Blame view
kernel/cpuset.c
74.5 KB
1da177e4c
|
1 2 3 4 5 6 |
/* * kernel/cpuset.c * * Processor and Memory placement constraints for sets of tasks. * * Copyright (C) 2003 BULL SA. |
029190c51
|
7 |
* Copyright (C) 2004-2007 Silicon Graphics, Inc. |
8793d854e
|
8 |
* Copyright (C) 2006 Google, Inc |
1da177e4c
|
9 10 11 |
* * Portions derived from Patrick Mochel's sysfs code. * sysfs is Copyright (c) 2001-3 Patrick Mochel |
1da177e4c
|
12 |
* |
825a46af5
|
13 |
* 2003-10-10 Written by Simon Derr. |
1da177e4c
|
14 |
* 2003-10-22 Updates by Stephen Hemminger. |
825a46af5
|
15 |
* 2004 May-July Rework by Paul Jackson. |
8793d854e
|
16 |
* 2006 Rework by Paul Menage to use generic cgroups |
cf417141c
|
17 18 |
* 2008 Rework of the scheduler domains and CPU hotplug handling * by Max Krasnyansky |
1da177e4c
|
19 20 21 22 23 |
* * 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. */ |
1da177e4c
|
24 25 26 27 28 29 30 31 32 33 34 35 |
#include <linux/cpu.h> #include <linux/cpumask.h> #include <linux/cpuset.h> #include <linux/err.h> #include <linux/errno.h> #include <linux/file.h> #include <linux/fs.h> #include <linux/init.h> #include <linux/interrupt.h> #include <linux/kernel.h> #include <linux/kmod.h> #include <linux/list.h> |
68860ec10
|
36 |
#include <linux/mempolicy.h> |
1da177e4c
|
37 |
#include <linux/mm.h> |
f481891fd
|
38 |
#include <linux/memory.h> |
9984de1a5
|
39 |
#include <linux/export.h> |
1da177e4c
|
40 41 42 43 |
#include <linux/mount.h> #include <linux/namei.h> #include <linux/pagemap.h> #include <linux/proc_fs.h> |
6b9c2603c
|
44 |
#include <linux/rcupdate.h> |
1da177e4c
|
45 46 |
#include <linux/sched.h> #include <linux/seq_file.h> |
22fb52dd7
|
47 |
#include <linux/security.h> |
1da177e4c
|
48 |
#include <linux/slab.h> |
1da177e4c
|
49 50 51 52 53 54 55 56 |
#include <linux/spinlock.h> #include <linux/stat.h> #include <linux/string.h> #include <linux/time.h> #include <linux/backing-dev.h> #include <linux/sort.h> #include <asm/uaccess.h> |
60063497a
|
57 |
#include <linux/atomic.h> |
3d3f26a7b
|
58 |
#include <linux/mutex.h> |
956db3ca0
|
59 60 |
#include <linux/workqueue.h> #include <linux/cgroup.h> |
1da177e4c
|
61 |
|
202f72d5d
|
62 |
/* |
f90d4118b
|
63 64 65 66 67 68 69 70 |
* Workqueue for cpuset related tasks. * * Using kevent workqueue may cause deadlock when memory_migrate * is set. So we create a separate workqueue thread for cpuset. */ static struct workqueue_struct *cpuset_wq; /* |
202f72d5d
|
71 72 73 74 |
* Tracks how many cpusets are currently defined in system. * When there is only one cpuset (the root cpuset) we can * short circuit some hooks. */ |
7edc59628
|
75 |
int number_of_cpusets __read_mostly; |
202f72d5d
|
76 |
|
2df167a30
|
77 |
/* Forward declare cgroup structures */ |
8793d854e
|
78 79 |
struct cgroup_subsys cpuset_subsys; struct cpuset; |
3e0d98b9f
|
80 81 82 83 84 85 86 87 |
/* See "Frequency meter" comments, below. */ struct fmeter { int cnt; /* unprocessed events count */ int val; /* most recent output value */ time_t time; /* clock (secs) when val computed */ spinlock_t lock; /* guards read or write of above */ }; |
1da177e4c
|
88 |
struct cpuset { |
8793d854e
|
89 |
struct cgroup_subsys_state css; |
1da177e4c
|
90 |
unsigned long flags; /* "unsigned long" so bitops work */ |
300ed6cbb
|
91 |
cpumask_var_t cpus_allowed; /* CPUs allowed to tasks in cpuset */ |
1da177e4c
|
92 |
nodemask_t mems_allowed; /* Memory Nodes allowed to tasks */ |
1da177e4c
|
93 |
struct cpuset *parent; /* my parent */ |
1da177e4c
|
94 |
|
3e0d98b9f
|
95 |
struct fmeter fmeter; /* memory_pressure filter */ |
029190c51
|
96 97 98 |
/* partition number for rebuild_sched_domains() */ int pn; |
956db3ca0
|
99 |
|
1d3504fcf
|
100 101 |
/* for custom sched domain */ int relax_domain_level; |
732bee7af
|
102 |
/* used for walking a cpuset hierarchy */ |
956db3ca0
|
103 |
struct list_head stack_list; |
1da177e4c
|
104 |
}; |
8793d854e
|
105 106 107 108 109 110 111 112 113 114 115 116 117 |
/* Retrieve the cpuset for a cgroup */ static inline struct cpuset *cgroup_cs(struct cgroup *cont) { return container_of(cgroup_subsys_state(cont, cpuset_subsys_id), struct cpuset, css); } /* Retrieve the cpuset for a task */ static inline struct cpuset *task_cs(struct task_struct *task) { return container_of(task_subsys_state(task, cpuset_subsys_id), struct cpuset, css); } |
8793d854e
|
118 |
|
b246272ec
|
119 120 121 122 123 124 125 126 127 128 129 |
#ifdef CONFIG_NUMA static inline bool task_has_mempolicy(struct task_struct *task) { return task->mempolicy; } #else static inline bool task_has_mempolicy(struct task_struct *task) { return false; } #endif |
1da177e4c
|
130 131 132 133 |
/* bits in struct cpuset flags field */ typedef enum { CS_CPU_EXCLUSIVE, CS_MEM_EXCLUSIVE, |
786083667
|
134 |
CS_MEM_HARDWALL, |
45b07ef31
|
135 |
CS_MEMORY_MIGRATE, |
029190c51
|
136 |
CS_SCHED_LOAD_BALANCE, |
825a46af5
|
137 138 |
CS_SPREAD_PAGE, CS_SPREAD_SLAB, |
1da177e4c
|
139 140 141 142 143 |
} cpuset_flagbits_t; /* convenient tests for these bits */ static inline int is_cpu_exclusive(const struct cpuset *cs) { |
7b5b9ef0e
|
144 |
return test_bit(CS_CPU_EXCLUSIVE, &cs->flags); |
1da177e4c
|
145 146 147 148 |
} static inline int is_mem_exclusive(const struct cpuset *cs) { |
7b5b9ef0e
|
149 |
return test_bit(CS_MEM_EXCLUSIVE, &cs->flags); |
1da177e4c
|
150 |
} |
786083667
|
151 152 153 154 |
static inline int is_mem_hardwall(const struct cpuset *cs) { return test_bit(CS_MEM_HARDWALL, &cs->flags); } |
029190c51
|
155 156 157 158 |
static inline int is_sched_load_balance(const struct cpuset *cs) { return test_bit(CS_SCHED_LOAD_BALANCE, &cs->flags); } |
45b07ef31
|
159 160 |
static inline int is_memory_migrate(const struct cpuset *cs) { |
7b5b9ef0e
|
161 |
return test_bit(CS_MEMORY_MIGRATE, &cs->flags); |
45b07ef31
|
162 |
} |
825a46af5
|
163 164 165 166 167 168 169 170 171 |
static inline int is_spread_page(const struct cpuset *cs) { return test_bit(CS_SPREAD_PAGE, &cs->flags); } static inline int is_spread_slab(const struct cpuset *cs) { return test_bit(CS_SPREAD_SLAB, &cs->flags); } |
1da177e4c
|
172 173 |
static struct cpuset top_cpuset = { .flags = ((1 << CS_CPU_EXCLUSIVE) | (1 << CS_MEM_EXCLUSIVE)), |
1da177e4c
|
174 |
}; |
1da177e4c
|
175 |
/* |
2df167a30
|
176 177 178 179 180 181 182 |
* There are two global mutexes guarding cpuset structures. The first * is the main control groups cgroup_mutex, accessed via * cgroup_lock()/cgroup_unlock(). The second is the cpuset-specific * callback_mutex, below. They can nest. It is ok to first take * cgroup_mutex, then nest callback_mutex. We also require taking * task_lock() when dereferencing a task's cpuset pointer. See "The * task_lock() exception", at the end of this comment. |
053199edf
|
183 |
* |
3d3f26a7b
|
184 |
* A task must hold both mutexes to modify cpusets. If a task |
2df167a30
|
185 |
* holds cgroup_mutex, then it blocks others wanting that mutex, |
3d3f26a7b
|
186 |
* ensuring that it is the only task able to also acquire callback_mutex |
053199edf
|
187 188 |
* and be able to modify cpusets. It can perform various checks on * the cpuset structure first, knowing nothing will change. It can |
2df167a30
|
189 |
* also allocate memory while just holding cgroup_mutex. While it is |
053199edf
|
190 |
* performing these checks, various callback routines can briefly |
3d3f26a7b
|
191 192 |
* acquire callback_mutex to query cpusets. Once it is ready to make * the changes, it takes callback_mutex, blocking everyone else. |
053199edf
|
193 194 |
* * Calls to the kernel memory allocator can not be made while holding |
3d3f26a7b
|
195 |
* callback_mutex, as that would risk double tripping on callback_mutex |
053199edf
|
196 197 198 |
* from one of the callbacks into the cpuset code from within * __alloc_pages(). * |
3d3f26a7b
|
199 |
* If a task is only holding callback_mutex, then it has read-only |
053199edf
|
200 201 |
* access to cpusets. * |
58568d2a8
|
202 203 204 |
* Now, the task_struct fields mems_allowed and mempolicy may be changed * by other task, we use alloc_lock in the task_struct fields to protect * them. |
053199edf
|
205 |
* |
3d3f26a7b
|
206 |
* The cpuset_common_file_read() handlers only hold callback_mutex across |
053199edf
|
207 208 209 |
* small pieces of code, such as when reading out possibly multi-word * cpumasks and nodemasks. * |
2df167a30
|
210 211 |
* Accessing a task's cpuset should be done in accordance with the * guidelines for accessing subsystem state in kernel/cgroup.c |
1da177e4c
|
212 |
*/ |
3d3f26a7b
|
213 |
static DEFINE_MUTEX(callback_mutex); |
4247bdc60
|
214 |
|
cf417141c
|
215 |
/* |
75aa19941
|
216 217 218 219 220 221 222 223 224 225 226 |
* cpuset_buffer_lock protects both the cpuset_name and cpuset_nodelist * buffers. They are statically allocated to prevent using excess stack * when calling cpuset_print_task_mems_allowed(). */ #define CPUSET_NAME_LEN (128) #define CPUSET_NODELIST_LEN (256) static char cpuset_name[CPUSET_NAME_LEN]; static char cpuset_nodelist[CPUSET_NODELIST_LEN]; static DEFINE_SPINLOCK(cpuset_buffer_lock); /* |
cf417141c
|
227 |
* This is ugly, but preserves the userspace API for existing cpuset |
8793d854e
|
228 |
* users. If someone tries to mount the "cpuset" filesystem, we |
cf417141c
|
229 230 |
* silently switch it to mount "cgroup" instead */ |
f7e835710
|
231 232 |
static struct dentry *cpuset_mount(struct file_system_type *fs_type, int flags, const char *unused_dev_name, void *data) |
1da177e4c
|
233 |
{ |
8793d854e
|
234 |
struct file_system_type *cgroup_fs = get_fs_type("cgroup"); |
f7e835710
|
235 |
struct dentry *ret = ERR_PTR(-ENODEV); |
8793d854e
|
236 237 238 239 |
if (cgroup_fs) { char mountopts[] = "cpuset,noprefix," "release_agent=/sbin/cpuset_release_agent"; |
f7e835710
|
240 241 |
ret = cgroup_fs->mount(cgroup_fs, flags, unused_dev_name, mountopts); |
8793d854e
|
242 243 244 |
put_filesystem(cgroup_fs); } return ret; |
1da177e4c
|
245 246 247 248 |
} static struct file_system_type cpuset_fs_type = { .name = "cpuset", |
f7e835710
|
249 |
.mount = cpuset_mount, |
1da177e4c
|
250 |
}; |
1da177e4c
|
251 |
/* |
300ed6cbb
|
252 |
* Return in pmask the portion of a cpusets's cpus_allowed that |
1da177e4c
|
253 254 255 256 257 258 259 260 261 |
* are online. If none are online, walk up the cpuset hierarchy * until we find one that does have some online cpus. If we get * all the way to the top and still haven't found any online cpus, * return cpu_online_map. Or if passed a NULL cs from an exit'ing * task, return cpu_online_map. * * One way or another, we guarantee to return some non-empty subset * of cpu_online_map. * |
3d3f26a7b
|
262 |
* Call with callback_mutex held. |
1da177e4c
|
263 |
*/ |
6af866af3
|
264 265 |
static void guarantee_online_cpus(const struct cpuset *cs, struct cpumask *pmask) |
1da177e4c
|
266 |
{ |
300ed6cbb
|
267 |
while (cs && !cpumask_intersects(cs->cpus_allowed, cpu_online_mask)) |
1da177e4c
|
268 269 |
cs = cs->parent; if (cs) |
300ed6cbb
|
270 |
cpumask_and(pmask, cs->cpus_allowed, cpu_online_mask); |
1da177e4c
|
271 |
else |
300ed6cbb
|
272 273 |
cpumask_copy(pmask, cpu_online_mask); BUG_ON(!cpumask_intersects(pmask, cpu_online_mask)); |
1da177e4c
|
274 275 276 277 |
} /* * Return in *pmask the portion of a cpusets's mems_allowed that |
0e1e7c7a7
|
278 279 280 281 |
* are online, with memory. If none are online with memory, walk * up the cpuset hierarchy until we find one that does have some * online mems. If we get all the way to the top and still haven't * found any online mems, return node_states[N_HIGH_MEMORY]. |
1da177e4c
|
282 283 |
* * One way or another, we guarantee to return some non-empty subset |
0e1e7c7a7
|
284 |
* of node_states[N_HIGH_MEMORY]. |
1da177e4c
|
285 |
* |
3d3f26a7b
|
286 |
* Call with callback_mutex held. |
1da177e4c
|
287 288 289 290 |
*/ static void guarantee_online_mems(const struct cpuset *cs, nodemask_t *pmask) { |
0e1e7c7a7
|
291 292 |
while (cs && !nodes_intersects(cs->mems_allowed, node_states[N_HIGH_MEMORY])) |
1da177e4c
|
293 294 |
cs = cs->parent; if (cs) |
0e1e7c7a7
|
295 296 |
nodes_and(*pmask, cs->mems_allowed, node_states[N_HIGH_MEMORY]); |
1da177e4c
|
297 |
else |
0e1e7c7a7
|
298 299 |
*pmask = node_states[N_HIGH_MEMORY]; BUG_ON(!nodes_intersects(*pmask, node_states[N_HIGH_MEMORY])); |
1da177e4c
|
300 |
} |
f3b39d47e
|
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 |
