memcontrol.c 61.2 KB
Newer Older
B
Balbir Singh 已提交
1 2 3 4 5
/* memcontrol.c - Memory Controller
 *
 * Copyright IBM Corporation, 2007
 * Author Balbir Singh <balbir@linux.vnet.ibm.com>
 *
6 7 8
 * Copyright 2007 OpenVZ SWsoft Inc
 * Author: Pavel Emelianov <xemul@openvz.org>
 *
B
Balbir Singh 已提交
9 10 11 12 13 14 15 16 17 18 19 20 21 22
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 */

#include <linux/res_counter.h>
#include <linux/memcontrol.h>
#include <linux/cgroup.h>
23
#include <linux/mm.h>
K
KAMEZAWA Hiroyuki 已提交
24
#include <linux/pagemap.h>
25
#include <linux/smp.h>
26
#include <linux/page-flags.h>
27
#include <linux/backing-dev.h>
28 29
#include <linux/bit_spinlock.h>
#include <linux/rcupdate.h>
30
#include <linux/limits.h>
31
#include <linux/mutex.h>
32
#include <linux/slab.h>
33 34 35
#include <linux/swap.h>
#include <linux/spinlock.h>
#include <linux/fs.h>
36
#include <linux/seq_file.h>
37
#include <linux/vmalloc.h>
38
#include <linux/mm_inline.h>
39
#include <linux/page_cgroup.h>
K
KAMEZAWA Hiroyuki 已提交
40
#include "internal.h"
B
Balbir Singh 已提交
41

42 43
#include <asm/uaccess.h>

44 45
struct cgroup_subsys mem_cgroup_subsys __read_mostly;
#define MEM_CGROUP_RECLAIM_RETRIES	5
B
Balbir Singh 已提交
46

47 48 49 50 51 52 53 54
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
/* Turned on only when memory cgroup is enabled && really_do_swap_account = 0 */
int do_swap_account __read_mostly;
static int really_do_swap_account __initdata = 1; /* for remember boot option*/
#else
#define do_swap_account		(0)
#endif

55
static DEFINE_MUTEX(memcg_tasklist);	/* can be hold under cgroup_mutex */
56

57 58 59 60 61 62 63 64 65
/*
 * Statistics for memory cgroup.
 */
enum mem_cgroup_stat_index {
	/*
	 * For MEM_CONTAINER_TYPE_ALL, usage = pagecache + rss.
	 */
	MEM_CGROUP_STAT_CACHE, 	   /* # of pages charged as cache */
	MEM_CGROUP_STAT_RSS,	   /* # of pages charged as rss */
66 67
	MEM_CGROUP_STAT_PGPGIN_COUNT,	/* # of pages paged in */
	MEM_CGROUP_STAT_PGPGOUT_COUNT,	/* # of pages paged out */
68 69 70 71 72 73 74 75 76

	MEM_CGROUP_STAT_NSTATS,
};

struct mem_cgroup_stat_cpu {
	s64 count[MEM_CGROUP_STAT_NSTATS];
} ____cacheline_aligned_in_smp;

struct mem_cgroup_stat {
77
	struct mem_cgroup_stat_cpu cpustat[0];
78 79 80 81 82
};

/*
 * For accounting under irq disable, no need for increment preempt count.
 */
83
static inline void __mem_cgroup_stat_add_safe(struct mem_cgroup_stat_cpu *stat,
84 85
		enum mem_cgroup_stat_index idx, int val)
{
86
	stat->count[idx] += val;
87 88 89 90 91 92 93 94 95 96 97 98
}

static s64 mem_cgroup_read_stat(struct mem_cgroup_stat *stat,
		enum mem_cgroup_stat_index idx)
{
	int cpu;
	s64 ret = 0;
	for_each_possible_cpu(cpu)
		ret += stat->cpustat[cpu].count[idx];
	return ret;
}

K
KAMEZAWA Hiroyuki 已提交
99 100 101 102 103 104 105 106 107
static s64 mem_cgroup_local_usage(struct mem_cgroup_stat *stat)
{
	s64 ret;

	ret = mem_cgroup_read_stat(stat, MEM_CGROUP_STAT_CACHE);
	ret += mem_cgroup_read_stat(stat, MEM_CGROUP_STAT_RSS);
	return ret;
}

108 109 110 111
/*
 * per-zone information in memory controller.
 */
struct mem_cgroup_per_zone {
112 113 114
	/*
	 * spin_lock to protect the per cgroup LRU
	 */
115 116
	struct list_head	lists[NR_LRU_LISTS];
	unsigned long		count[NR_LRU_LISTS];
K
KOSAKI Motohiro 已提交
117 118

	struct zone_reclaim_stat reclaim_stat;
119 120 121 122 123 124 125 126 127 128 129 130
};
/* Macro for accessing counter */
#define MEM_CGROUP_ZSTAT(mz, idx)	((mz)->count[(idx)])

struct mem_cgroup_per_node {
	struct mem_cgroup_per_zone zoneinfo[MAX_NR_ZONES];
};

struct mem_cgroup_lru_info {
	struct mem_cgroup_per_node *nodeinfo[MAX_NUMNODES];
};

B
Balbir Singh 已提交
131 132 133 134 135 136 137
/*
 * The memory controller data structure. The memory controller controls both
 * page cache and RSS per cgroup. We would eventually like to provide
 * statistics based on the statistics developed by Rik Van Riel for clock-pro,
 * to help the administrator determine what knobs to tune.
 *
 * TODO: Add a water mark for the memory controller. Reclaim will begin when
138 139 140
 * we hit the water mark. May be even add a low water mark, such that
 * no reclaim occurs from a cgroup at it's low water mark, this is
 * a feature that will be implemented much later in the future.
B
Balbir Singh 已提交
141 142 143 144 145 146 147
 */
struct mem_cgroup {
	struct cgroup_subsys_state css;
	/*
	 * the counter to account for memory usage
	 */
	struct res_counter res;
148 149 150 151
	/*
	 * the counter to account for mem+swap usage.
	 */
	struct res_counter memsw;
152 153 154 155
	/*
	 * Per cgroup active and inactive list, similar to the
	 * per zone LRU lists.
	 */
156
	struct mem_cgroup_lru_info info;
157

K
KOSAKI Motohiro 已提交
158 159 160 161 162
	/*
	  protect against reclaim related member.
	*/
	spinlock_t reclaim_param_lock;

163
	int	prev_priority;	/* for recording reclaim priority */
164 165 166

	/*
	 * While reclaiming in a hiearchy, we cache the last child we
K
KAMEZAWA Hiroyuki 已提交
167
	 * reclaimed from.
168
	 */
K
KAMEZAWA Hiroyuki 已提交
169
	int last_scanned_child;
170 171 172 173
	/*
	 * Should the accounting and control be hierarchical, per subtree?
	 */
	bool use_hierarchy;
174
	unsigned long	last_oom_jiffies;
175
	atomic_t	refcnt;
176

K
KOSAKI Motohiro 已提交
177 178
	unsigned int	swappiness;

179
	/*
180
	 * statistics. This must be placed at the end of memcg.
181 182
	 */
	struct mem_cgroup_stat stat;
B
Balbir Singh 已提交
183 184
};

185 186 187
enum charge_type {
	MEM_CGROUP_CHARGE_TYPE_CACHE = 0,
	MEM_CGROUP_CHARGE_TYPE_MAPPED,
188
	MEM_CGROUP_CHARGE_TYPE_SHMEM,	/* used by page migration of shmem */
189
	MEM_CGROUP_CHARGE_TYPE_FORCE,	/* used by force_empty */
K
KAMEZAWA Hiroyuki 已提交
190
	MEM_CGROUP_CHARGE_TYPE_SWAPOUT,	/* for accounting swapcache */
191 192 193
	NR_CHARGE_TYPE,
};

194 195 196 197
/* only for here (for easy reading.) */
#define PCGF_CACHE	(1UL << PCG_CACHE)
#define PCGF_USED	(1UL << PCG_USED)
#define PCGF_LOCK	(1UL << PCG_LOCK)
198 199
static const unsigned long
pcg_default_flags[NR_CHARGE_TYPE] = {
K
KAMEZAWA Hiroyuki 已提交
200 201 202
	PCGF_CACHE | PCGF_USED | PCGF_LOCK, /* File Cache */
	PCGF_USED | PCGF_LOCK, /* Anon */
	PCGF_CACHE | PCGF_USED | PCGF_LOCK, /* Shmem */
203
	0, /* FORCE */
204 205
};

206 207 208 209 210 211 212 213 214
/* for encoding cft->private value on file */
#define _MEM			(0)
#define _MEMSWAP		(1)
#define MEMFILE_PRIVATE(x, val)	(((x) << 16) | (val))
#define MEMFILE_TYPE(val)	(((val) >> 16) & 0xffff)
#define MEMFILE_ATTR(val)	((val) & 0xffff)

static void mem_cgroup_get(struct mem_cgroup *mem);
static void mem_cgroup_put(struct mem_cgroup *mem);
215
static struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *mem);
216

217 218 219
static void mem_cgroup_charge_statistics(struct mem_cgroup *mem,
					 struct page_cgroup *pc,
					 bool charge)
220 221 222
{
	int val = (charge)? 1 : -1;
	struct mem_cgroup_stat *stat = &mem->stat;
223
	struct mem_cgroup_stat_cpu *cpustat;
K
KAMEZAWA Hiroyuki 已提交
224
	int cpu = get_cpu();
225

K
KAMEZAWA Hiroyuki 已提交
226
	cpustat = &stat->cpustat[cpu];
227
	if (PageCgroupCache(pc))
228
		__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_CACHE, val);
229
	else
230
		__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_RSS, val);
231 232

	if (charge)
233
		__mem_cgroup_stat_add_safe(cpustat,
234 235
				MEM_CGROUP_STAT_PGPGIN_COUNT, 1);
	else
236
		__mem_cgroup_stat_add_safe(cpustat,
237
				MEM_CGROUP_STAT_PGPGOUT_COUNT, 1);
K
KAMEZAWA Hiroyuki 已提交
238
	put_cpu();
239 240
}

241
static struct mem_cgroup_per_zone *
242 243 244 245 246
mem_cgroup_zoneinfo(struct mem_cgroup *mem, int nid, int zid)
{
	return &mem->info.nodeinfo[nid]->zoneinfo[zid];
}

247
static struct mem_cgroup_per_zone *
248 249 250 251 252
page_cgroup_zoneinfo(struct page_cgroup *pc)
{
	struct mem_cgroup *mem = pc->mem_cgroup;
	int nid = page_cgroup_nid(pc);
	int zid = page_cgroup_zid(pc);
253

254 255 256
	if (!mem)
		return NULL;

257 258 259
	return mem_cgroup_zoneinfo(mem, nid, zid);
}

K
KAMEZAWA Hiroyuki 已提交
260
static unsigned long mem_cgroup_get_local_zonestat(struct mem_cgroup *mem,
261
					enum lru_list idx)
262 263 264 265 266 267 268 269 270 271 272
{
	int nid, zid;
	struct mem_cgroup_per_zone *mz;
	u64 total = 0;

	for_each_online_node(nid)
		for (zid = 0; zid < MAX_NR_ZONES; zid++) {
			mz = mem_cgroup_zoneinfo(mem, nid, zid);
			total += MEM_CGROUP_ZSTAT(mz, idx);
		}
	return total;
273 274
}

275
static struct mem_cgroup *mem_cgroup_from_cont(struct cgroup *cont)
B
Balbir Singh 已提交
276 277 278 279 280 281
{
	return container_of(cgroup_subsys_state(cont,
				mem_cgroup_subsys_id), struct mem_cgroup,
				css);
}

282
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p)
283
{
284 285 286 287 288 289 290 291
	/*
	 * mm_update_next_owner() may clear mm->owner to NULL
	 * if it races with swapoff, page migration, etc.
	 * So this can be called with p == NULL.
	 */
	if (unlikely(!p))
		return NULL;

292 293 294 295
	return container_of(task_subsys_state(p, mem_cgroup_subsys_id),
				struct mem_cgroup, css);
}

