ID is freed by RCU. (and lookup routine is RCU safe.) */ struct rcu_head rcu_head; /* * Hierarchy of CSS ID belongs to. */ unsigned short stack[0]; /* Array of Length (depth+1) */ }; |
ddbcc7e8e Task Control Grou... |
131 132 133 |
/* The list of hierarchy roots */ static LIST_HEAD(roots); |
817929ec2 Task Control Grou... |
134 |
static int root_count; |
ddbcc7e8e Task Control Grou... |
135 136 137 138 139 |
/* dummytop is a shorthand for the dummy hierarchy's top cgroup */ #define dummytop (&rootnode.top_cgroup) /* This flag indicates whether tasks in the fork and exit paths should |
a043e3b2c cgroup: fix comments |
140 141 142 |
* check for fork/exit handlers to call. This avoids us having to do * extra work in the fork/exit path if none of the subsystems need to * be called. |
ddbcc7e8e Task Control Grou... |
143 |
*/ |
8947f9d5b cgroups: annotate... |
144 |
static int need_forkexit_callback __read_mostly; |
ddbcc7e8e Task Control Grou... |
145 |
|
ddbcc7e8e Task Control Grou... |
146 |
/* convenient tests for these bits */ |
bd89aabc6 Control groups: R... |
147 |
inline int cgroup_is_removed(const struct cgroup *cgrp) |
ddbcc7e8e Task Control Grou... |
148 |
{ |
bd89aabc6 Control groups: R... |
149 |
return test_bit(CGRP_REMOVED, &cgrp->flags); |
ddbcc7e8e Task Control Grou... |
150 151 152 153 154 155 |
} /* bits in struct cgroupfs_root flags field */ enum { ROOT_NOPREFIX, /* mounted subsystems have no named prefix */ }; |
e9685a03c kernel/cgroup.c: ... |
156 |
static int cgroup_is_releasable(const struct cgroup *cgrp) |
81a6a5cdd Task Control Grou... |
157 158 |
{ const int bits = |
bd89aabc6 Control groups: R... |
159 160 161 |
(1 << CGRP_RELEASABLE) | (1 << CGRP_NOTIFY_ON_RELEASE); return (cgrp->flags & bits) == bits; |
81a6a5cdd Task Control Grou... |
162 |
} |
e9685a03c kernel/cgroup.c: ... |
163 |
static int notify_on_release(const struct cgroup *cgrp) |
81a6a5cdd Task Control Grou... |
164 |
{ |
bd89aabc6 Control groups: R... |
165 |
return test_bit(CGRP_NOTIFY_ON_RELEASE, &cgrp->flags); |
81a6a5cdd Task Control Grou... |
166 |
} |
ddbcc7e8e Task Control Grou... |
167 168 169 170 171 172 |
/* * for_each_subsys() allows you to iterate on each subsystem attached to * an active hierarchy */ #define for_each_subsys(_root, _ss) \ list_for_each_entry(_ss, &_root->subsys_list, sibling) |
e5f6a8609 cgroups: make roo... |
173 174 |
/* for_each_active_root() allows you to iterate across the active hierarchies */ #define for_each_active_root(_root) \ |
ddbcc7e8e Task Control Grou... |
175 |
list_for_each_entry(_root, &roots, root_list) |
81a6a5cdd Task Control Grou... |
176 177 178 179 180 181 |
/* the list of cgroups eligible for automatic release. Protected by * release_list_lock */ static LIST_HEAD(release_list); static DEFINE_SPINLOCK(release_list_lock); static void cgroup_release_agent(struct work_struct *work); static DECLARE_WORK(release_agent_work, cgroup_release_agent); |
bd89aabc6 Control groups: R... |
182 |
static void check_for_release(struct cgroup *cgrp); |
81a6a5cdd Task Control Grou... |
183 |
|
817929ec2 Task Control Grou... |
184 185 186 187 188 189 |
/* Link structure for associating css_set objects with cgroups */ struct cg_cgroup_link { /* * List running through cg_cgroup_links associated with a * cgroup, anchored on cgroup->css_sets */ |
bd89aabc6 Control groups: R... |
190 |
struct list_head cgrp_link_list; |
817929ec2 Task Control Grou... |
191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 |
/* * List running through cg_cgroup_links pointing at a * single css_set object, anchored on css_set->cg_links */ struct list_head cg_link_list; struct css_set *cg; }; /* The default css_set - used by init and its children prior to any * hierarchies being mounted. It contains a pointer to the root state * for each subsystem. Also used to anchor the list of css_sets. Not * reference-counted, to improve performance when child cgroups * haven't been created. */ static struct css_set init_css_set; static struct cg_cgroup_link init_css_set_link; |
38460b48d cgroup: CSS ID su... |
208 |
static int cgroup_subsys_init_idr(struct cgroup_subsys *ss); |
817929ec2 Task Control Grou... |
209 210 211 212 213 |
/* css_set_lock protects the list of css_set objects, and the * chain of tasks off each css_set. Nests outside task->alloc_lock * due to cgroup_iter_start() */ static DEFINE_RWLOCK(css_set_lock); static int css_set_count; |
472b1053f cgroups: use a ha... |
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 |
/* hash table for cgroup groups. This improves the performance to * find an existing css_set */ #define CSS_SET_HASH_BITS 7 #define CSS_SET_TABLE_SIZE (1 << CSS_SET_HASH_BITS) static struct hlist_head css_set_table[CSS_SET_TABLE_SIZE]; static struct hlist_head *css_set_hash(struct cgroup_subsys_state *css[]) { int i; int index; unsigned long tmp = 0UL; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) tmp += (unsigned long)css[i]; tmp = (tmp >> 16) ^ tmp; index = hash_long(tmp, CSS_SET_HASH_BITS); return &css_set_table[index]; } |
817929ec2 Task Control Grou... |
234 235 236 237 |
/* We don't maintain the lists running through each css_set to its * task until after the first call to cgroup_iter_start(). This * reduces the fork()/exit() overhead for people who have cgroups * compiled into their kernel but not actually in use */ |
8947f9d5b cgroups: annotate... |
238 |
static int use_task_css_set_links __read_mostly; |
817929ec2 Task Control Grou... |
239 240 241 242 243 244 245 |
/* When we create or destroy a css_set, the operation simply * takes/releases a reference count on all the cgroups referenced * by subsystems in this css_set. This can end up multiple-counting * some cgroups, but that's OK - the ref-count is just a * busy/not-busy indicator; ensuring that we only count each cgroup * once would require taking a global lock to ensure that no |
b4f48b636 Task Control Grou... |
246 247 248 249 250 251 252 |
* subsystems moved between hierarchies while we were doing so. * * Possible TODO: decide at boot time based on the number of * registered subsystems and the number of CPUs or NUMA nodes whether * it's better for performance to ref-count every subsystem, or to * take a global lock and only add one ref count to each hierarchy. */ |
817929ec2 Task Control Grou... |
253 254 255 256 |
/* * unlink a css_set from the list and free it */ |
81a6a5cdd Task Control Grou... |
257 |
static void unlink_css_set(struct css_set *cg) |
b4f48b636 Task Control Grou... |
258 |
{ |
71cbb949d cgroup: list_for_... |
259 260 |
struct cg_cgroup_link *link; struct cg_cgroup_link *saved_link; |
472b1053f cgroups: use a ha... |
261 |
hlist_del(&cg->hlist); |
817929ec2 Task Control Grou... |
262 |
css_set_count--; |
71cbb949d cgroup: list_for_... |
263 264 265 |
list_for_each_entry_safe(link, saved_link, &cg->cg_links, cg_link_list) { |
817929ec2 Task Control Grou... |
266 |
list_del(&link->cg_link_list); |
bd89aabc6 Control groups: R... |
267 |
list_del(&link->cgrp_link_list); |
817929ec2 Task Control Grou... |
268 269 |
kfree(link); } |
81a6a5cdd Task Control Grou... |
270 |
} |
146aa1bd0 cgroups: fix prob... |
271 |
static void __put_css_set(struct css_set *cg, int taskexit) |
81a6a5cdd Task Control Grou... |
272 273 |
{ int i; |
146aa1bd0 cgroups: fix prob... |
274 275 276 277 278 279 280 281 282 283 284 285 |
/* * Ensure that the refcount doesn't hit zero while any readers * can see it. Similar to atomic_dec_and_lock(), but for an * rwlock */ if (atomic_add_unless(&cg->refcount, -1, 1)) return; write_lock(&css_set_lock); if (!atomic_dec_and_test(&cg->refcount)) { write_unlock(&css_set_lock); return; } |
81a6a5cdd Task Control Grou... |
286 |
unlink_css_set(cg); |
146aa1bd0 cgroups: fix prob... |
287 |
write_unlock(&css_set_lock); |
81a6a5cdd Task Control Grou... |
288 289 290 |
rcu_read_lock(); for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { |
a47295e6b cgroups: make cgr... |
291 |
struct cgroup *cgrp = rcu_dereference(cg->subsys[i]->cgroup); |
bd89aabc6 Control groups: R... |
292 293 |
if (atomic_dec_and_test(&cgrp->count) && notify_on_release(cgrp)) { |
81a6a5cdd Task Control Grou... |
294 |
if (taskexit) |
bd89aabc6 Control groups: R... |
295 296 |
set_bit(CGRP_RELEASABLE, &cgrp->flags); check_for_release(cgrp); |
81a6a5cdd Task Control Grou... |
297 298 299 |
} } rcu_read_unlock(); |
817929ec2 Task Control Grou... |
300 |
kfree(cg); |
b4f48b636 Task Control Grou... |
301 |
} |
817929ec2 Task Control Grou... |
302 303 304 305 306 |
/* * refcounted get/put for css_set objects */ static inline void get_css_set(struct css_set *cg) { |
146aa1bd0 cgroups: fix prob... |
307 |
atomic_inc(&cg->refcount); |
817929ec2 Task Control Grou... |
308 309 310 311 |
} static inline void put_css_set(struct css_set *cg) { |
146aa1bd0 cgroups: fix prob... |
312 |
__put_css_set(cg, 0); |
817929ec2 Task Control Grou... |
313 |
} |
81a6a5cdd Task Control Grou... |
314 315 |
static inline void put_css_set_taskexit(struct css_set *cg) { |
146aa1bd0 cgroups: fix prob... |
316 |
__put_css_set(cg, 1); |
81a6a5cdd Task Control Grou... |
317 |
} |
817929ec2 Task Control Grou... |
318 319 320 |
/* * find_existing_css_set() is a helper for * find_css_set(), and checks to see whether an existing |
472b1053f cgroups: use a ha... |
321 |
* css_set is suitable. |
817929ec2 Task Control Grou... |
322 323 324 325 |
* * oldcg: the cgroup group that we're using before the cgroup * transition * |
bd89aabc6 Control groups: R... |
326 |
* cgrp: the cgroup that we're moving into |
817929ec2 Task Control Grou... |
327 328 329 330 |
* * template: location in which to build the desired set of subsystem * state objects for the new cgroup group */ |
817929ec2 Task Control Grou... |
331 332 |
static struct css_set *find_existing_css_set( struct css_set *oldcg, |
bd89aabc6 Control groups: R... |
333 |
struct cgroup *cgrp, |
817929ec2 Task Control Grou... |
334 |
struct cgroup_subsys_state *template[]) |
b4f48b636 Task Control Grou... |
335 336 |
{ int i; |
bd89aabc6 Control groups: R... |
337 |
struct cgroupfs_root *root = cgrp->root; |
472b1053f cgroups: use a ha... |
338 339 340 |
struct hlist_head *hhead; struct hlist_node *node; struct css_set *cg; |
817929ec2 Task Control Grou... |
341 342 343 344 |
/* Built the set of subsystem state objects that we want to * see in the new css_set */ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { |
8d53d55d2 cgroup: fix subsy... |
345 |
if (root->subsys_bits & (1UL << i)) { |
817929ec2 Task Control Grou... |
346 347 348 |
/* Subsystem is in this hierarchy. So we want * the subsystem state from the new * cgroup */ |
bd89aabc6 Control groups: R... |
349 |
template[i] = cgrp->subsys[i]; |
817929ec2 Task Control Grou... |
350 351 352 353 354 355 |
} else { /* Subsystem is not in this hierarchy, so we * don't want to change the subsystem state */ template[i] = oldcg->subsys[i]; } } |
472b1053f cgroups: use a ha... |
356 357 |
hhead = css_set_hash(template); hlist_for_each_entry(cg, node, hhead, hlist) { |
817929ec2 Task Control Grou... |
358 359 360 361 |
if (!memcmp(template, cg->subsys, sizeof(cg->subsys))) { /* All subsystems matched */ return cg; } |
472b1053f cgroups: use a ha... |
362 |
} |
817929ec2 Task Control Grou... |
363 364 365 366 |
/* No existing cgroup group matched */ return NULL; } |
36553434f cgroup: remove du... |
367 368 369 370 371 372 373 374 375 376 |
static void free_cg_links(struct list_head *tmp) { struct cg_cgroup_link *link; struct cg_cgroup_link *saved_link; list_for_each_entry_safe(link, saved_link, tmp, cgrp_link_list) { list_del(&link->cgrp_link_list); kfree(link); } } |
817929ec2 Task Control Grou... |
377 378 |
/* * allocate_cg_links() allocates "count" cg_cgroup_link structures |
bd89aabc6 Control groups: R... |
379 |
* and chains them on tmp through their cgrp_link_list fields. Returns 0 on |
817929ec2 Task Control Grou... |
380 381 |
* success or a negative error */ |
817929ec2 Task Control Grou... |
382 383 384 385 386 387 388 389 |
static int allocate_cg_links(int count, struct list_head *tmp) { struct cg_cgroup_link *link; int i; INIT_LIST_HEAD(tmp); for (i = 0; i < count; i++) { link = kmalloc(sizeof(*link), GFP_KERNEL); if (!link) { |
36553434f cgroup: remove du... |
390 |
free_cg_links(tmp); |
817929ec2 Task Control Grou... |
391 392 |
return -ENOMEM; } |
bd89aabc6 Control groups: R... |
393 |
list_add(&link->cgrp_link_list, tmp); |
817929ec2 Task Control Grou... |
394 395 396 |
} return 0; } |
c12f65d43 cgroups: introduc... |
397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 |
/** * link_css_set - a helper function to link a css_set to a cgroup * @tmp_cg_links: cg_cgroup_link objects allocated by allocate_cg_links() * @cg: the css_set to be linked * @cgrp: the destination cgroup */ static void link_css_set(struct list_head *tmp_cg_links, struct css_set *cg, struct cgroup *cgrp) { struct cg_cgroup_link *link; BUG_ON(list_empty(tmp_cg_links)); link = list_first_entry(tmp_cg_links, struct cg_cgroup_link, cgrp_link_list); link->cg = cg; list_move(&link->cgrp_link_list, &cgrp->css_sets); list_add(&link->cg_link_list, &cg->cg_links); } |
817929ec2 Task Control Grou... |
415 416 417 418 419 420 421 |
/* * find_css_set() takes an existing cgroup group and a * cgroup object, and returns a css_set object that's * equivalent to the old group, but with the given cgroup * substituted into the appropriate hierarchy. Must be called with * cgroup_mutex held */ |
817929ec2 Task Control Grou... |
422 |
static struct css_set *find_css_set( |
bd89aabc6 Control groups: R... |
423 |
struct css_set *oldcg, struct cgroup *cgrp) |
817929ec2 Task Control Grou... |
424 425 426 427 428 429 |
{ struct css_set *res; struct cgroup_subsys_state *template[CGROUP_SUBSYS_COUNT]; int i; struct list_head tmp_cg_links; |
817929ec2 Task Control Grou... |
430 |
|
472b1053f cgroups: use a ha... |
431 |
struct hlist_head *hhead; |
817929ec2 Task Control Grou... |
432 433 |
/* First see if we already have a cgroup group that matches * the desired set */ |
7e9abd89c cgroup: use read ... |
434 |
read_lock(&css_set_lock); |
bd89aabc6 Control groups: R... |
435 |
res = find_existing_css_set(oldcg, cgrp, template); |
817929ec2 Task Control Grou... |
436 437 |
if (res) get_css_set(res); |
7e9abd89c cgroup: use read ... |
438 |
read_unlock(&css_set_lock); |
817929ec2 Task Control Grou... |
439 440 441 442 443 444 445 446 447 448 449 450 451 |
if (res) return res; res = kmalloc(sizeof(*res), GFP_KERNEL); if (!res) return NULL; /* Allocate all the cg_cgroup_link objects that we'll need */ if (allocate_cg_links(root_count, &tmp_cg_links) < 0) { kfree(res); return NULL; } |
146aa1bd0 cgroups: fix prob... |
452 |
atomic_set(&res->refcount, 1); |
817929ec2 Task Control Grou... |
453 454 |
INIT_LIST_HEAD(&res->cg_links); INIT_LIST_HEAD(&res->tasks); |
472b1053f cgroups: use a ha... |
455 |
INIT_HLIST_NODE(&res->hlist); |
817929ec2 Task Control Grou... |
456 457 458 459 460 461 462 463 |
/* Copy the set of subsystem state objects generated in * find_existing_css_set() */ memcpy(res->subsys, template, sizeof(res->subsys)); write_lock(&css_set_lock); /* Add reference counts and links from the new css_set. */ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { |
bd89aabc6 Control groups: R... |
464 |
struct cgroup *cgrp = res->subsys[i]->cgroup; |
817929ec2 Task Control Grou... |
465 |
struct cgroup_subsys *ss = subsys[i]; |
bd89aabc6 Control groups: R... |
466 |
atomic_inc(&cgrp->count); |
817929ec2 Task Control Grou... |
467 468 469 470 471 |
/* * We want to add a link once per cgroup, so we * only do it for the first subsystem in each * hierarchy */ |
c12f65d43 cgroups: introduc... |
472 473 |
if (ss->root->subsys_list.next == &ss->sibling) link_css_set(&tmp_cg_links, res, cgrp); |
817929ec2 Task Control Grou... |
474 |
} |
c12f65d43 cgroups: introduc... |
475 476 |
if (list_empty(&rootnode.subsys_list)) link_css_set(&tmp_cg_links, res, dummytop); |
817929ec2 Task Control Grou... |
477 478 |
BUG_ON(!list_empty(&tmp_cg_links)); |
817929ec2 Task Control Grou... |
479 |
css_set_count++; |
472b1053f cgroups: use a ha... |
480 481 482 483 |
/* Add this cgroup group to the hash table */ hhead = css_set_hash(res->subsys); hlist_add_head(&res->hlist, hhead); |
817929ec2 Task Control Grou... |
484 485 486 |
write_unlock(&css_set_lock); return res; |
b4f48b636 Task Control Grou... |
487 |
} |
ddbcc7e8e Task Control Grou... |
488 489 490 491 492 493 494 495 496 497 |
/* * There is one global cgroup mutex. We also require taking * task_lock() when dereferencing a task's cgroup subsys pointers. * See "The task_lock() exception", at the end of this comment. * * A task must hold cgroup_mutex to modify cgroups. * * Any task can increment and decrement the count field without lock. * So in general, code holding cgroup_mutex can't rely on the count * field not changing. However, if the count goes to zero, then only |