/* * update task's spread flag if cpuset's page/slab spread flag is set * * Called with callback_mutex/cgroup_mutex held */ static void cpuset_update_task_spread_flag(struct cpuset *cs, struct task_struct *tsk) { if (is_spread_page(cs)) tsk->flags |= PF_SPREAD_PAGE; else tsk->flags &= ~PF_SPREAD_PAGE; if (is_spread_slab(cs)) tsk->flags |= PF_SPREAD_SLAB; else tsk->flags &= ~PF_SPREAD_SLAB; } |
1da177e4c
|
318 319 320 321 322 |
/* * is_cpuset_subset(p, q) - Is cpuset p a subset of cpuset q? * * One cpuset is a subset of another if all its allowed CPUs and * Memory Nodes are a subset of the other, and its exclusive flags |
2df167a30
|
323 |
* are only set if the other's are set. Call holding cgroup_mutex. |
1da177e4c
|
324 325 326 327 |
*/ static int is_cpuset_subset(const struct cpuset *p, const struct cpuset *q) { |
300ed6cbb
|
328 |
return cpumask_subset(p->cpus_allowed, q->cpus_allowed) && |
1da177e4c
|
329 330 331 332 |
nodes_subset(p->mems_allowed, q->mems_allowed) && is_cpu_exclusive(p) <= is_cpu_exclusive(q) && is_mem_exclusive(p) <= is_mem_exclusive(q); } |
645fcc9d2
|
333 334 335 336 337 338 |
/** * alloc_trial_cpuset - allocate a trial cpuset * @cs: the cpuset that the trial cpuset duplicates */ static struct cpuset *alloc_trial_cpuset(const struct cpuset *cs) { |
300ed6cbb
|
339 340 341 342 343 344 345 346 347 348 349 350 351 |
struct cpuset *trial; trial = kmemdup(cs, sizeof(*cs), GFP_KERNEL); if (!trial) return NULL; if (!alloc_cpumask_var(&trial->cpus_allowed, GFP_KERNEL)) { kfree(trial); return NULL; } cpumask_copy(trial->cpus_allowed, cs->cpus_allowed); return trial; |
645fcc9d2
|
352 353 354 355 356 357 358 359 |
} /** * free_trial_cpuset - free the trial cpuset * @trial: the trial cpuset to be freed */ static void free_trial_cpuset(struct cpuset *trial) { |
300ed6cbb
|
360 |
free_cpumask_var(trial->cpus_allowed); |
645fcc9d2
|
361 362 |
kfree(trial); } |
1da177e4c
|
363 364 365 366 367 368 369 |
/* * validate_change() - Used to validate that any proposed cpuset change * follows the structural rules for cpusets. * * If we replaced the flag and mask values of the current cpuset * (cur) with those values in the trial cpuset (trial), would * our various subset and exclusive rules still be valid? Presumes |
2df167a30
|
370 |
* cgroup_mutex held. |
1da177e4c
|
371 372 373 374 375 376 377 378 379 380 381 382 383 384 |
* * 'cur' is the address of an actual, in-use cpuset. Operations * such as list traversal that depend on the actual address of the * cpuset in the list must use cur below, not trial. * * 'trial' is the address of bulk structure copy of cur, with * perhaps one or more of the fields cpus_allowed, mems_allowed, * or flags changed to new, trial values. * * Return 0 if valid, -errno if not. */ static int validate_change(const struct cpuset *cur, const struct cpuset *trial) { |
8793d854e
|
385 |
struct cgroup *cont; |
1da177e4c
|
386 387 388 |
struct cpuset *c, *par; /* Each of our child cpusets must be a subset of us */ |
8793d854e
|
389 390 |
list_for_each_entry(cont, &cur->css.cgroup->children, sibling) { if (!is_cpuset_subset(cgroup_cs(cont), trial)) |
1da177e4c
|
391 392 393 394 |
return -EBUSY; } /* Remaining checks don't apply to root cpuset */ |
696040670
|
395 |
if (cur == &top_cpuset) |
1da177e4c
|
396 |
return 0; |
696040670
|
397 |
par = cur->parent; |
1da177e4c
|
398 399 400 |
/* We must be a subset of our parent cpuset */ if (!is_cpuset_subset(trial, par)) return -EACCES; |
2df167a30
|
401 402 403 404 |
/* * If either I or some sibling (!= me) is exclusive, we can't * overlap */ |
8793d854e
|
405 406 |
list_for_each_entry(cont, &par->css.cgroup->children, sibling) { c = cgroup_cs(cont); |
1da177e4c
|
407 408 |
if ((is_cpu_exclusive(trial) || is_cpu_exclusive(c)) && c != cur && |
300ed6cbb
|
409 |
cpumask_intersects(trial->cpus_allowed, c->cpus_allowed)) |
1da177e4c
|
410 411 412 413 414 415 |
return -EINVAL; if ((is_mem_exclusive(trial) || is_mem_exclusive(c)) && c != cur && nodes_intersects(trial->mems_allowed, c->mems_allowed)) return -EINVAL; } |
020958b62
|
416 417 |
/* Cpusets with tasks can't have empty cpus_allowed or mems_allowed */ if (cgroup_task_count(cur->css.cgroup)) { |
300ed6cbb
|
418 |
if (cpumask_empty(trial->cpus_allowed) || |
020958b62
|
419 420 421 422 |
nodes_empty(trial->mems_allowed)) { return -ENOSPC; } } |
1da177e4c
|
423 424 |
return 0; } |
db7f47cf4
|
425 |
#ifdef CONFIG_SMP |
85d7b9498
|
426 |
/* |
cf417141c
|
427 |
* Helper routine for generate_sched_domains(). |
029190c51
|
428 429 |
* Do cpusets a, b have overlapping cpus_allowed masks? */ |
029190c51
|
430 431 |
static int cpusets_overlap(struct cpuset *a, struct cpuset *b) { |
300ed6cbb
|
432 |
return cpumask_intersects(a->cpus_allowed, b->cpus_allowed); |
029190c51
|
433 |
} |
1d3504fcf
|
434 435 436 |
static void update_domain_attr(struct sched_domain_attr *dattr, struct cpuset *c) { |
1d3504fcf
|
437 438 439 440 |
if (dattr->relax_domain_level < c->relax_domain_level) dattr->relax_domain_level = c->relax_domain_level; return; } |
f5393693e
|
441 442 443 444 445 446 447 448 449 450 451 452 453 |
static void update_domain_attr_tree(struct sched_domain_attr *dattr, struct cpuset *c) { LIST_HEAD(q); list_add(&c->stack_list, &q); while (!list_empty(&q)) { struct cpuset *cp; struct cgroup *cont; struct cpuset *child; cp = list_first_entry(&q, struct cpuset, stack_list); list_del(q.next); |
300ed6cbb
|
454 |
if (cpumask_empty(cp->cpus_allowed)) |
f5393693e
|
455 456 457 458 459 460 461 462 463 464 465 |
continue; if (is_sched_load_balance(cp)) update_domain_attr(dattr, cp); list_for_each_entry(cont, &cp->css.cgroup->children, sibling) { child = cgroup_cs(cont); list_add_tail(&child->stack_list, &q); } } } |
029190c51
|
466 |
/* |
cf417141c
|
467 468 469 470 471 472 473 474 475 |
* generate_sched_domains() * * This function builds a partial partition of the systems CPUs * A 'partial partition' is a set of non-overlapping subsets whose * union is a subset of that set. * The output of this function needs to be passed to kernel/sched.c * partition_sched_domains() routine, which will rebuild the scheduler's * load balancing domains (sched domains) as specified by that partial * partition. |
029190c51
|
476 |
* |
45ce80fb6
|
477 |
* See "What is sched_load_balance" in Documentation/cgroups/cpusets.txt |
029190c51
|
478 479 480 481 482 483 484 |
* for a background explanation of this. * * Does not return errors, on the theory that the callers of this * routine would rather not worry about failures to rebuild sched * domains when operating in the severe memory shortage situations * that could cause allocation failures below. * |
cf417141c
|
485 |
* Must be called with cgroup_lock held. |
029190c51
|
486 487 |
* * The three key local variables below are: |
aeed68242
|
488 |
* q - a linked-list queue of cpuset pointers, used to implement a |
029190c51
|
489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 |
* top-down scan of all cpusets. This scan loads a pointer * to each cpuset marked is_sched_load_balance into the * array 'csa'. For our purposes, rebuilding the schedulers * sched domains, we can ignore !is_sched_load_balance cpusets. * csa - (for CpuSet Array) Array of pointers to all the cpusets * that need to be load balanced, for convenient iterative * access by the subsequent code that finds the best partition, * i.e the set of domains (subsets) of CPUs such that the * cpus_allowed of every cpuset marked is_sched_load_balance * is a subset of one of these domains, while there are as * many such domains as possible, each as small as possible. * doms - Conversion of 'csa' to an array of cpumasks, for passing to * the kernel/sched.c routine partition_sched_domains() in a * convenient format, that can be easily compared to the prior * value to determine what partition elements (sched domains) * were changed (added or removed.) * * Finding the best partition (set of domains): * The triple nested loops below over i, j, k scan over the * load balanced cpusets (using the array of cpuset pointers in * csa[]) looking for pairs of cpusets that have overlapping * cpus_allowed, but which don't have the same 'pn' partition * number and gives them in the same partition number. It keeps * looping on the 'restart' label until it can no longer find * any such pairs. * * The union of the cpus_allowed masks from the set of * all cpusets having the same 'pn' value then form the one * element of the partition (one sched domain) to be passed to * partition_sched_domains(). */ |
acc3f5d7c
|
520 |
static int generate_sched_domains(cpumask_var_t **domains, |
cf417141c
|
521 |
struct sched_domain_attr **attributes) |
029190c51
|
522 |
{ |
cf417141c
|
523 |
LIST_HEAD(q); /* queue of cpusets to be scanned */ |
029190c51
|
524 525 526 527 |
struct cpuset *cp; /* scans q */ struct cpuset **csa; /* array of all cpuset ptrs */ int csn; /* how many cpuset ptrs in csa so far */ int i, j, k; /* indices for partition finding loops */ |
acc3f5d7c
|
528 |
cpumask_var_t *doms; /* resulting partition; i.e. sched domains */ |
1d3504fcf
|
529 |
struct sched_domain_attr *dattr; /* attributes for custom domains */ |
1583715dd
|
530 |
int ndoms = 0; /* number of sched domains in result */ |
6af866af3
|
531 |
int nslot; /* next empty doms[] struct cpumask slot */ |
029190c51
|
532 |
|
029190c51
|
533 |
doms = NULL; |
1d3504fcf
|
534 |
dattr = NULL; |
cf417141c
|
535 |
csa = NULL; |
029190c51
|
536 537 538 |
/* Special case for the 99% of systems with one, full, sched domain */ if (is_sched_load_balance(&top_cpuset)) { |
acc3f5d7c
|
539 540 |
ndoms = 1; doms = alloc_sched_domains(ndoms); |
029190c51
|
541 |
if (!doms) |
cf417141c
|
542 |
goto done; |
1d3504fcf
|
543 544 545 |
dattr = kmalloc(sizeof(struct sched_domain_attr), GFP_KERNEL); if (dattr) { *dattr = SD_ATTR_INIT; |
93a655755
|
546 |
update_domain_attr_tree(dattr, &top_cpuset); |
1d3504fcf
|
547 |
} |
acc3f5d7c
|
548 |
cpumask_copy(doms[0], top_cpuset.cpus_allowed); |
cf417141c
|
549 |
|
cf417141c
|
550 |
goto done; |
029190c51
|
551 |
} |
029190c51
|
552 553 554 555 |
csa = kmalloc(number_of_cpusets * sizeof(cp), GFP_KERNEL); if (!csa) goto done; csn = 0; |
aeed68242
|
556 557 |
list_add(&top_cpuset.stack_list, &q); while (!list_empty(&q)) { |
029190c51
|
558 559 |
struct cgroup *cont; struct cpuset *child; /* scans child cpusets of cp */ |
489a5393a
|
560 |
|
aeed68242
|
561 562 |
cp = list_first_entry(&q, struct cpuset, stack_list); list_del(q.next); |
300ed6cbb
|
563 |
if (cpumask_empty(cp->cpus_allowed)) |
489a5393a
|
564 |
continue; |
f5393693e
|
565 566 567 568 569 570 571 |
/* * All child cpusets contain a subset of the parent's cpus, so * just skip them, and then we call update_domain_attr_tree() * to calc relax_domain_level of the corresponding sched * domain. */ if (is_sched_load_balance(cp)) { |
029190c51
|
572 |
csa[csn++] = cp; |
f5393693e
|
573 574 |
continue; } |
489a5393a
|
575 |
|
029190c51
|
576 577 |
list_for_each_entry(cont, &cp->css.cgroup->children, sibling) { child = cgroup_cs(cont); |
aeed68242
|
578 |
list_add_tail(&child->stack_list, &q); |
029190c51
|
579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 |
} } for (i = 0; i < csn; i++) csa[i]->pn = i; ndoms = csn; restart: /* Find the best partition (set of sched domains) */ for (i = 0; i < csn; i++) { struct cpuset *a = csa[i]; int apn = a->pn; for (j = 0; j < csn; j++) { struct cpuset *b = csa[j]; int bpn = b->pn; if (apn != bpn && cpusets_overlap(a, b)) { for (k = 0; k < csn; k++) { struct cpuset *c = csa[k]; if (c->pn == bpn) c->pn = apn; } ndoms--; /* one less element */ goto restart; } } } |
cf417141c
|
608 609 610 611 |
/* * Now we know how many domains to create. * Convert <csn, csa> to <ndoms, doms> and populate cpu masks. */ |
acc3f5d7c
|
612 |
doms = alloc_sched_domains(ndoms); |
700018e0a
|
613 |
if (!doms) |
cf417141c
|
614 |
goto done; |
cf417141c
|
615 616 617 618 619 |
/* * The rest of the code, including the scheduler, can deal with * dattr==NULL case. No need to abort if alloc fails. */ |
1d3504fcf
|
620 |
dattr = kmalloc(ndoms * sizeof(struct sched_domain_attr), GFP_KERNEL); |
029190c51
|
621 622 623 |
for (nslot = 0, i = 0; i < csn; i++) { struct cpuset *a = csa[i]; |
6af866af3
|
624 |
struct cpumask *dp; |
029190c51
|
625 |
int apn = a->pn; |
cf417141c
|
626 627 628 629 |
if (apn < 0) { /* Skip completed partitions */ continue; } |