296 297 298
static struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm)
{
	struct mem_cgroup *mem = NULL;
299 300 301

	if (!mm)
		return NULL;
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
	/*
	 * Because we have no locks, mm->owner's may be being moved to other
	 * cgroup. We use css_tryget() here even if this looks
	 * pessimistic (rather than adding locks here).
	 */
	rcu_read_lock();
	do {
		mem = mem_cgroup_from_task(rcu_dereference(mm->owner));
		if (unlikely(!mem))
			break;
	} while (!css_tryget(&mem->css));
	rcu_read_unlock();
	return mem;
}

static bool mem_cgroup_is_obsolete(struct mem_cgroup *mem)
{
	if (!mem)
		return true;
	return css_is_removed(&mem->css);
}

K
KAMEZAWA Hiroyuki 已提交
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359

/*
 * Call callback function against all cgroup under hierarchy tree.
 */
static int mem_cgroup_walk_tree(struct mem_cgroup *root, void *data,
			  int (*func)(struct mem_cgroup *, void *))
{
	int found, ret, nextid;
	struct cgroup_subsys_state *css;
	struct mem_cgroup *mem;

	if (!root->use_hierarchy)
		return (*func)(root, data);

	nextid = 1;
	do {
		ret = 0;
		mem = NULL;

		rcu_read_lock();
		css = css_get_next(&mem_cgroup_subsys, nextid, &root->css,
				   &found);
		if (css && css_tryget(css))
			mem = container_of(css, struct mem_cgroup, css);
		rcu_read_unlock();

		if (mem) {
			ret = (*func)(mem, data);
			css_put(&mem->css);
		}
		nextid = found + 1;
	} while (!ret && css);

	return ret;
}

K
KAMEZAWA Hiroyuki 已提交
360 361 362 363 364 365 366 367 368 369 370 371 372
/*
 * Following LRU functions are allowed to be used without PCG_LOCK.
 * Operations are called by routine of global LRU independently from memcg.
 * What we have to take care of here is validness of pc->mem_cgroup.
 *
 * Changes to pc->mem_cgroup happens when
 * 1. charge
 * 2. moving account
 * In typical case, "charge" is done before add-to-lru. Exception is SwapCache.
 * It is added to LRU before charge.
 * If PCG_USED bit is not set, page_cgroup is not added to this private LRU.
 * When moving account, the page is not on LRU. It's isolated.
 */
373

K
KAMEZAWA Hiroyuki 已提交
374 375 376 377 378
void mem_cgroup_del_lru_list(struct page *page, enum lru_list lru)
{
	struct page_cgroup *pc;
	struct mem_cgroup *mem;
	struct mem_cgroup_per_zone *mz;
379

380
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
381 382 383
		return;
	pc = lookup_page_cgroup(page);
	/* can happen while we handle swapcache. */
384
	if (list_empty(&pc->lru) || !pc->mem_cgroup)
K
KAMEZAWA Hiroyuki 已提交
385
		return;
386 387 388 389
	/*
	 * We don't check PCG_USED bit. It's cleared when the "page" is finally
	 * removed from global LRU.
	 */
K
KAMEZAWA Hiroyuki 已提交
390 391
	mz = page_cgroup_zoneinfo(pc);
	mem = pc->mem_cgroup;
392
	MEM_CGROUP_ZSTAT(mz, lru) -= 1;
K
KAMEZAWA Hiroyuki 已提交
393 394
	list_del_init(&pc->lru);
	return;
395 396
}

K
KAMEZAWA Hiroyuki 已提交
397
void mem_cgroup_del_lru(struct page *page)
398
{
K
KAMEZAWA Hiroyuki 已提交
399 400
	mem_cgroup_del_lru_list(page, page_lru(page));
}
401

K
KAMEZAWA Hiroyuki 已提交
402 403 404 405
void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru)
{
	struct mem_cgroup_per_zone *mz;
	struct page_cgroup *pc;
406

407
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
408
		return;
409

K
KAMEZAWA Hiroyuki 已提交
410
	pc = lookup_page_cgroup(page);
411 412 413 414
	/*
	 * Used bit is set without atomic ops but after smp_wmb().
	 * For making pc->mem_cgroup visible, insert smp_rmb() here.
	 */
K
KAMEZAWA Hiroyuki 已提交
415 416 417 418 419 420
	smp_rmb();
	/* unused page is not rotated. */
	if (!PageCgroupUsed(pc))
		return;
	mz = page_cgroup_zoneinfo(pc);
	list_move(&pc->lru, &mz->lists[lru]);
421 422
}

K
KAMEZAWA Hiroyuki 已提交
423
void mem_cgroup_add_lru_list(struct page *page, enum lru_list lru)
424
{
K
KAMEZAWA Hiroyuki 已提交
425 426
	struct page_cgroup *pc;
	struct mem_cgroup_per_zone *mz;
427

428
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
429 430
		return;
	pc = lookup_page_cgroup(page);
431 432 433 434
	/*
	 * Used bit is set without atomic ops but after smp_wmb().
	 * For making pc->mem_cgroup visible, insert smp_rmb() here.
	 */
K
KAMEZAWA Hiroyuki 已提交
435 436
	smp_rmb();
	if (!PageCgroupUsed(pc))
L
Lee Schermerhorn 已提交
437
		return;
438

K
KAMEZAWA Hiroyuki 已提交
439
	mz = page_cgroup_zoneinfo(pc);
440
	MEM_CGROUP_ZSTAT(mz, lru) += 1;
K
KAMEZAWA Hiroyuki 已提交
441 442
	list_add(&pc->lru, &mz->lists[lru]);
}
443

K
KAMEZAWA Hiroyuki 已提交
444
/*
445 446 447 448 449
 * At handling SwapCache, pc->mem_cgroup may be changed while it's linked to
 * lru because the page may.be reused after it's fully uncharged (because of
 * SwapCache behavior).To handle that, unlink page_cgroup from LRU when charge
 * it again. This function is only used to charge SwapCache. It's done under
 * lock_page and expected that zone->lru_lock is never held.
K
KAMEZAWA Hiroyuki 已提交
450
 */
451
static void mem_cgroup_lru_del_before_commit_swapcache(struct page *page)
K
KAMEZAWA Hiroyuki 已提交
452
{
453 454 455 456 457 458 459 460 461 462 463 464
	unsigned long flags;
	struct zone *zone = page_zone(page);
	struct page_cgroup *pc = lookup_page_cgroup(page);

	spin_lock_irqsave(&zone->lru_lock, flags);
	/*
	 * Forget old LRU when this page_cgroup is *not* used. This Used bit
	 * is guarded by lock_page() because the page is SwapCache.
	 */
	if (!PageCgroupUsed(pc))
		mem_cgroup_del_lru_list(page, page_lru(page));
	spin_unlock_irqrestore(&zone->lru_lock, flags);
K
KAMEZAWA Hiroyuki 已提交
465 466
}

467 468 469 470 471 472 473 474 475 476 477 478 479 480
static void mem_cgroup_lru_add_after_commit_swapcache(struct page *page)
{
	unsigned long flags;
	struct zone *zone = page_zone(page);
	struct page_cgroup *pc = lookup_page_cgroup(page);

	spin_lock_irqsave(&zone->lru_lock, flags);
	/* link when the page is linked to LRU but page_cgroup isn't */
	if (PageLRU(page) && list_empty(&pc->lru))
		mem_cgroup_add_lru_list(page, page_lru(page));
	spin_unlock_irqrestore(&zone->lru_lock, flags);
}


K
KAMEZAWA Hiroyuki 已提交
481 482 483
void mem_cgroup_move_lists(struct page *page,
			   enum lru_list from, enum lru_list to)
{
484
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
485 486 487
		return;
	mem_cgroup_del_lru_list(page, from);
	mem_cgroup_add_lru_list(page, to);
488 489
}

490 491 492
int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem)
{
	int ret;
493
	struct mem_cgroup *curr = NULL;
494 495

	task_lock(task);
496 497 498
	rcu_read_lock();
	curr = try_get_mem_cgroup_from_mm(task->mm);
	rcu_read_unlock();
499
	task_unlock(task);
500 501 502 503 504 505 506
	if (!curr)
		return 0;
	if (curr->use_hierarchy)
		ret = css_is_ancestor(&curr->css, &mem->css);
	else
		ret = (curr == mem);
	css_put(&curr->css);
507 508 509
	return ret;
}

510 511 512 513 514
/*
 * prev_priority control...this will be used in memory reclaim path.
 */
int mem_cgroup_get_reclaim_priority(struct mem_cgroup *mem)
{
K
KOSAKI Motohiro 已提交
515 516 517 518 519 520 521
	int prev_priority;

	spin_lock(&mem->reclaim_param_lock);
	prev_priority = mem->prev_priority;
	spin_unlock(&mem->reclaim_param_lock);

	return prev_priority;
522 523 524 525
}

void mem_cgroup_note_reclaim_priority(struct mem_cgroup *mem, int priority)
{
K
KOSAKI Motohiro 已提交
526
	spin_lock(&mem->reclaim_param_lock);
527 528
	if (priority < mem->prev_priority)
		mem->prev_priority = priority;
K
KOSAKI Motohiro 已提交
529
	spin_unlock(&mem->reclaim_param_lock);
530 531 532 533
}

void mem_cgroup_record_reclaim_priority(struct mem_cgroup *mem, int priority)
{
K
KOSAKI Motohiro 已提交
534
	spin_lock(&mem->reclaim_param_lock);
535
	mem->prev_priority = priority;
K
KOSAKI Motohiro 已提交
536
	spin_unlock(&mem->reclaim_param_lock);
537 538
}

539
static int calc_inactive_ratio(struct mem_cgroup *memcg, unsigned long *present_pages)
540 541 542
{
	unsigned long active;
	unsigned long inactive;
543 544
	unsigned long gb;
	unsigned long inactive_ratio;
545

K
KAMEZAWA Hiroyuki 已提交
546 547
	inactive = mem_cgroup_get_local_zonestat(memcg, LRU_INACTIVE_ANON);
	active = mem_cgroup_get_local_zonestat(memcg, LRU_ACTIVE_ANON);
548

549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575
	gb = (inactive + active) >> (30 - PAGE_SHIFT);
	if (gb)
		inactive_ratio = int_sqrt(10 * gb);
	else
		inactive_ratio = 1;

	if (present_pages) {
		present_pages[0] = inactive;
		present_pages[1] = active;
	}

	return inactive_ratio;
}

int mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg)
{
	unsigned long active;
	unsigned long inactive;
	unsigned long present_pages[2];
	unsigned long inactive_ratio;

	inactive_ratio = calc_inactive_ratio(memcg, present_pages);

	inactive = present_pages[0];
	active = present_pages[1];

	if (inactive * inactive_ratio < active)
576 577 578 579 580
		return 1;

	return 0;
}

581 582 583 584 585 586 587 588 589 590 591
unsigned long mem_cgroup_zone_nr_pages(struct mem_cgroup *memcg,
				       struct zone *zone,
				       enum lru_list lru)
{
	int nid = zone->zone_pgdat->node_id;
	int zid = zone_idx(zone);
	struct mem_cgroup_per_zone *mz = mem_cgroup_zoneinfo(memcg, nid, zid);

	return MEM_CGROUP_ZSTAT(mz, lru);
}

K
KOSAKI Motohiro 已提交
592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611
struct zone_reclaim_stat *mem_cgroup_get_reclaim_stat(struct mem_cgroup *memcg,
						      struct zone *zone)
{
	int nid = zone->zone_pgdat->node_id;
	int zid = zone_idx(zone);
	struct mem_cgroup_per_zone *mz = mem_cgroup_zoneinfo(memcg, nid, zid);

	return &mz->reclaim_stat;
}

struct zone_reclaim_stat *
mem_cgroup_get_reclaim_stat_from_page(struct page *page)
{
	struct page_cgroup *pc;
	struct mem_cgroup_per_zone *mz;

	if (mem_cgroup_disabled())
		return NULL;

	pc = lookup_page_cgroup(page);
612 613 614 615 616 617 618 619
	/*
	 * Used bit is set without atomic ops but after smp_wmb().
	 * For making pc->mem_cgroup visible, insert smp_rmb() here.
	 */
	smp_rmb();
	if (!PageCgroupUsed(pc))
		return NULL;