956db3ca0 hotplug cpu: move... |
498 |
* cgroup_attach_task() can increment it again. Because a count of zero |
ddbcc7e8e Task Control Grou... |
499 500 501 502 503 504 505 506 |
* means that no tasks are currently attached, therefore there is no * way a task attached to that cgroup can fork (the other way to * increment the count). So code holding cgroup_mutex can safely * assume that if the count is zero, it will stay zero. Similarly, if * a task holds cgroup_mutex on a cgroup with zero count, it * knows that the cgroup won't be removed, as cgroup_rmdir() * needs that mutex. * |
ddbcc7e8e Task Control Grou... |
507 508 509 510 511 |
* The fork and exit callbacks cgroup_fork() and cgroup_exit(), don't * (usually) take cgroup_mutex. These are the two most performance * critical pieces of code here. The exception occurs on cgroup_exit(), * when a task in a notify_on_release cgroup exits. Then cgroup_mutex * is taken, and if the cgroup count is zero, a usermode call made |
a043e3b2c cgroup: fix comments |
512 513 |
* to the release agent with the name of the cgroup (path relative to * the root of cgroup file system) as the argument. |
ddbcc7e8e Task Control Grou... |
514 515 516 517 518 519 520 521 522 523 524 |
* * A cgroup can only be deleted if both its 'count' of using tasks * is zero, and its list of 'children' cgroups is empty. Since all * tasks in the system use _some_ cgroup, and since there is always at * least one task in the system (init, pid == 1), therefore, top_cgroup * always has either children cgroups and/or using tasks. So we don't * need a special hack to ensure that top_cgroup cannot be deleted. * * The task_lock() exception * * The need for this exception arises from the action of |
956db3ca0 hotplug cpu: move... |
525 |
* cgroup_attach_task(), which overwrites one tasks cgroup pointer with |
a043e3b2c cgroup: fix comments |
526 |
* another. It does so using cgroup_mutex, however there are |
ddbcc7e8e Task Control Grou... |
527 528 529 |
* several performance critical places that need to reference * task->cgroup without the expense of grabbing a system global * mutex. Therefore except as noted below, when dereferencing or, as |
956db3ca0 hotplug cpu: move... |
530 |
* in cgroup_attach_task(), modifying a task'ss cgroup pointer we use |
ddbcc7e8e Task Control Grou... |
531 532 533 534 |
* task_lock(), which acts on a spinlock (task->alloc_lock) already in * the task_struct routinely used for such matters. * * P.S. One more locking exception. RCU is used to guard the |
956db3ca0 hotplug cpu: move... |
535 |
* update of a tasks cgroup pointer by cgroup_attach_task() |
ddbcc7e8e Task Control Grou... |
536 |
*/ |
ddbcc7e8e Task Control Grou... |
537 538 539 540 |
/** * cgroup_lock - lock out any changes to cgroup structures * */ |
ddbcc7e8e Task Control Grou... |
541 542 543 544 545 546 547 548 549 550 |
void cgroup_lock(void) { mutex_lock(&cgroup_mutex); } /** * cgroup_unlock - release lock on cgroup changes * * Undo the lock taken in a previous cgroup_lock() call. */ |
ddbcc7e8e Task Control Grou... |
551 552 553 554 555 556 557 558 559 560 561 562 563 564 |
void cgroup_unlock(void) { mutex_unlock(&cgroup_mutex); } /* * A couple of forward declarations required, due to cyclic reference loop: * cgroup_mkdir -> cgroup_create -> cgroup_populate_dir -> * cgroup_add_file -> cgroup_create_file -> cgroup_dir_inode_operations * -> cgroup_mkdir. */ static int cgroup_mkdir(struct inode *dir, struct dentry *dentry, int mode); static int cgroup_rmdir(struct inode *unused_dir, struct dentry *dentry); |
bd89aabc6 Control groups: R... |
565 |
static int cgroup_populate_dir(struct cgroup *cgrp); |
ddbcc7e8e Task Control Grou... |
566 |
static struct inode_operations cgroup_dir_inode_operations; |
a424316ca Task Control Grou... |
567 568 569 |
static struct file_operations proc_cgroupstats_operations; static struct backing_dev_info cgroup_backing_dev_info = { |
e4ad08fe6 mm: bdi: add sepa... |
570 |
.capabilities = BDI_CAP_NO_ACCT_AND_WRITEBACK, |
a424316ca Task Control Grou... |
571 |
}; |
ddbcc7e8e Task Control Grou... |
572 |
|
38460b48d cgroup: CSS ID su... |
573 574 |
static int alloc_css_id(struct cgroup_subsys *ss, struct cgroup *parent, struct cgroup *child); |
ddbcc7e8e Task Control Grou... |
575 576 577 |
static struct inode *cgroup_new_inode(mode_t mode, struct super_block *sb) { struct inode *inode = new_inode(sb); |
ddbcc7e8e Task Control Grou... |
578 579 580 |
if (inode) { inode->i_mode = mode; |
76aac0e9a CRED: Wrap task c... |
581 582 |
inode->i_uid = current_fsuid(); inode->i_gid = current_fsgid(); |
ddbcc7e8e Task Control Grou... |
583 584 585 586 587 |
inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; inode->i_mapping->backing_dev_info = &cgroup_backing_dev_info; } return inode; } |
4fca88c87 memory cgroup enh... |
588 589 590 591 |
/* * Call subsys's pre_destroy handler. * This is called before css refcnt check. */ |
ec64f5154 cgroup: fix frequ... |
592 |
static int cgroup_call_pre_destroy(struct cgroup *cgrp) |
4fca88c87 memory cgroup enh... |
593 594 |
{ struct cgroup_subsys *ss; |
ec64f5154 cgroup: fix frequ... |
595 |
int ret = 0; |
4fca88c87 memory cgroup enh... |
596 |
for_each_subsys(cgrp->root, ss) |
ec64f5154 cgroup: fix frequ... |
597 598 599 600 601 602 |
if (ss->pre_destroy) { ret = ss->pre_destroy(ss, cgrp); if (ret) break; } return ret; |
4fca88c87 memory cgroup enh... |
603 |
} |
a47295e6b cgroups: make cgr... |
604 605 606 607 608 609 |
static void free_cgroup_rcu(struct rcu_head *obj) { struct cgroup *cgrp = container_of(obj, struct cgroup, rcu_head); kfree(cgrp); } |
ddbcc7e8e Task Control Grou... |
610 611 612 613 |
static void cgroup_diput(struct dentry *dentry, struct inode *inode) { /* is dentry a directory ? if so, kfree() associated cgroup */ if (S_ISDIR(inode->i_mode)) { |
bd89aabc6 Control groups: R... |
614 |
struct cgroup *cgrp = dentry->d_fsdata; |
8dc4f3e17 cgroups: move cgr... |
615 |
struct cgroup_subsys *ss; |
bd89aabc6 Control groups: R... |
616 |
BUG_ON(!(cgroup_is_removed(cgrp))); |
81a6a5cdd Task Control Grou... |
617 618 619 620 621 622 623 |
/* It's possible for external users to be holding css * reference counts on a cgroup; css_put() needs to * be able to access the cgroup after decrementing * the reference count in order to know if it needs to * queue the cgroup to be handled by the release * agent */ synchronize_rcu(); |
8dc4f3e17 cgroups: move cgr... |
624 625 626 627 628 |
mutex_lock(&cgroup_mutex); /* * Release the subsystem state objects. */ |
75139b827 cgroups: remove s... |
629 630 |
for_each_subsys(cgrp->root, ss) ss->destroy(ss, cgrp); |
8dc4f3e17 cgroups: move cgr... |
631 632 633 |
cgrp->root->number_of_cgroups--; mutex_unlock(&cgroup_mutex); |
a47295e6b cgroups: make cgr... |
634 635 636 637 |
/* * Drop the active superblock reference that we took when we * created the cgroup */ |
8dc4f3e17 cgroups: move cgr... |
638 |
deactivate_super(cgrp->root->sb); |
a47295e6b cgroups: make cgr... |
639 |
call_rcu(&cgrp->rcu_head, free_cgroup_rcu); |
ddbcc7e8e Task Control Grou... |
640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 |
} iput(inode); } static void remove_dir(struct dentry *d) { struct dentry *parent = dget(d->d_parent); d_delete(d); simple_rmdir(parent->d_inode, d); dput(parent); } static void cgroup_clear_directory(struct dentry *dentry) { struct list_head *node; BUG_ON(!mutex_is_locked(&dentry->d_inode->i_mutex)); spin_lock(&dcache_lock); node = dentry->d_subdirs.next; while (node != &dentry->d_subdirs) { struct dentry *d = list_entry(node, struct dentry, d_u.d_child); list_del_init(node); if (d->d_inode) { /* This should never be called on a cgroup * directory with child cgroups */ BUG_ON(d->d_inode->i_mode & S_IFDIR); d = dget_locked(d); spin_unlock(&dcache_lock); d_delete(d); simple_unlink(dentry->d_inode, d); dput(d); spin_lock(&dcache_lock); } node = dentry->d_subdirs.next; } spin_unlock(&dcache_lock); } /* * NOTE : the dentry must have been dget()'ed */ static void cgroup_d_remove_dir(struct dentry *dentry) { cgroup_clear_directory(dentry); spin_lock(&dcache_lock); list_del_init(&dentry->d_u.d_child); spin_unlock(&dcache_lock); remove_dir(dentry); } |
ec64f5154 cgroup: fix frequ... |
691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 |
/* * A queue for waiters to do rmdir() cgroup. A tasks will sleep when * cgroup->count == 0 && list_empty(&cgroup->children) && subsys has some * reference to css->refcnt. In general, this refcnt is expected to goes down * to zero, soon. * * CGRP_WAIT_ON_RMDIR flag is modified under cgroup's inode->i_mutex; */ DECLARE_WAIT_QUEUE_HEAD(cgroup_rmdir_waitq); static void cgroup_wakeup_rmdir_waiters(const struct cgroup *cgrp) { if (unlikely(test_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags))) wake_up_all(&cgroup_rmdir_waitq); } |
ddbcc7e8e Task Control Grou... |
706 707 708 709 |
static int rebind_subsystems(struct cgroupfs_root *root, unsigned long final_bits) { unsigned long added_bits, removed_bits; |
bd89aabc6 Control groups: R... |
710 |
struct cgroup *cgrp = &root->top_cgroup; |
ddbcc7e8e Task Control Grou... |
711 712 713 714 715 716 |
int i; removed_bits = root->actual_subsys_bits & ~final_bits; added_bits = final_bits & ~root->actual_subsys_bits; /* Check that any added subsystems are currently free */ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { |
8d53d55d2 cgroup: fix subsy... |
717 |
unsigned long bit = 1UL << i; |
ddbcc7e8e Task Control Grou... |
718 719 720 721 722 723 724 725 726 727 728 729 730 |
struct cgroup_subsys *ss = subsys[i]; if (!(bit & added_bits)) continue; if (ss->root != &rootnode) { /* Subsystem isn't free */ return -EBUSY; } } /* Currently we don't handle adding/removing subsystems when * any child cgroups exist. This is theoretically supportable * but involves complex error handling, so it's being left until * later */ |
307257cf4 cgroups: fix a ra... |
731 |
if (root->number_of_cgroups > 1) |
ddbcc7e8e Task Control Grou... |
732 733 734 735 736 737 738 739 |
return -EBUSY; /* Process each subsystem */ for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; unsigned long bit = 1UL << i; if (bit & added_bits) { /* We're binding this subsystem to this hierarchy */ |
bd89aabc6 Control groups: R... |
740 |
BUG_ON(cgrp->subsys[i]); |
ddbcc7e8e Task Control Grou... |
741 742 |
BUG_ON(!dummytop->subsys[i]); BUG_ON(dummytop->subsys[i]->cgroup != dummytop); |
999cd8a45 cgroups: add a pe... |
743 |
mutex_lock(&ss->hierarchy_mutex); |
bd89aabc6 Control groups: R... |
744 745 |
cgrp->subsys[i] = dummytop->subsys[i]; cgrp->subsys[i]->cgroup = cgrp; |
33a68ac1c cgroups: add inac... |
746 |
list_move(&ss->sibling, &root->subsys_list); |
b2aa30f7b cgroups: don't pu... |
747 |
ss->root = root; |
ddbcc7e8e Task Control Grou... |
748 |
if (ss->bind) |
bd89aabc6 Control groups: R... |
749 |
ss->bind(ss, cgrp); |
999cd8a45 cgroups: add a pe... |
750 |
mutex_unlock(&ss->hierarchy_mutex); |
ddbcc7e8e Task Control Grou... |
751 752 |
} else if (bit & removed_bits) { /* We're removing this subsystem */ |
bd89aabc6 Control groups: R... |
753 754 |
BUG_ON(cgrp->subsys[i] != dummytop->subsys[i]); BUG_ON(cgrp->subsys[i]->cgroup != cgrp); |
999cd8a45 cgroups: add a pe... |
755 |
mutex_lock(&ss->hierarchy_mutex); |
ddbcc7e8e Task Control Grou... |
756 757 758 |
if (ss->bind) ss->bind(ss, dummytop); dummytop->subsys[i]->cgroup = dummytop; |
bd89aabc6 Control groups: R... |
759 |
cgrp->subsys[i] = NULL; |
b2aa30f7b cgroups: don't pu... |
760 |
subsys[i]->root = &rootnode; |
33a68ac1c cgroups: add inac... |
761 |
list_move(&ss->sibling, &rootnode.subsys_list); |
999cd8a45 cgroups: add a pe... |
762 |
mutex_unlock(&ss->hierarchy_mutex); |
ddbcc7e8e Task Control Grou... |
763 764 |
} else if (bit & final_bits) { /* Subsystem state should already exist */ |
bd89aabc6 Control groups: R... |
765 |
BUG_ON(!cgrp->subsys[i]); |
ddbcc7e8e Task Control Grou... |
766 767 |
} else { /* Subsystem state shouldn't exist */ |
bd89aabc6 Control groups: R... |
768 |
BUG_ON(cgrp->subsys[i]); |
ddbcc7e8e Task Control Grou... |
769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 |
} } root->subsys_bits = root->actual_subsys_bits = final_bits; synchronize_rcu(); return 0; } static int cgroup_show_options(struct seq_file *seq, struct vfsmount *vfs) { struct cgroupfs_root *root = vfs->mnt_sb->s_fs_info; struct cgroup_subsys *ss; mutex_lock(&cgroup_mutex); for_each_subsys(root, ss) seq_printf(seq, ",%s", ss->name); if (test_bit(ROOT_NOPREFIX, &root->flags)) seq_puts(seq, ",noprefix"); |
81a6a5cdd Task Control Grou... |
787 788 |
if (strlen(root->release_agent_path)) seq_printf(seq, ",release_agent=%s", root->release_agent_path); |
ddbcc7e8e Task Control Grou... |
789 790 791 792 793 794 795 |
mutex_unlock(&cgroup_mutex); return 0; } struct cgroup_sb_opts { unsigned long subsys_bits; unsigned long flags; |
81a6a5cdd Task Control Grou... |
796 |
char *release_agent; |
ddbcc7e8e Task Control Grou... |
797 798 799 800 801 802 803 804 |
}; /* Convert a hierarchy specifier into a bitmask of subsystems and * flags. */ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts) { char *token, *o = data ?: "all"; |
f9ab5b5b0 cgroups: forbid n... |
805 806 807 808 809 |
unsigned long mask = (unsigned long)-1; #ifdef CONFIG_CPUSETS mask = ~(1UL << cpuset_subsys_id); #endif |
ddbcc7e8e Task Control Grou... |
810 811 812 |
opts->subsys_bits = 0; opts->flags = 0; |
81a6a5cdd Task Control Grou... |
813 |
opts->release_agent = NULL; |
ddbcc7e8e Task Control Grou... |
814 815 816 817 818 |
while ((token = strsep(&o, ",")) != NULL) { if (!*token) return -EINVAL; if (!strcmp(token, "all")) { |
8bab8dded cgroups: add cgro... |
819 820 821 822 823 824 825 826 |
/* Add all non-disabled subsystems */ int i; opts->subsys_bits = 0; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; if (!ss->disabled) opts->subsys_bits |= 1ul << i; } |
ddbcc7e8e Task Control Grou... |
827 828 |
} else if (!strcmp(token, "noprefix")) { set_bit(ROOT_NOPREFIX, &opts->flags); |
81a6a5cdd Task Control Grou... |
829 830 831 832 833 834 835 836 837 |
} else if (!strncmp(token, "release_agent=", 14)) { /* Specifying two release agents is forbidden */ if (opts->release_agent) return -EINVAL; opts->release_agent = kzalloc(PATH_MAX, GFP_KERNEL); if (!opts->release_agent) return -ENOMEM; strncpy(opts->release_agent, token + 14, PATH_MAX - 1); opts->release_agent[PATH_MAX - 1] = 0; |
ddbcc7e8e Task Control Grou... |
838 839 840 841 842 843 |
} else { struct cgroup_subsys *ss; int i; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { ss = subsys[i]; if (!strcmp(token, ss->name)) { |
8bab8dded cgroups: add cgro... |
844 845 |
if (!ss->disabled) set_bit(i, &opts->subsys_bits); |
ddbcc7e8e Task Control Grou... |
846 847 848 849 850 851 852 |
break; } } if (i == CGROUP_SUBSYS_COUNT) return -ENOENT; } } |
f9ab5b5b0 cgroups: forbid n... |
853 854 855 856 857 858 859 860 |
/* * Option noprefix was introduced just for backward compatibility * with the old cpuset, so we allow noprefix only if mounting just * the cpuset subsystem. */ if (test_bit(ROOT_NOPREFIX, &opts->flags) && (opts->subsys_bits & mask)) return -EINVAL; |
ddbcc7e8e Task Control Grou... |
861 862 863 864 865 866 867 868 869 870 871 |
/* We can't have an empty hierarchy */ if (!opts->subsys_bits) return -EINVAL; return 0; } static int cgroup_remount(struct super_block *sb, int *flags, char *data) { int ret = 0; struct cgroupfs_root *root = sb->s_fs_info; |
bd89aabc6 Control groups: R... |
872 |
struct cgroup *cgrp = &root->top_cgroup; |
ddbcc7e8e Task Control Grou... |
873 |
struct cgroup_sb_opts opts; |
337eb00a2 Push BKL down int... |
874 |
lock_kernel(); |
bd89aabc6 Control groups: R... |
875 |
mutex_lock(&cgrp->dentry->d_inode->i_mutex); |
ddbcc7e8e Task Control Grou... |
876 877 878 879 880 881 882 883 884 885 886 887 888 889 |
mutex_lock(&cgroup_mutex); /* See what subsystems are wanted */ ret = parse_cgroupfs_options(data, &opts); if (ret) goto out_unlock; /* Don't allow flags to change at remount */ if (opts.flags != root->flags) { ret = -EINVAL; goto out_unlock; } ret = rebind_subsystems(root, opts.subsys_bits); |
0670e08bd cgroups: don't ch... |
890 891 |
if (ret) goto out_unlock; |
ddbcc7e8e Task Control Grou... |
892 893 |
/* (re)populate subsystem files */ |
0670e08bd cgroups: don't ch... |
894 |
cgroup_populate_dir(cgrp); |
ddbcc7e8e Task Control Grou... |
895 |
|
81a6a5cdd Task Control Grou... |
896 897 |
if (opts.release_agent) strcpy(root->release_agent_path, opts.release_agent); |
ddbcc7e8e Task Control Grou... |
898 |
out_unlock: |
66bdc9cfc kernel/cgroup.c: ... |