acc3f5d7c
|
630 |
dp = doms[nslot]; |
cf417141c
|
631 632 633 634 635 636 637 638 639 640 641 |
if (nslot == ndoms) { static int warnings = 10; if (warnings) { printk(KERN_WARNING "rebuild_sched_domains confused:" " nslot %d, ndoms %d, csn %d, i %d," " apn %d ", nslot, ndoms, csn, i, apn); warnings--; |
029190c51
|
642 |
} |
cf417141c
|
643 644 |
continue; } |
029190c51
|
645 |
|
6af866af3
|
646 |
cpumask_clear(dp); |
cf417141c
|
647 648 649 650 651 652 |
if (dattr) *(dattr + nslot) = SD_ATTR_INIT; for (j = i; j < csn; j++) { struct cpuset *b = csa[j]; if (apn == b->pn) { |
300ed6cbb
|
653 |
cpumask_or(dp, dp, b->cpus_allowed); |
cf417141c
|
654 655 656 657 658 |
if (dattr) update_domain_attr_tree(dattr + nslot, b); /* Done with this partition */ b->pn = -1; |
029190c51
|
659 |
} |
029190c51
|
660 |
} |
cf417141c
|
661 |
nslot++; |
029190c51
|
662 663 |
} BUG_ON(nslot != ndoms); |
cf417141c
|
664 665 |
done: kfree(csa); |
700018e0a
|
666 667 668 669 670 671 |
/* * Fallback to the default domain if kmalloc() failed. * See comments in partition_sched_domains(). */ if (doms == NULL) ndoms = 1; |
cf417141c
|
672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 |
*domains = doms; *attributes = dattr; return ndoms; } /* * Rebuild scheduler domains. * * Call with neither cgroup_mutex held nor within get_online_cpus(). * Takes both cgroup_mutex and get_online_cpus(). * * Cannot be directly called from cpuset code handling changes * to the cpuset pseudo-filesystem, because it cannot be called * from code that already holds cgroup_mutex. */ static void do_rebuild_sched_domains(struct work_struct *unused) { struct sched_domain_attr *attr; |
acc3f5d7c
|
690 |
cpumask_var_t *doms; |
cf417141c
|
691 |
int ndoms; |
86ef5c9a8
|
692 |
get_online_cpus(); |
cf417141c
|
693 694 695 696 697 698 699 700 |
/* Generate domain masks and attrs */ cgroup_lock(); ndoms = generate_sched_domains(&doms, &attr); cgroup_unlock(); /* Have scheduler rebuild the domains */ partition_sched_domains(ndoms, doms, attr); |
86ef5c9a8
|
701 |
put_online_cpus(); |
cf417141c
|
702 |
} |
db7f47cf4
|
703 704 705 706 |
#else /* !CONFIG_SMP */ static void do_rebuild_sched_domains(struct work_struct *unused) { } |
e1b8090bd
|
707 |
static int generate_sched_domains(cpumask_var_t **domains, |
db7f47cf4
|
708 709 710 711 712 713 |
struct sched_domain_attr **attributes) { *domains = NULL; return 1; } #endif /* CONFIG_SMP */ |
029190c51
|
714 |
|
cf417141c
|
715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 |
static DECLARE_WORK(rebuild_sched_domains_work, do_rebuild_sched_domains); /* * Rebuild scheduler domains, asynchronously via workqueue. * * If the flag 'sched_load_balance' of any cpuset with non-empty * 'cpus' changes, or if the 'cpus' allowed changes in any cpuset * which has that flag enabled, or if any cpuset with a non-empty * 'cpus' is removed, then call this routine to rebuild the * scheduler's dynamic sched domains. * * The rebuild_sched_domains() and partition_sched_domains() * routines must nest cgroup_lock() inside get_online_cpus(), * but such cpuset changes as these must nest that locking the * other way, holding cgroup_lock() for much of the code. * * So in order to avoid an ABBA deadlock, the cpuset code handling * these user changes delegates the actual sched domain rebuilding * to a separate workqueue thread, which ends up processing the * above do_rebuild_sched_domains() function. */ static void async_rebuild_sched_domains(void) { |
f90d4118b
|
738 |
queue_work(cpuset_wq, &rebuild_sched_domains_work); |
cf417141c
|
739 740 741 742 743 744 745 746 747 748 749 750 751 752 |
} /* * Accomplishes the same scheduler domain rebuild as the above * async_rebuild_sched_domains(), however it directly calls the * rebuild routine synchronously rather than calling it via an * asynchronous work thread. * * This can only be called from code that is not holding * cgroup_mutex (not nested in a cgroup_lock() call.) */ void rebuild_sched_domains(void) { do_rebuild_sched_domains(NULL); |
029190c51
|
753 |
} |
58f4790b7
|
754 755 756 757 758 |
/** * cpuset_test_cpumask - test a task's cpus_allowed versus its cpuset's * @tsk: task to test * @scan: struct cgroup_scanner contained in its struct cpuset_hotplug_scanner * |
2df167a30
|
759 |
* Call with cgroup_mutex held. May take callback_mutex during call. |
58f4790b7
|
760 761 762 |
* Called for each task in a cgroup by cgroup_scan_tasks(). * Return nonzero if this tasks's cpus_allowed mask should be changed (in other * words, if its mask is not equal to its cpuset's mask). |
053199edf
|
763 |
*/ |
9e0c914ca
|
764 765 |
static int cpuset_test_cpumask(struct task_struct *tsk, struct cgroup_scanner *scan) |
58f4790b7
|
766 |
{ |
300ed6cbb
|
767 |
return !cpumask_equal(&tsk->cpus_allowed, |
58f4790b7
|
768 769 |
(cgroup_cs(scan->cg))->cpus_allowed); } |
053199edf
|
770 |
|
58f4790b7
|
771 772 773 774 775 776 777 778 779 780 781 |
/** * cpuset_change_cpumask - make a task's cpus_allowed the same as its cpuset's * @tsk: task to test * @scan: struct cgroup_scanner containing the cgroup of the task * * Called by cgroup_scan_tasks() for each task in a cgroup whose * cpus_allowed mask needs to be changed. * * We don't need to re-check for the cgroup/cpuset membership, since we're * holding cgroup_lock() at this point. */ |
9e0c914ca
|
782 783 |
static void cpuset_change_cpumask(struct task_struct *tsk, struct cgroup_scanner *scan) |
58f4790b7
|
784 |
{ |
300ed6cbb
|
785 |
set_cpus_allowed_ptr(tsk, ((cgroup_cs(scan->cg))->cpus_allowed)); |
58f4790b7
|
786 787 788 |
} /** |
0b2f630a2
|
789 790 |
* update_tasks_cpumask - Update the cpumasks of tasks in the cpuset. * @cs: the cpuset in which each task's cpus_allowed mask needs to be changed |
4e74339af
|
791 |
* @heap: if NULL, defer allocating heap memory to cgroup_scan_tasks() |
0b2f630a2
|
792 793 794 795 796 797 |
* * Called with cgroup_mutex held * * The cgroup_scan_tasks() function will scan all the tasks in a cgroup, * calling callback functions for each. * |
4e74339af
|
798 799 |
* No return value. It's guaranteed that cgroup_scan_tasks() always returns 0 * if @heap != NULL. |
0b2f630a2
|
800 |
*/ |
4e74339af
|
801 |
static void update_tasks_cpumask(struct cpuset *cs, struct ptr_heap *heap) |
0b2f630a2
|
802 803 |
{ struct cgroup_scanner scan; |
0b2f630a2
|
804 805 806 807 |
scan.cg = cs->css.cgroup; scan.test_task = cpuset_test_cpumask; scan.process_task = cpuset_change_cpumask; |
4e74339af
|
808 809 |
scan.heap = heap; cgroup_scan_tasks(&scan); |
0b2f630a2
|
810 811 812 |
} /** |
58f4790b7
|
813 814 815 816 |
* update_cpumask - update the cpus_allowed mask of a cpuset and all tasks in it * @cs: the cpuset to consider * @buf: buffer of cpu numbers written to this cpuset */ |
645fcc9d2
|
817 818 |
static int update_cpumask(struct cpuset *cs, struct cpuset *trialcs, const char *buf) |
1da177e4c
|
819 |
{ |
4e74339af
|
820 |
struct ptr_heap heap; |
58f4790b7
|
821 822 |
int retval; int is_load_balanced; |
1da177e4c
|
823 |
|
4c4d50f7b
|
824 825 826 |
/* top_cpuset.cpus_allowed tracks cpu_online_map; it's read-only */ if (cs == &top_cpuset) return -EACCES; |
6f7f02e78
|
827 |
/* |
c8d9c90c7
|
828 |
* An empty cpus_allowed is ok only if the cpuset has no tasks. |
020958b62
|
829 830 831 |
* Since cpulist_parse() fails on an empty mask, we special case * that parsing. The validate_change() call ensures that cpusets * with tasks have cpus. |
6f7f02e78
|
832 |
*/ |
020958b62
|
833 |
if (!*buf) { |
300ed6cbb
|
834 |
cpumask_clear(trialcs->cpus_allowed); |
6f7f02e78
|
835 |
} else { |
300ed6cbb
|
836 |
retval = cpulist_parse(buf, trialcs->cpus_allowed); |
6f7f02e78
|
837 838 |
if (retval < 0) return retval; |
37340746a
|
839 |
|
6ad4c1888
|
840 |
if (!cpumask_subset(trialcs->cpus_allowed, cpu_active_mask)) |
37340746a
|
841 |
return -EINVAL; |
6f7f02e78
|
842 |
} |
645fcc9d2
|
843 |
retval = validate_change(cs, trialcs); |
85d7b9498
|
844 845 |
if (retval < 0) return retval; |
029190c51
|
846 |
|
8707d8b8c
|
847 |
/* Nothing to do if the cpus didn't change */ |
300ed6cbb
|
848 |
if (cpumask_equal(cs->cpus_allowed, trialcs->cpus_allowed)) |
8707d8b8c
|
849 |
return 0; |
58f4790b7
|
850 |
|
4e74339af
|
851 852 853 |
retval = heap_init(&heap, PAGE_SIZE, GFP_KERNEL, NULL); if (retval) return retval; |
645fcc9d2
|
854 |
is_load_balanced = is_sched_load_balance(trialcs); |
029190c51
|
855 |
|
3d3f26a7b
|
856 |
mutex_lock(&callback_mutex); |
300ed6cbb
|
857 |
cpumask_copy(cs->cpus_allowed, trialcs->cpus_allowed); |
3d3f26a7b
|
858 |
mutex_unlock(&callback_mutex); |
029190c51
|
859 |
|
8707d8b8c
|
860 861 |
/* * Scan tasks in the cpuset, and update the cpumasks of any |
58f4790b7
|
862 |
* that need an update. |
8707d8b8c
|
863 |
*/ |
4e74339af
|
864 865 866 |
update_tasks_cpumask(cs, &heap); heap_free(&heap); |
58f4790b7
|
867 |
|
8707d8b8c
|
868 |
if (is_load_balanced) |
cf417141c
|
869 |
async_rebuild_sched_domains(); |
85d7b9498
|
870 |
return 0; |
1da177e4c
|
871 |
} |
053199edf
|
872 |
/* |
e4e364e86
|
873 874 875 876 877 878 879 |
* cpuset_migrate_mm * * Migrate memory region from one set of nodes to another. * * Temporarilly set tasks mems_allowed to target nodes of migration, * so that the migration code can allocate pages on these nodes. * |
2df167a30
|
880 |
* Call holding cgroup_mutex, so current's cpuset won't change |
c8d9c90c7
|
881 |
* during this call, as manage_mutex holds off any cpuset_attach() |
e4e364e86
|
882 883 |
* calls. Therefore we don't need to take task_lock around the * call to guarantee_online_mems(), as we know no one is changing |
2df167a30
|
884 |
* our task's cpuset. |
e4e364e86
|
885 |
* |
e4e364e86
|
886 887 888 889 |
* While the mm_struct we are migrating is typically from some * other task, the task_struct mems_allowed that we are hacking * is for our current task, which must allocate new pages for that * migrating memory region. |
e4e364e86
|
890 891 892 893 894 895 |
*/ static void cpuset_migrate_mm(struct mm_struct *mm, const nodemask_t *from, const nodemask_t *to) { struct task_struct *tsk = current; |
e4e364e86
|
896 |
tsk->mems_allowed = *to; |
e4e364e86
|
897 898 |
do_migrate_pages(mm, from, to, MPOL_MF_MOVE_ALL); |
8793d854e
|
899 |
guarantee_online_mems(task_cs(tsk),&tsk->mems_allowed); |
e4e364e86
|
900 |
} |
3b6766fe6
|
901 |
/* |
58568d2a8
|
902 903 904 905 906 907 908 |
* cpuset_change_task_nodemask - change task's mems_allowed and mempolicy * @tsk: the task to change * @newmems: new nodes that the task will be set * * In order to avoid seeing no nodes if the old and new nodes are disjoint, * we structure updates as setting all new allowed nodes, then clearing newly * disallowed ones. |
58568d2a8
|
909 910 911 912 |
*/ static void cpuset_change_task_nodemask(struct task_struct *tsk, nodemask_t *newmems) { |
b246272ec
|
913 |
bool need_loop; |
89e8a244b
|
914 |
|
c0ff7453b
|
915 916 917 918 919 920 921 922 923 924 925 |
repeat: /* * Allow tasks that have access to memory reserves because they have * been OOM killed to get memory anywhere. */ if (unlikely(test_thread_flag(TIF_MEMDIE))) return; if (current->flags & PF_EXITING) /* Let dying task have memory */ return; task_lock(tsk); |
b246272ec
|
926 927 928 929 930 931 932 933 |
/* * Determine if a loop is necessary if another thread is doing * get_mems_allowed(). If at least one node remains unchanged and * tsk does not have a mempolicy, then an empty nodemask will not be * possible when mems_allowed is larger than a word. */ need_loop = task_has_mempolicy(tsk) || !nodes_intersects(*newmems, tsk->mems_allowed); |
58568d2a8
|
934 |
nodes_or(tsk->mems_allowed, tsk->mems_allowed, *newmems); |
c0ff7453b
|
935 |
mpol_rebind_task(tsk, newmems, MPOL_REBIND_STEP1); |
c0ff7453b
|
936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 |
/* * ensure checking ->mems_allowed_change_disable after setting all new * allowed nodes. * * the read-side task can see an nodemask with new allowed nodes and * old allowed nodes. and if it allocates page when cpuset clears newly * disallowed ones continuous, it can see the new allowed bits. * * And if setting all new allowed nodes is after the checking, setting * all new allowed nodes and clearing newly disallowed ones will be done * continuous, and the read-side task may find no node to alloc page. */ smp_mb(); /* * Allocation of memory is very fast, we needn't sleep when waiting |