K
KOSAKI Motohiro 已提交
620 621 622 623 624 625 626
	mz = page_cgroup_zoneinfo(pc);
	if (!mz)
		return NULL;

	return &mz->reclaim_stat;
}

627 628 629 630 631
unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan,
					struct list_head *dst,
					unsigned long *scanned, int order,
					int mode, struct zone *z,
					struct mem_cgroup *mem_cont,
632
					int active, int file)
633 634 635 636 637 638
{
	unsigned long nr_taken = 0;
	struct page *page;
	unsigned long scan;
	LIST_HEAD(pc_list);
	struct list_head *src;
639
	struct page_cgroup *pc, *tmp;
640 641 642
	int nid = z->zone_pgdat->node_id;
	int zid = zone_idx(z);
	struct mem_cgroup_per_zone *mz;
643
	int lru = LRU_FILE * !!file + !!active;
644

645
	BUG_ON(!mem_cont);
646
	mz = mem_cgroup_zoneinfo(mem_cont, nid, zid);
647
	src = &mz->lists[lru];
648

649 650
	scan = 0;
	list_for_each_entry_safe_reverse(pc, tmp, src, lru) {
H
Hugh Dickins 已提交
651
		if (scan >= nr_to_scan)
652
			break;
K
KAMEZAWA Hiroyuki 已提交
653 654

		page = pc->page;
655 656
		if (unlikely(!PageCgroupUsed(pc)))
			continue;
H
Hugh Dickins 已提交
657
		if (unlikely(!PageLRU(page)))
658 659
			continue;

H
Hugh Dickins 已提交
660
		scan++;
661
		if (__isolate_lru_page(page, mode, file) == 0) {
662 663 664 665 666 667 668 669 670
			list_move(&page->lru, dst);
			nr_taken++;
		}
	}

	*scanned = scan;
	return nr_taken;
}

671 672 673
#define mem_cgroup_from_res_counter(counter, member)	\
	container_of(counter, struct mem_cgroup, member)

674 675 676 677 678 679 680 681 682 683 684 685
static bool mem_cgroup_check_under_limit(struct mem_cgroup *mem)
{
	if (do_swap_account) {
		if (res_counter_check_under_limit(&mem->res) &&
			res_counter_check_under_limit(&mem->memsw))
			return true;
	} else
		if (res_counter_check_under_limit(&mem->res))
			return true;
	return false;
}

K
KOSAKI Motohiro 已提交
686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701
static unsigned int get_swappiness(struct mem_cgroup *memcg)
{
	struct cgroup *cgrp = memcg->css.cgroup;
	unsigned int swappiness;

	/* root ? */
	if (cgrp->parent == NULL)
		return vm_swappiness;

	spin_lock(&memcg->reclaim_param_lock);
	swappiness = memcg->swappiness;
	spin_unlock(&memcg->reclaim_param_lock);

	return swappiness;
}

702 703 704 705 706 707
static int mem_cgroup_count_children_cb(struct mem_cgroup *mem, void *data)
{
	int *val = data;
	(*val)++;
	return 0;
}
708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775

/**
 * mem_cgroup_print_mem_info: Called from OOM with tasklist_lock held in read mode.
 * @memcg: The memory cgroup that went over limit
 * @p: Task that is going to be killed
 *
 * NOTE: @memcg and @p's mem_cgroup can be different when hierarchy is
 * enabled
 */
void mem_cgroup_print_oom_info(struct mem_cgroup *memcg, struct task_struct *p)
{
	struct cgroup *task_cgrp;
	struct cgroup *mem_cgrp;
	/*
	 * Need a buffer in BSS, can't rely on allocations. The code relies
	 * on the assumption that OOM is serialized for memory controller.
	 * If this assumption is broken, revisit this code.
	 */
	static char memcg_name[PATH_MAX];
	int ret;

	if (!memcg)
		return;


	rcu_read_lock();

	mem_cgrp = memcg->css.cgroup;
	task_cgrp = task_cgroup(p, mem_cgroup_subsys_id);

	ret = cgroup_path(task_cgrp, memcg_name, PATH_MAX);
	if (ret < 0) {
		/*
		 * Unfortunately, we are unable to convert to a useful name
		 * But we'll still print out the usage information
		 */
		rcu_read_unlock();
		goto done;
	}
	rcu_read_unlock();

	printk(KERN_INFO "Task in %s killed", memcg_name);

	rcu_read_lock();
	ret = cgroup_path(mem_cgrp, memcg_name, PATH_MAX);
	if (ret < 0) {
		rcu_read_unlock();
		goto done;
	}
	rcu_read_unlock();

	/*
	 * Continues from above, so we don't need an KERN_ level
	 */
	printk(KERN_CONT " as a result of limit of %s\n", memcg_name);
done:

	printk(KERN_INFO "memory: usage %llukB, limit %llukB, failcnt %llu\n",
		res_counter_read_u64(&memcg->res, RES_USAGE) >> 10,
		res_counter_read_u64(&memcg->res, RES_LIMIT) >> 10,
		res_counter_read_u64(&memcg->res, RES_FAILCNT));
	printk(KERN_INFO "memory+swap: usage %llukB, limit %llukB, "
		"failcnt %llu\n",
		res_counter_read_u64(&memcg->memsw, RES_USAGE) >> 10,
		res_counter_read_u64(&memcg->memsw, RES_LIMIT) >> 10,
		res_counter_read_u64(&memcg->memsw, RES_FAILCNT));
}

776 777 778 779 780 781 782 783 784 785 786
/*
 * This function returns the number of memcg under hierarchy tree. Returns
 * 1(self count) if no children.
 */
static int mem_cgroup_count_children(struct mem_cgroup *mem)
{
	int num = 0;
 	mem_cgroup_walk_tree(mem, &num, mem_cgroup_count_children_cb);
	return num;
}

787
/*
K
KAMEZAWA Hiroyuki 已提交
788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829
 * Visit the first child (need not be the first child as per the ordering
 * of the cgroup list, since we track last_scanned_child) of @mem and use
 * that to reclaim free pages from.
 */
static struct mem_cgroup *
mem_cgroup_select_victim(struct mem_cgroup *root_mem)
{
	struct mem_cgroup *ret = NULL;
	struct cgroup_subsys_state *css;
	int nextid, found;

	if (!root_mem->use_hierarchy) {
		css_get(&root_mem->css);
		ret = root_mem;
	}

	while (!ret) {
		rcu_read_lock();
		nextid = root_mem->last_scanned_child + 1;
		css = css_get_next(&mem_cgroup_subsys, nextid, &root_mem->css,
				   &found);
		if (css && css_tryget(css))
			ret = container_of(css, struct mem_cgroup, css);

		rcu_read_unlock();
		/* Updates scanning parameter */
		spin_lock(&root_mem->reclaim_param_lock);
		if (!css) {
			/* this means start scan from ID:1 */
			root_mem->last_scanned_child = 0;
		} else
			root_mem->last_scanned_child = found;
		spin_unlock(&root_mem->reclaim_param_lock);
	}

	return ret;
}

/*
 * Scan the hierarchy if needed to reclaim memory. We remember the last child
 * we reclaimed from, so that we don't end up penalizing one child extensively
 * based on its position in the children list.
830 831
 *
 * root_mem is the original ancestor that we've been reclaim from.
K
KAMEZAWA Hiroyuki 已提交
832 833 834
 *
 * We give up and return to the caller when we visit root_mem twice.
 * (other groups can be removed while we're walking....)
835 836
 *
 * If shrink==true, for avoiding to free too much, this returns immedieately.
837 838
 */
static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
839
				   gfp_t gfp_mask, bool noswap, bool shrink)
840
{
K
KAMEZAWA Hiroyuki 已提交
841 842 843 844 845 846 847 848 849 850 851
	struct mem_cgroup *victim;
	int ret, total = 0;
	int loop = 0;

	while (loop < 2) {
		victim = mem_cgroup_select_victim(root_mem);
		if (victim == root_mem)
			loop++;
		if (!mem_cgroup_local_usage(&victim->stat)) {
			/* this cgroup's local usage == 0 */
			css_put(&victim->css);
852 853
			continue;
		}
K
KAMEZAWA Hiroyuki 已提交
854 855 856 857
		/* we use swappiness of local cgroup */
		ret = try_to_free_mem_cgroup_pages(victim, gfp_mask, noswap,
						   get_swappiness(victim));
		css_put(&victim->css);
858 859 860 861 862 863 864
		/*
		 * At shrinking usage, we can't check we should stop here or
		 * reclaim more. It's depends on callers. last_scanned_child
		 * will work enough for keeping fairness under tree.
		 */
		if (shrink)
			return ret;
K
KAMEZAWA Hiroyuki 已提交
865
		total += ret;
866
		if (mem_cgroup_check_under_limit(root_mem))
K
KAMEZAWA Hiroyuki 已提交
867
			return 1 + total;
868
	}
K
KAMEZAWA Hiroyuki 已提交
869
	return total;
870 871
}

872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887
bool mem_cgroup_oom_called(struct task_struct *task)
{
	bool ret = false;
	struct mem_cgroup *mem;
	struct mm_struct *mm;

	rcu_read_lock();
	mm = task->mm;
	if (!mm)
		mm = &init_mm;
	mem = mem_cgroup_from_task(rcu_dereference(mm->owner));
	if (mem && time_before(jiffies, mem->last_oom_jiffies + HZ/10))
		ret = true;
	rcu_read_unlock();
	return ret;
}
888 889 890 891 892 893 894 895 896 897 898 899 900

static int record_last_oom_cb(struct mem_cgroup *mem, void *data)
{
	mem->last_oom_jiffies = jiffies;
	return 0;
}

static void record_last_oom(struct mem_cgroup *mem)
{
	mem_cgroup_walk_tree(mem, NULL, record_last_oom_cb);
}


901 902 903
/*
 * Unlike exported interface, "oom" parameter is added. if oom==true,
 * oom-killer can be invoked.
904
 */
905
static int __mem_cgroup_try_charge(struct mm_struct *mm,
906 907
			gfp_t gfp_mask, struct mem_cgroup **memcg,
			bool oom)
908
{
909
	struct mem_cgroup *mem, *mem_over_limit;
910
	int nr_retries = MEM_CGROUP_RECLAIM_RETRIES;
911
	struct res_counter *fail_res;
912 913 914 915 916 917 918

	if (unlikely(test_thread_flag(TIF_MEMDIE))) {
		/* Don't account this! */
		*memcg = NULL;
		return 0;
	}

919
	/*
920 921
	 * We always charge the cgroup the mm_struct belongs to.
	 * The mm_struct's mem_cgroup changes on task migration if the
922 923 924
	 * thread group leader migrates. It's possible that mm is not
	 * set, if so charge the init_mm (happens for pagecache usage).
	 */
925 926 927
	mem = *memcg;
	if (likely(!mem)) {
		mem = try_get_mem_cgroup_from_mm(mm);
928
		*memcg = mem;
929
	} else {
930
		css_get(&mem->css);
931
	}
932 933 934
	if (unlikely(!mem))
		return 0;

935
	VM_BUG_ON(!mem || mem_cgroup_is_obsolete(mem));
936

937 938 939
	while (1) {
		int ret;
		bool noswap = false;
940

941
		ret = res_counter_charge(&mem->res, PAGE_SIZE, &fail_res);
942 943 944
		if (likely(!ret)) {
			if (!do_swap_account)
				break;
945 946
			ret = res_counter_charge(&mem->memsw, PAGE_SIZE,
							&fail_res);
947 948 949 950 951
			if (likely(!ret))
				break;
			/* mem+swap counter fails */
			res_counter_uncharge(&mem->res, PAGE_SIZE);
			noswap = true;
952 953 954 955 956 957 958
			mem_over_limit = mem_cgroup_from_res_counter(fail_res,
									memsw);
		} else
			/* mem counter fails */
			mem_over_limit = mem_cgroup_from_res_counter(fail_res,
									res);