899 |
kfree(opts.release_agent); |
ddbcc7e8e Task Control Grou... |
900 |
mutex_unlock(&cgroup_mutex); |
bd89aabc6 Control groups: R... |
901 |
mutex_unlock(&cgrp->dentry->d_inode->i_mutex); |
337eb00a2 Push BKL down int... |
902 |
unlock_kernel(); |
ddbcc7e8e Task Control Grou... |
903 904 905 906 907 908 909 910 911 |
return ret; } static struct super_operations cgroup_ops = { .statfs = simple_statfs, .drop_inode = generic_delete_inode, .show_options = cgroup_show_options, .remount_fs = cgroup_remount, }; |
cc31edcee cgroups: convert ... |
912 913 914 915 916 917 918 919 |
static void init_cgroup_housekeeping(struct cgroup *cgrp) { INIT_LIST_HEAD(&cgrp->sibling); INIT_LIST_HEAD(&cgrp->children); INIT_LIST_HEAD(&cgrp->css_sets); INIT_LIST_HEAD(&cgrp->release_list); init_rwsem(&cgrp->pids_mutex); } |
ddbcc7e8e Task Control Grou... |
920 921 |
static void init_cgroup_root(struct cgroupfs_root *root) { |
bd89aabc6 Control groups: R... |
922 |
struct cgroup *cgrp = &root->top_cgroup; |
ddbcc7e8e Task Control Grou... |
923 924 925 |
INIT_LIST_HEAD(&root->subsys_list); INIT_LIST_HEAD(&root->root_list); root->number_of_cgroups = 1; |
bd89aabc6 Control groups: R... |
926 927 |
cgrp->root = root; cgrp->top_cgroup = cgrp; |
cc31edcee cgroups: convert ... |
928 |
init_cgroup_housekeeping(cgrp); |
ddbcc7e8e Task Control Grou... |
929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 |
} static int cgroup_test_super(struct super_block *sb, void *data) { struct cgroupfs_root *new = data; struct cgroupfs_root *root = sb->s_fs_info; /* First check subsystems */ if (new->subsys_bits != root->subsys_bits) return 0; /* Next check flags */ if (new->flags != root->flags) return 0; return 1; } static int cgroup_set_super(struct super_block *sb, void *data) { int ret; struct cgroupfs_root *root = data; ret = set_anon_super(sb, NULL); if (ret) return ret; sb->s_fs_info = root; root->sb = sb; sb->s_blocksize = PAGE_CACHE_SIZE; sb->s_blocksize_bits = PAGE_CACHE_SHIFT; sb->s_magic = CGROUP_SUPER_MAGIC; sb->s_op = &cgroup_ops; return 0; } static int cgroup_get_rootdir(struct super_block *sb) { struct inode *inode = cgroup_new_inode(S_IFDIR | S_IRUGO | S_IXUGO | S_IWUSR, sb); struct dentry *dentry; if (!inode) return -ENOMEM; |
ddbcc7e8e Task Control Grou... |
975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 |
inode->i_fop = &simple_dir_operations; inode->i_op = &cgroup_dir_inode_operations; /* directories start off with i_nlink == 2 (for "." entry) */ inc_nlink(inode); dentry = d_alloc_root(inode); if (!dentry) { iput(inode); return -ENOMEM; } sb->s_root = dentry; return 0; } static int cgroup_get_sb(struct file_system_type *fs_type, int flags, const char *unused_dev_name, void *data, struct vfsmount *mnt) { struct cgroup_sb_opts opts; int ret = 0; struct super_block *sb; struct cgroupfs_root *root; |
28fd5dfc1 cgroups: remove t... |
996 |
struct list_head tmp_cg_links; |
ddbcc7e8e Task Control Grou... |
997 998 999 |
/* First find the desired set of subsystems */ ret = parse_cgroupfs_options(data, &opts); |
81a6a5cdd Task Control Grou... |
1000 |
if (ret) { |
66bdc9cfc kernel/cgroup.c: ... |
1001 |
kfree(opts.release_agent); |
ddbcc7e8e Task Control Grou... |
1002 |
return ret; |
81a6a5cdd Task Control Grou... |
1003 |
} |
ddbcc7e8e Task Control Grou... |
1004 1005 |
root = kzalloc(sizeof(*root), GFP_KERNEL); |
f77707384 cgroup: fix memor... |
1006 |
if (!root) { |
66bdc9cfc kernel/cgroup.c: ... |
1007 |
kfree(opts.release_agent); |
ddbcc7e8e Task Control Grou... |
1008 |
return -ENOMEM; |
f77707384 cgroup: fix memor... |
1009 |
} |
ddbcc7e8e Task Control Grou... |
1010 1011 1012 1013 |
init_cgroup_root(root); root->subsys_bits = opts.subsys_bits; root->flags = opts.flags; |
81a6a5cdd Task Control Grou... |
1014 1015 1016 1017 |
if (opts.release_agent) { strcpy(root->release_agent_path, opts.release_agent); kfree(opts.release_agent); } |
ddbcc7e8e Task Control Grou... |
1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 |
sb = sget(fs_type, cgroup_test_super, cgroup_set_super, root); if (IS_ERR(sb)) { kfree(root); return PTR_ERR(sb); } if (sb->s_fs_info != root) { /* Reusing an existing superblock */ BUG_ON(sb->s_root == NULL); kfree(root); root = NULL; } else { /* New superblock */ |
c12f65d43 cgroups: introduc... |
1033 |
struct cgroup *root_cgrp = &root->top_cgroup; |
817929ec2 Task Control Grou... |
1034 |
struct inode *inode; |
28fd5dfc1 cgroups: remove t... |
1035 |
int i; |
ddbcc7e8e Task Control Grou... |
1036 1037 1038 1039 1040 1041 |
BUG_ON(sb->s_root != NULL); ret = cgroup_get_rootdir(sb); if (ret) goto drop_new_super; |
817929ec2 Task Control Grou... |
1042 |
inode = sb->s_root->d_inode; |
ddbcc7e8e Task Control Grou... |
1043 |
|
817929ec2 Task Control Grou... |
1044 |
mutex_lock(&inode->i_mutex); |
ddbcc7e8e Task Control Grou... |
1045 |
mutex_lock(&cgroup_mutex); |
817929ec2 Task Control Grou... |
1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 |
/* * We're accessing css_set_count without locking * css_set_lock here, but that's OK - it can only be * increased by someone holding cgroup_lock, and * that's us. The worst that can happen is that we * have some link structures left over */ ret = allocate_cg_links(css_set_count, &tmp_cg_links); if (ret) { mutex_unlock(&cgroup_mutex); mutex_unlock(&inode->i_mutex); goto drop_new_super; } |
ddbcc7e8e Task Control Grou... |
1059 1060 1061 |
ret = rebind_subsystems(root, root->subsys_bits); if (ret == -EBUSY) { mutex_unlock(&cgroup_mutex); |
817929ec2 Task Control Grou... |
1062 |
mutex_unlock(&inode->i_mutex); |
20ca9b3f4 cgroups: avoid ac... |
1063 |
goto free_cg_links; |
ddbcc7e8e Task Control Grou... |
1064 1065 1066 1067 1068 1069 |
} /* EBUSY should be the only error here */ BUG_ON(ret); list_add(&root->root_list, &roots); |
817929ec2 Task Control Grou... |
1070 |
root_count++; |
ddbcc7e8e Task Control Grou... |
1071 |
|
c12f65d43 cgroups: introduc... |
1072 |
sb->s_root->d_fsdata = root_cgrp; |
ddbcc7e8e Task Control Grou... |
1073 |
root->top_cgroup.dentry = sb->s_root; |
817929ec2 Task Control Grou... |
1074 1075 1076 |
/* Link the top cgroup in this hierarchy into all * the css_set objects */ write_lock(&css_set_lock); |
28fd5dfc1 cgroups: remove t... |
1077 1078 1079 |
for (i = 0; i < CSS_SET_TABLE_SIZE; i++) { struct hlist_head *hhead = &css_set_table[i]; struct hlist_node *node; |
817929ec2 Task Control Grou... |
1080 |
struct css_set *cg; |
28fd5dfc1 cgroups: remove t... |
1081 |
|
c12f65d43 cgroups: introduc... |
1082 1083 |
hlist_for_each_entry(cg, node, hhead, hlist) link_css_set(&tmp_cg_links, cg, root_cgrp); |
28fd5dfc1 cgroups: remove t... |
1084 |
} |
817929ec2 Task Control Grou... |
1085 1086 1087 |
write_unlock(&css_set_lock); free_cg_links(&tmp_cg_links); |
c12f65d43 cgroups: introduc... |
1088 1089 |
BUG_ON(!list_empty(&root_cgrp->sibling)); BUG_ON(!list_empty(&root_cgrp->children)); |
ddbcc7e8e Task Control Grou... |
1090 |
BUG_ON(root->number_of_cgroups != 1); |
c12f65d43 cgroups: introduc... |
1091 |
cgroup_populate_dir(root_cgrp); |
817929ec2 Task Control Grou... |
1092 |
mutex_unlock(&inode->i_mutex); |
ddbcc7e8e Task Control Grou... |
1093 1094 |
mutex_unlock(&cgroup_mutex); } |
a3ec947c8 vfs: simple_set_m... |
1095 1096 |
simple_set_mnt(mnt, sb); return 0; |
ddbcc7e8e Task Control Grou... |
1097 |
|
20ca9b3f4 cgroups: avoid ac... |
1098 1099 |
free_cg_links: free_cg_links(&tmp_cg_links); |
ddbcc7e8e Task Control Grou... |
1100 |
drop_new_super: |
6f5bbff9a Convert obvious p... |
1101 |
deactivate_locked_super(sb); |
ddbcc7e8e Task Control Grou... |
1102 1103 1104 1105 1106 |
return ret; } static void cgroup_kill_sb(struct super_block *sb) { struct cgroupfs_root *root = sb->s_fs_info; |
bd89aabc6 Control groups: R... |
1107 |
struct cgroup *cgrp = &root->top_cgroup; |
ddbcc7e8e Task Control Grou... |
1108 |
int ret; |
71cbb949d cgroup: list_for_... |
1109 1110 |
struct cg_cgroup_link *link; struct cg_cgroup_link *saved_link; |
ddbcc7e8e Task Control Grou... |
1111 1112 1113 1114 |
BUG_ON(!root); BUG_ON(root->number_of_cgroups != 1); |
bd89aabc6 Control groups: R... |
1115 1116 |
BUG_ON(!list_empty(&cgrp->children)); BUG_ON(!list_empty(&cgrp->sibling)); |
ddbcc7e8e Task Control Grou... |
1117 1118 1119 1120 1121 1122 1123 |
mutex_lock(&cgroup_mutex); /* Rebind all subsystems back to the default hierarchy */ ret = rebind_subsystems(root, 0); /* Shouldn't be able to fail ... */ BUG_ON(ret); |
817929ec2 Task Control Grou... |
1124 1125 1126 1127 1128 |
/* * Release all the links from css_sets to this hierarchy's * root cgroup */ write_lock(&css_set_lock); |
71cbb949d cgroup: list_for_... |
1129 1130 1131 |
list_for_each_entry_safe(link, saved_link, &cgrp->css_sets, cgrp_link_list) { |
817929ec2 Task Control Grou... |
1132 |
list_del(&link->cg_link_list); |
bd89aabc6 Control groups: R... |
1133 |
list_del(&link->cgrp_link_list); |
817929ec2 Task Control Grou... |
1134 1135 1136 |
kfree(link); } write_unlock(&css_set_lock); |
839ec5452 cgroup: fix root_... |
1137 1138 1139 1140 |
if (!list_empty(&root->root_list)) { list_del(&root->root_list); root_count--; } |
e5f6a8609 cgroups: make roo... |
1141 |
|
ddbcc7e8e Task Control Grou... |
1142 |
mutex_unlock(&cgroup_mutex); |
ddbcc7e8e Task Control Grou... |
1143 |
kill_litter_super(sb); |
67e055d14 cgroups: fix poss... |
1144 |
kfree(root); |
ddbcc7e8e Task Control Grou... |
1145 1146 1147 1148 1149 1150 1151 |
} static struct file_system_type cgroup_fs_type = { .name = "cgroup", .get_sb = cgroup_get_sb, .kill_sb = cgroup_kill_sb, }; |
bd89aabc6 Control groups: R... |
1152 |
static inline struct cgroup *__d_cgrp(struct dentry *dentry) |
ddbcc7e8e Task Control Grou... |
1153 1154 1155 1156 1157 1158 1159 1160 |
{ return dentry->d_fsdata; } static inline struct cftype *__d_cft(struct dentry *dentry) { return dentry->d_fsdata; } |
a043e3b2c cgroup: fix comments |
1161 1162 1163 1164 1165 1166 |
/** * cgroup_path - generate the path of a cgroup * @cgrp: the cgroup in question * @buf: the buffer to write the path into * @buflen: the length of the buffer * |
a47295e6b cgroups: make cgr... |
1167 1168 1169 |
* Called with cgroup_mutex held or else with an RCU-protected cgroup * reference. Writes path of cgroup into buf. Returns 0 on success, * -errno on error. |
ddbcc7e8e Task Control Grou... |
1170 |
*/ |
bd89aabc6 Control groups: R... |
1171 |
int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen) |
ddbcc7e8e Task Control Grou... |
1172 1173 |
{ char *start; |
a47295e6b cgroups: make cgr... |
1174 |
struct dentry *dentry = rcu_dereference(cgrp->dentry); |
ddbcc7e8e Task Control Grou... |
1175 |
|
a47295e6b cgroups: make cgr... |
1176 |
if (!dentry || cgrp == dummytop) { |
ddbcc7e8e Task Control Grou... |
1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 |
/* * Inactive subsystems have no dentry for their root * cgroup */ strcpy(buf, "/"); return 0; } start = buf + buflen; *--start = '\0'; for (;;) { |
a47295e6b cgroups: make cgr... |
1189 |
int len = dentry->d_name.len; |
ddbcc7e8e Task Control Grou... |
1190 1191 |
if ((start -= len) < buf) return -ENAMETOOLONG; |
bd89aabc6 Control groups: R... |
1192 1193 1194 |
memcpy(start, cgrp->dentry->d_name.name, len); cgrp = cgrp->parent; if (!cgrp) |
ddbcc7e8e Task Control Grou... |
1195 |
break; |
a47295e6b cgroups: make cgr... |
1196 |
dentry = rcu_dereference(cgrp->dentry); |
bd89aabc6 Control groups: R... |
1197 |
if (!cgrp->parent) |
ddbcc7e8e Task Control Grou... |
1198 1199 1200 1201 1202 1203 1204 1205 |
continue; if (--start < buf) return -ENAMETOOLONG; *start = '/'; } memmove(buf, start, buf + buflen - start); return 0; } |
bbcb81d09 Task Control Grou... |
1206 1207 1208 1209 |
/* * Return the first subsystem attached to a cgroup's hierarchy, and * its subsystem id. */ |
bd89aabc6 Control groups: R... |
1210 |
static void get_first_subsys(const struct cgroup *cgrp, |
bbcb81d09 Task Control Grou... |
1211 1212 |
struct cgroup_subsys_state **css, int *subsys_id) { |
bd89aabc6 Control groups: R... |
1213 |
const struct cgroupfs_root *root = cgrp->root; |
bbcb81d09 Task Control Grou... |
1214 1215 1216 1217 1218 |
const struct cgroup_subsys *test_ss; BUG_ON(list_empty(&root->subsys_list)); test_ss = list_entry(root->subsys_list.next, struct cgroup_subsys, sibling); if (css) { |
bd89aabc6 Control groups: R... |
1219 |
*css = cgrp->subsys[test_ss->subsys_id]; |
bbcb81d09 Task Control Grou... |
1220 1221 1222 1223 1224 |
BUG_ON(!*css); } if (subsys_id) *subsys_id = test_ss->subsys_id; } |
a043e3b2c cgroup: fix comments |
1225 1226 1227 1228 |
/** * cgroup_attach_task - attach task 'tsk' to cgroup 'cgrp' * @cgrp: the cgroup the task is attaching to * @tsk: the task to be attached |
bbcb81d09 Task Control Grou... |
1229 |
* |
a043e3b2c cgroup: fix comments |
1230 1231 |
* Call holding cgroup_mutex. May take task_lock of * the task 'tsk' during call. |
bbcb81d09 Task Control Grou... |
1232 |
*/ |
956db3ca0 hotplug cpu: move... |
1233 |
int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk) |
bbcb81d09 Task Control Grou... |
1234 1235 1236 |
{ int retval = 0; struct cgroup_subsys *ss; |
bd89aabc6 Control groups: R... |
1237 |
struct cgroup *oldcgrp; |
77efecd9e cgroups: call fin... |
1238 |
struct css_set *cg; |
817929ec2 Task Control Grou... |
1239 |
struct css_set *newcg; |
bd89aabc6 Control groups: R... |
1240 |
struct cgroupfs_root *root = cgrp->root; |
bbcb81d09 Task Control Grou... |
1241 |
int subsys_id; |
bd89aabc6 Control groups: R... |
1242 |
get_first_subsys(cgrp, NULL, &subsys_id); |
bbcb81d09 Task Control Grou... |
1243 1244 |
/* Nothing to do if the task is already in that cgroup */ |
bd89aabc6 Control groups: R... |
1245 1246 |
oldcgrp = task_cgroup(tsk, subsys_id); if (cgrp == oldcgrp) |
bbcb81d09 Task Control Grou... |
1247 1248 1249 1250 |
return 0; for_each_subsys(root, ss) { if (ss->can_attach) { |
bd89aabc6 Control groups: R... |
1251 |
retval = ss->can_attach(ss, cgrp, tsk); |
e18f6318e cgroup brace codi... |
1252 |
if (retval) |
bbcb81d09 Task Control Grou... |
1253 |
return retval; |
bbcb81d09 Task Control Grou... |
1254 1255 |
} } |
77efecd9e cgroups: call fin... |
1256 1257 1258 1259 |
task_lock(tsk); cg = tsk->cgroups; get_css_set(cg); task_unlock(tsk); |
817929ec2 Task Control Grou... |
1260 1261 1262 1263 |
/* * Locate or allocate a new css_set for this task, * based on its final set of cgroups */ |
bd89aabc6 Control groups: R... |
1264 |
newcg = find_css_set(cg, cgrp); |
77efecd9e cgroups: call fin... |
1265 |
put_css_set(cg); |
e18f6318e cgroup brace codi... |
1266 |
if (!newcg) |
817929ec2 Task Control Grou... |
1267 |
return -ENOMEM; |
817929ec2 Task Control Grou... |
1268 |
|
bbcb81d09 Task Control Grou... |
1269 1270 1271 |
task_lock(tsk); if (tsk->flags & PF_EXITING) { task_unlock(tsk); |
817929ec2 Task Control Grou... |
1272 |
put_css_set(newcg); |
bbcb81d09 Task Control Grou... |
1273 1274 |
return -ESRCH; } |
817929ec2 Task Control Grou... |
1275 |
rcu_assign_pointer(tsk->cgroups, newcg); |
bbcb81d09 Task Control Grou... |
1276 |
task_unlock(tsk); |
817929ec2 Task Control Grou... |
1277 1278 1279 1280 1281 1282 1283 |
/* Update the css_set linked lists if we're using them */ write_lock(&css_set_lock); if (!list_empty(&tsk->cg_list)) { list_del(&tsk->cg_list); list_add(&tsk->cg_list, &newcg->tasks); } write_unlock(&css_set_lock); |
bbcb81d09 Task Control Grou... |
1284 |
for_each_subsys(root, ss) { |
e18f6318e cgroup brace codi... |
1285 |
if (ss->attach) |
bd89aabc6 Control groups: R... |
1286 |
ss->attach(ss, cgrp, oldcgrp, tsk); |
bbcb81d09 Task Control Grou... |
1287 |
} |
bd89aabc6 Control groups: R... |
1288 |
set_bit(CGRP_RELEASABLE, &oldcgrp->flags); |
bbcb81d09 Task Control Grou... |
1289 |
synchronize_rcu(); |
817929ec2 Task Control Grou... |
1290 |
put_css_set(cg); |
ec64f5154 cgroup: fix frequ... |
1291 1292 1293 1294 1295 1296 |
/* * wake up rmdir() waiter. the rmdir should fail since the cgroup * is no longer empty. */ cgroup_wakeup_rmdir_waiters(cgrp); |
bbcb81d09 Task Control Grou... |
1297 1298 1299 1300 |
return 0; } /* |
af351026a cgroup files: tur... |
1301 1302 |
* Attach task with pid 'pid' to cgroup 'cgrp'. Call with cgroup_mutex * held. May take task_lock of task |
bbcb81d09 Task Control Grou... |
1303 |
*/ |
af351026a cgroup files: tur... |
1304 |
static int attach_task_by_pid(struct cgroup *cgrp, u64 pid) |
bbcb81d09 Task Control Grou... |
1305 |
{ |
bbcb81d09 Task Control Grou... |
1306 |
struct task_struct *tsk; |
c69e8d9c0 CRED: Use RCU to ... |
1307 |
const struct cred *cred = current_cred(), *tcred; |
bbcb81d09 Task Control Grou... |
1308 |
int ret; |
bbcb81d09 Task Control Grou... |
1309 1310 |
if (pid) { rcu_read_lock(); |
73507f335 Handle pid namesp... |
1311 |
tsk = find_task_by_vpid(pid); |
bbcb81d09 Task Control Grou... |