b246272ec
|
952 |
* for the read-side. |
c0ff7453b
|
953 |
*/ |
b246272ec
|
954 |
while (need_loop && ACCESS_ONCE(tsk->mems_allowed_change_disable)) { |
c0ff7453b
|
955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 |
task_unlock(tsk); if (!task_curr(tsk)) yield(); goto repeat; } /* * ensure checking ->mems_allowed_change_disable before clearing all new * disallowed nodes. * * if clearing newly disallowed bits before the checking, the read-side * task may find no node to alloc page. */ smp_mb(); mpol_rebind_task(tsk, newmems, MPOL_REBIND_STEP2); |
58568d2a8
|
971 |
tsk->mems_allowed = *newmems; |
c0ff7453b
|
972 |
task_unlock(tsk); |
58568d2a8
|
973 974 975 976 977 978 |
} /* * Update task's mems_allowed and rebind its mempolicy and vmas' mempolicy * of it to cpuset's new mems_allowed, and migrate pages to new nodes if * memory_migrate flag is set. Called with cgroup_mutex held. |
3b6766fe6
|
979 980 981 982 983 984 985 986 |
*/ static void cpuset_change_nodemask(struct task_struct *p, struct cgroup_scanner *scan) { struct mm_struct *mm; struct cpuset *cs; int migrate; const nodemask_t *oldmem = scan->data; |
ee24d3797
|
987 |
static nodemask_t newmems; /* protected by cgroup_mutex */ |
58568d2a8
|
988 989 |
cs = cgroup_cs(scan->cg); |
ee24d3797
|
990 |
guarantee_online_mems(cs, &newmems); |
58568d2a8
|
991 |
|
ee24d3797
|
992 |
cpuset_change_task_nodemask(p, &newmems); |
53feb2976
|
993 |
|
3b6766fe6
|
994 995 996 |
mm = get_task_mm(p); if (!mm) return; |
3b6766fe6
|
997 998 999 1000 1001 1002 1003 |
migrate = is_memory_migrate(cs); mpol_rebind_mm(mm, &cs->mems_allowed); if (migrate) cpuset_migrate_mm(mm, oldmem, &cs->mems_allowed); mmput(mm); } |
8793d854e
|
1004 |
static void *cpuset_being_rebound; |
0b2f630a2
|
1005 1006 1007 1008 |
/** * update_tasks_nodemask - Update the nodemasks of tasks in the cpuset. * @cs: the cpuset in which each task's mems_allowed mask needs to be changed * @oldmem: old mems_allowed of cpuset cs |
010cfac4c
|
1009 |
* @heap: if NULL, defer allocating heap memory to cgroup_scan_tasks() |
0b2f630a2
|
1010 1011 |
* * Called with cgroup_mutex held |
010cfac4c
|
1012 1013 |
* No return value. It's guaranteed that cgroup_scan_tasks() always returns 0 * if @heap != NULL. |
0b2f630a2
|
1014 |
*/ |
010cfac4c
|
1015 1016 |
static void update_tasks_nodemask(struct cpuset *cs, const nodemask_t *oldmem, struct ptr_heap *heap) |
1da177e4c
|
1017 |
{ |
3b6766fe6
|
1018 |
struct cgroup_scanner scan; |
59dac16fb
|
1019 |
|
846a16bf0
|
1020 |
cpuset_being_rebound = cs; /* causes mpol_dup() rebind */ |
4225399a6
|
1021 |
|
3b6766fe6
|
1022 1023 1024 |
scan.cg = cs->css.cgroup; scan.test_task = NULL; scan.process_task = cpuset_change_nodemask; |
010cfac4c
|
1025 |
scan.heap = heap; |
3b6766fe6
|
1026 |
scan.data = (nodemask_t *)oldmem; |
4225399a6
|
1027 1028 |
/* |
3b6766fe6
|
1029 1030 1031 1032 1033 1034 |
* The mpol_rebind_mm() call takes mmap_sem, which we couldn't * take while holding tasklist_lock. Forks can happen - the * mpol_dup() cpuset_being_rebound check will catch such forks, * and rebind their vma mempolicies too. Because we still hold * the global cgroup_mutex, we know that no other rebind effort * will be contending for the global variable cpuset_being_rebound. |
4225399a6
|
1035 |
* It's ok if we rebind the same mm twice; mpol_rebind_mm() |
04c19fa6f
|
1036 |
* is idempotent. Also migrate pages in each mm to new nodes. |
4225399a6
|
1037 |
*/ |
010cfac4c
|
1038 |
cgroup_scan_tasks(&scan); |
4225399a6
|
1039 |
|
2df167a30
|
1040 |
/* We're done rebinding vmas to this cpuset's new mems_allowed. */ |
8793d854e
|
1041 |
cpuset_being_rebound = NULL; |
1da177e4c
|
1042 |
} |
0b2f630a2
|
1043 1044 1045 |
/* * Handle user request to change the 'mems' memory placement * of a cpuset. Needs to validate the request, update the |
58568d2a8
|
1046 1047 1048 1049 |
* cpusets mems_allowed, and for each task in the cpuset, * update mems_allowed and rebind task's mempolicy and any vma * mempolicies and if the cpuset is marked 'memory_migrate', * migrate the tasks pages to the new memory. |
0b2f630a2
|
1050 1051 1052 1053 1054 1055 |
* * Call with cgroup_mutex held. May take callback_mutex during call. * Will take tasklist_lock, scan tasklist for tasks in cpuset cs, * lock each such tasks mm->mmap_sem, scan its vma's and rebind * their mempolicies to the cpusets new mems_allowed. */ |
645fcc9d2
|
1056 1057 |
static int update_nodemask(struct cpuset *cs, struct cpuset *trialcs, const char *buf) |
0b2f630a2
|
1058 |
{ |
53feb2976
|
1059 |
NODEMASK_ALLOC(nodemask_t, oldmem, GFP_KERNEL); |
0b2f630a2
|
1060 |
int retval; |
010cfac4c
|
1061 |
struct ptr_heap heap; |
0b2f630a2
|
1062 |
|
53feb2976
|
1063 1064 |
if (!oldmem) return -ENOMEM; |
0b2f630a2
|
1065 1066 1067 1068 |
/* * top_cpuset.mems_allowed tracks node_stats[N_HIGH_MEMORY]; * it's read-only */ |
53feb2976
|
1069 1070 1071 1072 |
if (cs == &top_cpuset) { retval = -EACCES; goto done; } |
0b2f630a2
|
1073 |
|
0b2f630a2
|
1074 1075 1076 1077 1078 1079 1080 |
/* * An empty mems_allowed is ok iff there are no tasks in the cpuset. * Since nodelist_parse() fails on an empty mask, we special case * that parsing. The validate_change() call ensures that cpusets * with tasks have memory. */ if (!*buf) { |
645fcc9d2
|
1081 |
nodes_clear(trialcs->mems_allowed); |
0b2f630a2
|
1082 |
} else { |
645fcc9d2
|
1083 |
retval = nodelist_parse(buf, trialcs->mems_allowed); |
0b2f630a2
|
1084 1085 |
if (retval < 0) goto done; |
645fcc9d2
|
1086 |
if (!nodes_subset(trialcs->mems_allowed, |
53feb2976
|
1087 1088 1089 1090 |
node_states[N_HIGH_MEMORY])) { retval = -EINVAL; goto done; } |
0b2f630a2
|
1091 |
} |
53feb2976
|
1092 1093 |
*oldmem = cs->mems_allowed; if (nodes_equal(*oldmem, trialcs->mems_allowed)) { |
0b2f630a2
|
1094 1095 1096 |
retval = 0; /* Too easy - nothing to do */ goto done; } |
645fcc9d2
|
1097 |
retval = validate_change(cs, trialcs); |
0b2f630a2
|
1098 1099 |
if (retval < 0) goto done; |
010cfac4c
|
1100 1101 1102 |
retval = heap_init(&heap, PAGE_SIZE, GFP_KERNEL, NULL); if (retval < 0) goto done; |
0b2f630a2
|
1103 |
mutex_lock(&callback_mutex); |
645fcc9d2
|
1104 |
cs->mems_allowed = trialcs->mems_allowed; |
0b2f630a2
|
1105 |
mutex_unlock(&callback_mutex); |
53feb2976
|
1106 |
update_tasks_nodemask(cs, oldmem, &heap); |
010cfac4c
|
1107 1108 |
heap_free(&heap); |
0b2f630a2
|
1109 |
done: |
53feb2976
|
1110 |
NODEMASK_FREE(oldmem); |
0b2f630a2
|
1111 1112 |
return retval; } |
8793d854e
|
1113 1114 1115 1116 |
int current_cpuset_is_being_rebound(void) { return task_cs(current) == cpuset_being_rebound; } |
5be7a4792
|
1117 |
static int update_relax_domain_level(struct cpuset *cs, s64 val) |
1d3504fcf
|
1118 |
{ |
db7f47cf4
|
1119 |
#ifdef CONFIG_SMP |
60495e776
|
1120 |
if (val < -1 || val >= sched_domain_level_max) |
30e0e1781
|
1121 |
return -EINVAL; |
db7f47cf4
|
1122 |
#endif |
1d3504fcf
|
1123 1124 1125 |
if (val != cs->relax_domain_level) { cs->relax_domain_level = val; |
300ed6cbb
|
1126 1127 |
if (!cpumask_empty(cs->cpus_allowed) && is_sched_load_balance(cs)) |
cf417141c
|
1128 |
async_rebuild_sched_domains(); |
1d3504fcf
|
1129 1130 1131 1132 |
} return 0; } |
3e0d98b9f
|
1133 |
/* |
950592f7b
|
1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 |
* cpuset_change_flag - make a task's spread flags the same as its cpuset's * @tsk: task to be updated * @scan: struct cgroup_scanner containing the cgroup of the task * * Called by cgroup_scan_tasks() for each task in a cgroup. * * We don't need to re-check for the cgroup/cpuset membership, since we're * holding cgroup_lock() at this point. */ static void cpuset_change_flag(struct task_struct *tsk, struct cgroup_scanner *scan) { cpuset_update_task_spread_flag(cgroup_cs(scan->cg), tsk); } /* * update_tasks_flags - update the spread flags of tasks in the cpuset. * @cs: the cpuset in which each task's spread flags needs to be changed * @heap: if NULL, defer allocating heap memory to cgroup_scan_tasks() * * Called with cgroup_mutex held * * The cgroup_scan_tasks() function will scan all the tasks in a cgroup, * calling callback functions for each. * * No return value. It's guaranteed that cgroup_scan_tasks() always returns 0 * if @heap != NULL. */ static void update_tasks_flags(struct cpuset *cs, struct ptr_heap *heap) { struct cgroup_scanner scan; scan.cg = cs->css.cgroup; scan.test_task = NULL; scan.process_task = cpuset_change_flag; scan.heap = heap; cgroup_scan_tasks(&scan); } /* |
1da177e4c
|
1174 |
* update_flag - read a 0 or a 1 in a file and update associated flag |
786083667
|
1175 1176 1177 |
* bit: the bit to update (see cpuset_flagbits_t) * cs: the cpuset to update * turning_on: whether the flag is being set or cleared |
053199edf
|
1178 |
* |
2df167a30
|
1179 |
* Call with cgroup_mutex held. |
1da177e4c
|
1180 |
*/ |
700fe1ab9
|
1181 1182 |
static int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, int turning_on) |
1da177e4c
|
1183 |
{ |
645fcc9d2
|
1184 |
struct cpuset *trialcs; |
40b6a7623
|
1185 |
int balance_flag_changed; |
950592f7b
|
1186 1187 1188 |
int spread_flag_changed; struct ptr_heap heap; int err; |
1da177e4c
|
1189 |
|
645fcc9d2
|
1190 1191 1192 |
trialcs = alloc_trial_cpuset(cs); if (!trialcs) return -ENOMEM; |
1da177e4c
|
1193 |
if (turning_on) |
645fcc9d2
|
1194 |
set_bit(bit, &trialcs->flags); |
1da177e4c
|
1195 |
else |
645fcc9d2
|
1196 |
clear_bit(bit, &trialcs->flags); |
1da177e4c
|
1197 |
|
645fcc9d2
|
1198 |
err = validate_change(cs, trialcs); |
85d7b9498
|
1199 |
if (err < 0) |
645fcc9d2
|
1200 |
goto out; |
029190c51
|
1201 |
|
950592f7b
|
1202 1203 1204 |
err = heap_init(&heap, PAGE_SIZE, GFP_KERNEL, NULL); if (err < 0) goto out; |
029190c51
|
1205 |
balance_flag_changed = (is_sched_load_balance(cs) != |
645fcc9d2
|
1206 |
is_sched_load_balance(trialcs)); |
029190c51
|
1207 |
|
950592f7b
|
1208 1209 |
spread_flag_changed = ((is_spread_slab(cs) != is_spread_slab(trialcs)) || (is_spread_page(cs) != is_spread_page(trialcs))); |
3d3f26a7b
|
1210 |
mutex_lock(&callback_mutex); |
645fcc9d2
|
1211 |
cs->flags = trialcs->flags; |
3d3f26a7b
|
1212 |
mutex_unlock(&callback_mutex); |
85d7b9498
|
1213 |
|
300ed6cbb
|
1214 |
if (!cpumask_empty(trialcs->cpus_allowed) && balance_flag_changed) |
cf417141c
|
1215 |
async_rebuild_sched_domains(); |
029190c51
|
1216 |
|
950592f7b
|
1217 1218 1219 |
if (spread_flag_changed) update_tasks_flags(cs, &heap); heap_free(&heap); |
645fcc9d2
|
1220 1221 1222 |
out: free_trial_cpuset(trialcs); return err; |
1da177e4c
|
1223 |
} |
053199edf
|
1224 |
/* |
80f7228b5
|
1225 |
* Frequency meter - How fast is some event occurring? |
3e0d98b9f
|
1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 |
* * These routines manage a digitally filtered, constant time based, * event frequency meter. There are four routines: * fmeter_init() - initialize a frequency meter. * fmeter_markevent() - called each time the event happens. * fmeter_getrate() - returns the recent rate of such events. * fmeter_update() - internal routine used to update fmeter. * * A common data structure is passed to each of these routines, * which is used to keep track of the state required to manage the * frequency meter and its digital filter. * * The filter works on the number of events marked per unit time. * The filter is single-pole low-pass recursive (IIR). The time unit * is 1 second. Arithmetic is done using 32-bit integers scaled to * simulate 3 decimal digits of precision (multiplied by 1000). * * With an FM_COEF of 933, and a time base of 1 second, the filter * has a half-life of 10 seconds, meaning that if the events quit * happening, then the rate returned from the fmeter_getrate() * will be cut in half each 10 seconds, until it converges to zero. * * It is not worth doing a real infinitely recursive filter. If more * than FM_MAXTICKS ticks have elapsed since the last filter event, * just compute FM_MAXTICKS ticks worth, by which point the level * will be stable. * * Limit the count of unprocessed events to FM_MAXCNT, so as to avoid * arithmetic overflow in the fmeter_update() routine. * * Given the simple 32 bit integer arithmetic used, this meter works * best for reporting rates between one per millisecond (msec) and * one per 32 (approx) seconds. At constant rates faster than one * per msec it maxes out at values just under 1,000,000. At constant * rates between one per msec, and one per second it will stabilize * to a value N*1000, where N is the rate of events per second. * At constant rates between one per second and one per 32 seconds, * it will be choppy, moving up on the seconds that have an event, * and then decaying until the next event. At rates slower than * about one in 32 seconds, it decays all the way back to zero between * each event. */ #define FM_COEF 933 /* coefficient for half-life of 10 secs */ #define FM_MAXTICKS ((time_t)99) /* useless computing more ticks than this */ #define FM_MAXCNT 1000000 /* limit cnt to avoid overflow */ #define FM_SCALE 1000 /* faux fixed point scale */ /* Initialize a frequency meter */ static void fmeter_init(struct fmeter *fmp) { fmp->cnt = 0; fmp->val = 0; fmp->time = 0; spin_lock_init(&fmp->lock); } /* Internal meter update - process cnt events and update value */ static void fmeter_update(struct fmeter *fmp) { time_t now = get_seconds(); time_t ticks = now - fmp->time; if (ticks == 0) return; ticks = min(FM_MAXTICKS, ticks); while (ticks-- > 0) fmp->val = (FM_COEF * fmp->val) / FM_SCALE; fmp->time = now; fmp->val += ((FM_SCALE - FM_COEF) * fmp->cnt) / FM_SCALE; fmp->cnt = 0; } /* Process any previous ticks, then bump cnt by one (times scale). */ static void fmeter_markevent(struct fmeter *fmp) { spin_lock(&fmp->lock); fmeter_update(fmp); fmp->cnt = min(FM_MAXCNT, fmp->cnt + FM_SCALE); spin_unlock(&fmp->lock); } /* Process any previous ticks, then return current value. */ static int fmeter_getrate(struct fmeter *fmp) { int val; spin_lock(&fmp->lock); fmeter_update(fmp); val = fmp->val; spin_unlock(&fmp->lock); return val; } |