959
		if (!(gfp_mask & __GFP_WAIT))
960
			goto nomem;
961

962
		ret = mem_cgroup_hierarchical_reclaim(mem_over_limit, gfp_mask,
963
							noswap, false);
964 965
		if (ret)
			continue;
966 967

		/*
968 969 970 971 972
		 * try_to_free_mem_cgroup_pages() might not give us a full
		 * picture of reclaim. Some pages are reclaimed and might be
		 * moved to swap cache or just unmapped from the cgroup.
		 * Check the limit again to see if the reclaim reduced the
		 * current usage of the cgroup before giving up
973
		 *
974
		 */
975 976
		if (mem_cgroup_check_under_limit(mem_over_limit))
			continue;
977 978

		if (!nr_retries--) {
979
			if (oom) {
980
				mutex_lock(&memcg_tasklist);
981
				mem_cgroup_out_of_memory(mem_over_limit, gfp_mask);
982
				mutex_unlock(&memcg_tasklist);
983
				record_last_oom(mem_over_limit);
984
			}
985
			goto nomem;
986
		}
987
	}
988 989 990 991 992
	return 0;
nomem:
	css_put(&mem->css);
	return -ENOMEM;
}
993

994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013

/*
 * A helper function to get mem_cgroup from ID. must be called under
 * rcu_read_lock(). The caller must check css_is_removed() or some if
 * it's concern. (dropping refcnt from swap can be called against removed
 * memcg.)
 */
static struct mem_cgroup *mem_cgroup_lookup(unsigned short id)
{
	struct cgroup_subsys_state *css;

	/* ID 0 is unused ID */
	if (!id)
		return NULL;
	css = css_lookup(&mem_cgroup_subsys, id);
	if (!css)
		return NULL;
	return container_of(css, struct mem_cgroup, css);
}

1014 1015 1016
static struct mem_cgroup *try_get_mem_cgroup_from_swapcache(struct page *page)
{
	struct mem_cgroup *mem;
1017
	struct page_cgroup *pc;
1018
	unsigned short id;
1019 1020
	swp_entry_t ent;

1021 1022
	VM_BUG_ON(!PageLocked(page));

1023 1024 1025
	if (!PageSwapCache(page))
		return NULL;

1026
	pc = lookup_page_cgroup(page);
1027
	lock_page_cgroup(pc);
1028
	if (PageCgroupUsed(pc)) {
1029
		mem = pc->mem_cgroup;
1030 1031 1032
		if (mem && !css_tryget(&mem->css))
			mem = NULL;
	} else {
1033
		ent.val = page_private(page);
1034 1035 1036 1037 1038 1039
		id = lookup_swap_cgroup(ent);
		rcu_read_lock();
		mem = mem_cgroup_lookup(id);
		if (mem && !css_tryget(&mem->css))
			mem = NULL;
		rcu_read_unlock();
1040
	}
1041
	unlock_page_cgroup(pc);
1042 1043 1044
	return mem;
}

1045
/*
1046
 * commit a charge got by __mem_cgroup_try_charge() and makes page_cgroup to be
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056
 * USED state. If already USED, uncharge and return.
 */

static void __mem_cgroup_commit_charge(struct mem_cgroup *mem,
				     struct page_cgroup *pc,
				     enum charge_type ctype)
{
	/* try_charge() can return NULL to *memcg, taking care of it. */
	if (!mem)
		return;
1057 1058 1059 1060 1061

	lock_page_cgroup(pc);
	if (unlikely(PageCgroupUsed(pc))) {
		unlock_page_cgroup(pc);
		res_counter_uncharge(&mem->res, PAGE_SIZE);
1062 1063
		if (do_swap_account)
			res_counter_uncharge(&mem->memsw, PAGE_SIZE);
1064
		css_put(&mem->css);
1065
		return;
1066
	}
1067
	pc->mem_cgroup = mem;
K
KAMEZAWA Hiroyuki 已提交
1068
	smp_wmb();
1069
	pc->flags = pcg_default_flags[ctype];
1070

K
KAMEZAWA Hiroyuki 已提交
1071
	mem_cgroup_charge_statistics(mem, pc, true);
1072 1073

	unlock_page_cgroup(pc);
1074
}
1075

1076 1077 1078 1079 1080 1081 1082
/**
 * mem_cgroup_move_account - move account of the page
 * @pc:	page_cgroup of the page.
 * @from: mem_cgroup which the page is moved from.
 * @to:	mem_cgroup which the page is moved to. @from != @to.
 *
 * The caller must confirm following.
K
KAMEZAWA Hiroyuki 已提交
1083
 * - page is not on LRU (isolate_page() is useful.)
1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099
 *
 * returns 0 at success,
 * returns -EBUSY when lock is busy or "pc" is unstable.
 *
 * This function does "uncharge" from old cgroup but doesn't do "charge" to
 * new cgroup. It should be done by a caller.
 */

static int mem_cgroup_move_account(struct page_cgroup *pc,
	struct mem_cgroup *from, struct mem_cgroup *to)
{
	struct mem_cgroup_per_zone *from_mz, *to_mz;
	int nid, zid;
	int ret = -EBUSY;

	VM_BUG_ON(from == to);
K
KAMEZAWA Hiroyuki 已提交
1100
	VM_BUG_ON(PageLRU(pc->page));
1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115

	nid = page_cgroup_nid(pc);
	zid = page_cgroup_zid(pc);
	from_mz =  mem_cgroup_zoneinfo(from, nid, zid);
	to_mz =  mem_cgroup_zoneinfo(to, nid, zid);

	if (!trylock_page_cgroup(pc))
		return ret;

	if (!PageCgroupUsed(pc))
		goto out;

	if (pc->mem_cgroup != from)
		goto out;

K
KAMEZAWA Hiroyuki 已提交
1116 1117 1118 1119
	res_counter_uncharge(&from->res, PAGE_SIZE);
	mem_cgroup_charge_statistics(from, pc, false);
	if (do_swap_account)
		res_counter_uncharge(&from->memsw, PAGE_SIZE);
1120 1121 1122
	css_put(&from->css);

	css_get(&to->css);
K
KAMEZAWA Hiroyuki 已提交
1123 1124 1125
	pc->mem_cgroup = to;
	mem_cgroup_charge_statistics(to, pc, true);
	ret = 0;
1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138
out:
	unlock_page_cgroup(pc);
	return ret;
}

/*
 * move charges to its parent.
 */

static int mem_cgroup_move_parent(struct page_cgroup *pc,
				  struct mem_cgroup *child,
				  gfp_t gfp_mask)
{
K
KAMEZAWA Hiroyuki 已提交
1139
	struct page *page = pc->page;
1140 1141 1142 1143 1144 1145 1146 1147 1148
	struct cgroup *cg = child->css.cgroup;
	struct cgroup *pcg = cg->parent;
	struct mem_cgroup *parent;
	int ret;

	/* Is ROOT ? */
	if (!pcg)
		return -EINVAL;

K
KAMEZAWA Hiroyuki 已提交
1149

1150 1151
	parent = mem_cgroup_from_cont(pcg);

K
KAMEZAWA Hiroyuki 已提交
1152

1153
	ret = __mem_cgroup_try_charge(NULL, gfp_mask, &parent, false);
1154
	if (ret || !parent)
1155 1156
		return ret;

1157 1158 1159 1160
	if (!get_page_unless_zero(page)) {
		ret = -EBUSY;
		goto uncharge;
	}
K
KAMEZAWA Hiroyuki 已提交
1161 1162 1163 1164 1165

	ret = isolate_lru_page(page);

	if (ret)
		goto cancel;
1166 1167 1168

	ret = mem_cgroup_move_account(pc, child, parent);

K
KAMEZAWA Hiroyuki 已提交
1169 1170 1171
	putback_lru_page(page);
	if (!ret) {
		put_page(page);
1172 1173
		/* drop extra refcnt by try_charge() */
		css_put(&parent->css);
K
KAMEZAWA Hiroyuki 已提交
1174
		return 0;
1175
	}
1176

K
KAMEZAWA Hiroyuki 已提交
1177
cancel:
1178 1179 1180 1181 1182
	put_page(page);
uncharge:
	/* drop extra refcnt by try_charge() */
	css_put(&parent->css);
	/* uncharge if move fails */
K
KAMEZAWA Hiroyuki 已提交
1183 1184 1185
	res_counter_uncharge(&parent->res, PAGE_SIZE);
	if (do_swap_account)
		res_counter_uncharge(&parent->memsw, PAGE_SIZE);
1186 1187 1188
	return ret;
}

1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209
/*
 * Charge the memory controller for page usage.
 * Return
 * 0 if the charge was successful
 * < 0 if the cgroup is over its limit
 */
static int mem_cgroup_charge_common(struct page *page, struct mm_struct *mm,
				gfp_t gfp_mask, enum charge_type ctype,
				struct mem_cgroup *memcg)
{
	struct mem_cgroup *mem;
	struct page_cgroup *pc;
	int ret;

	pc = lookup_page_cgroup(page);
	/* can happen at boot */
	if (unlikely(!pc))
		return 0;
	prefetchw(pc);

	mem = memcg;
1210
	ret = __mem_cgroup_try_charge(mm, gfp_mask, &mem, true);
1211
	if (ret || !mem)
1212 1213 1214
		return ret;

	__mem_cgroup_commit_charge(mem, pc, ctype);
1215 1216 1217
	return 0;
}

1218 1219
int mem_cgroup_newpage_charge(struct page *page,
			      struct mm_struct *mm, gfp_t gfp_mask)
1220
{
1221
	if (mem_cgroup_disabled())
1222
		return 0;
1223 1224
	if (PageCompound(page))
		return 0;
1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235
	/*
	 * If already mapped, we don't have to account.
	 * If page cache, page->mapping has address_space.
	 * But page->mapping may have out-of-use anon_vma pointer,
	 * detecit it by PageAnon() check. newly-mapped-anon's page->mapping
	 * is NULL.
  	 */
	if (page_mapped(page) || (page->mapping && !PageAnon(page)))
		return 0;
	if (unlikely(!mm))
		mm = &init_mm;
1236
	return mem_cgroup_charge_common(page, mm, gfp_mask,
1237
				MEM_CGROUP_CHARGE_TYPE_MAPPED, NULL);
1238 1239
}

D
Daisuke Nishimura 已提交
1240 1241 1242 1243
static void
__mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr,
					enum charge_type ctype);

1244 1245
int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
				gfp_t gfp_mask)
1246
{
1247 1248 1249
	struct mem_cgroup *mem = NULL;
	int ret;

1250
	if (mem_cgroup_disabled())
1251
		return 0;
1252 1253
	if (PageCompound(page))
		return 0;
1254 1255 1256 1257 1258 1259 1260 1261
	/*
	 * Corner case handling. This is called from add_to_page_cache()
	 * in usual. But some FS (shmem) precharges this page before calling it
	 * and call add_to_page_cache() with GFP_NOWAIT.
	 *
	 * For GFP_NOWAIT case, the page may be pre-charged before calling
	 * add_to_page_cache(). (See shmem.c) check it here and avoid to call
	 * charge twice. (It works but has to pay a bit larger cost.)
1262 1263
	 * And when the page is SwapCache, it should take swap information
	 * into account. This is under lock_page() now.
1264 1265 1266 1267
	 */
	if (!(gfp_mask & __GFP_WAIT)) {
		struct page_cgroup *pc;

1268 1269 1270 1271 1272 1273 1274

		pc = lookup_page_cgroup(page);
		if (!pc)
			return 0;
		lock_page_cgroup(pc);
		if (PageCgroupUsed(pc)) {
			unlock_page_cgroup(pc);
1275 1276
			return 0;
		}
1277
		unlock_page_cgroup(pc);
1278 1279
	}

1280
	if (unlikely(!mm && !mem))
1281
		mm = &init_mm;
1282

1283 1284
	if (page_is_file_cache(page))
		return mem_cgroup_charge_common(page, mm, gfp_mask,
1285
				MEM_CGROUP_CHARGE_TYPE_CACHE, NULL);
1286