1312 1313 1314 1315 |
if (!tsk || tsk->flags & PF_EXITING) { rcu_read_unlock(); return -ESRCH; } |
bbcb81d09 Task Control Grou... |
1316 |
|
c69e8d9c0 CRED: Use RCU to ... |
1317 1318 1319 1320 1321 |
tcred = __task_cred(tsk); if (cred->euid && cred->euid != tcred->uid && cred->euid != tcred->suid) { rcu_read_unlock(); |
bbcb81d09 Task Control Grou... |
1322 1323 |
return -EACCES; } |
c69e8d9c0 CRED: Use RCU to ... |
1324 1325 |
get_task_struct(tsk); rcu_read_unlock(); |
bbcb81d09 Task Control Grou... |
1326 1327 1328 1329 |
} else { tsk = current; get_task_struct(tsk); } |
956db3ca0 hotplug cpu: move... |
1330 |
ret = cgroup_attach_task(cgrp, tsk); |
bbcb81d09 Task Control Grou... |
1331 1332 1333 |
put_task_struct(tsk); return ret; } |
af351026a cgroup files: tur... |
1334 1335 1336 1337 1338 1339 1340 1341 1342 |
static int cgroup_tasks_write(struct cgroup *cgrp, struct cftype *cft, u64 pid) { int ret; if (!cgroup_lock_live_group(cgrp)) return -ENODEV; ret = attach_task_by_pid(cgrp, pid); cgroup_unlock(); return ret; } |
ddbcc7e8e Task Control Grou... |
1343 |
/* The various types of files and directories in a cgroup file system */ |
ddbcc7e8e Task Control Grou... |
1344 1345 1346 1347 |
enum cgroup_filetype { FILE_ROOT, FILE_DIR, FILE_TASKLIST, |
81a6a5cdd Task Control Grou... |
1348 |
FILE_NOTIFY_ON_RELEASE, |
81a6a5cdd Task Control Grou... |
1349 |
FILE_RELEASE_AGENT, |
ddbcc7e8e Task Control Grou... |
1350 |
}; |
e788e066c cgroup files: mov... |
1351 1352 1353 1354 |
/** * cgroup_lock_live_group - take cgroup_mutex and check that cgrp is alive. * @cgrp: the cgroup to be checked for liveness * |
84eea8428 cgroups: misc cle... |
1355 1356 |
* On success, returns true; the lock should be later released with * cgroup_unlock(). On failure returns false with no lock held. |
e788e066c cgroup files: mov... |
1357 |
*/ |
84eea8428 cgroups: misc cle... |
1358 |
bool cgroup_lock_live_group(struct cgroup *cgrp) |
e788e066c cgroup files: mov... |
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 |
{ mutex_lock(&cgroup_mutex); if (cgroup_is_removed(cgrp)) { mutex_unlock(&cgroup_mutex); return false; } return true; } static int cgroup_release_agent_write(struct cgroup *cgrp, struct cftype *cft, const char *buffer) { BUILD_BUG_ON(sizeof(cgrp->root->release_agent_path) < PATH_MAX); if (!cgroup_lock_live_group(cgrp)) return -ENODEV; strcpy(cgrp->root->release_agent_path, buffer); |
84eea8428 cgroups: misc cle... |
1375 |
cgroup_unlock(); |
e788e066c cgroup files: mov... |
1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 |
return 0; } static int cgroup_release_agent_show(struct cgroup *cgrp, struct cftype *cft, struct seq_file *seq) { if (!cgroup_lock_live_group(cgrp)) return -ENODEV; seq_puts(seq, cgrp->root->release_agent_path); seq_putc(seq, ' '); |
84eea8428 cgroups: misc cle... |
1387 |
cgroup_unlock(); |
e788e066c cgroup files: mov... |
1388 1389 |
return 0; } |
84eea8428 cgroups: misc cle... |
1390 1391 |
/* A buffer size big enough for numbers or short strings */ #define CGROUP_LOCAL_BUFFER_SIZE 64 |
e73d2c61d CGroups _s64 file... |
1392 |
static ssize_t cgroup_write_X64(struct cgroup *cgrp, struct cftype *cft, |
f4c753b7e CGroup API files:... |
1393 1394 1395 |
struct file *file, const char __user *userbuf, size_t nbytes, loff_t *unused_ppos) |
355e0c48b Add cgroup write_... |
1396 |
{ |
84eea8428 cgroups: misc cle... |
1397 |
char buffer[CGROUP_LOCAL_BUFFER_SIZE]; |
355e0c48b Add cgroup write_... |
1398 |
int retval = 0; |
355e0c48b Add cgroup write_... |
1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 |
char *end; if (!nbytes) return -EINVAL; if (nbytes >= sizeof(buffer)) return -E2BIG; if (copy_from_user(buffer, userbuf, nbytes)) return -EFAULT; buffer[nbytes] = 0; /* nul-terminate */ |
b7269dfc8 CGroup API files:... |
1409 |
strstrip(buffer); |
e73d2c61d CGroups _s64 file... |
1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 |
if (cft->write_u64) { u64 val = simple_strtoull(buffer, &end, 0); if (*end) return -EINVAL; retval = cft->write_u64(cgrp, cft, val); } else { s64 val = simple_strtoll(buffer, &end, 0); if (*end) return -EINVAL; retval = cft->write_s64(cgrp, cft, val); } |
355e0c48b Add cgroup write_... |
1421 1422 1423 1424 |
if (!retval) retval = nbytes; return retval; } |
db3b14978 cgroup files: add... |
1425 1426 1427 1428 1429 |
static ssize_t cgroup_write_string(struct cgroup *cgrp, struct cftype *cft, struct file *file, const char __user *userbuf, size_t nbytes, loff_t *unused_ppos) { |
84eea8428 cgroups: misc cle... |
1430 |
char local_buffer[CGROUP_LOCAL_BUFFER_SIZE]; |
db3b14978 cgroup files: add... |
1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 |
int retval = 0; size_t max_bytes = cft->max_write_len; char *buffer = local_buffer; if (!max_bytes) max_bytes = sizeof(local_buffer) - 1; if (nbytes >= max_bytes) return -E2BIG; /* Allocate a dynamic buffer if we need one */ if (nbytes >= sizeof(local_buffer)) { buffer = kmalloc(nbytes + 1, GFP_KERNEL); if (buffer == NULL) return -ENOMEM; } |
5a3eb9f6b cgroup: fix possi... |
1445 1446 1447 1448 |
if (nbytes && copy_from_user(buffer, userbuf, nbytes)) { retval = -EFAULT; goto out; } |
db3b14978 cgroup files: add... |
1449 1450 1451 1452 1453 1454 |
buffer[nbytes] = 0; /* nul-terminate */ strstrip(buffer); retval = cft->write_string(cgrp, cft, buffer); if (!retval) retval = nbytes; |
5a3eb9f6b cgroup: fix possi... |
1455 |
out: |
db3b14978 cgroup files: add... |
1456 1457 1458 1459 |
if (buffer != local_buffer) kfree(buffer); return retval; } |
ddbcc7e8e Task Control Grou... |
1460 1461 1462 1463 |
static ssize_t cgroup_file_write(struct file *file, const char __user *buf, size_t nbytes, loff_t *ppos) { struct cftype *cft = __d_cft(file->f_dentry); |
bd89aabc6 Control groups: R... |
1464 |
struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); |
ddbcc7e8e Task Control Grou... |
1465 |
|
75139b827 cgroups: remove s... |
1466 |
if (cgroup_is_removed(cgrp)) |
ddbcc7e8e Task Control Grou... |
1467 |
return -ENODEV; |
355e0c48b Add cgroup write_... |
1468 |
if (cft->write) |
bd89aabc6 Control groups: R... |
1469 |
return cft->write(cgrp, cft, file, buf, nbytes, ppos); |
e73d2c61d CGroups _s64 file... |
1470 1471 |
if (cft->write_u64 || cft->write_s64) return cgroup_write_X64(cgrp, cft, file, buf, nbytes, ppos); |
db3b14978 cgroup files: add... |
1472 1473 |
if (cft->write_string) return cgroup_write_string(cgrp, cft, file, buf, nbytes, ppos); |
d447ea2f3 cgroups: add the ... |
1474 1475 1476 1477 |
if (cft->trigger) { int ret = cft->trigger(cgrp, (unsigned int)cft->private); return ret ? ret : nbytes; } |
355e0c48b Add cgroup write_... |
1478 |
return -EINVAL; |
ddbcc7e8e Task Control Grou... |
1479 |
} |
f4c753b7e CGroup API files:... |
1480 1481 1482 1483 |
static ssize_t cgroup_read_u64(struct cgroup *cgrp, struct cftype *cft, struct file *file, char __user *buf, size_t nbytes, loff_t *ppos) |
ddbcc7e8e Task Control Grou... |
1484 |
{ |
84eea8428 cgroups: misc cle... |
1485 |
char tmp[CGROUP_LOCAL_BUFFER_SIZE]; |
f4c753b7e CGroup API files:... |
1486 |
u64 val = cft->read_u64(cgrp, cft); |
ddbcc7e8e Task Control Grou... |
1487 1488 1489 1490 1491 |
int len = sprintf(tmp, "%llu ", (unsigned long long) val); return simple_read_from_buffer(buf, nbytes, ppos, tmp, len); } |
e73d2c61d CGroups _s64 file... |
1492 1493 1494 1495 1496 |
static ssize_t cgroup_read_s64(struct cgroup *cgrp, struct cftype *cft, struct file *file, char __user *buf, size_t nbytes, loff_t *ppos) { |
84eea8428 cgroups: misc cle... |
1497 |
char tmp[CGROUP_LOCAL_BUFFER_SIZE]; |
e73d2c61d CGroups _s64 file... |
1498 1499 1500 1501 1502 1503 |
s64 val = cft->read_s64(cgrp, cft); int len = sprintf(tmp, "%lld ", (long long) val); return simple_read_from_buffer(buf, nbytes, ppos, tmp, len); } |
ddbcc7e8e Task Control Grou... |
1504 1505 1506 1507 |
static ssize_t cgroup_file_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos) { struct cftype *cft = __d_cft(file->f_dentry); |
bd89aabc6 Control groups: R... |
1508 |
struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); |
ddbcc7e8e Task Control Grou... |
1509 |
|
75139b827 cgroups: remove s... |
1510 |
if (cgroup_is_removed(cgrp)) |
ddbcc7e8e Task Control Grou... |
1511 1512 1513 |
return -ENODEV; if (cft->read) |
bd89aabc6 Control groups: R... |
1514 |
return cft->read(cgrp, cft, file, buf, nbytes, ppos); |
f4c753b7e CGroup API files:... |
1515 1516 |
if (cft->read_u64) return cgroup_read_u64(cgrp, cft, file, buf, nbytes, ppos); |
e73d2c61d CGroups _s64 file... |
1517 1518 |
if (cft->read_s64) return cgroup_read_s64(cgrp, cft, file, buf, nbytes, ppos); |
ddbcc7e8e Task Control Grou... |
1519 1520 |
return -EINVAL; } |
917965696 CGroup API files:... |
1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 |
/* * seqfile ops/methods for returning structured data. Currently just * supports string->u64 maps, but can be extended in future. */ struct cgroup_seqfile_state { struct cftype *cft; struct cgroup *cgroup; }; static int cgroup_map_add(struct cgroup_map_cb *cb, const char *key, u64 value) { struct seq_file *sf = cb->state; return seq_printf(sf, "%s %llu ", key, (unsigned long long)value); } static int cgroup_seqfile_show(struct seq_file *m, void *arg) { struct cgroup_seqfile_state *state = m->private; struct cftype *cft = state->cft; |
29486df32 cgroups: introduc... |
1542 1543 1544 1545 1546 1547 1548 1549 |
if (cft->read_map) { struct cgroup_map_cb cb = { .fill = cgroup_map_add, .state = m, }; return cft->read_map(state->cgroup, cft, &cb); } return cft->read_seq_string(state->cgroup, cft, m); |
917965696 CGroup API files:... |
1550 |
} |
96930a636 make cgroup_seqfi... |
1551 |
static int cgroup_seqfile_release(struct inode *inode, struct file *file) |
917965696 CGroup API files:... |
1552 1553 1554 1555 1556 1557 1558 1559 |
{ struct seq_file *seq = file->private_data; kfree(seq->private); return single_release(inode, file); } static struct file_operations cgroup_seqfile_operations = { .read = seq_read, |
e788e066c cgroup files: mov... |
1560 |
.write = cgroup_file_write, |
917965696 CGroup API files:... |
1561 1562 1563 |
.llseek = seq_lseek, .release = cgroup_seqfile_release, }; |
ddbcc7e8e Task Control Grou... |
1564 1565 1566 1567 1568 1569 1570 1571 |
static int cgroup_file_open(struct inode *inode, struct file *file) { int err; struct cftype *cft; err = generic_file_open(inode, file); if (err) return err; |
ddbcc7e8e Task Control Grou... |
1572 |
cft = __d_cft(file->f_dentry); |
75139b827 cgroups: remove s... |
1573 |
|
29486df32 cgroups: introduc... |
1574 |
if (cft->read_map || cft->read_seq_string) { |
917965696 CGroup API files:... |
1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 |
struct cgroup_seqfile_state *state = kzalloc(sizeof(*state), GFP_USER); if (!state) return -ENOMEM; state->cft = cft; state->cgroup = __d_cgrp(file->f_dentry->d_parent); file->f_op = &cgroup_seqfile_operations; err = single_open(file, cgroup_seqfile_show, state); if (err < 0) kfree(state); } else if (cft->open) |
ddbcc7e8e Task Control Grou... |
1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 |
err = cft->open(inode, file); else err = 0; return err; } static int cgroup_file_release(struct inode *inode, struct file *file) { struct cftype *cft = __d_cft(file->f_dentry); if (cft->release) return cft->release(inode, file); return 0; } /* * cgroup_rename - Only allow simple rename of directories in place. */ static int cgroup_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { if (!S_ISDIR(old_dentry->d_inode->i_mode)) return -ENOTDIR; if (new_dentry->d_inode) return -EEXIST; if (old_dir != new_dir) return -EIO; return simple_rename(old_dir, old_dentry, new_dir, new_dentry); } static struct file_operations cgroup_file_operations = { .read = cgroup_file_read, .write = cgroup_file_write, .llseek = generic_file_llseek, .open = cgroup_file_open, .release = cgroup_file_release, }; static struct inode_operations cgroup_dir_inode_operations = { .lookup = simple_lookup, .mkdir = cgroup_mkdir, .rmdir = cgroup_rmdir, .rename = cgroup_rename, }; |
099fca322 cgroups: show cor... |
1630 |
static int cgroup_create_file(struct dentry *dentry, mode_t mode, |
ddbcc7e8e Task Control Grou... |
1631 1632 |
struct super_block *sb) { |
3ba13d179 constify dentry_o... |
1633 |
static const struct dentry_operations cgroup_dops = { |
ddbcc7e8e Task Control Grou... |
1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 |
.d_iput = cgroup_diput, }; struct inode *inode; if (!dentry) return -ENOENT; if (dentry->d_inode) return -EEXIST; inode = cgroup_new_inode(mode, sb); if (!inode) return -ENOMEM; if (S_ISDIR(mode)) { inode->i_op = &cgroup_dir_inode_operations; inode->i_fop = &simple_dir_operations; /* start off with i_nlink == 2 (for "." entry) */ inc_nlink(inode); /* start with the directory inode held, so that we can * populate it without racing with another mkdir */ |
817929ec2 Task Control Grou... |
1657 |
mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD); |
ddbcc7e8e Task Control Grou... |
1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 |
} else if (S_ISREG(mode)) { inode->i_size = 0; inode->i_fop = &cgroup_file_operations; } dentry->d_op = &cgroup_dops; d_instantiate(dentry, inode); dget(dentry); /* Extra count - pin the dentry in core */ return 0; } /* |
a043e3b2c cgroup: fix comments |
1669 1670 1671 1672 1673 |
* cgroup_create_dir - create a directory for an object. * @cgrp: the cgroup we create the directory for. It must have a valid * ->parent field. And we are going to fill its ->dentry field. * @dentry: dentry of the new cgroup * @mode: mode to set on new directory. |
ddbcc7e8e Task Control Grou... |
1674 |
*/ |
bd89aabc6 Control groups: R... |
1675 |
static int cgroup_create_dir(struct cgroup *cgrp, struct dentry *dentry, |
099fca322 cgroups: show cor... |
1676 |
mode_t mode) |
ddbcc7e8e Task Control Grou... |
1677 1678 1679 |
{ struct dentry *parent; int error = 0; |
bd89aabc6 Control groups: R... |
1680 1681 |
parent = cgrp->parent->dentry; error = cgroup_create_file(dentry, S_IFDIR | mode, cgrp->root->sb); |
ddbcc7e8e Task Control Grou... |
1682 |
if (!error) { |
bd89aabc6 Control groups: R... |
1683 |
dentry->d_fsdata = cgrp; |
ddbcc7e8e Task Control Grou... |
1684 |
inc_nlink(parent->d_inode); |
a47295e6b cgroups: make cgr... |
1685 |
rcu_assign_pointer(cgrp->dentry, dentry); |
ddbcc7e8e Task Control Grou... |
1686 1687 1688 1689 1690 1691 |
dget(dentry); } dput(dentry); return error; } |
099fca322 cgroups: show cor... |
1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 |
/** * cgroup_file_mode - deduce file mode of a control file * @cft: the control file in question * * returns cft->mode if ->mode is not 0 * returns S_IRUGO|S_IWUSR if it has both a read and a write handler * returns S_IRUGO if it has only a read handler * returns S_IWUSR if it has only a write hander */ static mode_t cgroup_file_mode(const struct cftype *cft) { mode_t mode = 0; if (cft->mode) return cft->mode; if (cft->read || cft->read_u64 || cft->read_s64 || cft->read_map || cft->read_seq_string) mode |= S_IRUGO; if (cft->write || cft->write_u64 || cft->write_s64 || cft->write_string || cft->trigger) mode |= S_IWUSR; return mode; } |
bd89aabc6 Control groups: R... |
1718 |
int cgroup_add_file(struct cgroup *cgrp, |
ddbcc7e8e Task Control Grou... |
1719 1720 1721 |
struct cgroup_subsys *subsys, const struct cftype *cft) { |
bd89aabc6 Control groups: R... |
1722 |
struct dentry *dir = cgrp->dentry; |
ddbcc7e8e Task Control Grou... |
1723 1724 |
struct dentry *dentry; int error; |
099fca322 cgroups: show cor... |
1725 |
mode_t mode; |
ddbcc7e8e Task Control Grou... |
1726 1727 |
char name[MAX_CGROUP_TYPE_NAMELEN + MAX_CFTYPE_NAME + 2] = { 0 }; |
bd89aabc6 Control groups: R... |
1728 |
if (subsys && !test_bit(ROOT_NOPREFIX, &cgrp->root->flags)) { |
ddbcc7e8e Task Control Grou... |
1729 1730 1731 1732 1733 1734 1735 |
strcpy(name, subsys->name); strcat(name, "."); } strcat(name, cft->name); BUG_ON(!mutex_is_locked(&dir->d_inode->i_mutex)); dentry = lookup_one_len(name, dir, strlen(name)); if (!IS_ERR(dentry)) { |
099fca322 cgroups: show cor... |
1736 1737 |
mode = cgroup_file_mode(cft); error = cgroup_create_file(dentry, mode | S_IFREG, |
bd89aabc6 Control groups: R... |
1738 |
cgrp->root->sb); |
ddbcc7e8e Task Control Grou... |
1739 1740 1741 1742 1743 1744 1745 |
if (!error) dentry->d_fsdata = (void *)cft; dput(dentry); } else error = PTR_ERR(dentry); return error; } |
bd89aabc6 Control groups: R... |
1746 |
int cgroup_add_files(struct cgroup *cgrp, |
ddbcc7e8e Task Control Grou... |
1747 1748 1749 1750 1751 1752 |
struct cgroup_subsys *subsys, const struct cftype cft[], int count) { int i, err; for (i = 0; i < count; i++) { |
bd89aabc6 Control groups: R... |