f780bdb7c
|
1321 1322 |
/* * Protected by cgroup_lock. The nodemasks must be stored globally because |
94196f51c
|
1323 1324 |
* dynamically allocating them is not allowed in can_attach, and they must * persist until attach. |
f780bdb7c
|
1325 1326 1327 1328 |
*/ static cpumask_var_t cpus_attach; static nodemask_t cpuset_attach_nodemask_from; static nodemask_t cpuset_attach_nodemask_to; |
2df167a30
|
1329 |
/* Called by cgroups to determine if a cpuset is usable; cgroup_mutex held */ |
2f7ee5691
|
1330 1331 |
static int cpuset_can_attach(struct cgroup_subsys *ss, struct cgroup *cgrp, struct cgroup_taskset *tset) |
f780bdb7c
|
1332 |
{ |
2f7ee5691
|
1333 |
struct cpuset *cs = cgroup_cs(cgrp); |
bb9d97b6d
|
1334 1335 |
struct task_struct *task; int ret; |
1da177e4c
|
1336 |
|
300ed6cbb
|
1337 |
if (cpumask_empty(cs->cpus_allowed) || nodes_empty(cs->mems_allowed)) |
1da177e4c
|
1338 |
return -ENOSPC; |
9985b0bab
|
1339 |
|
bb9d97b6d
|
1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 |
cgroup_taskset_for_each(task, cgrp, tset) { /* * Kthreads bound to specific cpus cannot be moved to a new * cpuset; we cannot change their cpu affinity and * isolating such threads by their set of allowed nodes is * unnecessary. Thus, cpusets are not applicable for such * threads. This prevents checking for success of * set_cpus_allowed_ptr() on all attached tasks before * cpus_allowed may be changed. */ if (task->flags & PF_THREAD_BOUND) return -EINVAL; if ((ret = security_task_setscheduler(task))) return ret; } |
f780bdb7c
|
1355 |
|
94196f51c
|
1356 |
/* prepare for attach */ |
f780bdb7c
|
1357 1358 1359 1360 1361 1362 |
if (cs == &top_cpuset) cpumask_copy(cpus_attach, cpu_possible_mask); else guarantee_online_cpus(cs, cpus_attach); guarantee_online_mems(cs, &cpuset_attach_nodemask_to); |
f780bdb7c
|
1363 |
|
94196f51c
|
1364 |
return 0; |
8793d854e
|
1365 |
} |
1da177e4c
|
1366 |
|
2f7ee5691
|
1367 1368 |
static void cpuset_attach(struct cgroup_subsys *ss, struct cgroup *cgrp, struct cgroup_taskset *tset) |
8793d854e
|
1369 |
{ |
8793d854e
|
1370 |
struct mm_struct *mm; |
bb9d97b6d
|
1371 1372 |
struct task_struct *task; struct task_struct *leader = cgroup_taskset_first(tset); |
2f7ee5691
|
1373 1374 1375 |
struct cgroup *oldcgrp = cgroup_taskset_cur_cgroup(tset); struct cpuset *cs = cgroup_cs(cgrp); struct cpuset *oldcs = cgroup_cs(oldcgrp); |
22fb52dd7
|
1376 |
|
bb9d97b6d
|
1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 |
cgroup_taskset_for_each(task, cgrp, tset) { /* * can_attach beforehand should guarantee that this doesn't * fail. TODO: have a better way to handle failure here */ WARN_ON_ONCE(set_cpus_allowed_ptr(task, cpus_attach)); cpuset_change_task_nodemask(task, &cpuset_attach_nodemask_to); cpuset_update_task_spread_flag(cs, task); } |
22fb52dd7
|
1387 |
|
f780bdb7c
|
1388 1389 1390 1391 1392 1393 |
/* * Change mm, possibly for multiple threads in a threadgroup. This is * expensive and may sleep. */ cpuset_attach_nodemask_from = oldcs->mems_allowed; cpuset_attach_nodemask_to = cs->mems_allowed; |
bb9d97b6d
|
1394 |
mm = get_task_mm(leader); |
4225399a6
|
1395 |
if (mm) { |
f780bdb7c
|
1396 |
mpol_rebind_mm(mm, &cpuset_attach_nodemask_to); |
2741a559a
|
1397 |
if (is_memory_migrate(cs)) |
f780bdb7c
|
1398 1399 |
cpuset_migrate_mm(mm, &cpuset_attach_nodemask_from, &cpuset_attach_nodemask_to); |
4225399a6
|
1400 1401 |
mmput(mm); } |
1da177e4c
|
1402 1403 1404 1405 1406 |
} /* The various types of files and directories in a cpuset file system */ typedef enum { |
45b07ef31
|
1407 |
FILE_MEMORY_MIGRATE, |
1da177e4c
|
1408 1409 1410 1411 |
FILE_CPULIST, FILE_MEMLIST, FILE_CPU_EXCLUSIVE, FILE_MEM_EXCLUSIVE, |
786083667
|
1412 |
FILE_MEM_HARDWALL, |
029190c51
|
1413 |
FILE_SCHED_LOAD_BALANCE, |
1d3504fcf
|
1414 |
FILE_SCHED_RELAX_DOMAIN_LEVEL, |
3e0d98b9f
|
1415 1416 |
FILE_MEMORY_PRESSURE_ENABLED, FILE_MEMORY_PRESSURE, |
825a46af5
|
1417 1418 |
FILE_SPREAD_PAGE, FILE_SPREAD_SLAB, |
1da177e4c
|
1419 |
} cpuset_filetype_t; |
700fe1ab9
|
1420 1421 1422 1423 1424 |
static int cpuset_write_u64(struct cgroup *cgrp, struct cftype *cft, u64 val) { int retval = 0; struct cpuset *cs = cgroup_cs(cgrp); cpuset_filetype_t type = cft->private; |
e37123953
|
1425 |
if (!cgroup_lock_live_group(cgrp)) |
700fe1ab9
|
1426 |
return -ENODEV; |
700fe1ab9
|
1427 1428 |
switch (type) { |
1da177e4c
|
1429 |
case FILE_CPU_EXCLUSIVE: |
700fe1ab9
|
1430 |
retval = update_flag(CS_CPU_EXCLUSIVE, cs, val); |
1da177e4c
|
1431 1432 |
break; case FILE_MEM_EXCLUSIVE: |
700fe1ab9
|
1433 |
retval = update_flag(CS_MEM_EXCLUSIVE, cs, val); |
1da177e4c
|
1434 |
break; |
786083667
|
1435 1436 1437 |
case FILE_MEM_HARDWALL: retval = update_flag(CS_MEM_HARDWALL, cs, val); break; |
029190c51
|
1438 |
case FILE_SCHED_LOAD_BALANCE: |
700fe1ab9
|
1439 |
retval = update_flag(CS_SCHED_LOAD_BALANCE, cs, val); |
1d3504fcf
|
1440 |
break; |
45b07ef31
|
1441 |
case FILE_MEMORY_MIGRATE: |
700fe1ab9
|
1442 |
retval = update_flag(CS_MEMORY_MIGRATE, cs, val); |
45b07ef31
|
1443 |
break; |
3e0d98b9f
|
1444 |
case FILE_MEMORY_PRESSURE_ENABLED: |
700fe1ab9
|
1445 |
cpuset_memory_pressure_enabled = !!val; |
3e0d98b9f
|
1446 1447 1448 1449 |
break; case FILE_MEMORY_PRESSURE: retval = -EACCES; break; |
825a46af5
|
1450 |
case FILE_SPREAD_PAGE: |
700fe1ab9
|
1451 |
retval = update_flag(CS_SPREAD_PAGE, cs, val); |
825a46af5
|
1452 1453 |
break; case FILE_SPREAD_SLAB: |
700fe1ab9
|
1454 |
retval = update_flag(CS_SPREAD_SLAB, cs, val); |
825a46af5
|
1455 |
break; |
1da177e4c
|
1456 1457 |
default: retval = -EINVAL; |
700fe1ab9
|
1458 |
break; |
1da177e4c
|
1459 |
} |
8793d854e
|
1460 |
cgroup_unlock(); |
1da177e4c
|
1461 1462 |
return retval; } |
5be7a4792
|
1463 1464 1465 1466 1467 |
static int cpuset_write_s64(struct cgroup *cgrp, struct cftype *cft, s64 val) { int retval = 0; struct cpuset *cs = cgroup_cs(cgrp); cpuset_filetype_t type = cft->private; |
e37123953
|
1468 |
if (!cgroup_lock_live_group(cgrp)) |
5be7a4792
|
1469 |
return -ENODEV; |
e37123953
|
1470 |
|
5be7a4792
|
1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 |
switch (type) { case FILE_SCHED_RELAX_DOMAIN_LEVEL: retval = update_relax_domain_level(cs, val); break; default: retval = -EINVAL; break; } cgroup_unlock(); return retval; } |
1da177e4c
|
1482 |
/* |
e37123953
|
1483 1484 1485 1486 1487 1488 |
* Common handling for a write to a "cpus" or "mems" file. */ static int cpuset_write_resmask(struct cgroup *cgrp, struct cftype *cft, const char *buf) { int retval = 0; |
645fcc9d2
|
1489 1490 |
struct cpuset *cs = cgroup_cs(cgrp); struct cpuset *trialcs; |
e37123953
|
1491 1492 1493 |
if (!cgroup_lock_live_group(cgrp)) return -ENODEV; |
645fcc9d2
|
1494 |
trialcs = alloc_trial_cpuset(cs); |
b75f38d65
|
1495 1496 1497 1498 |
if (!trialcs) { retval = -ENOMEM; goto out; } |
645fcc9d2
|
1499 |
|
e37123953
|
1500 1501 |
switch (cft->private) { case FILE_CPULIST: |
645fcc9d2
|
1502 |
retval = update_cpumask(cs, trialcs, buf); |
e37123953
|
1503 1504 |
break; case FILE_MEMLIST: |
645fcc9d2
|
1505 |
retval = update_nodemask(cs, trialcs, buf); |
e37123953
|
1506 1507 1508 1509 1510 |
break; default: retval = -EINVAL; break; } |
645fcc9d2
|
1511 1512 |
free_trial_cpuset(trialcs); |
b75f38d65
|
1513 |
out: |
e37123953
|
1514 1515 1516 1517 1518 |
cgroup_unlock(); return retval; } /* |
1da177e4c
|
1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 |
* These ascii lists should be read in a single call, by using a user * buffer large enough to hold the entire map. If read in smaller * chunks, there is no guarantee of atomicity. Since the display format * used, list of ranges of sequential numbers, is variable length, * and since these maps can change value dynamically, one could read * gibberish by doing partial reads while a list was changing. * A single large read to a buffer that crosses a page boundary is * ok, because the result being copied to user land is not recomputed * across a page fault. */ |
9303e0c48
|
1529 |
static size_t cpuset_sprintf_cpulist(char *page, struct cpuset *cs) |
1da177e4c
|
1530 |
{ |
9303e0c48
|
1531 |
size_t count; |
1da177e4c
|
1532 |
|
3d3f26a7b
|
1533 |
mutex_lock(&callback_mutex); |
9303e0c48
|
1534 |
count = cpulist_scnprintf(page, PAGE_SIZE, cs->cpus_allowed); |
3d3f26a7b
|
1535 |
mutex_unlock(&callback_mutex); |
1da177e4c
|
1536 |
|
9303e0c48
|
1537 |
return count; |
1da177e4c
|
1538 |
} |
9303e0c48
|
1539 |
static size_t cpuset_sprintf_memlist(char *page, struct cpuset *cs) |
1da177e4c
|
1540 |
{ |
9303e0c48
|
1541 |
size_t count; |
1da177e4c
|
1542 |
|
3d3f26a7b
|
1543 |
mutex_lock(&callback_mutex); |
9303e0c48
|
1544 |
count = nodelist_scnprintf(page, PAGE_SIZE, cs->mems_allowed); |
3d3f26a7b
|
1545 |
mutex_unlock(&callback_mutex); |
1da177e4c
|
1546 |
|
9303e0c48
|
1547 |
return count; |
1da177e4c
|
1548 |
} |
8793d854e
|
1549 1550 1551 1552 1553 |
static ssize_t cpuset_common_file_read(struct cgroup *cont, struct cftype *cft, struct file *file, char __user *buf, size_t nbytes, loff_t *ppos) |
1da177e4c
|
1554 |
{ |
8793d854e
|
1555 |
struct cpuset *cs = cgroup_cs(cont); |
1da177e4c
|
1556 1557 1558 1559 |
cpuset_filetype_t type = cft->private; char *page; ssize_t retval = 0; char *s; |
1da177e4c
|
1560 |
|
e12ba74d8
|
1561 |
if (!(page = (char *)__get_free_page(GFP_TEMPORARY))) |
1da177e4c
|
1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 |
return -ENOMEM; s = page; switch (type) { case FILE_CPULIST: s += cpuset_sprintf_cpulist(s, cs); break; case FILE_MEMLIST: s += cpuset_sprintf_memlist(s, cs); break; |
1da177e4c
|
1573 1574 1575 1576 1577 1578 |
default: retval = -EINVAL; goto out; } *s++ = ' '; |
1da177e4c
|
1579 |
|
eacaa1f5a
|
1580 |
retval = simple_read_from_buffer(buf, nbytes, ppos, page, s - page); |
1da177e4c
|
1581 1582 1583 1584 |
out: free_page((unsigned long)page); return retval; } |
700fe1ab9
|
1585 1586 1587 1588 1589 1590 1591 1592 1593 |
static u64 cpuset_read_u64(struct cgroup *cont, struct cftype *cft) { struct cpuset *cs = cgroup_cs(cont); cpuset_filetype_t type = cft->private; switch (type) { case FILE_CPU_EXCLUSIVE: return is_cpu_exclusive(cs); case FILE_MEM_EXCLUSIVE: return is_mem_exclusive(cs); |
786083667
|
1594 1595 |
case FILE_MEM_HARDWALL: return is_mem_hardwall(cs); |
700fe1ab9
|
1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 |
case FILE_SCHED_LOAD_BALANCE: return is_sched_load_balance(cs); case FILE_MEMORY_MIGRATE: return is_memory_migrate(cs); case FILE_MEMORY_PRESSURE_ENABLED: return cpuset_memory_pressure_enabled; case FILE_MEMORY_PRESSURE: return fmeter_getrate(&cs->fmeter); case FILE_SPREAD_PAGE: return is_spread_page(cs); case FILE_SPREAD_SLAB: return is_spread_slab(cs); default: BUG(); } |
cf417141c
|
1611 1612 1613 |