D
Daisuke Nishimura 已提交
1287 1288 1289 1290 1291 1292 1293 1294 1295
	/* shmem */
	if (PageSwapCache(page)) {
		ret = mem_cgroup_try_charge_swapin(mm, page, gfp_mask, &mem);
		if (!ret)
			__mem_cgroup_commit_charge_swapin(page, mem,
					MEM_CGROUP_CHARGE_TYPE_SHMEM);
	} else
		ret = mem_cgroup_charge_common(page, mm, gfp_mask,
					MEM_CGROUP_CHARGE_TYPE_SHMEM, mem);
1296 1297

	return ret;
1298 1299
}

1300 1301 1302 1303 1304 1305
/*
 * While swap-in, try_charge -> commit or cancel, the page is locked.
 * And when try_charge() successfully returns, one refcnt to memcg without
 * struct page_cgroup is aquired. This refcnt will be cumsumed by
 * "commit()" or removed by "cancel()"
 */
1306 1307 1308 1309 1310
int mem_cgroup_try_charge_swapin(struct mm_struct *mm,
				 struct page *page,
				 gfp_t mask, struct mem_cgroup **ptr)
{
	struct mem_cgroup *mem;
1311
	int ret;
1312

1313
	if (mem_cgroup_disabled())
1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
		return 0;

	if (!do_swap_account)
		goto charge_cur_mm;
	/*
	 * A racing thread's fault, or swapoff, may have already updated
	 * the pte, and even removed page from swap cache: return success
	 * to go on to do_swap_page()'s pte_same() test, which should fail.
	 */
	if (!PageSwapCache(page))
		return 0;
1325
	mem = try_get_mem_cgroup_from_swapcache(page);
1326 1327
	if (!mem)
		goto charge_cur_mm;
1328
	*ptr = mem;
1329 1330 1331 1332
	ret = __mem_cgroup_try_charge(NULL, mask, ptr, true);
	/* drop extra refcnt from tryget */
	css_put(&mem->css);
	return ret;
1333 1334 1335 1336 1337 1338
charge_cur_mm:
	if (unlikely(!mm))
		mm = &init_mm;
	return __mem_cgroup_try_charge(mm, mask, ptr, true);
}

D
Daisuke Nishimura 已提交
1339 1340 1341
static void
__mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr,
					enum charge_type ctype)
1342 1343 1344
{
	struct page_cgroup *pc;

1345
	if (mem_cgroup_disabled())
1346 1347 1348 1349
		return;
	if (!ptr)
		return;
	pc = lookup_page_cgroup(page);
1350
	mem_cgroup_lru_del_before_commit_swapcache(page);
D
Daisuke Nishimura 已提交
1351
	__mem_cgroup_commit_charge(ptr, pc, ctype);
1352
	mem_cgroup_lru_add_after_commit_swapcache(page);
1353 1354 1355
	/*
	 * Now swap is on-memory. This means this page may be
	 * counted both as mem and swap....double count.
1356 1357 1358
	 * Fix it by uncharging from memsw. Basically, this SwapCache is stable
	 * under lock_page(). But in do_swap_page()::memory.c, reuse_swap_page()
	 * may call delete_from_swap_cache() before reach here.
1359
	 */
1360
	if (do_swap_account && PageSwapCache(page)) {
1361
		swp_entry_t ent = {.val = page_private(page)};
1362
		unsigned short id;
1363
		struct mem_cgroup *memcg;
1364 1365 1366 1367

		id = swap_cgroup_record(ent, 0);
		rcu_read_lock();
		memcg = mem_cgroup_lookup(id);
1368
		if (memcg) {
1369 1370 1371 1372
			/*
			 * This recorded memcg can be obsolete one. So, avoid
			 * calling css_tryget
			 */
1373 1374 1375
			res_counter_uncharge(&memcg->memsw, PAGE_SIZE);
			mem_cgroup_put(memcg);
		}
1376
		rcu_read_unlock();
1377
	}
K
KAMEZAWA Hiroyuki 已提交
1378
	/* add this page(page_cgroup) to the LRU we want. */
1379

1380 1381
}

D
Daisuke Nishimura 已提交
1382 1383 1384 1385 1386 1387
void mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr)
{
	__mem_cgroup_commit_charge_swapin(page, ptr,
					MEM_CGROUP_CHARGE_TYPE_MAPPED);
}

1388 1389
void mem_cgroup_cancel_charge_swapin(struct mem_cgroup *mem)
{
1390
	if (mem_cgroup_disabled())
1391 1392 1393 1394
		return;
	if (!mem)
		return;
	res_counter_uncharge(&mem->res, PAGE_SIZE);
1395 1396
	if (do_swap_account)
		res_counter_uncharge(&mem->memsw, PAGE_SIZE);
1397 1398 1399 1400
	css_put(&mem->css);
}


1401
/*
1402
 * uncharge if !page_mapped(page)
1403
 */
1404
static struct mem_cgroup *
1405
__mem_cgroup_uncharge_common(struct page *page, enum charge_type ctype)
1406
{
H
Hugh Dickins 已提交
1407
	struct page_cgroup *pc;
1408
	struct mem_cgroup *mem = NULL;
1409
	struct mem_cgroup_per_zone *mz;
1410

1411
	if (mem_cgroup_disabled())
1412
		return NULL;
1413

K
KAMEZAWA Hiroyuki 已提交
1414
	if (PageSwapCache(page))
1415
		return NULL;
K
KAMEZAWA Hiroyuki 已提交
1416

1417
	/*
1418
	 * Check if our page_cgroup is valid
1419
	 */
1420 1421
	pc = lookup_page_cgroup(page);
	if (unlikely(!pc || !PageCgroupUsed(pc)))
1422
		return NULL;
1423

1424
	lock_page_cgroup(pc);
K
KAMEZAWA Hiroyuki 已提交
1425

1426 1427
	mem = pc->mem_cgroup;

K
KAMEZAWA Hiroyuki 已提交
1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444
	if (!PageCgroupUsed(pc))
		goto unlock_out;

	switch (ctype) {
	case MEM_CGROUP_CHARGE_TYPE_MAPPED:
		if (page_mapped(page))
			goto unlock_out;
		break;
	case MEM_CGROUP_CHARGE_TYPE_SWAPOUT:
		if (!PageAnon(page)) {	/* Shared memory */
			if (page->mapping && !page_is_file_cache(page))
				goto unlock_out;
		} else if (page_mapped(page)) /* Anon */
				goto unlock_out;
		break;
	default:
		break;
1445
	}
K
KAMEZAWA Hiroyuki 已提交
1446

1447 1448 1449
	res_counter_uncharge(&mem->res, PAGE_SIZE);
	if (do_swap_account && (ctype != MEM_CGROUP_CHARGE_TYPE_SWAPOUT))
		res_counter_uncharge(&mem->memsw, PAGE_SIZE);
K
KAMEZAWA Hiroyuki 已提交
1450
	mem_cgroup_charge_statistics(mem, pc, false);
K
KAMEZAWA Hiroyuki 已提交
1451

1452
	ClearPageCgroupUsed(pc);
1453 1454 1455 1456 1457 1458
	/*
	 * pc->mem_cgroup is not cleared here. It will be accessed when it's
	 * freed from LRU. This is safe because uncharged page is expected not
	 * to be reused (freed soon). Exception is SwapCache, it's handled by
	 * special functions.
	 */
1459

1460
	mz = page_cgroup_zoneinfo(pc);
1461
	unlock_page_cgroup(pc);
H
Hugh Dickins 已提交
1462

K
KAMEZAWA Hiroyuki 已提交
1463 1464 1465
	/* at swapout, this memcg will be accessed to record to swap */
	if (ctype != MEM_CGROUP_CHARGE_TYPE_SWAPOUT)
		css_put(&mem->css);
1466

1467
	return mem;
K
KAMEZAWA Hiroyuki 已提交
1468 1469 1470

unlock_out:
	unlock_page_cgroup(pc);
1471
	return NULL;
1472 1473
}

1474 1475
void mem_cgroup_uncharge_page(struct page *page)
{
1476 1477 1478 1479 1480
	/* early check. */
	if (page_mapped(page))
		return;
	if (page->mapping && !PageAnon(page))
		return;
1481 1482 1483 1484 1485 1486
	__mem_cgroup_uncharge_common(page, MEM_CGROUP_CHARGE_TYPE_MAPPED);
}

void mem_cgroup_uncharge_cache_page(struct page *page)
{
	VM_BUG_ON(page_mapped(page));
1487
	VM_BUG_ON(page->mapping);
1488 1489 1490
	__mem_cgroup_uncharge_common(page, MEM_CGROUP_CHARGE_TYPE_CACHE);
}

1491
#ifdef CONFIG_SWAP
1492
/*
1493
 * called after __delete_from_swap_cache() and drop "page" account.
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
 * memcg information is recorded to swap_cgroup of "ent"
 */
void mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent)
{
	struct mem_cgroup *memcg;

	memcg = __mem_cgroup_uncharge_common(page,
					MEM_CGROUP_CHARGE_TYPE_SWAPOUT);
	/* record memcg information */
	if (do_swap_account && memcg) {
1504
		swap_cgroup_record(ent, css_id(&memcg->css));
1505 1506
		mem_cgroup_get(memcg);
	}
K
KAMEZAWA Hiroyuki 已提交
1507 1508
	if (memcg)
		css_put(&memcg->css);
1509
}
1510
#endif
1511 1512 1513 1514 1515 1516 1517

#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
/*
 * called from swap_entry_free(). remove record in swap_cgroup and
 * uncharge "memsw" account.
 */
void mem_cgroup_uncharge_swap(swp_entry_t ent)
K
KAMEZAWA Hiroyuki 已提交
1518
{
1519
	struct mem_cgroup *memcg;
1520
	unsigned short id;
1521 1522 1523 1524

	if (!do_swap_account)
		return;

1525 1526 1527
	id = swap_cgroup_record(ent, 0);
	rcu_read_lock();
	memcg = mem_cgroup_lookup(id);
1528
	if (memcg) {
1529 1530 1531 1532
		/*
		 * We uncharge this because swap is freed.
		 * This memcg can be obsolete one. We avoid calling css_tryget
		 */
1533 1534 1535
		res_counter_uncharge(&memcg->memsw, PAGE_SIZE);
		mem_cgroup_put(memcg);
	}
1536
	rcu_read_unlock();
K
KAMEZAWA Hiroyuki 已提交
1537
}
1538
#endif
K
KAMEZAWA Hiroyuki 已提交
1539

1540
/*
1541 1542
 * Before starting migration, account PAGE_SIZE to mem_cgroup that the old
 * page belongs to.
1543
 */
1544
int mem_cgroup_prepare_migration(struct page *page, struct mem_cgroup **ptr)
1545 1546
{
	struct page_cgroup *pc;
1547 1548
	struct mem_cgroup *mem = NULL;
	int ret = 0;
1549

1550
	if (mem_cgroup_disabled())
1551 1552
		return 0;

1553 1554 1555
	pc = lookup_page_cgroup(page);
	lock_page_cgroup(pc);
	if (PageCgroupUsed(pc)) {
1556 1557 1558
		mem = pc->mem_cgroup;
		css_get(&mem->css);
	}
1559
	unlock_page_cgroup(pc);
1560

1561
	if (mem) {
1562
		ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, &mem, false);
1563 1564
		css_put(&mem->css);
	}
1565
	*ptr = mem;
1566
	return ret;
1567
}
1568

1569
/* remove redundant charge if migration failed*/
1570 1571
void mem_cgroup_end_migration(struct mem_cgroup *mem,
		struct page *oldpage, struct page *newpage)
1572
{
1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596
	struct page *target, *unused;
	struct page_cgroup *pc;
	enum charge_type ctype;

	if (!mem)
		return;

	/* at migration success, oldpage->mapping is NULL. */
	if (oldpage->mapping) {
		target = oldpage;
		unused = NULL;
	} else {
		target = newpage;
		unused = oldpage;
	}

	if (PageAnon(target))
		ctype = MEM_CGROUP_CHARGE_TYPE_MAPPED;
	else if (page_is_file_cache(target))
		ctype = MEM_CGROUP_CHARGE_TYPE_CACHE;
	else
		ctype = MEM_CGROUP_CHARGE_TYPE_SHMEM;