1753 |
err = cgroup_add_file(cgrp, subsys, &cft[i]); |
ddbcc7e8e Task Control Grou... |
1754 1755 1756 1757 1758 |
if (err) return err; } return 0; } |
a043e3b2c cgroup: fix comments |
1759 1760 1761 1762 1763 1764 |
/** * cgroup_task_count - count the number of tasks in a cgroup. * @cgrp: the cgroup in question * * Return the number of tasks in the cgroup. */ |
bd89aabc6 Control groups: R... |
1765 |
int cgroup_task_count(const struct cgroup *cgrp) |
bbcb81d09 Task Control Grou... |
1766 1767 |
{ int count = 0; |
71cbb949d cgroup: list_for_... |
1768 |
struct cg_cgroup_link *link; |
817929ec2 Task Control Grou... |
1769 1770 |
read_lock(&css_set_lock); |
71cbb949d cgroup: list_for_... |
1771 |
list_for_each_entry(link, &cgrp->css_sets, cgrp_link_list) { |
146aa1bd0 cgroups: fix prob... |
1772 |
count += atomic_read(&link->cg->refcount); |
817929ec2 Task Control Grou... |
1773 1774 |
} read_unlock(&css_set_lock); |
bbcb81d09 Task Control Grou... |
1775 1776 1777 1778 |
return count; } /* |
817929ec2 Task Control Grou... |
1779 1780 1781 |
* Advance a list_head iterator. The iterator should be positioned at * the start of a css_set */ |
bd89aabc6 Control groups: R... |
1782 |
static void cgroup_advance_iter(struct cgroup *cgrp, |
817929ec2 Task Control Grou... |
1783 1784 1785 1786 1787 1788 1789 1790 1791 |
struct cgroup_iter *it) { struct list_head *l = it->cg_link; struct cg_cgroup_link *link; struct css_set *cg; /* Advance to the next non-empty css_set */ do { l = l->next; |
bd89aabc6 Control groups: R... |
1792 |
if (l == &cgrp->css_sets) { |
817929ec2 Task Control Grou... |
1793 1794 1795 |
it->cg_link = NULL; return; } |
bd89aabc6 Control groups: R... |
1796 |
link = list_entry(l, struct cg_cgroup_link, cgrp_link_list); |
817929ec2 Task Control Grou... |
1797 1798 1799 1800 1801 |
cg = link->cg; } while (list_empty(&cg->tasks)); it->cg_link = l; it->task = cg->tasks.next; } |
31a7df01f cgroups: mechanis... |
1802 1803 1804 1805 1806 1807 1808 1809 1810 |
/* * To reduce the fork() overhead for systems that are not actually * using their cgroups capability, we don't maintain the lists running * through each css_set to its tasks until we see the list actually * used - in other words after the first call to cgroup_iter_start(). * * The tasklist_lock is not held here, as do_each_thread() and * while_each_thread() are protected by RCU. */ |
3df91fe30 make cgroup_enabl... |
1811 |
static void cgroup_enable_task_cg_lists(void) |
31a7df01f cgroups: mechanis... |
1812 1813 1814 1815 1816 1817 |
{ struct task_struct *p, *g; write_lock(&css_set_lock); use_task_css_set_links = 1; do_each_thread(g, p) { task_lock(p); |
0e04388f0 cgroup: fix a rac... |
1818 1819 1820 1821 1822 1823 |
/* * We should check if the process is exiting, otherwise * it will race with cgroup_exit() in that the list * entry won't be deleted though the process has exited. */ if (!(p->flags & PF_EXITING) && list_empty(&p->cg_list)) |
31a7df01f cgroups: mechanis... |
1824 1825 1826 1827 1828 |
list_add(&p->cg_list, &p->cgroups->tasks); task_unlock(p); } while_each_thread(g, p); write_unlock(&css_set_lock); } |
bd89aabc6 Control groups: R... |
1829 |
void cgroup_iter_start(struct cgroup *cgrp, struct cgroup_iter *it) |
817929ec2 Task Control Grou... |
1830 1831 1832 1833 1834 1835 |
{ /* * The first time anyone tries to iterate across a cgroup, * we need to enable the list linking each css_set to its * tasks, and fix up all existing tasks. */ |
31a7df01f cgroups: mechanis... |
1836 1837 |
if (!use_task_css_set_links) cgroup_enable_task_cg_lists(); |
817929ec2 Task Control Grou... |
1838 |
read_lock(&css_set_lock); |
bd89aabc6 Control groups: R... |
1839 1840 |
it->cg_link = &cgrp->css_sets; cgroup_advance_iter(cgrp, it); |
817929ec2 Task Control Grou... |
1841 |
} |
bd89aabc6 Control groups: R... |
1842 |
struct task_struct *cgroup_iter_next(struct cgroup *cgrp, |
817929ec2 Task Control Grou... |
1843 1844 1845 1846 |
struct cgroup_iter *it) { struct task_struct *res; struct list_head *l = it->task; |
2019f634c cgroups: fix cgro... |
1847 |
struct cg_cgroup_link *link; |
817929ec2 Task Control Grou... |
1848 1849 1850 1851 1852 1853 1854 |
/* If the iterator cg is NULL, we have no tasks */ if (!it->cg_link) return NULL; res = list_entry(l, struct task_struct, cg_list); /* Advance iterator to find next entry */ l = l->next; |
2019f634c cgroups: fix cgro... |
1855 1856 |
link = list_entry(it->cg_link, struct cg_cgroup_link, cgrp_link_list); if (l == &link->cg->tasks) { |
817929ec2 Task Control Grou... |
1857 1858 |
/* We reached the end of this task list - move on to * the next cg_cgroup_link */ |
bd89aabc6 Control groups: R... |
1859 |
cgroup_advance_iter(cgrp, it); |
817929ec2 Task Control Grou... |
1860 1861 1862 1863 1864 |
} else { it->task = l; } return res; } |
bd89aabc6 Control groups: R... |
1865 |
void cgroup_iter_end(struct cgroup *cgrp, struct cgroup_iter *it) |
817929ec2 Task Control Grou... |
1866 1867 1868 |
{ read_unlock(&css_set_lock); } |
31a7df01f cgroups: mechanis... |
1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 |
static inline int started_after_time(struct task_struct *t1, struct timespec *time, struct task_struct *t2) { int start_diff = timespec_compare(&t1->start_time, time); if (start_diff > 0) { return 1; } else if (start_diff < 0) { return 0; } else { /* * Arbitrarily, if two processes started at the same * time, we'll say that the lower pointer value * started first. Note that t2 may have exited by now * so this may not be a valid pointer any longer, but * that's fine - it still serves to distinguish * between two tasks started (effectively) simultaneously. */ return t1 > t2; } } /* * This function is a callback from heap_insert() and is used to order * the heap. * In this case we order the heap in descending task start time. */ static inline int started_after(void *p1, void *p2) { struct task_struct *t1 = p1; struct task_struct *t2 = p2; return started_after_time(t1, &t2->start_time, t2); } /** * cgroup_scan_tasks - iterate though all the tasks in a cgroup * @scan: struct cgroup_scanner containing arguments for the scan * * Arguments include pointers to callback functions test_task() and * process_task(). * Iterate through all the tasks in a cgroup, calling test_task() for each, * and if it returns true, call process_task() for it also. * The test_task pointer may be NULL, meaning always true (select all tasks). * Effectively duplicates cgroup_iter_{start,next,end}() * but does not lock css_set_lock for the call to process_task(). * The struct cgroup_scanner may be embedded in any structure of the caller's * creation. * It is guaranteed that process_task() will act on every task that * is a member of the cgroup for the duration of this call. This * function may or may not call process_task() for tasks that exit * or move to a different cgroup during the call, or are forked or * move into the cgroup during the call. * * Note that test_task() may be called with locks held, and may in some * situations be called multiple times for the same task, so it should * be cheap. * If the heap pointer in the struct cgroup_scanner is non-NULL, a heap has been * pre-allocated and will be used for heap operations (and its "gt" member will * be overwritten), else a temporary heap will be used (allocation of which * may cause this function to fail). */ int cgroup_scan_tasks(struct cgroup_scanner *scan) { int retval, i; struct cgroup_iter it; struct task_struct *p, *dropped; /* Never dereference latest_task, since it's not refcounted */ struct task_struct *latest_task = NULL; struct ptr_heap tmp_heap; struct ptr_heap *heap; struct timespec latest_time = { 0, 0 }; if (scan->heap) { /* The caller supplied our heap and pre-allocated its memory */ heap = scan->heap; heap->gt = &started_after; } else { /* We need to allocate our own heap memory */ heap = &tmp_heap; retval = heap_init(heap, PAGE_SIZE, GFP_KERNEL, &started_after); if (retval) /* cannot allocate the heap */ return retval; } again: /* * Scan tasks in the cgroup, using the scanner's "test_task" callback * to determine which are of interest, and using the scanner's * "process_task" callback to process any of them that need an update. * Since we don't want to hold any locks during the task updates, * gather tasks to be processed in a heap structure. * The heap is sorted by descending task start time. * If the statically-sized heap fills up, we overflow tasks that * started later, and in future iterations only consider tasks that * started after the latest task in the previous pass. This * guarantees forward progress and that we don't miss any tasks. */ heap->size = 0; cgroup_iter_start(scan->cg, &it); while ((p = cgroup_iter_next(scan->cg, &it))) { /* * Only affect tasks that qualify per the caller's callback, * if he provided one */ if (scan->test_task && !scan->test_task(p, scan)) continue; /* * Only process tasks that started after the last task * we processed */ if (!started_after_time(p, &latest_time, latest_task)) continue; dropped = heap_insert(heap, p); if (dropped == NULL) { /* * The new task was inserted; the heap wasn't * previously full */ get_task_struct(p); } else if (dropped != p) { /* * The new task was inserted, and pushed out a * different task */ get_task_struct(p); put_task_struct(dropped); } /* * Else the new task was newer than anything already in * the heap and wasn't inserted */ } cgroup_iter_end(scan->cg, &it); if (heap->size) { for (i = 0; i < heap->size; i++) { |
4fe91d518 cgroup: fix spars... |
2006 |
struct task_struct *q = heap->ptrs[i]; |
31a7df01f cgroups: mechanis... |
2007 |
if (i == 0) { |
4fe91d518 cgroup: fix spars... |
2008 2009 |
latest_time = q->start_time; latest_task = q; |
31a7df01f cgroups: mechanis... |
2010 2011 |
} /* Process the task per the caller's callback */ |
4fe91d518 cgroup: fix spars... |
2012 2013 |
scan->process_task(q, scan); put_task_struct(q); |
31a7df01f cgroups: mechanis... |
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 |
} /* * If we had to process any tasks at all, scan again * in case some of them were in the middle of forking * children that didn't get processed. * Not the most efficient way to do it, but it avoids * having to take callback_mutex in the fork path */ goto again; } if (heap == &tmp_heap) heap_free(&tmp_heap); return 0; } |
817929ec2 Task Control Grou... |
2028 |
/* |
bbcb81d09 Task Control Grou... |
2029 2030 2031 2032 2033 2034 2035 |
* Stuff for reading the 'tasks' file. * * Reading this file can return large amounts of data if a cgroup has * *lots* of attached tasks. So it may need several calls to read(), * but we cannot guarantee that the information we produce is correct * unless we produce it entirely atomically. * |
bbcb81d09 Task Control Grou... |
2036 |
*/ |
bbcb81d09 Task Control Grou... |
2037 2038 2039 |
/* * Load into 'pidarray' up to 'npids' of the tasks using cgroup |
bd89aabc6 Control groups: R... |
2040 |
* 'cgrp'. Return actual number of pids loaded. No need to |
bbcb81d09 Task Control Grou... |
2041 2042 2043 2044 |
* task_lock(p) when reading out p->cgroup, since we're in an RCU * read section, so the css_set can't go away, and is * immutable after creation. */ |
bd89aabc6 Control groups: R... |
2045 |
static int pid_array_load(pid_t *pidarray, int npids, struct cgroup *cgrp) |
bbcb81d09 Task Control Grou... |
2046 |
{ |
e7b80bb69 cgroups: skip pro... |
2047 |
int n = 0, pid; |
817929ec2 Task Control Grou... |
2048 2049 |
struct cgroup_iter it; struct task_struct *tsk; |
bd89aabc6 Control groups: R... |
2050 2051 |
cgroup_iter_start(cgrp, &it); while ((tsk = cgroup_iter_next(cgrp, &it))) { |
817929ec2 Task Control Grou... |
2052 2053 |
if (unlikely(n == npids)) break; |
e7b80bb69 cgroups: skip pro... |
2054 2055 2056 |
pid = task_pid_vnr(tsk); if (pid > 0) pidarray[n++] = pid; |
817929ec2 Task Control Grou... |
2057 |
} |
bd89aabc6 Control groups: R... |
2058 |
cgroup_iter_end(cgrp, &it); |
bbcb81d09 Task Control Grou... |
2059 2060 |
return n; } |
846c7bb05 Add cgroupstats |
2061 |
/** |
a043e3b2c cgroup: fix comments |
2062 |
* cgroupstats_build - build and fill cgroupstats |
846c7bb05 Add cgroupstats |
2063 2064 2065 |
* @stats: cgroupstats to fill information into * @dentry: A dentry entry belonging to the cgroup for which stats have * been requested. |
a043e3b2c cgroup: fix comments |
2066 2067 2068 |
* * Build and fill cgroupstats so that taskstats can export it to user * space. |
846c7bb05 Add cgroupstats |
2069 2070 2071 2072 |
*/ int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry) { int ret = -EINVAL; |
bd89aabc6 Control groups: R... |
2073 |
struct cgroup *cgrp; |
846c7bb05 Add cgroupstats |
2074 2075 |
struct cgroup_iter it; struct task_struct *tsk; |
33d283bef cgroups: fix a se... |
2076 |
|
846c7bb05 Add cgroupstats |
2077 |
/* |
33d283bef cgroups: fix a se... |
2078 2079 |
* Validate dentry by checking the superblock operations, * and make sure it's a directory. |
846c7bb05 Add cgroupstats |
2080 |
*/ |
33d283bef cgroups: fix a se... |
2081 2082 |
if (dentry->d_sb->s_op != &cgroup_ops || !S_ISDIR(dentry->d_inode->i_mode)) |
846c7bb05 Add cgroupstats |
2083 2084 2085 |
goto err; ret = 0; |
bd89aabc6 Control groups: R... |
2086 |
cgrp = dentry->d_fsdata; |
846c7bb05 Add cgroupstats |
2087 |
|
bd89aabc6 Control groups: R... |
2088 2089 |
cgroup_iter_start(cgrp, &it); while ((tsk = cgroup_iter_next(cgrp, &it))) { |
846c7bb05 Add cgroupstats |
2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 |
switch (tsk->state) { case TASK_RUNNING: stats->nr_running++; break; case TASK_INTERRUPTIBLE: stats->nr_sleeping++; break; case TASK_UNINTERRUPTIBLE: stats->nr_uninterruptible++; break; case TASK_STOPPED: stats->nr_stopped++; break; default: if (delayacct_is_task_waiting_on_io(tsk)) stats->nr_io_wait++; break; } } |
bd89aabc6 Control groups: R... |
2109 |
cgroup_iter_end(cgrp, &it); |
846c7bb05 Add cgroupstats |
2110 |
|
846c7bb05 Add cgroupstats |
2111 2112 2113 |
err: return ret; } |
bbcb81d09 Task Control Grou... |
2114 2115 2116 2117 |
static int cmppid(const void *a, const void *b) { return *(pid_t *)a - *(pid_t *)b; } |
cc31edcee cgroups: convert ... |
2118 |
|
bbcb81d09 Task Control Grou... |
2119 |
/* |
cc31edcee cgroups: convert ... |
2120 2121 2122 |
* seq_file methods for the "tasks" file. The seq_file position is the * next pid to display; the seq_file iterator is a pointer to the pid * in the cgroup->tasks_pids array. |
bbcb81d09 Task Control Grou... |
2123 |
*/ |
cc31edcee cgroups: convert ... |
2124 2125 |
static void *cgroup_tasks_start(struct seq_file *s, loff_t *pos) |
bbcb81d09 Task Control Grou... |
2126 |
{ |
cc31edcee cgroups: convert ... |
2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 |
/* * Initially we receive a position value that corresponds to * one more than the last pid shown (or 0 on the first call or * after a seek to the start). Use a binary-search to find the * next pid to display, if any */ struct cgroup *cgrp = s->private; int index = 0, pid = *pos; int *iter; down_read(&cgrp->pids_mutex); if (pid) { int end = cgrp->pids_length; |
207777664 cgroup: remove un... |
2140 |
|
cc31edcee cgroups: convert ... |
2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 |
while (index < end) { int mid = (index + end) / 2; if (cgrp->tasks_pids[mid] == pid) { index = mid; break; } else if (cgrp->tasks_pids[mid] <= pid) index = mid + 1; else end = mid; } } /* If we're off the end of the array, we're done */ if (index >= cgrp->pids_length) return NULL; /* Update the abstract position to be the actual pid that we found */ iter = cgrp->tasks_pids + index; *pos = *iter; return iter; } static void cgroup_tasks_stop(struct seq_file *s, void *v) { struct cgroup *cgrp = s->private; up_read(&cgrp->pids_mutex); } static void *cgroup_tasks_next(struct seq_file *s, void *v, loff_t *pos) { struct cgroup *cgrp = s->private; int *p = v; int *end = cgrp->tasks_pids + cgrp->pids_length; /* * Advance to the next pid in the array. If this goes off the * end, we're done */ p++; if (p >= end) { return NULL; } else { *pos = *p; return p; } } static int cgroup_tasks_show(struct seq_file *s, void *v) { return seq_printf(s, "%d ", *(int *)v); } |
bbcb81d09 Task Control Grou... |
2191 |
|
cc31edcee cgroups: convert ... |
2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 |
static struct seq_operations cgroup_tasks_seq_operations = { .start = cgroup_tasks_start, .stop = cgroup_tasks_stop, .next = cgroup_tasks_next, .show = cgroup_tasks_show, }; static void release_cgroup_pid_array(struct cgroup *cgrp) { down_write(&cgrp->pids_mutex); BUG_ON(!cgrp->pids_use_count); if (!--cgrp->pids_use_count) { kfree(cgrp->tasks_pids); cgrp->tasks_pids = NULL; cgrp->pids_length = 0; } up_write(&cgrp->pids_mutex); |