/* Unreachable but makes gcc happy */ return 0; |
700fe1ab9
|
1614 |
} |
1da177e4c
|
1615 |
|
5be7a4792
|
1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 |
static s64 cpuset_read_s64(struct cgroup *cont, struct cftype *cft) { struct cpuset *cs = cgroup_cs(cont); cpuset_filetype_t type = cft->private; switch (type) { case FILE_SCHED_RELAX_DOMAIN_LEVEL: return cs->relax_domain_level; default: BUG(); } |
cf417141c
|
1626 1627 1628 |
/* Unrechable but makes gcc happy */ return 0; |
5be7a4792
|
1629 |
} |
1da177e4c
|
1630 1631 1632 1633 |
/* * for the common functions, 'private' gives the type of file */ |
addf2c739
|
1634 1635 1636 1637 |
static struct cftype files[] = { { .name = "cpus", .read = cpuset_common_file_read, |
e37123953
|
1638 1639 |
.write_string = cpuset_write_resmask, .max_write_len = (100U + 6 * NR_CPUS), |
addf2c739
|
1640 1641 1642 1643 1644 1645 |
.private = FILE_CPULIST, }, { .name = "mems", .read = cpuset_common_file_read, |
e37123953
|
1646 1647 |
.write_string = cpuset_write_resmask, .max_write_len = (100U + 6 * MAX_NUMNODES), |
addf2c739
|
1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 |
.private = FILE_MEMLIST, }, { .name = "cpu_exclusive", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_CPU_EXCLUSIVE, }, { .name = "mem_exclusive", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_MEM_EXCLUSIVE, }, { |
786083667
|
1666 1667 1668 1669 1670 1671 1672 |
.name = "mem_hardwall", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_MEM_HARDWALL, }, { |
addf2c739
|
1673 1674 1675 1676 1677 1678 1679 1680 |
.name = "sched_load_balance", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_SCHED_LOAD_BALANCE, }, { .name = "sched_relax_domain_level", |
5be7a4792
|
1681 1682 |
.read_s64 = cpuset_read_s64, .write_s64 = cpuset_write_s64, |
addf2c739
|
1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 |
.private = FILE_SCHED_RELAX_DOMAIN_LEVEL, }, { .name = "memory_migrate", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_MEMORY_MIGRATE, }, { .name = "memory_pressure", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_MEMORY_PRESSURE, |
099fca322
|
1698 |
.mode = S_IRUGO, |
addf2c739
|
1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 |
}, { .name = "memory_spread_page", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_SPREAD_PAGE, }, { .name = "memory_spread_slab", .read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, .private = FILE_SPREAD_SLAB, }, |
45b07ef31
|
1714 |
}; |
3e0d98b9f
|
1715 1716 |
static struct cftype cft_memory_pressure_enabled = { .name = "memory_pressure_enabled", |
700fe1ab9
|
1717 1718 |
.read_u64 = cpuset_read_u64, .write_u64 = cpuset_write_u64, |
3e0d98b9f
|
1719 1720 |
.private = FILE_MEMORY_PRESSURE_ENABLED, }; |
8793d854e
|
1721 |
static int cpuset_populate(struct cgroup_subsys *ss, struct cgroup *cont) |
1da177e4c
|
1722 1723 |
{ int err; |
addf2c739
|
1724 1725 |
err = cgroup_add_files(cont, ss, files, ARRAY_SIZE(files)); if (err) |
1da177e4c
|
1726 |
return err; |
8793d854e
|
1727 |
/* memory_pressure_enabled is in root cpuset only */ |
addf2c739
|
1728 |
if (!cont->parent) |
8793d854e
|
1729 |
err = cgroup_add_file(cont, ss, |
addf2c739
|
1730 1731 |
&cft_memory_pressure_enabled); return err; |
1da177e4c
|
1732 1733 1734 |
} /* |
a77aea920
|
1735 1736 1737 |
* post_clone() is called during cgroup_create() when the * clone_children mount argument was specified. The cgroup * can not yet have any tasks. |
8793d854e
|
1738 1739 1740 1741 1742 1743 1744 1745 1746 |
* * Currently we refuse to set up the cgroup - thereby * refusing the task to be entered, and as a result refusing * the sys_unshare() or clone() which initiated it - if any * sibling cpusets have exclusive cpus or mem. * * If this becomes a problem for some users who wish to * allow that scenario, then cpuset_post_clone() could be * changed to grant parent->cpus_allowed-sibling_cpus_exclusive |
2df167a30
|
1747 1748 |
* (and likewise for mems) to the new cgroup. Called with cgroup_mutex * held. |
8793d854e
|
1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 |
*/ static void cpuset_post_clone(struct cgroup_subsys *ss, struct cgroup *cgroup) { struct cgroup *parent, *child; struct cpuset *cs, *parent_cs; parent = cgroup->parent; list_for_each_entry(child, &parent->children, sibling) { cs = cgroup_cs(child); if (is_mem_exclusive(cs) || is_cpu_exclusive(cs)) return; } cs = cgroup_cs(cgroup); parent_cs = cgroup_cs(parent); |
523fb486b
|
1764 |
mutex_lock(&callback_mutex); |
8793d854e
|
1765 |
cs->mems_allowed = parent_cs->mems_allowed; |
300ed6cbb
|
1766 |
cpumask_copy(cs->cpus_allowed, parent_cs->cpus_allowed); |
523fb486b
|
1767 |
mutex_unlock(&callback_mutex); |
8793d854e
|
1768 1769 1770 1771 |
return; } /* |
1da177e4c
|
1772 |
* cpuset_create - create a cpuset |
2df167a30
|
1773 1774 |
* ss: cpuset cgroup subsystem * cont: control group that the new cpuset will be part of |
1da177e4c
|
1775 |
*/ |
8793d854e
|
1776 1777 1778 |
static struct cgroup_subsys_state *cpuset_create( struct cgroup_subsys *ss, struct cgroup *cont) |
1da177e4c
|
1779 1780 |
{ struct cpuset *cs; |
8793d854e
|
1781 |
struct cpuset *parent; |
1da177e4c
|
1782 |
|
8793d854e
|
1783 |
if (!cont->parent) { |
8793d854e
|
1784 1785 1786 |
return &top_cpuset.css; } parent = cgroup_cs(cont->parent); |
1da177e4c
|
1787 1788 |
cs = kmalloc(sizeof(*cs), GFP_KERNEL); if (!cs) |
8793d854e
|
1789 |
return ERR_PTR(-ENOMEM); |
300ed6cbb
|
1790 1791 1792 1793 |
if (!alloc_cpumask_var(&cs->cpus_allowed, GFP_KERNEL)) { kfree(cs); return ERR_PTR(-ENOMEM); } |
1da177e4c
|
1794 |
|
1da177e4c
|
1795 |
cs->flags = 0; |
825a46af5
|
1796 1797 1798 1799 |
if (is_spread_page(parent)) set_bit(CS_SPREAD_PAGE, &cs->flags); if (is_spread_slab(parent)) set_bit(CS_SPREAD_SLAB, &cs->flags); |
029190c51
|
1800 |
set_bit(CS_SCHED_LOAD_BALANCE, &cs->flags); |
300ed6cbb
|
1801 |
cpumask_clear(cs->cpus_allowed); |
f9a86fcbb
|
1802 |
nodes_clear(cs->mems_allowed); |
3e0d98b9f
|
1803 |
fmeter_init(&cs->fmeter); |
1d3504fcf
|
1804 |
cs->relax_domain_level = -1; |
1da177e4c
|
1805 1806 |
cs->parent = parent; |
202f72d5d
|
1807 |
number_of_cpusets++; |
8793d854e
|
1808 |
return &cs->css ; |
1da177e4c
|
1809 |
} |
029190c51
|
1810 |
/* |
029190c51
|
1811 1812 |
* If the cpuset being removed has its flag 'sched_load_balance' * enabled, then simulate turning sched_load_balance off, which |
cf417141c
|
1813 |
* will call async_rebuild_sched_domains(). |
029190c51
|
1814 |
*/ |
8793d854e
|
1815 |
static void cpuset_destroy(struct cgroup_subsys *ss, struct cgroup *cont) |
1da177e4c
|
1816 |
{ |
8793d854e
|
1817 |
struct cpuset *cs = cgroup_cs(cont); |
1da177e4c
|
1818 |
|
029190c51
|
1819 |
if (is_sched_load_balance(cs)) |
700fe1ab9
|
1820 |
update_flag(CS_SCHED_LOAD_BALANCE, cs, 0); |
029190c51
|
1821 |
|
202f72d5d
|
1822 |
number_of_cpusets--; |
300ed6cbb
|
1823 |
free_cpumask_var(cs->cpus_allowed); |
8793d854e
|
1824 |
kfree(cs); |
1da177e4c
|
1825 |
} |
8793d854e
|
1826 1827 1828 |
struct cgroup_subsys cpuset_subsys = { .name = "cpuset", .create = cpuset_create, |
cf417141c
|
1829 |
.destroy = cpuset_destroy, |
8793d854e
|
1830 1831 1832 1833 1834 1835 1836 |
.can_attach = cpuset_can_attach, .attach = cpuset_attach, .populate = cpuset_populate, .post_clone = cpuset_post_clone, .subsys_id = cpuset_subsys_id, .early_init = 1, }; |
1da177e4c
|
1837 1838 1839 1840 1841 1842 1843 1844 |
/** * cpuset_init - initialize cpusets at system boot * * Description: Initialize top_cpuset and the cpuset internal file system, **/ int __init cpuset_init(void) { |
8793d854e
|
1845 |
int err = 0; |
1da177e4c
|
1846 |
|
58568d2a8
|
1847 1848 |
if (!alloc_cpumask_var(&top_cpuset.cpus_allowed, GFP_KERNEL)) BUG(); |
300ed6cbb
|
1849 |
cpumask_setall(top_cpuset.cpus_allowed); |
f9a86fcbb
|
1850 |
nodes_setall(top_cpuset.mems_allowed); |
1da177e4c
|
1851 |
|
3e0d98b9f
|
1852 |
fmeter_init(&top_cpuset.fmeter); |
029190c51
|
1853 |
set_bit(CS_SCHED_LOAD_BALANCE, &top_cpuset.flags); |
1d3504fcf
|
1854 |
top_cpuset.relax_domain_level = -1; |
1da177e4c
|
1855 |
|
1da177e4c
|
1856 1857 |
err = register_filesystem(&cpuset_fs_type); if (err < 0) |
8793d854e
|
1858 |
return err; |
2341d1b65
|
1859 1860 |
if (!alloc_cpumask_var(&cpus_attach, GFP_KERNEL)) BUG(); |
202f72d5d
|
1861 |
number_of_cpusets = 1; |
8793d854e
|
1862 |
return 0; |
1da177e4c
|
1863 |
} |
956db3ca0
|
1864 1865 1866 1867 1868 1869 1870 1871 |
/** * cpuset_do_move_task - move a given task to another cpuset * @tsk: pointer to task_struct the task to move * @scan: struct cgroup_scanner contained in its struct cpuset_hotplug_scanner * * Called by cgroup_scan_tasks() for each task in a cgroup. * Return nonzero to stop the walk through the tasks. */ |
9e0c914ca
|
1872 1873 |
static void cpuset_do_move_task(struct task_struct *tsk, struct cgroup_scanner *scan) |
956db3ca0
|
1874 |
{ |
7f81b1ae1
|
1875 |
struct cgroup *new_cgroup = scan->data; |
956db3ca0
|
1876 |
|
7f81b1ae1
|
1877 |
cgroup_attach_task(new_cgroup, tsk); |
956db3ca0
|
1878 1879 1880 1881 1882 1883 1884 |
} /** * move_member_tasks_to_cpuset - move tasks from one cpuset to another * @from: cpuset in which the tasks currently reside * @to: cpuset to which the tasks will be moved * |
c8d9c90c7
|
1885 1886 |
* Called with cgroup_mutex held * callback_mutex must not be held, as cpuset_attach() will take it. |
956db3ca0
|
1887 1888 1889 1890 1891 1892 |
* * The cgroup_scan_tasks() function will scan all the tasks in a cgroup, * calling callback functions for each. */ static void move_member_tasks_to_cpuset(struct cpuset *from, struct cpuset *to) { |
7f81b1ae1
|
1893 |
struct cgroup_scanner scan; |
956db3ca0
|
1894 |
|
7f81b1ae1
|
1895 1896 1897 1898 1899 |
scan.cg = from->css.cgroup; scan.test_task = NULL; /* select all tasks in cgroup */ scan.process_task = cpuset_do_move_task; scan.heap = NULL; scan.data = to->css.cgroup; |
956db3ca0
|
1900 |
|
7f81b1ae1
|
1901 |
if (cgroup_scan_tasks(&scan)) |
956db3ca0
|
1902 1903 1904 1905 |
printk(KERN_ERR "move_member_tasks_to_cpuset: " "cgroup_scan_tasks failed "); } |
b1aac8bb8
|
1906 |
/* |
cf417141c
|
1907 |
* If CPU and/or memory hotplug handlers, below, unplug any CPUs |
b1aac8bb8
|
1908 1909 |
* or memory nodes, we need to walk over the cpuset hierarchy, * removing that CPU or node from all cpusets. If this removes the |
956db3ca0
|
1910 1911 |
* last CPU or node from a cpuset, then move the tasks in the empty * cpuset to its next-highest non-empty parent. |
b1aac8bb8
|
1912 |
* |
c8d9c90c7
|
1913 1914 |
* Called with cgroup_mutex held * callback_mutex must not be held, as cpuset_attach() will take it. |
b1aac8bb8
|
1915 |
*/ |
956db3ca0
|
1916 1917 1918 |
static void remove_tasks_in_empty_cpuset(struct cpuset *cs) { struct cpuset *parent; |
c8d9c90c7
|
1919 1920 1921 1922 1923 |
/* * The cgroup's css_sets list is in use if there are tasks * in the cpuset; the list is empty if there are none; * the cs->css.refcnt seems always 0. */ |
956db3ca0
|
1924 1925 |
if (list_empty(&cs->css.cgroup->css_sets)) return; |
b1aac8bb8
|
1926 |
|
956db3ca0
|
1927 1928 1929 1930 1931 |
/* * Find its next-highest non-empty parent, (top cpuset * has online cpus, so can't be empty). */ parent = cs->parent; |
300ed6cbb
|
1932 |
while (cpumask_empty(parent->cpus_allowed) || |
b45012955
|
1933 |
nodes_empty(parent->mems_allowed)) |
956db3ca0
|
1934 |
parent = parent->parent; |
956db3ca0
|
1935 1936 1937 1938 1939 1940 1941 1942 |
move_member_tasks_to_cpuset(cs, parent); } /* * Walk the specified cpuset subtree and look for empty cpusets. * The tasks of such cpuset must be moved to a parent cpuset. * |
2df167a30
|
1943 |
* Called with cgroup_mutex held. We take callback_mutex to modify |
956db3ca0
|
1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 |
* cpus_allowed and mems_allowed. * * This walk processes the tree from top to bottom, completing one layer * before dropping down to the next. It always processes a node before * any of its children. * * For now, since we lack memory hot unplug, we'll never see a cpuset * that has tasks along with an empty 'mems'. But if we did see such * a cpuset, we'd handle it just like we do if its 'cpus' was empty. */ |
d294eb83d
|
1954 |
static void scan_for_empty_cpusets(struct cpuset *root) |
b1aac8bb8
|
1955 |
{ |
8d1e6266f
|
1956 |
LIST_HEAD(queue); |
956db3ca0
|
1957 1958 |
struct cpuset *cp; /* scans cpusets being updated */ struct cpuset *child; /* scans child cpusets of cp */ |