	/* unused page is not on radix-tree now. */
K
KAMEZAWA Hiroyuki 已提交
1597
	if (unused)
1598 1599 1600
		__mem_cgroup_uncharge_common(unused, ctype);

	pc = lookup_page_cgroup(target);
1601
	/*
1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615
	 * __mem_cgroup_commit_charge() check PCG_USED bit of page_cgroup.
	 * So, double-counting is effectively avoided.
	 */
	__mem_cgroup_commit_charge(mem, pc, ctype);

	/*
	 * Both of oldpage and newpage are still under lock_page().
	 * Then, we don't have to care about race in radix-tree.
	 * But we have to be careful that this page is unmapped or not.
	 *
	 * There is a case for !page_mapped(). At the start of
	 * migration, oldpage was mapped. But now, it's zapped.
	 * But we know *target* page is not freed/reused under us.
	 * mem_cgroup_uncharge_page() does all necessary checks.
1616
	 */
1617 1618
	if (ctype == MEM_CGROUP_CHARGE_TYPE_MAPPED)
		mem_cgroup_uncharge_page(target);
1619
}
1620

1621
/*
1622 1623 1624 1625 1626 1627
 * A call to try to shrink memory usage on charge failure at shmem's swapin.
 * Calling hierarchical_reclaim is not enough because we should update
 * last_oom_jiffies to prevent pagefault_out_of_memory from invoking global OOM.
 * Moreover considering hierarchy, we should reclaim from the mem_over_limit,
 * not from the memcg which this page would be charged to.
 * try_charge_swapin does all of these works properly.
1628
 */
1629
int mem_cgroup_shmem_charge_fallback(struct page *page,
1630 1631
			    struct mm_struct *mm,
			    gfp_t gfp_mask)
1632
{
1633
	struct mem_cgroup *mem = NULL;
1634
	int ret;
1635

1636
	if (mem_cgroup_disabled())
1637
		return 0;
1638

1639 1640 1641
	ret = mem_cgroup_try_charge_swapin(mm, page, gfp_mask, &mem);
	if (!ret)
		mem_cgroup_cancel_charge_swapin(mem); /* it does !mem check */
1642

1643
	return ret;
1644 1645
}

1646 1647
static DEFINE_MUTEX(set_limit_mutex);

1648
static int mem_cgroup_resize_limit(struct mem_cgroup *memcg,
1649
				unsigned long long val)
1650
{
1651
	int retry_count;
1652
	int progress;
1653
	u64 memswlimit;
1654
	int ret = 0;
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665
	int children = mem_cgroup_count_children(memcg);
	u64 curusage, oldusage;

	/*
	 * For keeping hierarchical_reclaim simple, how long we should retry
	 * is depends on callers. We set our retry-count to be function
	 * of # of children which we should visit in this loop.
	 */
	retry_count = MEM_CGROUP_RECLAIM_RETRIES * children;

	oldusage = res_counter_read_u64(&memcg->res, RES_USAGE);
1666

1667
	while (retry_count) {
1668 1669 1670 1671
		if (signal_pending(current)) {
			ret = -EINTR;
			break;
		}
1672 1673 1674 1675 1676 1677 1678 1679 1680 1681
		/*
		 * Rather than hide all in some function, I do this in
		 * open coded manner. You see what this really does.
		 * We have to guarantee mem->res.limit < mem->memsw.limit.
		 */
		mutex_lock(&set_limit_mutex);
		memswlimit = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
		if (memswlimit < val) {
			ret = -EINVAL;
			mutex_unlock(&set_limit_mutex);
1682 1683
			break;
		}
1684 1685 1686 1687 1688 1689
		ret = res_counter_set_limit(&memcg->res, val);
		mutex_unlock(&set_limit_mutex);

		if (!ret)
			break;

1690
		progress = mem_cgroup_hierarchical_reclaim(memcg, GFP_KERNEL,
1691 1692 1693 1694 1695 1696 1697
						   false, true);
		curusage = res_counter_read_u64(&memcg->res, RES_USAGE);
		/* Usage is reduced ? */
  		if (curusage >= oldusage)
			retry_count--;
		else
			oldusage = curusage;
1698
	}
1699

1700 1701 1702 1703 1704 1705
	return ret;
}

int mem_cgroup_resize_memsw_limit(struct mem_cgroup *memcg,
				unsigned long long val)
{
1706
	int retry_count;
1707
	u64 memlimit, oldusage, curusage;
1708 1709
	int children = mem_cgroup_count_children(memcg);
	int ret = -EBUSY;
1710 1711 1712

	if (!do_swap_account)
		return -EINVAL;
1713 1714 1715
	/* see mem_cgroup_resize_res_limit */
 	retry_count = children * MEM_CGROUP_RECLAIM_RETRIES;
	oldusage = res_counter_read_u64(&memcg->memsw, RES_USAGE);
1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738
	while (retry_count) {
		if (signal_pending(current)) {
			ret = -EINTR;
			break;
		}
		/*
		 * Rather than hide all in some function, I do this in
		 * open coded manner. You see what this really does.
		 * We have to guarantee mem->res.limit < mem->memsw.limit.
		 */
		mutex_lock(&set_limit_mutex);
		memlimit = res_counter_read_u64(&memcg->res, RES_LIMIT);
		if (memlimit > val) {
			ret = -EINVAL;
			mutex_unlock(&set_limit_mutex);
			break;
		}
		ret = res_counter_set_limit(&memcg->memsw, val);
		mutex_unlock(&set_limit_mutex);

		if (!ret)
			break;

1739
		mem_cgroup_hierarchical_reclaim(memcg, GFP_KERNEL, true, true);
1740
		curusage = res_counter_read_u64(&memcg->memsw, RES_USAGE);
1741
		/* Usage is reduced ? */
1742
		if (curusage >= oldusage)
1743
			retry_count--;
1744 1745
		else
			oldusage = curusage;
1746 1747 1748 1749
	}
	return ret;
}

1750 1751 1752 1753
/*
 * This routine traverse page_cgroup in given list and drop them all.
 * *And* this routine doesn't reclaim page itself, just removes page_cgroup.
 */
1754
static int mem_cgroup_force_empty_list(struct mem_cgroup *mem,
K
KAMEZAWA Hiroyuki 已提交
1755
				int node, int zid, enum lru_list lru)
1756
{
K
KAMEZAWA Hiroyuki 已提交
1757 1758
	struct zone *zone;
	struct mem_cgroup_per_zone *mz;
1759
	struct page_cgroup *pc, *busy;
K
KAMEZAWA Hiroyuki 已提交
1760
	unsigned long flags, loop;
1761
	struct list_head *list;
1762
	int ret = 0;
1763

K
KAMEZAWA Hiroyuki 已提交
1764 1765
	zone = &NODE_DATA(node)->node_zones[zid];
	mz = mem_cgroup_zoneinfo(mem, node, zid);
1766
	list = &mz->lists[lru];
1767

1768 1769 1770 1771 1772 1773
	loop = MEM_CGROUP_ZSTAT(mz, lru);
	/* give some margin against EBUSY etc...*/
	loop += 256;
	busy = NULL;
	while (loop--) {
		ret = 0;
K
KAMEZAWA Hiroyuki 已提交
1774
		spin_lock_irqsave(&zone->lru_lock, flags);
1775
		if (list_empty(list)) {
K
KAMEZAWA Hiroyuki 已提交
1776
			spin_unlock_irqrestore(&zone->lru_lock, flags);
1777
			break;
1778 1779 1780 1781 1782
		}
		pc = list_entry(list->prev, struct page_cgroup, lru);
		if (busy == pc) {
			list_move(&pc->lru, list);
			busy = 0;
K
KAMEZAWA Hiroyuki 已提交
1783
			spin_unlock_irqrestore(&zone->lru_lock, flags);
1784 1785
			continue;
		}
K
KAMEZAWA Hiroyuki 已提交
1786
		spin_unlock_irqrestore(&zone->lru_lock, flags);
1787

K
KAMEZAWA Hiroyuki 已提交
1788
		ret = mem_cgroup_move_parent(pc, mem, GFP_KERNEL);
1789
		if (ret == -ENOMEM)
1790
			break;
1791 1792 1793 1794 1795 1796 1797

		if (ret == -EBUSY || ret == -EINVAL) {
			/* found lock contention or "pc" is obsolete. */
			busy = pc;
			cond_resched();
		} else
			busy = NULL;
1798
	}
K
KAMEZAWA Hiroyuki 已提交
1799

1800 1801 1802
	if (!ret && !list_empty(list))
		return -EBUSY;
	return ret;
1803 1804 1805 1806 1807 1808
}

/*
 * make mem_cgroup's charge to be 0 if there is no task.
 * This enables deleting this mem_cgroup.
 */
1809
static int mem_cgroup_force_empty(struct mem_cgroup *mem, bool free_all)
1810
{
1811 1812 1813
	int ret;
	int node, zid, shrink;
	int nr_retries = MEM_CGROUP_RECLAIM_RETRIES;
1814
	struct cgroup *cgrp = mem->css.cgroup;
1815

1816
	css_get(&mem->css);
1817 1818

	shrink = 0;
1819 1820 1821
	/* should free all ? */
	if (free_all)
		goto try_to_free;
1822
move_account:
1823
	while (mem->res.usage > 0) {
1824
		ret = -EBUSY;
1825 1826 1827 1828
		if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children))
			goto out;
		ret = -EINTR;
		if (signal_pending(current))
1829
			goto out;
1830 1831
		/* This is for making all *used* pages to be on LRU. */
		lru_add_drain_all();
1832
		ret = 0;
1833
		for_each_node_state(node, N_HIGH_MEMORY) {
1834
			for (zid = 0; !ret && zid < MAX_NR_ZONES; zid++) {
1835
				enum lru_list l;
1836 1837
				for_each_lru(l) {
					ret = mem_cgroup_force_empty_list(mem,
K
KAMEZAWA Hiroyuki 已提交
1838
							node, zid, l);
1839 1840 1841
					if (ret)
						break;
				}
1842
			}
1843 1844 1845 1846 1847 1848
			if (ret)
				break;
		}
		/* it seems parent cgroup doesn't have enough mem */
		if (ret == -ENOMEM)
			goto try_to_free;
1849
		cond_resched();
1850 1851 1852 1853 1854
	}
	ret = 0;
out:
	css_put(&mem->css);
	return ret;
1855 1856

try_to_free:
1857 1858
	/* returns EBUSY if there is a task or if we come here twice. */
	if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children) || shrink) {
1859 1860 1861
		ret = -EBUSY;
		goto out;
	}
1862 1863
	/* we call try-to-free pages for make this cgroup empty */
	lru_add_drain_all();
1864 1865 1866 1867
	/* try to free all pages in this cgroup */
	shrink = 1;
	while (nr_retries && mem->res.usage > 0) {
		int progress;
1868 1869 1870 1871 1872

		if (signal_pending(current)) {
			ret = -EINTR;
			goto out;
		}
K
KOSAKI Motohiro 已提交
1873 1874
		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
						false, get_swappiness(mem));
1875
		if (!progress) {
1876
			nr_retries--;
1877 1878 1879
			/* maybe some writeback is necessary */
			congestion_wait(WRITE, HZ/10);
		}
1880 1881

	}
K
KAMEZAWA Hiroyuki 已提交
1882
	lru_add_drain();
1883 1884 1885 1886 1887
	/* try move_account...there may be some *locked* pages. */
	if (mem->res.usage)
		goto move_account;
	ret = 0;
	goto out;
1888 1889
}

1890 1891 1892 1893 1894 1895
int mem_cgroup_force_empty_write(struct cgroup *cont, unsigned int event)
{
	return mem_cgroup_force_empty(mem_cgroup_from_cont(cont), true);
}