bbcb81d09 Task Control Grou... |
2209 |
} |
cc31edcee cgroups: convert ... |
2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 |
static int cgroup_tasks_release(struct inode *inode, struct file *file) { struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); if (!(file->f_mode & FMODE_READ)) return 0; release_cgroup_pid_array(cgrp); return seq_release(inode, file); } static struct file_operations cgroup_tasks_operations = { .read = seq_read, .llseek = seq_lseek, .write = cgroup_file_write, .release = cgroup_tasks_release, }; |
bbcb81d09 Task Control Grou... |
2227 |
/* |
cc31edcee cgroups: convert ... |
2228 |
* Handle an open on 'tasks' file. Prepare an array containing the |
bbcb81d09 Task Control Grou... |
2229 |
* process id's of tasks currently attached to the cgroup being opened. |
bbcb81d09 Task Control Grou... |
2230 |
*/ |
cc31edcee cgroups: convert ... |
2231 |
|
bbcb81d09 Task Control Grou... |
2232 2233 |
static int cgroup_tasks_open(struct inode *unused, struct file *file) { |
bd89aabc6 Control groups: R... |
2234 |
struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); |
bbcb81d09 Task Control Grou... |
2235 2236 |
pid_t *pidarray; int npids; |
cc31edcee cgroups: convert ... |
2237 |
int retval; |
bbcb81d09 Task Control Grou... |
2238 |
|
cc31edcee cgroups: convert ... |
2239 |
/* Nothing to do for write-only files */ |
bbcb81d09 Task Control Grou... |
2240 2241 |
if (!(file->f_mode & FMODE_READ)) return 0; |
bbcb81d09 Task Control Grou... |
2242 2243 2244 2245 2246 2247 |
/* * If cgroup gets more users after we read count, we won't have * enough space - tough. This race is indistinguishable to the * caller from the case that the additional cgroup users didn't * show up until sometime later on. */ |
bd89aabc6 Control groups: R... |
2248 |
npids = cgroup_task_count(cgrp); |
cc31edcee cgroups: convert ... |
2249 2250 2251 2252 2253 |
pidarray = kmalloc(npids * sizeof(pid_t), GFP_KERNEL); if (!pidarray) return -ENOMEM; npids = pid_array_load(pidarray, npids, cgrp); sort(pidarray, npids, sizeof(pid_t), cmppid, NULL); |
bbcb81d09 Task Control Grou... |
2254 |
|
cc31edcee cgroups: convert ... |
2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 |
/* * Store the array in the cgroup, freeing the old * array if necessary */ down_write(&cgrp->pids_mutex); kfree(cgrp->tasks_pids); cgrp->tasks_pids = pidarray; cgrp->pids_length = npids; cgrp->pids_use_count++; up_write(&cgrp->pids_mutex); file->f_op = &cgroup_tasks_operations; retval = seq_open(file, &cgroup_tasks_seq_operations); if (retval) { release_cgroup_pid_array(cgrp); return retval; |
bbcb81d09 Task Control Grou... |
2272 |
} |
cc31edcee cgroups: convert ... |
2273 |
((struct seq_file *)file->private_data)->private = cgrp; |
bbcb81d09 Task Control Grou... |
2274 2275 |
return 0; } |
bd89aabc6 Control groups: R... |
2276 |
static u64 cgroup_read_notify_on_release(struct cgroup *cgrp, |
81a6a5cdd Task Control Grou... |
2277 2278 |
struct cftype *cft) { |
bd89aabc6 Control groups: R... |
2279 |
return notify_on_release(cgrp); |
81a6a5cdd Task Control Grou... |
2280 |
} |
6379c1061 cgroup files: mov... |
2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 |
static int cgroup_write_notify_on_release(struct cgroup *cgrp, struct cftype *cft, u64 val) { clear_bit(CGRP_RELEASABLE, &cgrp->flags); if (val) set_bit(CGRP_NOTIFY_ON_RELEASE, &cgrp->flags); else clear_bit(CGRP_NOTIFY_ON_RELEASE, &cgrp->flags); return 0; } |
bbcb81d09 Task Control Grou... |
2292 2293 2294 |
/* * for the common functions, 'private' gives the type of file */ |
81a6a5cdd Task Control Grou... |
2295 2296 2297 2298 |
static struct cftype files[] = { { .name = "tasks", .open = cgroup_tasks_open, |
af351026a cgroup files: tur... |
2299 |
.write_u64 = cgroup_tasks_write, |
81a6a5cdd Task Control Grou... |
2300 2301 |
.release = cgroup_tasks_release, .private = FILE_TASKLIST, |
099fca322 cgroups: show cor... |
2302 |
.mode = S_IRUGO | S_IWUSR, |
81a6a5cdd Task Control Grou... |
2303 2304 2305 2306 |
}, { .name = "notify_on_release", |
f4c753b7e CGroup API files:... |
2307 |
.read_u64 = cgroup_read_notify_on_release, |
6379c1061 cgroup files: mov... |
2308 |
.write_u64 = cgroup_write_notify_on_release, |
81a6a5cdd Task Control Grou... |
2309 2310 |
.private = FILE_NOTIFY_ON_RELEASE, }, |
81a6a5cdd Task Control Grou... |
2311 2312 2313 2314 |
}; static struct cftype cft_release_agent = { .name = "release_agent", |
e788e066c cgroup files: mov... |
2315 2316 2317 |
.read_seq_string = cgroup_release_agent_show, .write_string = cgroup_release_agent_write, .max_write_len = PATH_MAX, |
81a6a5cdd Task Control Grou... |
2318 |
.private = FILE_RELEASE_AGENT, |
bbcb81d09 Task Control Grou... |
2319 |
}; |
bd89aabc6 Control groups: R... |
2320 |
static int cgroup_populate_dir(struct cgroup *cgrp) |
ddbcc7e8e Task Control Grou... |
2321 2322 2323 2324 2325 |
{ int err; struct cgroup_subsys *ss; /* First clear out any existing files */ |
bd89aabc6 Control groups: R... |
2326 |
cgroup_clear_directory(cgrp->dentry); |
ddbcc7e8e Task Control Grou... |
2327 |
|
bd89aabc6 Control groups: R... |
2328 |
err = cgroup_add_files(cgrp, NULL, files, ARRAY_SIZE(files)); |
bbcb81d09 Task Control Grou... |
2329 2330 |
if (err < 0) return err; |
bd89aabc6 Control groups: R... |
2331 2332 |
if (cgrp == cgrp->top_cgroup) { if ((err = cgroup_add_file(cgrp, NULL, &cft_release_agent)) < 0) |
81a6a5cdd Task Control Grou... |
2333 2334 |
return err; } |
bd89aabc6 Control groups: R... |
2335 2336 |
for_each_subsys(cgrp->root, ss) { if (ss->populate && (err = ss->populate(ss, cgrp)) < 0) |
ddbcc7e8e Task Control Grou... |
2337 2338 |
return err; } |
38460b48d cgroup: CSS ID su... |
2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 |
/* This cgroup is ready now */ for_each_subsys(cgrp->root, ss) { struct cgroup_subsys_state *css = cgrp->subsys[ss->subsys_id]; /* * Update id->css pointer and make this css visible from * CSS ID functions. This pointer will be dereferened * from RCU-read-side without locks. */ if (css->id) rcu_assign_pointer(css->id->css, css); } |
ddbcc7e8e Task Control Grou... |
2350 2351 2352 2353 2354 2355 |
return 0; } static void init_cgroup_css(struct cgroup_subsys_state *css, struct cgroup_subsys *ss, |
bd89aabc6 Control groups: R... |
2356 |
struct cgroup *cgrp) |
ddbcc7e8e Task Control Grou... |
2357 |
{ |
bd89aabc6 Control groups: R... |
2358 |
css->cgroup = cgrp; |
e7c5ec919 cgroups: add css_... |
2359 |
atomic_set(&css->refcnt, 1); |
ddbcc7e8e Task Control Grou... |
2360 |
css->flags = 0; |
38460b48d cgroup: CSS ID su... |
2361 |
css->id = NULL; |
bd89aabc6 Control groups: R... |
2362 |
if (cgrp == dummytop) |
ddbcc7e8e Task Control Grou... |
2363 |
set_bit(CSS_ROOT, &css->flags); |
bd89aabc6 Control groups: R... |
2364 2365 |
BUG_ON(cgrp->subsys[ss->subsys_id]); cgrp->subsys[ss->subsys_id] = css; |
ddbcc7e8e Task Control Grou... |
2366 |
} |
999cd8a45 cgroups: add a pe... |
2367 2368 2369 2370 2371 2372 2373 2374 |
static void cgroup_lock_hierarchy(struct cgroupfs_root *root) { /* We need to take each hierarchy_mutex in a consistent order */ int i; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; if (ss->root == root) |
cfebe563b cgroups: fix lock... |
2375 |
mutex_lock(&ss->hierarchy_mutex); |
999cd8a45 cgroups: add a pe... |
2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 |
} } static void cgroup_unlock_hierarchy(struct cgroupfs_root *root) { int i; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; if (ss->root == root) mutex_unlock(&ss->hierarchy_mutex); } } |
ddbcc7e8e Task Control Grou... |
2389 |
/* |
a043e3b2c cgroup: fix comments |
2390 2391 2392 2393 |
* cgroup_create - create a cgroup * @parent: cgroup that will be parent of the new cgroup * @dentry: dentry of the new cgroup * @mode: mode to set on new inode |
ddbcc7e8e Task Control Grou... |
2394 |
* |
a043e3b2c cgroup: fix comments |
2395 |
* Must be called with the mutex on the parent inode held |
ddbcc7e8e Task Control Grou... |
2396 |
*/ |
ddbcc7e8e Task Control Grou... |
2397 |
static long cgroup_create(struct cgroup *parent, struct dentry *dentry, |
099fca322 cgroups: show cor... |
2398 |
mode_t mode) |
ddbcc7e8e Task Control Grou... |
2399 |
{ |
bd89aabc6 Control groups: R... |
2400 |
struct cgroup *cgrp; |
ddbcc7e8e Task Control Grou... |
2401 2402 2403 2404 |
struct cgroupfs_root *root = parent->root; int err = 0; struct cgroup_subsys *ss; struct super_block *sb = root->sb; |
bd89aabc6 Control groups: R... |
2405 2406 |
cgrp = kzalloc(sizeof(*cgrp), GFP_KERNEL); if (!cgrp) |
ddbcc7e8e Task Control Grou... |
2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 |
return -ENOMEM; /* Grab a reference on the superblock so the hierarchy doesn't * get deleted on unmount if there are child cgroups. This * can be done outside cgroup_mutex, since the sb can't * disappear while someone has an open control file on the * fs */ atomic_inc(&sb->s_active); mutex_lock(&cgroup_mutex); |
cc31edcee cgroups: convert ... |
2417 |
init_cgroup_housekeeping(cgrp); |
ddbcc7e8e Task Control Grou... |
2418 |
|
bd89aabc6 Control groups: R... |
2419 2420 2421 |
cgrp->parent = parent; cgrp->root = parent->root; cgrp->top_cgroup = parent->top_cgroup; |
ddbcc7e8e Task Control Grou... |
2422 |
|
b6abdb0e6 cgroup: fix defau... |
2423 2424 |
if (notify_on_release(parent)) set_bit(CGRP_NOTIFY_ON_RELEASE, &cgrp->flags); |
ddbcc7e8e Task Control Grou... |
2425 |
for_each_subsys(root, ss) { |
bd89aabc6 Control groups: R... |
2426 |
struct cgroup_subsys_state *css = ss->create(ss, cgrp); |
ddbcc7e8e Task Control Grou... |
2427 2428 2429 2430 |
if (IS_ERR(css)) { err = PTR_ERR(css); goto err_destroy; } |
bd89aabc6 Control groups: R... |
2431 |
init_cgroup_css(css, ss, cgrp); |
38460b48d cgroup: CSS ID su... |
2432 2433 2434 2435 |
if (ss->use_id) if (alloc_css_id(ss, parent, cgrp)) goto err_destroy; /* At error, ->destroy() callback has to free assigned ID. */ |
ddbcc7e8e Task Control Grou... |
2436 |
} |
999cd8a45 cgroups: add a pe... |
2437 |
cgroup_lock_hierarchy(root); |
bd89aabc6 Control groups: R... |
2438 |
list_add(&cgrp->sibling, &cgrp->parent->children); |
999cd8a45 cgroups: add a pe... |
2439 |
cgroup_unlock_hierarchy(root); |
ddbcc7e8e Task Control Grou... |
2440 |
root->number_of_cgroups++; |
bd89aabc6 Control groups: R... |
2441 |
err = cgroup_create_dir(cgrp, dentry, mode); |
ddbcc7e8e Task Control Grou... |
2442 2443 2444 2445 |
if (err < 0) goto err_remove; /* The cgroup directory was pre-locked for us */ |
bd89aabc6 Control groups: R... |
2446 |
BUG_ON(!mutex_is_locked(&cgrp->dentry->d_inode->i_mutex)); |
ddbcc7e8e Task Control Grou... |
2447 |
|
bd89aabc6 Control groups: R... |
2448 |
err = cgroup_populate_dir(cgrp); |
ddbcc7e8e Task Control Grou... |
2449 2450 2451 |
/* If err < 0, we have a half-filled directory - oh well ;) */ mutex_unlock(&cgroup_mutex); |
bd89aabc6 Control groups: R... |
2452 |
mutex_unlock(&cgrp->dentry->d_inode->i_mutex); |
ddbcc7e8e Task Control Grou... |
2453 2454 2455 2456 |
return 0; err_remove: |
baef99a08 cgroups: use hier... |
2457 |
cgroup_lock_hierarchy(root); |
bd89aabc6 Control groups: R... |
2458 |
list_del(&cgrp->sibling); |
baef99a08 cgroups: use hier... |
2459 |
cgroup_unlock_hierarchy(root); |
ddbcc7e8e Task Control Grou... |
2460 2461 2462 2463 2464 |
root->number_of_cgroups--; err_destroy: for_each_subsys(root, ss) { |
bd89aabc6 Control groups: R... |
2465 2466 |
if (cgrp->subsys[ss->subsys_id]) ss->destroy(ss, cgrp); |
ddbcc7e8e Task Control Grou... |
2467 2468 2469 2470 2471 2472 |
} mutex_unlock(&cgroup_mutex); /* Release the reference count that we took on the superblock */ deactivate_super(sb); |
bd89aabc6 Control groups: R... |
2473 |
kfree(cgrp); |
ddbcc7e8e Task Control Grou... |
2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 |
return err; } static int cgroup_mkdir(struct inode *dir, struct dentry *dentry, int mode) { struct cgroup *c_parent = dentry->d_parent->d_fsdata; /* the vfs holds inode->i_mutex already */ return cgroup_create(c_parent, dentry, mode | S_IFDIR); } |
55b6fd016 cgroup: uninline ... |
2484 |
static int cgroup_has_css_refs(struct cgroup *cgrp) |
81a6a5cdd Task Control Grou... |
2485 2486 2487 |
{ /* Check the reference count on each subsystem. Since we * already established that there are no tasks in the |
e7c5ec919 cgroups: add css_... |
2488 |
* cgroup, if the css refcount is also 1, then there should |
81a6a5cdd Task Control Grou... |
2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 |
* be no outstanding references, so the subsystem is safe to * destroy. We scan across all subsystems rather than using * the per-hierarchy linked list of mounted subsystems since * we can be called via check_for_release() with no * synchronization other than RCU, and the subsystem linked * list isn't RCU-safe */ int i; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; struct cgroup_subsys_state *css; /* Skip subsystems not in this hierarchy */ |
bd89aabc6 Control groups: R... |
2500 |
if (ss->root != cgrp->root) |
81a6a5cdd Task Control Grou... |
2501 |
continue; |
bd89aabc6 Control groups: R... |
2502 |
css = cgrp->subsys[ss->subsys_id]; |
81a6a5cdd Task Control Grou... |
2503 2504 2505 2506 2507 2508 |
/* When called from check_for_release() it's possible * that by this point the cgroup has been removed * and the css deleted. But a false-positive doesn't * matter, since it can only happen if the cgroup * has been deleted and hence no longer needs the * release agent to be called anyway. */ |
e7c5ec919 cgroups: add css_... |
2509 |
if (css && (atomic_read(&css->refcnt) > 1)) |
81a6a5cdd Task Control Grou... |
2510 |
return 1; |
81a6a5cdd Task Control Grou... |
2511 2512 2513 |
} return 0; } |
e7c5ec919 cgroups: add css_... |
2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 |
/* * Atomically mark all (or else none) of the cgroup's CSS objects as * CSS_REMOVED. Return true on success, or false if the cgroup has * busy subsystems. Call with cgroup_mutex held */ static int cgroup_clear_css_refs(struct cgroup *cgrp) { struct cgroup_subsys *ss; unsigned long flags; bool failed = false; local_irq_save(flags); for_each_subsys(cgrp->root, ss) { struct cgroup_subsys_state *css = cgrp->subsys[ss->subsys_id]; int refcnt; |
804b3c28a cgroups: add cpu_... |
2529 |
while (1) { |
e7c5ec919 cgroups: add css_... |
2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 |
/* We can only remove a CSS with a refcnt==1 */ refcnt = atomic_read(&css->refcnt); if (refcnt > 1) { failed = true; goto done; } BUG_ON(!refcnt); /* * Drop the refcnt to 0 while we check other * subsystems. This will cause any racing * css_tryget() to spin until we set the * CSS_REMOVED bits or abort */ |
804b3c28a cgroups: add cpu_... |
2543 2544 2545 2546 |
if (atomic_cmpxchg(&css->refcnt, refcnt, 0) == refcnt) break; cpu_relax(); } |
e7c5ec919 cgroups: add css_... |
2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 |
} done: for_each_subsys(cgrp->root, ss) { struct cgroup_subsys_state *css = cgrp->subsys[ss->subsys_id]; if (failed) { /* * Restore old refcnt if we previously managed * to clear it from 1 to 0 */ if (!atomic_read(&css->refcnt)) atomic_set(&css->refcnt, 1); } else { /* Commit the fact that the CSS is removed */ set_bit(CSS_REMOVED, &css->flags); } } local_irq_restore(flags); return !failed; } |
ddbcc7e8e Task Control Grou... |
2566 2567 |
static int cgroup_rmdir(struct inode *unused_dir, struct dentry *dentry) { |
bd89aabc6 Control groups: R... |
2568 |
struct cgroup *cgrp = dentry->d_fsdata; |
ddbcc7e8e Task Control Grou... |
2569 2570 |
struct dentry *d; struct cgroup *parent; |
ec64f5154 cgroup: fix frequ... |
2571 2572 |
DEFINE_WAIT(wait); int ret; |
ddbcc7e8e Task Control Grou... |
2573 2574 |
/* the vfs holds both inode->i_mutex already */ |
ec64f5154 cgroup: fix frequ... |
2575 |
again: |
ddbcc7e8e Task Control Grou... |
2576 |
mutex_lock(&cgroup_mutex); |
bd89aabc6 Control groups: R... |
2577 |
if (atomic_read(&cgrp->count) != 0) { |
ddbcc7e8e Task Control Grou... |
2578 2579 2580 |
mutex_unlock(&cgroup_mutex); return -EBUSY; } |
bd89aabc6 Control groups: R... |
2581 |
if (!list_empty(&cgrp->children)) { |
ddbcc7e8e Task Control Grou... |
2582 2583 2584 |
mutex_unlock(&cgroup_mutex); return -EBUSY; } |
3fa59dfbc cgroup: fix poten... |
2585 |
mutex_unlock(&cgroup_mutex); |
a043e3b2c cgroup: fix comments |
2586 |
|
4fca88c87 memory cgroup enh... |
2587 |
/* |
a043e3b2c cgroup: fix comments |
2588 2589 |
* Call pre_destroy handlers of subsys. Notify subsystems * that rmdir() request comes. |
4fca88c87 memory cgroup enh... |
2590 |
*/ |
ec64f5154 cgroup: fix frequ... |
2591 2592 2593 |
ret = cgroup_call_pre_destroy(cgrp); if (ret) return ret; |
ddbcc7e8e Task Control Grou... |
2594 |
|