8793d854e
|
1959 |
struct cgroup *cont; |
ee24d3797
|
1960 |
static nodemask_t oldmems; /* protected by cgroup_mutex */ |
b1aac8bb8
|
1961 |
|
956db3ca0
|
1962 |
list_add_tail((struct list_head *)&root->stack_list, &queue); |
956db3ca0
|
1963 |
while (!list_empty(&queue)) { |
8d1e6266f
|
1964 |
cp = list_first_entry(&queue, struct cpuset, stack_list); |
956db3ca0
|
1965 1966 1967 1968 1969 |
list_del(queue.next); list_for_each_entry(cont, &cp->css.cgroup->children, sibling) { child = cgroup_cs(cont); list_add_tail(&child->stack_list, &queue); } |
b45012955
|
1970 1971 |
/* Continue past cpusets with all cpus, mems online */ |
6ad4c1888
|
1972 |
if (cpumask_subset(cp->cpus_allowed, cpu_active_mask) && |
b45012955
|
1973 1974 |
nodes_subset(cp->mems_allowed, node_states[N_HIGH_MEMORY])) continue; |
ee24d3797
|
1975 |
oldmems = cp->mems_allowed; |
f9b4fb8da
|
1976 |
|
956db3ca0
|
1977 |
/* Remove offline cpus and mems from this cpuset. */ |
b45012955
|
1978 |
mutex_lock(&callback_mutex); |
300ed6cbb
|
1979 |
cpumask_and(cp->cpus_allowed, cp->cpus_allowed, |
6ad4c1888
|
1980 |
cpu_active_mask); |
956db3ca0
|
1981 1982 |
nodes_and(cp->mems_allowed, cp->mems_allowed, node_states[N_HIGH_MEMORY]); |
b45012955
|
1983 1984 1985 |
mutex_unlock(&callback_mutex); /* Move tasks from the empty cpuset to a parent */ |
300ed6cbb
|
1986 |
if (cpumask_empty(cp->cpus_allowed) || |
b45012955
|
1987 |
nodes_empty(cp->mems_allowed)) |
956db3ca0
|
1988 |
remove_tasks_in_empty_cpuset(cp); |
f9b4fb8da
|
1989 |
else { |
4e74339af
|
1990 |
update_tasks_cpumask(cp, NULL); |
ee24d3797
|
1991 |
update_tasks_nodemask(cp, &oldmems, NULL); |
f9b4fb8da
|
1992 |
} |
b1aac8bb8
|
1993 1994 1995 1996 |
} } /* |
4c4d50f7b
|
1997 1998 1999 2000 2001 |
* The top_cpuset tracks what CPUs and Memory Nodes are online, * period. This is necessary in order to make cpusets transparent * (of no affect) on systems that are actively using CPU hotplug * but making no active use of cpusets. * |
38837fc75
|
2002 |
* This routine ensures that top_cpuset.cpus_allowed tracks |
3a101d054
|
2003 |
* cpu_active_mask on each CPU hotplug (cpuhp) event. |
cf417141c
|
2004 2005 2006 |
* * Called within get_online_cpus(). Needs to call cgroup_lock() * before calling generate_sched_domains(). |
4c4d50f7b
|
2007 |
*/ |
0b2e918aa
|
2008 |
void cpuset_update_active_cpus(void) |
4c4d50f7b
|
2009 |
{ |
cf417141c
|
2010 |
struct sched_domain_attr *attr; |
acc3f5d7c
|
2011 |
cpumask_var_t *doms; |
cf417141c
|
2012 |
int ndoms; |
cf417141c
|
2013 |
cgroup_lock(); |
0b4217b3f
|
2014 |
mutex_lock(&callback_mutex); |
6ad4c1888
|
2015 |
cpumask_copy(top_cpuset.cpus_allowed, cpu_active_mask); |
0b4217b3f
|
2016 |
mutex_unlock(&callback_mutex); |
cf417141c
|
2017 2018 2019 2020 2021 2022 |
scan_for_empty_cpusets(&top_cpuset); ndoms = generate_sched_domains(&doms, &attr); cgroup_unlock(); /* Have scheduler rebuild the domains */ partition_sched_domains(ndoms, doms, attr); |
4c4d50f7b
|
2023 |
} |
4c4d50f7b
|
2024 |
|
b1aac8bb8
|
2025 |
#ifdef CONFIG_MEMORY_HOTPLUG |
38837fc75
|
2026 |
/* |
0e1e7c7a7
|
2027 |
* Keep top_cpuset.mems_allowed tracking node_states[N_HIGH_MEMORY]. |
cf417141c
|
2028 2029 |
* Call this routine anytime after node_states[N_HIGH_MEMORY] changes. * See also the previous routine cpuset_track_online_cpus(). |
38837fc75
|
2030 |
*/ |
f481891fd
|
2031 2032 |
static int cpuset_track_online_nodes(struct notifier_block *self, unsigned long action, void *arg) |
38837fc75
|
2033 |
{ |
ee24d3797
|
2034 |
static nodemask_t oldmems; /* protected by cgroup_mutex */ |
5ab116c93
|
2035 |
|
cf417141c
|
2036 |
cgroup_lock(); |
f481891fd
|
2037 2038 |
switch (action) { case MEM_ONLINE: |
ee24d3797
|
2039 |
oldmems = top_cpuset.mems_allowed; |
0b4217b3f
|
2040 |
mutex_lock(&callback_mutex); |
f481891fd
|
2041 |
top_cpuset.mems_allowed = node_states[N_HIGH_MEMORY]; |
0b4217b3f
|
2042 |
mutex_unlock(&callback_mutex); |
ee24d3797
|
2043 |
update_tasks_nodemask(&top_cpuset, &oldmems, NULL); |
5ab116c93
|
2044 2045 2046 2047 2048 2049 2050 |
break; case MEM_OFFLINE: /* * needn't update top_cpuset.mems_allowed explicitly because * scan_for_empty_cpusets() will update it. */ scan_for_empty_cpusets(&top_cpuset); |
f481891fd
|
2051 2052 2053 2054 |
break; default: break; } |
cf417141c
|
2055 |
cgroup_unlock(); |
53feb2976
|
2056 |
|
f481891fd
|
2057 |
return NOTIFY_OK; |
38837fc75
|
2058 2059 |
} #endif |
1da177e4c
|
2060 2061 2062 2063 2064 2065 2066 2067 |
/** * cpuset_init_smp - initialize cpus_allowed * * Description: Finish top cpuset after cpu, node maps are initialized **/ void __init cpuset_init_smp(void) { |
6ad4c1888
|
2068 |
cpumask_copy(top_cpuset.cpus_allowed, cpu_active_mask); |
0e1e7c7a7
|
2069 |
top_cpuset.mems_allowed = node_states[N_HIGH_MEMORY]; |
4c4d50f7b
|
2070 |
|
f481891fd
|
2071 |
hotplug_memory_notifier(cpuset_track_online_nodes, 10); |
f90d4118b
|
2072 2073 2074 |
cpuset_wq = create_singlethread_workqueue("cpuset"); BUG_ON(!cpuset_wq); |
1da177e4c
|
2075 2076 2077 |
} /** |
1da177e4c
|
2078 2079 |
* cpuset_cpus_allowed - return cpus_allowed mask from a tasks cpuset. * @tsk: pointer to task_struct from which to obtain cpuset->cpus_allowed. |
6af866af3
|
2080 |
* @pmask: pointer to struct cpumask variable to receive cpus_allowed set. |
1da177e4c
|
2081 |
* |
300ed6cbb
|
2082 |
* Description: Returns the cpumask_var_t cpus_allowed of the cpuset |
1da177e4c
|
2083 2084 2085 2086 |
* attached to the specified @tsk. Guaranteed to return some non-empty * subset of cpu_online_map, even if this means going outside the * tasks cpuset. **/ |
6af866af3
|
2087 |
void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask) |
1da177e4c
|
2088 |
{ |
3d3f26a7b
|
2089 |
mutex_lock(&callback_mutex); |
909d75a3b
|
2090 |
task_lock(tsk); |
f9a86fcbb
|
2091 |
guarantee_online_cpus(task_cs(tsk), pmask); |
909d75a3b
|
2092 |
task_unlock(tsk); |
897f0b3c3
|
2093 |
mutex_unlock(&callback_mutex); |
1da177e4c
|
2094 |
} |
9084bb824
|
2095 2096 2097 2098 2099 2100 2101 2102 |
int cpuset_cpus_allowed_fallback(struct task_struct *tsk) { const struct cpuset *cs; int cpu; rcu_read_lock(); cs = task_cs(tsk); if (cs) |
1e1b6c511
|
2103 |
do_set_cpus_allowed(tsk, cs->cpus_allowed); |
9084bb824
|
2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 |
rcu_read_unlock(); /* * We own tsk->cpus_allowed, nobody can change it under us. * * But we used cs && cs->cpus_allowed lockless and thus can * race with cgroup_attach_task() or update_cpumask() and get * the wrong tsk->cpus_allowed. However, both cases imply the * subsequent cpuset_change_cpumask()->set_cpus_allowed_ptr() * which takes task_rq_lock(). * * If we are called after it dropped the lock we must see all * changes in tsk_cs()->cpus_allowed. Otherwise we can temporary * set any mask even if it is not right from task_cs() pov, * the pending set_cpus_allowed_ptr() will fix things. */ cpu = cpumask_any_and(&tsk->cpus_allowed, cpu_active_mask); if (cpu >= nr_cpu_ids) { /* * Either tsk->cpus_allowed is wrong (see above) or it * is actually empty. The latter case is only possible * if we are racing with remove_tasks_in_empty_cpuset(). * Like above we can temporary set any mask and rely on * set_cpus_allowed_ptr() as synchronization point. */ |
1e1b6c511
|
2130 |
do_set_cpus_allowed(tsk, cpu_possible_mask); |
9084bb824
|
2131 2132 2133 2134 2135 |
cpu = cpumask_any(cpu_active_mask); } return cpu; } |
1da177e4c
|
2136 2137 |
void cpuset_init_current_mems_allowed(void) { |
f9a86fcbb
|
2138 |
nodes_setall(current->mems_allowed); |
1da177e4c
|
2139 |
} |
d9fd8a6d4
|
2140 |
/** |
909d75a3b
|
2141 2142 2143 2144 2145 |
* cpuset_mems_allowed - return mems_allowed mask from a tasks cpuset. * @tsk: pointer to task_struct from which to obtain cpuset->mems_allowed. * * Description: Returns the nodemask_t mems_allowed of the cpuset * attached to the specified @tsk. Guaranteed to return some non-empty |
0e1e7c7a7
|
2146 |
* subset of node_states[N_HIGH_MEMORY], even if this means going outside the |
909d75a3b
|
2147 2148 2149 2150 2151 2152 |
* tasks cpuset. **/ nodemask_t cpuset_mems_allowed(struct task_struct *tsk) { nodemask_t mask; |
3d3f26a7b
|
2153 |
mutex_lock(&callback_mutex); |
909d75a3b
|
2154 |
task_lock(tsk); |
8793d854e
|
2155 |
guarantee_online_mems(task_cs(tsk), &mask); |
909d75a3b
|
2156 |
task_unlock(tsk); |
3d3f26a7b
|
2157 |
mutex_unlock(&callback_mutex); |
909d75a3b
|
2158 2159 2160 2161 2162 |
return mask; } /** |
19770b326
|
2163 2164 |
* cpuset_nodemask_valid_mems_allowed - check nodemask vs. curremt mems_allowed * @nodemask: the nodemask to be checked |
d9fd8a6d4
|
2165 |
* |
19770b326
|
2166 |
* Are any of the nodes in the nodemask allowed in current->mems_allowed? |
1da177e4c
|
2167 |
*/ |
19770b326
|
2168 |
int cpuset_nodemask_valid_mems_allowed(nodemask_t *nodemask) |
1da177e4c
|
2169 |
{ |
19770b326
|
2170 |
return nodes_intersects(*nodemask, current->mems_allowed); |
1da177e4c
|
2171 |
} |
9bf2229f8
|
2172 |
/* |
786083667
|
2173 2174 2175 2176 |
* nearest_hardwall_ancestor() - Returns the nearest mem_exclusive or * mem_hardwall ancestor to the specified cpuset. Call holding * callback_mutex. If no ancestor is mem_exclusive or mem_hardwall * (an unusual configuration), then returns the root cpuset. |
9bf2229f8
|
2177 |
*/ |
786083667
|
2178 |
static const struct cpuset *nearest_hardwall_ancestor(const struct cpuset *cs) |
9bf2229f8
|
2179 |
{ |
786083667
|
2180 |
while (!(is_mem_exclusive(cs) || is_mem_hardwall(cs)) && cs->parent) |
9bf2229f8
|
2181 2182 2183 |
cs = cs->parent; return cs; } |
d9fd8a6d4
|
2184 |
/** |
a1bc5a4ee
|
2185 2186 |
* cpuset_node_allowed_softwall - Can we allocate on a memory node? * @node: is this an allowed node? |
02a0e53d8
|
2187 |
* @gfp_mask: memory allocation flags |
d9fd8a6d4
|
2188 |
* |
a1bc5a4ee
|
2189 2190 2191 2192 2193 2194 |
* If we're in interrupt, yes, we can always allocate. If __GFP_THISNODE is * set, yes, we can always allocate. If node is in our task's mems_allowed, * yes. If it's not a __GFP_HARDWALL request and this node is in the nearest * hardwalled cpuset ancestor to this task's cpuset, yes. If the task has been * OOM killed and has access to memory reserves as specified by the TIF_MEMDIE * flag, yes. |
9bf2229f8
|
2195 2196 |
* Otherwise, no. * |
a1bc5a4ee
|
2197 2198 2199 |
* If __GFP_HARDWALL is set, cpuset_node_allowed_softwall() reduces to * cpuset_node_allowed_hardwall(). Otherwise, cpuset_node_allowed_softwall() * might sleep, and might allow a node from an enclosing cpuset. |
02a0e53d8
|
2200 |
* |
a1bc5a4ee
|
2201 2202 |
* cpuset_node_allowed_hardwall() only handles the simpler case of hardwall * cpusets, and never sleeps. |
02a0e53d8
|
2203 2204 2205 2206 2207 2208 2209 |
* * The __GFP_THISNODE placement logic is really handled elsewhere, * by forcibly using a zonelist starting at a specified node, and by * (in get_page_from_freelist()) refusing to consider the zones for * any node on the zonelist except the first. By the time any such * calls get to this routine, we should just shut up and say 'yes'. * |
9bf2229f8
|
2210 |
* GFP_USER allocations are marked with the __GFP_HARDWALL bit, |
c596d9f32
|
2211 2212 |
* and do not allow allocations outside the current tasks cpuset * unless the task has been OOM killed as is marked TIF_MEMDIE. |
9bf2229f8
|
2213 |
* GFP_KERNEL allocations are not so marked, so can escape to the |
786083667
|
2214 |
* nearest enclosing hardwalled ancestor cpuset. |
9bf2229f8
|
2215 |
* |
02a0e53d8
|
2216 2217 2218 2219 2220 2221 2222 |
* Scanning up parent cpusets requires callback_mutex. The * __alloc_pages() routine only calls here with __GFP_HARDWALL bit * _not_ set if it's a GFP_KERNEL allocation, and all nodes in the * current tasks mems_allowed came up empty on the first pass over * the zonelist. So only GFP_KERNEL allocations, if all nodes in the * cpuset are short of memory, might require taking the callback_mutex * mutex. |
9bf2229f8
|
2223 |
* |
36be57ffe
|
2224 |
* The first call here from mm/page_alloc:get_page_from_freelist() |
02a0e53d8
|
2225 2226 2227 |
* has __GFP_HARDWALL set in gfp_mask, enforcing hardwall cpusets, * so no allocation on a node outside the cpuset is allowed (unless * in interrupt, of course). |
36be57ffe
|
2228 2229 2230 2231 2232 2233 |