1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933
static u64 mem_cgroup_hierarchy_read(struct cgroup *cont, struct cftype *cft)
{
	return mem_cgroup_from_cont(cont)->use_hierarchy;
}

static int mem_cgroup_hierarchy_write(struct cgroup *cont, struct cftype *cft,
					u64 val)
{
	int retval = 0;
	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);
	struct cgroup *parent = cont->parent;
	struct mem_cgroup *parent_mem = NULL;

	if (parent)
		parent_mem = mem_cgroup_from_cont(parent);

	cgroup_lock();
	/*
	 * If parent's use_hiearchy is set, we can't make any modifications
	 * in the child subtrees. If it is unset, then the change can
	 * occur, provided the current cgroup has no children.
	 *
	 * For the root cgroup, parent_mem is NULL, we allow value to be
	 * set if there are no children.
	 */
	if ((!parent_mem || !parent_mem->use_hierarchy) &&
				(val == 1 || val == 0)) {
		if (list_empty(&cont->children))
			mem->use_hierarchy = val;
		else
			retval = -EBUSY;
	} else
		retval = -EINVAL;
	cgroup_unlock();

	return retval;
}

1934
static u64 mem_cgroup_read(struct cgroup *cont, struct cftype *cft)
B
Balbir Singh 已提交
1935
{
1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954
	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);
	u64 val = 0;
	int type, name;

	type = MEMFILE_TYPE(cft->private);
	name = MEMFILE_ATTR(cft->private);
	switch (type) {
	case _MEM:
		val = res_counter_read_u64(&mem->res, name);
		break;
	case _MEMSWAP:
		if (do_swap_account)
			val = res_counter_read_u64(&mem->memsw, name);
		break;
	default:
		BUG();
		break;
	}
	return val;
B
Balbir Singh 已提交
1955
}
1956 1957 1958 1959
/*
 * The user of this function is...
 * RES_LIMIT.
 */
1960 1961
static int mem_cgroup_write(struct cgroup *cont, struct cftype *cft,
			    const char *buffer)
B
Balbir Singh 已提交
1962
{
1963
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cont);
1964
	int type, name;
1965 1966 1967
	unsigned long long val;
	int ret;

1968 1969 1970
	type = MEMFILE_TYPE(cft->private);
	name = MEMFILE_ATTR(cft->private);
	switch (name) {
1971 1972 1973
	case RES_LIMIT:
		/* This function does all necessary parse...reuse it */
		ret = res_counter_memparse_write_strategy(buffer, &val);
1974 1975 1976
		if (ret)
			break;
		if (type == _MEM)
1977
			ret = mem_cgroup_resize_limit(memcg, val);
1978 1979
		else
			ret = mem_cgroup_resize_memsw_limit(memcg, val);
1980 1981 1982 1983 1984 1985
		break;
	default:
		ret = -EINVAL; /* should be BUG() ? */
		break;
	}
	return ret;
B
Balbir Singh 已提交
1986 1987
}

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
static void memcg_get_hierarchical_limit(struct mem_cgroup *memcg,
		unsigned long long *mem_limit, unsigned long long *memsw_limit)
{
	struct cgroup *cgroup;
	unsigned long long min_limit, min_memsw_limit, tmp;

	min_limit = res_counter_read_u64(&memcg->res, RES_LIMIT);
	min_memsw_limit = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
	cgroup = memcg->css.cgroup;
	if (!memcg->use_hierarchy)
		goto out;

	while (cgroup->parent) {
		cgroup = cgroup->parent;
		memcg = mem_cgroup_from_cont(cgroup);
		if (!memcg->use_hierarchy)
			break;
		tmp = res_counter_read_u64(&memcg->res, RES_LIMIT);
		min_limit = min(min_limit, tmp);
		tmp = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
		min_memsw_limit = min(min_memsw_limit, tmp);
	}
out:
	*mem_limit = min_limit;
	*memsw_limit = min_memsw_limit;
	return;
}

2016
static int mem_cgroup_reset(struct cgroup *cont, unsigned int event)
2017 2018
{
	struct mem_cgroup *mem;
2019
	int type, name;
2020 2021

	mem = mem_cgroup_from_cont(cont);
2022 2023 2024
	type = MEMFILE_TYPE(event);
	name = MEMFILE_ATTR(event);
	switch (name) {
2025
	case RES_MAX_USAGE:
2026 2027 2028 2029
		if (type == _MEM)
			res_counter_reset_max(&mem->res);
		else
			res_counter_reset_max(&mem->memsw);
2030 2031
		break;
	case RES_FAILCNT:
2032 2033 2034 2035
		if (type == _MEM)
			res_counter_reset_failcnt(&mem->res);
		else
			res_counter_reset_failcnt(&mem->memsw);
2036 2037
		break;
	}
2038
	return 0;
2039 2040
}

K
KAMEZAWA Hiroyuki 已提交
2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057

/* For read statistics */
enum {
	MCS_CACHE,
	MCS_RSS,
	MCS_PGPGIN,
	MCS_PGPGOUT,
	MCS_INACTIVE_ANON,
	MCS_ACTIVE_ANON,
	MCS_INACTIVE_FILE,
	MCS_ACTIVE_FILE,
	MCS_UNEVICTABLE,
	NR_MCS_STAT,
};

struct mcs_total_stat {
	s64 stat[NR_MCS_STAT];
2058 2059
};

K
KAMEZAWA Hiroyuki 已提交
2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110
struct {
	char *local_name;
	char *total_name;
} memcg_stat_strings[NR_MCS_STAT] = {
	{"cache", "total_cache"},
	{"rss", "total_rss"},
	{"pgpgin", "total_pgpgin"},
	{"pgpgout", "total_pgpgout"},
	{"inactive_anon", "total_inactive_anon"},
	{"active_anon", "total_active_anon"},
	{"inactive_file", "total_inactive_file"},
	{"active_file", "total_active_file"},
	{"unevictable", "total_unevictable"}
};


static int mem_cgroup_get_local_stat(struct mem_cgroup *mem, void *data)
{
	struct mcs_total_stat *s = data;
	s64 val;

	/* per cpu stat */
	val = mem_cgroup_read_stat(&mem->stat, MEM_CGROUP_STAT_CACHE);
	s->stat[MCS_CACHE] += val * PAGE_SIZE;
	val = mem_cgroup_read_stat(&mem->stat, MEM_CGROUP_STAT_RSS);
	s->stat[MCS_RSS] += val * PAGE_SIZE;
	val = mem_cgroup_read_stat(&mem->stat, MEM_CGROUP_STAT_PGPGIN_COUNT);
	s->stat[MCS_PGPGIN] += val;
	val = mem_cgroup_read_stat(&mem->stat, MEM_CGROUP_STAT_PGPGOUT_COUNT);
	s->stat[MCS_PGPGOUT] += val;

	/* per zone stat */
	val = mem_cgroup_get_local_zonestat(mem, LRU_INACTIVE_ANON);
	s->stat[MCS_INACTIVE_ANON] += val * PAGE_SIZE;
	val = mem_cgroup_get_local_zonestat(mem, LRU_ACTIVE_ANON);
	s->stat[MCS_ACTIVE_ANON] += val * PAGE_SIZE;
	val = mem_cgroup_get_local_zonestat(mem, LRU_INACTIVE_FILE);
	s->stat[MCS_INACTIVE_FILE] += val * PAGE_SIZE;
	val = mem_cgroup_get_local_zonestat(mem, LRU_ACTIVE_FILE);
	s->stat[MCS_ACTIVE_FILE] += val * PAGE_SIZE;
	val = mem_cgroup_get_local_zonestat(mem, LRU_UNEVICTABLE);
	s->stat[MCS_UNEVICTABLE] += val * PAGE_SIZE;
	return 0;
}

static void
mem_cgroup_get_total_stat(struct mem_cgroup *mem, struct mcs_total_stat *s)
{
	mem_cgroup_walk_tree(mem, s, mem_cgroup_get_local_stat);
}

2111 2112
static int mem_control_stat_show(struct cgroup *cont, struct cftype *cft,
				 struct cgroup_map_cb *cb)
2113 2114
{
	struct mem_cgroup *mem_cont = mem_cgroup_from_cont(cont);
K
KAMEZAWA Hiroyuki 已提交
2115
	struct mcs_total_stat mystat;
2116 2117
	int i;

K
KAMEZAWA Hiroyuki 已提交
2118 2119
	memset(&mystat, 0, sizeof(mystat));
	mem_cgroup_get_local_stat(mem_cont, &mystat);
2120

K
KAMEZAWA Hiroyuki 已提交
2121 2122
	for (i = 0; i < NR_MCS_STAT; i++)
		cb->fill(cb, memcg_stat_strings[i].local_name, mystat.stat[i]);
L
Lee Schermerhorn 已提交
2123

K
KAMEZAWA Hiroyuki 已提交
2124
	/* Hierarchical information */
2125 2126 2127 2128 2129 2130 2131
	{
		unsigned long long limit, memsw_limit;
		memcg_get_hierarchical_limit(mem_cont, &limit, &memsw_limit);
		cb->fill(cb, "hierarchical_memory_limit", limit);
		if (do_swap_account)
			cb->fill(cb, "hierarchical_memsw_limit", memsw_limit);
	}
K
KOSAKI Motohiro 已提交
2132

K
KAMEZAWA Hiroyuki 已提交
2133 2134 2135 2136 2137 2138
	memset(&mystat, 0, sizeof(mystat));
	mem_cgroup_get_total_stat(mem_cont, &mystat);
	for (i = 0; i < NR_MCS_STAT; i++)
		cb->fill(cb, memcg_stat_strings[i].total_name, mystat.stat[i]);


K
KOSAKI Motohiro 已提交
2139
#ifdef CONFIG_DEBUG_VM
2140
	cb->fill(cb, "inactive_ratio", calc_inactive_ratio(mem_cont, NULL));
K
KOSAKI Motohiro 已提交
2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167

	{
		int nid, zid;
		struct mem_cgroup_per_zone *mz;
		unsigned long recent_rotated[2] = {0, 0};
		unsigned long recent_scanned[2] = {0, 0};

		for_each_online_node(nid)
			for (zid = 0; zid < MAX_NR_ZONES; zid++) {
				mz = mem_cgroup_zoneinfo(mem_cont, nid, zid);

				recent_rotated[0] +=
					mz->reclaim_stat.recent_rotated[0];
				recent_rotated[1] +=
					mz->reclaim_stat.recent_rotated[1];
				recent_scanned[0] +=
					mz->reclaim_stat.recent_scanned[0];
				recent_scanned[1] +=
					mz->reclaim_stat.recent_scanned[1];
			}
		cb->fill(cb, "recent_rotated_anon", recent_rotated[0]);
		cb->fill(cb, "recent_rotated_file", recent_rotated[1]);
		cb->fill(cb, "recent_scanned_anon", recent_scanned[0]);
		cb->fill(cb, "recent_scanned_file", recent_scanned[1]);
	}
#endif

2168 2169 2170
	return 0;
}

K
KOSAKI Motohiro 已提交
2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182
static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft)
{
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);

	return get_swappiness(memcg);
}

static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft,
				       u64 val)
{
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
	struct mem_cgroup *parent;
2183

K
KOSAKI Motohiro 已提交
2184 2185 2186 2187 2188 2189 2190
	if (val > 100)
		return -EINVAL;

	if (cgrp->parent == NULL)
		return -EINVAL;

	parent = mem_cgroup_from_cont(cgrp->parent);
2191 2192 2193

	cgroup_lock();

K
KOSAKI Motohiro 已提交
2194 2195
	/* If under hierarchy, only empty-root can set this value */
	if ((parent->use_hierarchy) ||
2196 2197
	    (memcg->use_hierarchy && !list_empty(&cgrp->children))) {
		cgroup_unlock();
K
KOSAKI Motohiro 已提交
2198
		return -EINVAL;
2199
	}
K
KOSAKI Motohiro 已提交
2200 2201 2202 2203 2204

	spin_lock(&memcg->reclaim_param_lock);
	memcg->swappiness = val;
	spin_unlock(&memcg->reclaim_param_lock);

2205 2206
	cgroup_unlock();