3fa59dfbc cgroup: fix poten... |
2595 2596 |
mutex_lock(&cgroup_mutex); parent = cgrp->parent; |
ec64f5154 cgroup: fix frequ... |
2597 |
if (atomic_read(&cgrp->count) || !list_empty(&cgrp->children)) { |
ddbcc7e8e Task Control Grou... |
2598 2599 2600 |
mutex_unlock(&cgroup_mutex); return -EBUSY; } |
ec64f5154 cgroup: fix frequ... |
2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 |
/* * css_put/get is provided for subsys to grab refcnt to css. In typical * case, subsystem has no reference after pre_destroy(). But, under * hierarchy management, some *temporal* refcnt can be hold. * To avoid returning -EBUSY to a user, waitqueue is used. If subsys * is really busy, it should return -EBUSY at pre_destroy(). wake_up * is called when css_put() is called and refcnt goes down to 0. */ set_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); prepare_to_wait(&cgroup_rmdir_waitq, &wait, TASK_INTERRUPTIBLE); if (!cgroup_clear_css_refs(cgrp)) { mutex_unlock(&cgroup_mutex); schedule(); finish_wait(&cgroup_rmdir_waitq, &wait); clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); if (signal_pending(current)) return -EINTR; goto again; } /* NO css_tryget() can success after here. */ finish_wait(&cgroup_rmdir_waitq, &wait); clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); |
ddbcc7e8e Task Control Grou... |
2624 |
|
81a6a5cdd Task Control Grou... |
2625 |
spin_lock(&release_list_lock); |
bd89aabc6 Control groups: R... |
2626 2627 2628 |
set_bit(CGRP_REMOVED, &cgrp->flags); if (!list_empty(&cgrp->release_list)) list_del(&cgrp->release_list); |
81a6a5cdd Task Control Grou... |
2629 |
spin_unlock(&release_list_lock); |
999cd8a45 cgroups: add a pe... |
2630 2631 2632 |
cgroup_lock_hierarchy(cgrp->root); /* delete this cgroup from parent->children */ |
bd89aabc6 Control groups: R... |
2633 |
list_del(&cgrp->sibling); |
999cd8a45 cgroups: add a pe... |
2634 |
cgroup_unlock_hierarchy(cgrp->root); |
bd89aabc6 Control groups: R... |
2635 2636 |
spin_lock(&cgrp->dentry->d_lock); d = dget(cgrp->dentry); |
ddbcc7e8e Task Control Grou... |
2637 2638 2639 2640 |
spin_unlock(&d->d_lock); cgroup_d_remove_dir(d); dput(d); |
ddbcc7e8e Task Control Grou... |
2641 |
|
bd89aabc6 Control groups: R... |
2642 |
set_bit(CGRP_RELEASABLE, &parent->flags); |
81a6a5cdd Task Control Grou... |
2643 |
check_for_release(parent); |
ddbcc7e8e Task Control Grou... |
2644 |
mutex_unlock(&cgroup_mutex); |
ddbcc7e8e Task Control Grou... |
2645 2646 |
return 0; } |
06a119204 cgroup: annotate ... |
2647 |
static void __init cgroup_init_subsys(struct cgroup_subsys *ss) |
ddbcc7e8e Task Control Grou... |
2648 |
{ |
ddbcc7e8e Task Control Grou... |
2649 |
struct cgroup_subsys_state *css; |
cfe36bde5 Improve cgroup pr... |
2650 2651 2652 |
printk(KERN_INFO "Initializing cgroup subsys %s ", ss->name); |
ddbcc7e8e Task Control Grou... |
2653 2654 |
/* Create the top cgroup state for this subsystem */ |
33a68ac1c cgroups: add inac... |
2655 |
list_add(&ss->sibling, &rootnode.subsys_list); |
ddbcc7e8e Task Control Grou... |
2656 2657 2658 2659 2660 |
ss->root = &rootnode; css = ss->create(ss, dummytop); /* We don't handle early failures gracefully */ BUG_ON(IS_ERR(css)); init_cgroup_css(css, ss, dummytop); |
e8d55fdeb cgroups: simplify... |
2661 |
/* Update the init_css_set to contain a subsys |
817929ec2 Task Control Grou... |
2662 |
* pointer to this state - since the subsystem is |
e8d55fdeb cgroups: simplify... |
2663 2664 2665 |
* newly registered, all tasks and hence the * init_css_set is in the subsystem's top cgroup. */ init_css_set.subsys[ss->subsys_id] = dummytop->subsys[ss->subsys_id]; |
ddbcc7e8e Task Control Grou... |
2666 2667 |
need_forkexit_callback |= ss->fork || ss->exit; |
e8d55fdeb cgroups: simplify... |
2668 2669 2670 2671 |
/* At system boot, before all subsystems have been * registered, no tasks have been forked, so we don't * need to invoke fork callbacks here. */ BUG_ON(!list_empty(&init_task.tasks)); |
999cd8a45 cgroups: add a pe... |
2672 |
mutex_init(&ss->hierarchy_mutex); |
cfebe563b cgroups: fix lock... |
2673 |
lockdep_set_class(&ss->hierarchy_mutex, &ss->subsys_key); |
ddbcc7e8e Task Control Grou... |
2674 2675 2676 2677 |
ss->active = 1; } /** |
a043e3b2c cgroup: fix comments |
2678 2679 2680 2681 |
* cgroup_init_early - cgroup initialization at system boot * * Initialize cgroups at system boot, and initialize any * subsystems that request early init. |
ddbcc7e8e Task Control Grou... |
2682 2683 2684 2685 |
*/ int __init cgroup_init_early(void) { int i; |
146aa1bd0 cgroups: fix prob... |
2686 |
atomic_set(&init_css_set.refcount, 1); |
817929ec2 Task Control Grou... |
2687 2688 |
INIT_LIST_HEAD(&init_css_set.cg_links); INIT_LIST_HEAD(&init_css_set.tasks); |
472b1053f cgroups: use a ha... |
2689 |
INIT_HLIST_NODE(&init_css_set.hlist); |
817929ec2 Task Control Grou... |
2690 |
css_set_count = 1; |
ddbcc7e8e Task Control Grou... |
2691 |
init_cgroup_root(&rootnode); |
817929ec2 Task Control Grou... |
2692 2693 2694 2695 |
root_count = 1; init_task.cgroups = &init_css_set; init_css_set_link.cg = &init_css_set; |
bd89aabc6 Control groups: R... |
2696 |
list_add(&init_css_set_link.cgrp_link_list, |
817929ec2 Task Control Grou... |
2697 2698 2699 |
&rootnode.top_cgroup.css_sets); list_add(&init_css_set_link.cg_link_list, &init_css_set.cg_links); |
ddbcc7e8e Task Control Grou... |
2700 |
|
472b1053f cgroups: use a ha... |
2701 2702 |
for (i = 0; i < CSS_SET_TABLE_SIZE; i++) INIT_HLIST_HEAD(&css_set_table[i]); |
ddbcc7e8e Task Control Grou... |
2703 2704 2705 2706 2707 2708 2709 2710 |
for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; BUG_ON(!ss->name); BUG_ON(strlen(ss->name) > MAX_CGROUP_TYPE_NAMELEN); BUG_ON(!ss->create); BUG_ON(!ss->destroy); if (ss->subsys_id != i) { |
cfe36bde5 Improve cgroup pr... |
2711 2712 |
printk(KERN_ERR "cgroup: Subsys %s id == %d ", |
ddbcc7e8e Task Control Grou... |
2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 |
ss->name, ss->subsys_id); BUG(); } if (ss->early_init) cgroup_init_subsys(ss); } return 0; } /** |
a043e3b2c cgroup: fix comments |
2724 2725 2726 2727 |
* cgroup_init - cgroup initialization * * Register cgroup filesystem and /proc file, and initialize * any subsystems that didn't request early init. |
ddbcc7e8e Task Control Grou... |
2728 2729 2730 2731 2732 |
*/ int __init cgroup_init(void) { int err; int i; |
472b1053f cgroups: use a ha... |
2733 |
struct hlist_head *hhead; |
a424316ca Task Control Grou... |
2734 2735 2736 2737 |
err = bdi_init(&cgroup_backing_dev_info); if (err) return err; |
ddbcc7e8e Task Control Grou... |
2738 2739 2740 2741 2742 |
for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; if (!ss->early_init) cgroup_init_subsys(ss); |
38460b48d cgroup: CSS ID su... |
2743 2744 |
if (ss->use_id) cgroup_subsys_init_idr(ss); |
ddbcc7e8e Task Control Grou... |
2745 |
} |
472b1053f cgroups: use a ha... |
2746 2747 2748 |
/* Add init_css_set to the hash table */ hhead = css_set_hash(init_css_set.subsys); hlist_add_head(&init_css_set.hlist, hhead); |
ddbcc7e8e Task Control Grou... |
2749 2750 2751 |
err = register_filesystem(&cgroup_fs_type); if (err < 0) goto out; |
46ae220be cgroup: switch to... |
2752 |
proc_create("cgroups", 0, NULL, &proc_cgroupstats_operations); |
a424316ca Task Control Grou... |
2753 |
|
ddbcc7e8e Task Control Grou... |
2754 |
out: |
a424316ca Task Control Grou... |
2755 2756 |
if (err) bdi_destroy(&cgroup_backing_dev_info); |
ddbcc7e8e Task Control Grou... |
2757 2758 |
return err; } |
b4f48b636 Task Control Grou... |
2759 |
|
a424316ca Task Control Grou... |
2760 2761 2762 2763 2764 2765 |
/* * proc_cgroup_show() * - Print task's cgroup paths into seq_file, one line for each hierarchy * - Used for /proc/<pid>/cgroup. * - No need to task_lock(tsk) on this tsk->cgroup reference, as it * doesn't really matter if tsk->cgroup changes after we read it, |
956db3ca0 hotplug cpu: move... |
2766 |
* and we take cgroup_mutex, keeping cgroup_attach_task() from changing it |
a424316ca Task Control Grou... |
2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 |
* anyway. No need to check that tsk->cgroup != NULL, thanks to * the_top_cgroup_hack in cgroup_exit(), which sets an exiting tasks * cgroup to top_cgroup. */ /* TODO: Use a proper seq_file iterator */ static int proc_cgroup_show(struct seq_file *m, void *v) { struct pid *pid; struct task_struct *tsk; char *buf; int retval; struct cgroupfs_root *root; retval = -ENOMEM; buf = kmalloc(PAGE_SIZE, GFP_KERNEL); if (!buf) goto out; retval = -ESRCH; pid = m->private; tsk = get_pid_task(pid, PIDTYPE_PID); if (!tsk) goto out_free; retval = 0; mutex_lock(&cgroup_mutex); |
e5f6a8609 cgroups: make roo... |
2795 |
for_each_active_root(root) { |
a424316ca Task Control Grou... |
2796 |
struct cgroup_subsys *ss; |
bd89aabc6 Control groups: R... |
2797 |
struct cgroup *cgrp; |
a424316ca Task Control Grou... |
2798 2799 |
int subsys_id; int count = 0; |
b6c3006d2 cgroups: include ... |
2800 |
seq_printf(m, "%lu:", root->subsys_bits); |
a424316ca Task Control Grou... |
2801 2802 2803 2804 |
for_each_subsys(root, ss) seq_printf(m, "%s%s", count++ ? "," : "", ss->name); seq_putc(m, ':'); get_first_subsys(&root->top_cgroup, NULL, &subsys_id); |
bd89aabc6 Control groups: R... |
2805 2806 |
cgrp = task_cgroup(tsk, subsys_id); retval = cgroup_path(cgrp, buf, PAGE_SIZE); |
a424316ca Task Control Grou... |
2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 |
if (retval < 0) goto out_unlock; seq_puts(m, buf); seq_putc(m, ' '); } out_unlock: mutex_unlock(&cgroup_mutex); put_task_struct(tsk); out_free: kfree(buf); out: return retval; } static int cgroup_open(struct inode *inode, struct file *file) { struct pid *pid = PROC_I(inode)->pid; return single_open(file, proc_cgroup_show, pid); } struct file_operations proc_cgroup_operations = { .open = cgroup_open, .read = seq_read, .llseek = seq_lseek, .release = single_release, }; /* Display information about each subsystem and each hierarchy */ static int proc_cgroupstats_show(struct seq_file *m, void *v) { int i; |
a424316ca Task Control Grou... |
2840 |
|
8bab8dded cgroups: add cgro... |
2841 2842 |
seq_puts(m, "#subsys_name\thierarchy\tnum_cgroups\tenabled "); |
a424316ca Task Control Grou... |
2843 |
mutex_lock(&cgroup_mutex); |
a424316ca Task Control Grou... |
2844 2845 |
for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; |
8bab8dded cgroups: add cgro... |
2846 2847 |
seq_printf(m, "%s\t%lu\t%d\t%d ", |
817929ec2 Task Control Grou... |
2848 |
ss->name, ss->root->subsys_bits, |
8bab8dded cgroups: add cgro... |
2849 |
ss->root->number_of_cgroups, !ss->disabled); |
a424316ca Task Control Grou... |
2850 2851 2852 2853 2854 2855 2856 |
} mutex_unlock(&cgroup_mutex); return 0; } static int cgroupstats_open(struct inode *inode, struct file *file) { |
9dce07f1a NULL noise: fs/*,... |
2857 |
return single_open(file, proc_cgroupstats_show, NULL); |
a424316ca Task Control Grou... |
2858 2859 2860 2861 2862 2863 2864 2865 |
} static struct file_operations proc_cgroupstats_operations = { .open = cgroupstats_open, .read = seq_read, .llseek = seq_lseek, .release = single_release, }; |
b4f48b636 Task Control Grou... |
2866 2867 |
/** * cgroup_fork - attach newly forked task to its parents cgroup. |
a043e3b2c cgroup: fix comments |
2868 |
* @child: pointer to task_struct of forking parent process. |
b4f48b636 Task Control Grou... |
2869 2870 2871 2872 2873 2874 |
* * Description: A task inherits its parent's cgroup at fork(). * * A pointer to the shared css_set was automatically copied in * fork.c by dup_task_struct(). However, we ignore that copy, since * it was not made under the protection of RCU or cgroup_mutex, so |
956db3ca0 hotplug cpu: move... |
2875 |
* might no longer be a valid cgroup pointer. cgroup_attach_task() might |
817929ec2 Task Control Grou... |
2876 2877 |
* have already changed current->cgroups, allowing the previously * referenced cgroup group to be removed and freed. |
b4f48b636 Task Control Grou... |
2878 2879 2880 2881 2882 2883 |
* * At the point that cgroup_fork() is called, 'current' is the parent * task, and the passed argument 'child' points to the child task. */ void cgroup_fork(struct task_struct *child) { |
817929ec2 Task Control Grou... |
2884 2885 2886 2887 2888 |
task_lock(current); child->cgroups = current->cgroups; get_css_set(child->cgroups); task_unlock(current); INIT_LIST_HEAD(&child->cg_list); |
b4f48b636 Task Control Grou... |
2889 2890 2891 |
} /** |
a043e3b2c cgroup: fix comments |
2892 2893 2894 2895 2896 2897 |
* cgroup_fork_callbacks - run fork callbacks * @child: the new task * * Called on a new task very soon before adding it to the * tasklist. No need to take any locks since no-one can * be operating on this task. |
b4f48b636 Task Control Grou... |
2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 |
*/ void cgroup_fork_callbacks(struct task_struct *child) { if (need_forkexit_callback) { int i; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; if (ss->fork) ss->fork(ss, child); } } } /** |
a043e3b2c cgroup: fix comments |
2912 2913 2914 2915 2916 2917 2918 2919 |
* cgroup_post_fork - called on a new task after adding it to the task list * @child: the task in question * * Adds the task to the list running through its css_set if necessary. * Has to be after the task is visible on the task list in case we race * with the first call to cgroup_iter_start() - to guarantee that the * new task ends up on its list. */ |
817929ec2 Task Control Grou... |
2920 2921 2922 2923 |
void cgroup_post_fork(struct task_struct *child) { if (use_task_css_set_links) { write_lock(&css_set_lock); |
b12b533fa cgroups: add lock... |
2924 |
task_lock(child); |
817929ec2 Task Control Grou... |
2925 2926 |
if (list_empty(&child->cg_list)) list_add(&child->cg_list, &child->cgroups->tasks); |
b12b533fa cgroups: add lock... |
2927 |
task_unlock(child); |
817929ec2 Task Control Grou... |
2928 2929 2930 2931 |
write_unlock(&css_set_lock); } } /** |
b4f48b636 Task Control Grou... |
2932 2933 |
* cgroup_exit - detach cgroup from exiting task * @tsk: pointer to task_struct of exiting process |
a043e3b2c cgroup: fix comments |
2934 |
* @run_callback: run exit callbacks? |
b4f48b636 Task Control Grou... |
2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 |
* * Description: Detach cgroup from @tsk and release it. * * Note that cgroups marked notify_on_release force every task in * them to take the global cgroup_mutex mutex when exiting. * This could impact scaling on very large systems. Be reluctant to * use notify_on_release cgroups where very high task exit scaling * is required on large systems. * * the_top_cgroup_hack: * * Set the exiting tasks cgroup to the root cgroup (top_cgroup). * * We call cgroup_exit() while the task is still competent to * handle notify_on_release(), then leave the task attached to the * root cgroup in each hierarchy for the remainder of its exit. * * To do this properly, we would increment the reference count on * top_cgroup, and near the very end of the kernel/exit.c do_exit() * code we would add a second cgroup function call, to drop that * reference. This would just create an unnecessary hot spot on * the top_cgroup reference count, to no avail. * * Normally, holding a reference to a cgroup without bumping its * count is unsafe. The cgroup could go away, or someone could * attach us to a different cgroup, decrementing the count on * the first cgroup that we never incremented. But in this case, * top_cgroup isn't going away, and either task has PF_EXITING set, |
956db3ca0 hotplug cpu: move... |
2963 2964 |
* which wards off any cgroup_attach_task() attempts, or task is a failed * fork, never visible to cgroup_attach_task. |
b4f48b636 Task Control Grou... |
2965 2966 2967 2968 |
*/ void cgroup_exit(struct task_struct *tsk, int run_callbacks) { int i; |
817929ec2 Task Control Grou... |
2969 |
struct css_set *cg; |
b4f48b636 Task Control Grou... |
2970 2971 2972 2973 2974 2975 2976 2977 |
if (run_callbacks && need_forkexit_callback) { for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; if (ss->exit) ss->exit(ss, tsk); } } |
817929ec2 Task Control Grou... |
2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 |
/* * Unlink from the css_set task list if necessary. * Optimistically check cg_list before taking * css_set_lock */ if (!list_empty(&tsk->cg_list)) { write_lock(&css_set_lock); if (!list_empty(&tsk->cg_list)) list_del(&tsk->cg_list); write_unlock(&css_set_lock); } |
b4f48b636 Task Control Grou... |
2990 2991 |
/* Reassign the task to the init_css_set. */ task_lock(tsk); |
817929ec2 Task Control Grou... |
2992 2993 |
cg = tsk->cgroups; tsk->cgroups = &init_css_set; |
b4f48b636 Task Control Grou... |
2994 |
task_unlock(tsk); |
817929ec2 Task Control Grou... |
2995 |