* * The second pass through get_page_from_freelist() doesn't even call * here for GFP_ATOMIC calls. For those calls, the __alloc_pages() * variable 'wait' is not set, and the bit ALLOC_CPUSET is not set * in alloc_flags. That logic and the checks below have the combined * affect that: |
9bf2229f8
|
2234 2235 |
* in_interrupt - any node ok (current task context irrelevant) * GFP_ATOMIC - any node ok |
c596d9f32
|
2236 |
* TIF_MEMDIE - any node ok |
786083667
|
2237 |
* GFP_KERNEL - any node in enclosing hardwalled cpuset ok |
9bf2229f8
|
2238 |
* GFP_USER - only nodes in current tasks mems allowed ok. |
36be57ffe
|
2239 2240 |
* * Rule: |
a1bc5a4ee
|
2241 |
* Don't call cpuset_node_allowed_softwall if you can't sleep, unless you |
36be57ffe
|
2242 2243 |
* pass in the __GFP_HARDWALL flag set in gfp_flag, which disables * the code that might scan up ancestor cpusets and sleep. |
02a0e53d8
|
2244 |
*/ |
a1bc5a4ee
|
2245 |
int __cpuset_node_allowed_softwall(int node, gfp_t gfp_mask) |
1da177e4c
|
2246 |
{ |
9bf2229f8
|
2247 |
const struct cpuset *cs; /* current cpuset ancestors */ |
29afd49b7
|
2248 |
int allowed; /* is allocation in zone z allowed? */ |
9bf2229f8
|
2249 |
|
9b819d204
|
2250 |
if (in_interrupt() || (gfp_mask & __GFP_THISNODE)) |
9bf2229f8
|
2251 |
return 1; |
92d1dbd27
|
2252 |
might_sleep_if(!(gfp_mask & __GFP_HARDWALL)); |
9bf2229f8
|
2253 2254 |
if (node_isset(node, current->mems_allowed)) return 1; |
c596d9f32
|
2255 2256 2257 2258 2259 2260 |
/* * Allow tasks that have access to memory reserves because they have * been OOM killed to get memory anywhere. */ if (unlikely(test_thread_flag(TIF_MEMDIE))) return 1; |
9bf2229f8
|
2261 2262 |
if (gfp_mask & __GFP_HARDWALL) /* If hardwall request, stop here */ return 0; |
5563e7707
|
2263 2264 |
if (current->flags & PF_EXITING) /* Let dying task have memory */ return 1; |
9bf2229f8
|
2265 |
/* Not hardwall and node outside mems_allowed: scan up cpusets */ |
3d3f26a7b
|
2266 |
mutex_lock(&callback_mutex); |
053199edf
|
2267 |
|
053199edf
|
2268 |
task_lock(current); |
786083667
|
2269 |
cs = nearest_hardwall_ancestor(task_cs(current)); |
053199edf
|
2270 |
task_unlock(current); |
9bf2229f8
|
2271 |
allowed = node_isset(node, cs->mems_allowed); |
3d3f26a7b
|
2272 |
mutex_unlock(&callback_mutex); |
9bf2229f8
|
2273 |
return allowed; |
1da177e4c
|
2274 |
} |
02a0e53d8
|
2275 |
/* |
a1bc5a4ee
|
2276 2277 |
* cpuset_node_allowed_hardwall - Can we allocate on a memory node? * @node: is this an allowed node? |
02a0e53d8
|
2278 2279 |
* @gfp_mask: memory allocation flags * |
a1bc5a4ee
|
2280 2281 2282 2283 2284 |
* If we're in interrupt, yes, we can always allocate. If __GFP_THISNODE is * set, yes, we can always allocate. If node is in our task's mems_allowed, * yes. If the task has been OOM killed and has access to memory reserves as * specified by the TIF_MEMDIE flag, yes. * Otherwise, no. |
02a0e53d8
|
2285 2286 2287 2288 2289 2290 2291 |
* * The __GFP_THISNODE placement logic is really handled elsewhere, * by forcibly using a zonelist starting at a specified node, and by * (in get_page_from_freelist()) refusing to consider the zones for * any node on the zonelist except the first. By the time any such * calls get to this routine, we should just shut up and say 'yes'. * |
a1bc5a4ee
|
2292 2293 |
* Unlike the cpuset_node_allowed_softwall() variant, above, * this variant requires that the node be in the current task's |
02a0e53d8
|
2294 2295 2296 2297 |
* mems_allowed or that we're in interrupt. It does not scan up the * cpuset hierarchy for the nearest enclosing mem_exclusive cpuset. * It never sleeps. */ |
a1bc5a4ee
|
2298 |
int __cpuset_node_allowed_hardwall(int node, gfp_t gfp_mask) |
02a0e53d8
|
2299 |
{ |
02a0e53d8
|
2300 2301 |
if (in_interrupt() || (gfp_mask & __GFP_THISNODE)) return 1; |
02a0e53d8
|
2302 2303 |
if (node_isset(node, current->mems_allowed)) return 1; |
dedf8b79e
|
2304 2305 2306 2307 2308 2309 |
/* * Allow tasks that have access to memory reserves because they have * been OOM killed to get memory anywhere. */ if (unlikely(test_thread_flag(TIF_MEMDIE))) return 1; |
02a0e53d8
|
2310 2311 |
return 0; } |
ef08e3b49
|
2312 |
/** |
505970b96
|
2313 2314 2315 2316 2317 2318 2319 |
* cpuset_unlock - release lock on cpuset changes * * Undo the lock taken in a previous cpuset_lock() call. */ void cpuset_unlock(void) { |
3d3f26a7b
|
2320 |
mutex_unlock(&callback_mutex); |
505970b96
|
2321 2322 2323 |
} /** |
6adef3ebe
|
2324 2325 |
* cpuset_mem_spread_node() - On which node to begin search for a file page * cpuset_slab_spread_node() - On which node to begin search for a slab page |
825a46af5
|
2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 |
* * If a task is marked PF_SPREAD_PAGE or PF_SPREAD_SLAB (as for * tasks in a cpuset with is_spread_page or is_spread_slab set), * and if the memory allocation used cpuset_mem_spread_node() * to determine on which node to start looking, as it will for * certain page cache or slab cache pages such as used for file * system buffers and inode caches, then instead of starting on the * local node to look for a free page, rather spread the starting * node around the tasks mems_allowed nodes. * * We don't have to worry about the returned node being offline * because "it can't happen", and even if it did, it would be ok. * * The routines calling guarantee_online_mems() are careful to * only set nodes in task->mems_allowed that are online. So it * should not be possible for the following code to return an * offline node. But if it did, that would be ok, as this routine * is not returning the node where the allocation must be, only * the node where the search should start. The zonelist passed to * __alloc_pages() will include all nodes. If the slab allocator * is passed an offline node, it will fall back to the local node. * See kmem_cache_alloc_node(). */ |
6adef3ebe
|
2349 |
static int cpuset_spread_node(int *rotor) |
825a46af5
|
2350 2351 |
{ int node; |
6adef3ebe
|
2352 |
node = next_node(*rotor, current->mems_allowed); |
825a46af5
|
2353 2354 |
if (node == MAX_NUMNODES) node = first_node(current->mems_allowed); |
6adef3ebe
|
2355 |
*rotor = node; |
825a46af5
|
2356 2357 |
return node; } |
6adef3ebe
|
2358 2359 2360 |
int cpuset_mem_spread_node(void) { |
778d3b0ff
|
2361 2362 2363 |
if (current->cpuset_mem_spread_rotor == NUMA_NO_NODE) current->cpuset_mem_spread_rotor = node_random(¤t->mems_allowed); |
6adef3ebe
|
2364 2365 2366 2367 2368 |
return cpuset_spread_node(¤t->cpuset_mem_spread_rotor); } int cpuset_slab_spread_node(void) { |
778d3b0ff
|
2369 2370 2371 |
if (current->cpuset_slab_spread_rotor == NUMA_NO_NODE) current->cpuset_slab_spread_rotor = node_random(¤t->mems_allowed); |
6adef3ebe
|
2372 2373 |
return cpuset_spread_node(¤t->cpuset_slab_spread_rotor); } |
825a46af5
|
2374 2375 2376 |
EXPORT_SYMBOL_GPL(cpuset_mem_spread_node); /** |
bbe373f2c
|
2377 2378 2379 2380 2381 2382 2383 2384 |
* cpuset_mems_allowed_intersects - Does @tsk1's mems_allowed intersect @tsk2's? * @tsk1: pointer to task_struct of some task. * @tsk2: pointer to task_struct of some other task. * * Description: Return true if @tsk1's mems_allowed intersects the * mems_allowed of @tsk2. Used by the OOM killer to determine if * one of the task's memory usage might impact the memory available * to the other. |
ef08e3b49
|
2385 |
**/ |
bbe373f2c
|
2386 2387 |
int cpuset_mems_allowed_intersects(const struct task_struct *tsk1, const struct task_struct *tsk2) |
ef08e3b49
|
2388 |
{ |
bbe373f2c
|
2389 |
return nodes_intersects(tsk1->mems_allowed, tsk2->mems_allowed); |
ef08e3b49
|
2390 |
} |
75aa19941
|
2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 |
/** * cpuset_print_task_mems_allowed - prints task's cpuset and mems_allowed * @task: pointer to task_struct of some task. * * Description: Prints @task's name, cpuset name, and cached copy of its * mems_allowed to the kernel log. Must hold task_lock(task) to allow * dereferencing task_cs(task). */ void cpuset_print_task_mems_allowed(struct task_struct *tsk) { struct dentry *dentry; dentry = task_cs(tsk)->css.cgroup->dentry; spin_lock(&cpuset_buffer_lock); snprintf(cpuset_name, CPUSET_NAME_LEN, dentry ? (const char *)dentry->d_name.name : "/"); nodelist_scnprintf(cpuset_nodelist, CPUSET_NODELIST_LEN, tsk->mems_allowed); printk(KERN_INFO "%s cpuset=%s mems_allowed=%s ", tsk->comm, cpuset_name, cpuset_nodelist); spin_unlock(&cpuset_buffer_lock); } |
1da177e4c
|
2414 |
/* |
3e0d98b9f
|
2415 2416 2417 2418 |
* Collection of memory_pressure is suppressed unless * this flag is enabled by writing "1" to the special * cpuset file 'memory_pressure_enabled' in the root cpuset. */ |
c5b2aff89
|
2419 |
int cpuset_memory_pressure_enabled __read_mostly; |
3e0d98b9f
|
2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 |
/** * cpuset_memory_pressure_bump - keep stats of per-cpuset reclaims. * * Keep a running average of the rate of synchronous (direct) * page reclaim efforts initiated by tasks in each cpuset. * * This represents the rate at which some task in the cpuset * ran low on memory on all nodes it was allowed to use, and * had to enter the kernels page reclaim code in an effort to * create more free memory by tossing clean pages or swapping * or writing dirty pages. * * Display to user space in the per-cpuset read-only file * "memory_pressure". Value displayed is an integer * representing the recent rate of entry into the synchronous * (direct) page reclaim by any task attached to the cpuset. **/ void __cpuset_memory_pressure_bump(void) { |
3e0d98b9f
|
2441 |
task_lock(current); |
8793d854e
|
2442 |
fmeter_markevent(&task_cs(current)->fmeter); |
3e0d98b9f
|
2443 2444 |
task_unlock(current); } |
8793d854e
|
2445 |
#ifdef CONFIG_PROC_PID_CPUSET |
3e0d98b9f
|
2446 |
/* |
1da177e4c
|
2447 2448 2449 |
* proc_cpuset_show() * - Print tasks cpuset path into seq_file. * - Used for /proc/<pid>/cpuset. |
053199edf
|
2450 2451 |
* - No need to task_lock(tsk) on this tsk->cpuset reference, as it * doesn't really matter if tsk->cpuset changes after we read it, |
c8d9c90c7
|
2452 |
* and we take cgroup_mutex, keeping cpuset_attach() from changing it |
2df167a30
|
2453 |
* anyway. |
1da177e4c
|
2454 |
*/ |
029190c51
|
2455 |
static int proc_cpuset_show(struct seq_file *m, void *unused_v) |
1da177e4c
|
2456 |
{ |
13b41b094
|
2457 |
struct pid *pid; |
1da177e4c
|
2458 2459 |
struct task_struct *tsk; char *buf; |
8793d854e
|
2460 |
struct cgroup_subsys_state *css; |
99f895518
|
2461 |
int retval; |
1da177e4c
|
2462 |
|
99f895518
|
2463 |
retval = -ENOMEM; |
1da177e4c
|
2464 2465 |
buf = kmalloc(PAGE_SIZE, GFP_KERNEL); if (!buf) |
99f895518
|
2466 2467 2468 |
goto out; retval = -ESRCH; |
13b41b094
|
2469 2470 |
pid = m->private; tsk = get_pid_task(pid, PIDTYPE_PID); |
99f895518
|
2471 2472 |
if (!tsk) goto out_free; |
1da177e4c
|
2473 |
|
99f895518
|
2474 |
retval = -EINVAL; |
8793d854e
|
2475 2476 2477 |
cgroup_lock(); css = task_subsys_state(tsk, cpuset_subsys_id); retval = cgroup_path(css->cgroup, buf, PAGE_SIZE); |
1da177e4c
|
2478 |
if (retval < 0) |
99f895518
|
2479 |
goto out_unlock; |
1da177e4c
|
2480 2481 2482 |
seq_puts(m, buf); seq_putc(m, ' '); |
99f895518
|
2483 |
out_unlock: |
8793d854e
|
2484 |
cgroup_unlock(); |
99f895518
|
2485 2486 |
put_task_struct(tsk); out_free: |
1da177e4c
|
2487 |
kfree(buf); |
99f895518
|
2488 |
out: |
1da177e4c
|
2489 2490 2491 2492 2493 |
return retval; } static int cpuset_open(struct inode *inode, struct file *file) { |
13b41b094
|
2494 2495 |
struct pid *pid = PROC_I(inode)->pid; return single_open(file, proc_cpuset_show, pid); |
1da177e4c
|
2496 |
} |
9a32144e9
|
2497 |
const struct file_operations proc_cpuset_operations = { |
1da177e4c
|
2498 2499 2500 2501 2502 |
.open = cpuset_open, .read = seq_read, .llseek = seq_lseek, .release = single_release, }; |
8793d854e
|
2503 |
#endif /* CONFIG_PROC_PID_CPUSET */ |
1da177e4c
|
2504 |
|
d01d48278
|
2505 |
/* Display task mems_allowed in /proc/<pid>/status file. */ |
df5f8314c
|
2506 2507 |
void cpuset_task_status_allowed(struct seq_file *m, struct task_struct *task) { |
df5f8314c
|
2508 |
seq_printf(m, "Mems_allowed:\t"); |
30e8e1360
|
2509 |
seq_nodemask(m, &task->mems_allowed); |
df5f8314c
|
2510 2511 |
seq_printf(m, " "); |
39106dcf8
|
2512 |
seq_printf(m, "Mems_allowed_list:\t"); |
30e8e1360
|
2513 |
seq_nodemask_list(m, &task->mems_allowed); |
39106dcf8
|
2514 2515 |
seq_printf(m, " "); |
1da177e4c
|
2516 |
} |