K
KOSAKI Motohiro 已提交
2207 2208 2209
	return 0;
}

2210

B
Balbir Singh 已提交
2211 2212
static struct cftype mem_cgroup_files[] = {
	{
2213
		.name = "usage_in_bytes",
2214
		.private = MEMFILE_PRIVATE(_MEM, RES_USAGE),
2215
		.read_u64 = mem_cgroup_read,
B
Balbir Singh 已提交
2216
	},
2217 2218
	{
		.name = "max_usage_in_bytes",
2219
		.private = MEMFILE_PRIVATE(_MEM, RES_MAX_USAGE),
2220
		.trigger = mem_cgroup_reset,
2221 2222
		.read_u64 = mem_cgroup_read,
	},
B
Balbir Singh 已提交
2223
	{
2224
		.name = "limit_in_bytes",
2225
		.private = MEMFILE_PRIVATE(_MEM, RES_LIMIT),
2226
		.write_string = mem_cgroup_write,
2227
		.read_u64 = mem_cgroup_read,
B
Balbir Singh 已提交
2228 2229 2230
	},
	{
		.name = "failcnt",
2231
		.private = MEMFILE_PRIVATE(_MEM, RES_FAILCNT),
2232
		.trigger = mem_cgroup_reset,
2233
		.read_u64 = mem_cgroup_read,
B
Balbir Singh 已提交
2234
	},
2235 2236
	{
		.name = "stat",
2237
		.read_map = mem_control_stat_show,
2238
	},
2239 2240 2241 2242
	{
		.name = "force_empty",
		.trigger = mem_cgroup_force_empty_write,
	},
2243 2244 2245 2246 2247
	{
		.name = "use_hierarchy",
		.write_u64 = mem_cgroup_hierarchy_write,
		.read_u64 = mem_cgroup_hierarchy_read,
	},
K
KOSAKI Motohiro 已提交
2248 2249 2250 2251 2252
	{
		.name = "swappiness",
		.read_u64 = mem_cgroup_swappiness_read,
		.write_u64 = mem_cgroup_swappiness_write,
	},
B
Balbir Singh 已提交
2253 2254
};

2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
static struct cftype memsw_cgroup_files[] = {
	{
		.name = "memsw.usage_in_bytes",
		.private = MEMFILE_PRIVATE(_MEMSWAP, RES_USAGE),
		.read_u64 = mem_cgroup_read,
	},
	{
		.name = "memsw.max_usage_in_bytes",
		.private = MEMFILE_PRIVATE(_MEMSWAP, RES_MAX_USAGE),
		.trigger = mem_cgroup_reset,
		.read_u64 = mem_cgroup_read,
	},
	{
		.name = "memsw.limit_in_bytes",
		.private = MEMFILE_PRIVATE(_MEMSWAP, RES_LIMIT),
		.write_string = mem_cgroup_write,
		.read_u64 = mem_cgroup_read,
	},
	{
		.name = "memsw.failcnt",
		.private = MEMFILE_PRIVATE(_MEMSWAP, RES_FAILCNT),
		.trigger = mem_cgroup_reset,
		.read_u64 = mem_cgroup_read,
	},
};

static int register_memsw_files(struct cgroup *cont, struct cgroup_subsys *ss)
{
	if (!do_swap_account)
		return 0;
	return cgroup_add_files(cont, ss, memsw_cgroup_files,
				ARRAY_SIZE(memsw_cgroup_files));
};
#else
static int register_memsw_files(struct cgroup *cont, struct cgroup_subsys *ss)
{
	return 0;
}
#endif

2296 2297 2298
static int alloc_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node)
{
	struct mem_cgroup_per_node *pn;
2299
	struct mem_cgroup_per_zone *mz;
2300
	enum lru_list l;
2301
	int zone, tmp = node;
2302 2303 2304 2305 2306 2307 2308 2309
	/*
	 * This routine is called against possible nodes.
	 * But it's BUG to call kmalloc() against offline node.
	 *
	 * TODO: this routine can waste much memory for nodes which will
	 *       never be onlined. It's better to use memory hotplug callback
	 *       function.
	 */
2310 2311 2312
	if (!node_state(node, N_NORMAL_MEMORY))
		tmp = -1;
	pn = kmalloc_node(sizeof(*pn), GFP_KERNEL, tmp);
2313 2314
	if (!pn)
		return 1;
2315

2316 2317
	mem->info.nodeinfo[node] = pn;
	memset(pn, 0, sizeof(*pn));
2318 2319 2320

	for (zone = 0; zone < MAX_NR_ZONES; zone++) {
		mz = &pn->zoneinfo[zone];
2321 2322
		for_each_lru(l)
			INIT_LIST_HEAD(&mz->lists[l]);
2323
	}
2324 2325 2326
	return 0;
}

2327 2328 2329 2330 2331
static void free_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node)
{
	kfree(mem->info.nodeinfo[node]);
}

2332 2333 2334 2335 2336 2337
static int mem_cgroup_size(void)
{
	int cpustat_size = nr_cpu_ids * sizeof(struct mem_cgroup_stat_cpu);
	return sizeof(struct mem_cgroup) + cpustat_size;
}

2338 2339 2340
static struct mem_cgroup *mem_cgroup_alloc(void)
{
	struct mem_cgroup *mem;
2341
	int size = mem_cgroup_size();
2342

2343 2344
	if (size < PAGE_SIZE)
		mem = kmalloc(size, GFP_KERNEL);
2345
	else
2346
		mem = vmalloc(size);
2347 2348

	if (mem)
2349
		memset(mem, 0, size);
2350 2351 2352
	return mem;
}

2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363
/*
 * At destroying mem_cgroup, references from swap_cgroup can remain.
 * (scanning all at force_empty is too costly...)
 *
 * Instead of clearing all references at force_empty, we remember
 * the number of reference from swap_cgroup and free mem_cgroup when
 * it goes down to 0.
 *
 * Removal of cgroup itself succeeds regardless of refs from swap.
 */

2364
static void __mem_cgroup_free(struct mem_cgroup *mem)
2365
{
K
KAMEZAWA Hiroyuki 已提交
2366 2367
	int node;

K
KAMEZAWA Hiroyuki 已提交
2368 2369
	free_css_id(&mem_cgroup_subsys, &mem->css);

K
KAMEZAWA Hiroyuki 已提交
2370 2371 2372
	for_each_node_state(node, N_POSSIBLE)
		free_mem_cgroup_per_zone_info(mem, node);

2373
	if (mem_cgroup_size() < PAGE_SIZE)
2374 2375 2376 2377 2378
		kfree(mem);
	else
		vfree(mem);
}

2379 2380 2381 2382 2383 2384 2385
static void mem_cgroup_get(struct mem_cgroup *mem)
{
	atomic_inc(&mem->refcnt);
}

static void mem_cgroup_put(struct mem_cgroup *mem)
{
2386 2387
	if (atomic_dec_and_test(&mem->refcnt)) {
		struct mem_cgroup *parent = parent_mem_cgroup(mem);
2388
		__mem_cgroup_free(mem);
2389 2390 2391
		if (parent)
			mem_cgroup_put(parent);
	}
2392 2393
}

2394 2395 2396 2397 2398 2399 2400 2401 2402
/*
 * Returns the parent mem_cgroup in memcgroup hierarchy with hierarchy enabled.
 */
static struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *mem)
{
	if (!mem->res.parent)
		return NULL;
	return mem_cgroup_from_res_counter(mem->res.parent, res);
}
2403

2404 2405 2406
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
static void __init enable_swap_cgroup(void)
{
2407
	if (!mem_cgroup_disabled() && really_do_swap_account)
2408 2409 2410 2411 2412 2413 2414 2415
		do_swap_account = 1;
}
#else
static void __init enable_swap_cgroup(void)
{
}
#endif

L
Li Zefan 已提交
2416
static struct cgroup_subsys_state * __ref
B
Balbir Singh 已提交
2417 2418
mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
{
2419
	struct mem_cgroup *mem, *parent;
K
KAMEZAWA Hiroyuki 已提交
2420
	long error = -ENOMEM;
2421
	int node;
B
Balbir Singh 已提交
2422

2423 2424
	mem = mem_cgroup_alloc();
	if (!mem)
K
KAMEZAWA Hiroyuki 已提交
2425
		return ERR_PTR(error);
2426

2427 2428 2429
	for_each_node_state(node, N_POSSIBLE)
		if (alloc_mem_cgroup_per_zone_info(mem, node))
			goto free_out;
2430
	/* root ? */
2431
	if (cont->parent == NULL) {
2432
		enable_swap_cgroup();
2433
		parent = NULL;
2434
	} else {
2435
		parent = mem_cgroup_from_cont(cont->parent);
2436 2437
		mem->use_hierarchy = parent->use_hierarchy;
	}
2438

2439 2440 2441
	if (parent && parent->use_hierarchy) {
		res_counter_init(&mem->res, &parent->res);
		res_counter_init(&mem->memsw, &parent->memsw);
2442 2443 2444 2445 2446 2447 2448
		/*
		 * We increment refcnt of the parent to ensure that we can
		 * safely access it on res_counter_charge/uncharge.
		 * This refcnt will be decremented when freeing this
		 * mem_cgroup(see mem_cgroup_put).
		 */
		mem_cgroup_get(parent);
2449 2450 2451 2452
	} else {
		res_counter_init(&mem->res, NULL);
		res_counter_init(&mem->memsw, NULL);
	}
K
KAMEZAWA Hiroyuki 已提交
2453
	mem->last_scanned_child = 0;
K
KOSAKI Motohiro 已提交
2454
	spin_lock_init(&mem->reclaim_param_lock);
2455

K
KOSAKI Motohiro 已提交
2456 2457
	if (parent)
		mem->swappiness = get_swappiness(parent);
2458
	atomic_set(&mem->refcnt, 1);
B
Balbir Singh 已提交
2459
	return &mem->css;
2460
free_out:
2461
	__mem_cgroup_free(mem);
K
KAMEZAWA Hiroyuki 已提交
2462
	return ERR_PTR(error);
B
Balbir Singh 已提交
2463 2464
}

2465
static int mem_cgroup_pre_destroy(struct cgroup_subsys *ss,
2466 2467 2468
					struct cgroup *cont)
{
	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);
2469 2470

	return mem_cgroup_force_empty(mem, false);
2471 2472
}

B
Balbir Singh 已提交
2473 2474 2475
static void mem_cgroup_destroy(struct cgroup_subsys *ss,
				struct cgroup *cont)
{
2476 2477 2478
	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);

	mem_cgroup_put(mem);
B
Balbir Singh 已提交
2479 2480 2481 2482 2483
}

static int mem_cgroup_populate(struct cgroup_subsys *ss,
				struct cgroup *cont)
{
2484 2485 2486 2487 2488 2489 2490 2491
	int ret;

	ret = cgroup_add_files(cont, ss, mem_cgroup_files,
				ARRAY_SIZE(mem_cgroup_files));

	if (!ret)
		ret = register_memsw_files(cont, ss);
	return ret;
B
Balbir Singh 已提交
2492 2493
}

B
Balbir Singh 已提交
2494 2495 2496 2497 2498
static void mem_cgroup_move_task(struct cgroup_subsys *ss,
				struct cgroup *cont,
				struct cgroup *old_cont,
				struct task_struct *p)
{
2499
	mutex_lock(&memcg_tasklist);
B
Balbir Singh 已提交
2500
	/*
2501 2502
	 * FIXME: It's better to move charges of this process from old
	 * memcg to new memcg. But it's just on TODO-List now.
B
Balbir Singh 已提交
2503
	 */
2504
	mutex_unlock(&memcg_tasklist);
B
Balbir Singh 已提交
2505 2506
}

B
Balbir Singh 已提交
2507 2508 2509 2510
struct cgroup_subsys mem_cgroup_subsys = {
	.name = "memory",
	.subsys_id = mem_cgroup_subsys_id,
	.create = mem_cgroup_create,
2511
	.pre_destroy = mem_cgroup_pre_destroy,
B
Balbir Singh 已提交
2512 2513
	.destroy = mem_cgroup_destroy,
	.populate = mem_cgroup_populate,
B
Balbir Singh 已提交
2514
	.attach = mem_cgroup_move_task,
2515
	.early_init = 0,
K
KAMEZAWA Hiroyuki 已提交
2516
	.use_id = 1,
B
Balbir Singh 已提交
2517
};
2518 2519 2520 2521 2522 2523 2524 2525 2526 2527

#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP

static int __init disable_swap_account(char *s)
{
	really_do_swap_account = 0;
	return 1;
}
__setup("noswapaccount", disable_swap_account);
#endif