if (cg) |
81a6a5cdd Task Control Grou... |
2996 |
put_css_set_taskexit(cg); |
b4f48b636 Task Control Grou... |
2997 |
} |
697f41610 Task Control Grou... |
2998 2999 |
/** |
a043e3b2c cgroup: fix comments |
3000 3001 3002 |
* cgroup_clone - clone the cgroup the given subsystem is attached to * @tsk: the task to be moved * @subsys: the given subsystem |
e885dcde7 cgroup_clone: use... |
3003 |
* @nodename: the name for the new cgroup |
a043e3b2c cgroup: fix comments |
3004 3005 3006 3007 |
* * Duplicate the current cgroup in the hierarchy that the given * subsystem is attached to, and move this task into the new * child. |
697f41610 Task Control Grou... |
3008 |
*/ |
e885dcde7 cgroup_clone: use... |
3009 3010 |
int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *subsys, char *nodename) |
697f41610 Task Control Grou... |
3011 3012 3013 |
{ struct dentry *dentry; int ret = 0; |
697f41610 Task Control Grou... |
3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 |
struct cgroup *parent, *child; struct inode *inode; struct css_set *cg; struct cgroupfs_root *root; struct cgroup_subsys *ss; /* We shouldn't be called by an unregistered subsystem */ BUG_ON(!subsys->active); /* First figure out what hierarchy and cgroup we're dealing * with, and pin them so we can drop cgroup_mutex */ mutex_lock(&cgroup_mutex); again: root = subsys->root; if (root == &rootnode) { |
697f41610 Task Control Grou... |
3029 3030 3031 |
mutex_unlock(&cgroup_mutex); return 0; } |
697f41610 Task Control Grou... |
3032 |
|
697f41610 Task Control Grou... |
3033 |
/* Pin the hierarchy */ |
1404f0656 cgroups: fix lock... |
3034 |
if (!atomic_inc_not_zero(&root->sb->s_active)) { |
7b574b7b0 cgroups: fix a ra... |
3035 3036 3037 3038 |
/* We race with the final deactivate_super() */ mutex_unlock(&cgroup_mutex); return 0; } |
697f41610 Task Control Grou... |
3039 |
|
817929ec2 Task Control Grou... |
3040 |
/* Keep the cgroup alive */ |
1404f0656 cgroups: fix lock... |
3041 3042 3043 |
task_lock(tsk); parent = task_cgroup(tsk, subsys->subsys_id); cg = tsk->cgroups; |
817929ec2 Task Control Grou... |
3044 |
get_css_set(cg); |
104cbd553 cgroups: use task... |
3045 |
task_unlock(tsk); |
1404f0656 cgroups: fix lock... |
3046 |
|
697f41610 Task Control Grou... |
3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 |
mutex_unlock(&cgroup_mutex); /* Now do the VFS work to create a cgroup */ inode = parent->dentry->d_inode; /* Hold the parent directory mutex across this operation to * stop anyone else deleting the new cgroup */ mutex_lock(&inode->i_mutex); dentry = lookup_one_len(nodename, parent->dentry, strlen(nodename)); if (IS_ERR(dentry)) { printk(KERN_INFO |
cfe36bde5 Improve cgroup pr... |
3058 3059 |
"cgroup: Couldn't allocate dentry for %s: %ld ", nodename, |
697f41610 Task Control Grou... |
3060 3061 3062 3063 3064 3065 |
PTR_ERR(dentry)); ret = PTR_ERR(dentry); goto out_release; } /* Create the cgroup directory, which also creates the cgroup */ |
75139b827 cgroups: remove s... |
3066 |
ret = vfs_mkdir(inode, dentry, 0755); |
bd89aabc6 Control groups: R... |
3067 |
child = __d_cgrp(dentry); |
697f41610 Task Control Grou... |
3068 3069 3070 3071 3072 3073 3074 3075 |
dput(dentry); if (ret) { printk(KERN_INFO "Failed to create cgroup %s: %d ", nodename, ret); goto out_release; } |
697f41610 Task Control Grou... |
3076 3077 3078 3079 3080 3081 3082 3083 |
/* The cgroup now exists. Retake cgroup_mutex and check * that we're still in the same state that we thought we * were. */ mutex_lock(&cgroup_mutex); if ((root != subsys->root) || (parent != task_cgroup(tsk, subsys->subsys_id))) { /* Aargh, we raced ... */ mutex_unlock(&inode->i_mutex); |
817929ec2 Task Control Grou... |
3084 |
put_css_set(cg); |
697f41610 Task Control Grou... |
3085 |
|
1404f0656 cgroups: fix lock... |
3086 |
deactivate_super(root->sb); |
697f41610 Task Control Grou... |
3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 |
/* The cgroup is still accessible in the VFS, but * we're not going to try to rmdir() it at this * point. */ printk(KERN_INFO "Race in cgroup_clone() - leaking cgroup %s ", nodename); goto again; } /* do any required auto-setup */ for_each_subsys(root, ss) { if (ss->post_clone) ss->post_clone(ss, child); } /* All seems fine. Finish by moving the task into the new cgroup */ |
956db3ca0 hotplug cpu: move... |
3104 |
ret = cgroup_attach_task(child, tsk); |
697f41610 Task Control Grou... |
3105 3106 3107 3108 |
mutex_unlock(&cgroup_mutex); out_release: mutex_unlock(&inode->i_mutex); |
81a6a5cdd Task Control Grou... |
3109 3110 |
mutex_lock(&cgroup_mutex); |
817929ec2 Task Control Grou... |
3111 |
put_css_set(cg); |
81a6a5cdd Task Control Grou... |
3112 |
mutex_unlock(&cgroup_mutex); |
1404f0656 cgroups: fix lock... |
3113 |
deactivate_super(root->sb); |
697f41610 Task Control Grou... |
3114 3115 |
return ret; } |
a043e3b2c cgroup: fix comments |
3116 |
/** |
313e924c0 cgroups: relax ns... |
3117 |
* cgroup_is_descendant - see if @cgrp is a descendant of @task's cgrp |
a043e3b2c cgroup: fix comments |
3118 |
* @cgrp: the cgroup in question |
313e924c0 cgroups: relax ns... |
3119 |
* @task: the task in question |
a043e3b2c cgroup: fix comments |
3120 |
* |
313e924c0 cgroups: relax ns... |
3121 3122 |
* See if @cgrp is a descendant of @task's cgroup in the appropriate * hierarchy. |
697f41610 Task Control Grou... |
3123 3124 3125 3126 3127 3128 |
* * If we are sending in dummytop, then presumably we are creating * the top cgroup in the subsystem. * * Called only by the ns (nsproxy) cgroup. */ |
313e924c0 cgroups: relax ns... |
3129 |
int cgroup_is_descendant(const struct cgroup *cgrp, struct task_struct *task) |
697f41610 Task Control Grou... |
3130 3131 3132 3133 |
{ int ret; struct cgroup *target; int subsys_id; |
bd89aabc6 Control groups: R... |
3134 |
if (cgrp == dummytop) |
697f41610 Task Control Grou... |
3135 |
return 1; |
bd89aabc6 Control groups: R... |
3136 |
get_first_subsys(cgrp, NULL, &subsys_id); |
313e924c0 cgroups: relax ns... |
3137 |
target = task_cgroup(task, subsys_id); |
bd89aabc6 Control groups: R... |
3138 3139 3140 |
while (cgrp != target && cgrp!= cgrp->top_cgroup) cgrp = cgrp->parent; ret = (cgrp == target); |
697f41610 Task Control Grou... |
3141 3142 |
return ret; } |
81a6a5cdd Task Control Grou... |
3143 |
|
bd89aabc6 Control groups: R... |
3144 |
static void check_for_release(struct cgroup *cgrp) |
81a6a5cdd Task Control Grou... |
3145 3146 3147 |
{ /* All of these checks rely on RCU to keep the cgroup * structure alive */ |
bd89aabc6 Control groups: R... |
3148 3149 |
if (cgroup_is_releasable(cgrp) && !atomic_read(&cgrp->count) && list_empty(&cgrp->children) && !cgroup_has_css_refs(cgrp)) { |
81a6a5cdd Task Control Grou... |
3150 3151 3152 3153 3154 |
/* Control Group is currently removeable. If it's not * already queued for a userspace notification, queue * it now */ int need_schedule_work = 0; spin_lock(&release_list_lock); |
bd89aabc6 Control groups: R... |
3155 3156 3157 |
if (!cgroup_is_removed(cgrp) && list_empty(&cgrp->release_list)) { list_add(&cgrp->release_list, &release_list); |
81a6a5cdd Task Control Grou... |
3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 |
need_schedule_work = 1; } spin_unlock(&release_list_lock); if (need_schedule_work) schedule_work(&release_agent_work); } } void __css_put(struct cgroup_subsys_state *css) { |
bd89aabc6 Control groups: R... |
3168 |
struct cgroup *cgrp = css->cgroup; |
81a6a5cdd Task Control Grou... |
3169 |
rcu_read_lock(); |
ec64f5154 cgroup: fix frequ... |
3170 3171 3172 3173 3174 3175 |
if (atomic_dec_return(&css->refcnt) == 1) { if (notify_on_release(cgrp)) { set_bit(CGRP_RELEASABLE, &cgrp->flags); check_for_release(cgrp); } cgroup_wakeup_rmdir_waiters(cgrp); |
81a6a5cdd Task Control Grou... |
3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 |
} rcu_read_unlock(); } /* * Notify userspace when a cgroup is released, by running the * configured release agent with the name of the cgroup (path * relative to the root of cgroup file system) as the argument. * * Most likely, this user command will try to rmdir this cgroup. * * This races with the possibility that some other task will be * attached to this cgroup before it is removed, or that some other * user task will 'mkdir' a child cgroup of this cgroup. That's ok. * The presumed 'rmdir' will fail quietly if this cgroup is no longer * unused, and this cgroup will be reprieved from its death sentence, * to continue to serve a useful existence. Next time it's released, * we will get notified again, if it still has 'notify_on_release' set. * * The final arg to call_usermodehelper() is UMH_WAIT_EXEC, which * means only wait until the task is successfully execve()'d. The * separate release agent task is forked by call_usermodehelper(), * then control in this thread returns here, without waiting for the * release agent task. We don't bother to wait because the caller of * this routine has no use for the exit status of the release agent * task, so no sense holding our caller up for that. |
81a6a5cdd Task Control Grou... |
3202 |
*/ |
81a6a5cdd Task Control Grou... |
3203 3204 3205 3206 3207 3208 3209 3210 |
static void cgroup_release_agent(struct work_struct *work) { BUG_ON(work != &release_agent_work); mutex_lock(&cgroup_mutex); spin_lock(&release_list_lock); while (!list_empty(&release_list)) { char *argv[3], *envp[3]; int i; |
e788e066c cgroup files: mov... |
3211 |
char *pathbuf = NULL, *agentbuf = NULL; |
bd89aabc6 Control groups: R... |
3212 |
struct cgroup *cgrp = list_entry(release_list.next, |
81a6a5cdd Task Control Grou... |
3213 3214 |
struct cgroup, release_list); |
bd89aabc6 Control groups: R... |
3215 |
list_del_init(&cgrp->release_list); |
81a6a5cdd Task Control Grou... |
3216 3217 |
spin_unlock(&release_list_lock); pathbuf = kmalloc(PAGE_SIZE, GFP_KERNEL); |
e788e066c cgroup files: mov... |
3218 3219 3220 3221 3222 3223 3224 |
if (!pathbuf) goto continue_free; if (cgroup_path(cgrp, pathbuf, PAGE_SIZE) < 0) goto continue_free; agentbuf = kstrdup(cgrp->root->release_agent_path, GFP_KERNEL); if (!agentbuf) goto continue_free; |
81a6a5cdd Task Control Grou... |
3225 3226 |
i = 0; |
e788e066c cgroup files: mov... |
3227 3228 |
argv[i++] = agentbuf; argv[i++] = pathbuf; |
81a6a5cdd Task Control Grou... |
3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 |
argv[i] = NULL; i = 0; /* minimal command environment */ envp[i++] = "HOME=/"; envp[i++] = "PATH=/sbin:/bin:/usr/sbin:/usr/bin"; envp[i] = NULL; /* Drop the lock while we invoke the usermode helper, * since the exec could involve hitting disk and hence * be a slow process */ mutex_unlock(&cgroup_mutex); call_usermodehelper(argv[0], argv, envp, UMH_WAIT_EXEC); |
81a6a5cdd Task Control Grou... |
3242 |
mutex_lock(&cgroup_mutex); |
e788e066c cgroup files: mov... |
3243 3244 3245 |
continue_free: kfree(pathbuf); kfree(agentbuf); |
81a6a5cdd Task Control Grou... |
3246 3247 3248 3249 3250 |
spin_lock(&release_list_lock); } spin_unlock(&release_list_lock); mutex_unlock(&cgroup_mutex); } |
8bab8dded cgroups: add cgro... |
3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 |
static int __init cgroup_disable(char *str) { int i; char *token; while ((token = strsep(&str, ",")) != NULL) { if (!*token) continue; for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) { struct cgroup_subsys *ss = subsys[i]; if (!strcmp(token, ss->name)) { ss->disabled = 1; printk(KERN_INFO "Disabling %s control group" " subsystem ", ss->name); break; } } } return 1; } __setup("cgroup_disable=", cgroup_disable); |
38460b48d cgroup: CSS ID su... |
3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 |
/* * Functons for CSS ID. */ /* *To get ID other than 0, this should be called when !cgroup_is_removed(). */ unsigned short css_id(struct cgroup_subsys_state *css) { struct css_id *cssid = rcu_dereference(css->id); if (cssid) return cssid->id; return 0; } unsigned short css_depth(struct cgroup_subsys_state *css) { struct css_id *cssid = rcu_dereference(css->id); if (cssid) return cssid->depth; return 0; } bool css_is_ancestor(struct cgroup_subsys_state *child, |
0b7f569e4 memcg: fix OOM ki... |
3303 |
const struct cgroup_subsys_state *root) |
38460b48d cgroup: CSS ID su... |
3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 |
{ struct css_id *child_id = rcu_dereference(child->id); struct css_id *root_id = rcu_dereference(root->id); if (!child_id || !root_id || (child_id->depth < root_id->depth)) return false; return child_id->stack[root_id->depth] == root_id->id; } static void __free_css_id_cb(struct rcu_head *head) { struct css_id *id; id = container_of(head, struct css_id, rcu_head); kfree(id); } void free_css_id(struct cgroup_subsys *ss, struct cgroup_subsys_state *css) { struct css_id *id = css->id; /* When this is called before css_id initialization, id can be NULL */ if (!id) return; BUG_ON(!ss->use_id); rcu_assign_pointer(id->css, NULL); rcu_assign_pointer(css->id, NULL); spin_lock(&ss->id_lock); idr_remove(&ss->idr, id->id); spin_unlock(&ss->id_lock); call_rcu(&id->rcu_head, __free_css_id_cb); } /* * This is called by init or create(). Then, calls to this function are * always serialized (By cgroup_mutex() at create()). */ static struct css_id *get_new_cssid(struct cgroup_subsys *ss, int depth) { struct css_id *newid; int myid, error, size; BUG_ON(!ss->use_id); size = sizeof(*newid) + sizeof(unsigned short) * (depth + 1); newid = kzalloc(size, GFP_KERNEL); if (!newid) return ERR_PTR(-ENOMEM); /* get id */ if (unlikely(!idr_pre_get(&ss->idr, GFP_KERNEL))) { error = -ENOMEM; goto err_out; } spin_lock(&ss->id_lock); /* Don't use 0. allocates an ID of 1-65535 */ error = idr_get_new_above(&ss->idr, newid, 1, &myid); spin_unlock(&ss->id_lock); /* Returns error when there are no free spaces for new ID.*/ if (error) { error = -ENOSPC; goto err_out; } if (myid > CSS_ID_MAX) goto remove_idr; newid->id = myid; newid->depth = depth; return newid; remove_idr: error = -ENOSPC; spin_lock(&ss->id_lock); idr_remove(&ss->idr, myid); spin_unlock(&ss->id_lock); err_out: kfree(newid); return ERR_PTR(error); } static int __init cgroup_subsys_init_idr(struct cgroup_subsys *ss) { struct css_id *newid; struct cgroup_subsys_state *rootcss; spin_lock_init(&ss->id_lock); idr_init(&ss->idr); rootcss = init_css_set.subsys[ss->subsys_id]; newid = get_new_cssid(ss, 0); if (IS_ERR(newid)) return PTR_ERR(newid); newid->stack[0] = newid->id; newid->css = rootcss; rootcss->id = newid; return 0; } static int alloc_css_id(struct cgroup_subsys *ss, struct cgroup *parent, struct cgroup *child) { int subsys_id, i, depth = 0; struct cgroup_subsys_state *parent_css, *child_css; struct css_id *child_id, *parent_id = NULL; subsys_id = ss->subsys_id; parent_css = parent->subsys[subsys_id]; child_css = child->subsys[subsys_id]; depth = css_depth(parent_css) + 1; parent_id = parent_css->id; child_id = get_new_cssid(ss, depth); if (IS_ERR(child_id)) return PTR_ERR(child_id); for (i = 0; i < depth; i++) child_id->stack[i] = parent_id->stack[i]; child_id->stack[depth] = child_id->id; /* * child_id->css pointer will be set after this cgroup is available * see cgroup_populate_dir() */ rcu_assign_pointer(child_css->id, child_id); return 0; } /** * css_lookup - lookup css by id * @ss: cgroup subsys to be looked into. * @id: the id * * Returns pointer to cgroup_subsys_state if there is valid one with id. * NULL if not. Should be called under rcu_read_lock() */ struct cgroup_subsys_state *css_lookup(struct cgroup_subsys *ss, int id) { struct css_id *cssid = NULL; BUG_ON(!ss->use_id); cssid = idr_find(&ss->idr, id); if (unlikely(!cssid)) return NULL; return rcu_dereference(cssid->css); } /** * css_get_next - lookup next cgroup under specified hierarchy. * @ss: pointer to subsystem * @id: current position of iteration. * @root: pointer to css. search tree under this. * @foundid: position of found object. * * Search next css under the specified hierarchy of rootid. Calling under * rcu_read_lock() is necessary. Returns NULL if it reaches the end. */ struct cgroup_subsys_state * css_get_next(struct cgroup_subsys *ss, int id, struct cgroup_subsys_state *root, int *foundid) { struct cgroup_subsys_state *ret = NULL; struct css_id *tmp; int tmpid; int rootid = css_id(root); int depth = css_depth(root); if (!rootid) return NULL; BUG_ON(!ss->use_id); /* fill start point for scan */ tmpid = id; while (1) { /* * scan next entry from bitmap(tree), tmpid is updated after * idr_get_next(). */ spin_lock(&ss->id_lock); tmp = idr_get_next(&ss->idr, &tmpid); spin_unlock(&ss->id_lock); if (!tmp) break; if (tmp->depth >= depth && tmp->stack[depth] == rootid) { ret = rcu_dereference(tmp->css); if (ret) { *foundid = tmpid; break; } } /* continue to scan from next id */ tmpid = tmpid + 1; } return ret; } |