memcontrol.c 97.3 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>
24
#include <linux/hugetlb.h>
K
KAMEZAWA Hiroyuki 已提交
25
#include <linux/pagemap.h>
26
#include <linux/smp.h>
27
#include <linux/page-flags.h>
28
#include <linux/backing-dev.h>
29 30
#include <linux/bit_spinlock.h>
#include <linux/rcupdate.h>
31
#include <linux/limits.h>
32
#include <linux/mutex.h>
33
#include <linux/rbtree.h>
34
#include <linux/slab.h>
35 36 37
#include <linux/swap.h>
#include <linux/spinlock.h>
#include <linux/fs.h>
38
#include <linux/seq_file.h>
39
#include <linux/vmalloc.h>
40
#include <linux/mm_inline.h>
41
#include <linux/page_cgroup.h>
42
#include <linux/cpu.h>
K
KAMEZAWA Hiroyuki 已提交
43
#include "internal.h"
B
Balbir Singh 已提交
44

45 46
#include <asm/uaccess.h>

47 48
struct cgroup_subsys mem_cgroup_subsys __read_mostly;
#define MEM_CGROUP_RECLAIM_RETRIES	5
49
struct mem_cgroup *root_mem_cgroup __read_mostly;
B
Balbir Singh 已提交
50

51
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
L
Li Zefan 已提交
52
/* Turned on only when memory cgroup is enabled && really_do_swap_account = 1 */
53 54 55 56 57 58
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

59
#define SOFTLIMIT_EVENTS_THRESH (1000)
60

61 62 63 64 65 66 67 68
/*
 * 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 */
69
	MEM_CGROUP_STAT_RSS,	   /* # of pages charged as anon rss */
70
	MEM_CGROUP_STAT_FILE_MAPPED,  /* # of pages charged as file rss */
71 72
	MEM_CGROUP_STAT_PGPGIN_COUNT,	/* # of pages paged in */
	MEM_CGROUP_STAT_PGPGOUT_COUNT,	/* # of pages paged out */
73
	MEM_CGROUP_STAT_EVENTS,	/* sum of pagein + pageout for internal use */
74
	MEM_CGROUP_STAT_SWAPOUT, /* # of pages, swapped out */
75 76 77 78 79 80 81 82 83

	MEM_CGROUP_STAT_NSTATS,
};

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

struct mem_cgroup_stat {
84
	struct mem_cgroup_stat_cpu cpustat[0];
85 86
};

87 88 89 90 91 92 93 94 95 96 97 98 99 100
static inline void
__mem_cgroup_stat_reset_safe(struct mem_cgroup_stat_cpu *stat,
				enum mem_cgroup_stat_index idx)
{
	stat->count[idx] = 0;
}

static inline s64
__mem_cgroup_stat_read_local(struct mem_cgroup_stat_cpu *stat,
				enum mem_cgroup_stat_index idx)
{
	return stat->count[idx];
}

101 102 103
/*
 * For accounting under irq disable, no need for increment preempt count.
 */
104
static inline void __mem_cgroup_stat_add_safe(struct mem_cgroup_stat_cpu *stat,
105 106
		enum mem_cgroup_stat_index idx, int val)
{
107
	stat->count[idx] += val;
108 109 110 111 112 113 114 115 116 117 118 119
}

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
120 121 122 123 124 125 126 127 128
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;
}

129 130 131 132
/*
 * per-zone information in memory controller.
 */
struct mem_cgroup_per_zone {
133 134 135
	/*
	 * spin_lock to protect the per cgroup LRU
	 */
136 137
	struct list_head	lists[NR_LRU_LISTS];
	unsigned long		count[NR_LRU_LISTS];
K
KOSAKI Motohiro 已提交
138 139

	struct zone_reclaim_stat reclaim_stat;
140 141 142 143
	struct rb_node		tree_node;	/* RB tree node */
	unsigned long long	usage_in_excess;/* Set to the value by which */
						/* the soft limit is exceeded*/
	bool			on_tree;
144 145
	struct mem_cgroup	*mem;		/* Back pointer, we cannot */
						/* use container_of	   */
146 147 148 149 150 151 152 153 154 155 156 157
};
/* 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];
};

158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177
/*
 * Cgroups above their limits are maintained in a RB-Tree, independent of
 * their hierarchy representation
 */

struct mem_cgroup_tree_per_zone {
	struct rb_root rb_root;
	spinlock_t lock;
};

struct mem_cgroup_tree_per_node {
	struct mem_cgroup_tree_per_zone rb_tree_per_zone[MAX_NR_ZONES];
};

struct mem_cgroup_tree {
	struct mem_cgroup_tree_per_node *rb_tree_per_node[MAX_NUMNODES];
};

static struct mem_cgroup_tree soft_limit_tree __read_mostly;

B
Balbir Singh 已提交
178 179 180 181 182 183 184
/*
 * 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
185 186 187
 * 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
188 189 190 191 192 193 194
 */
struct mem_cgroup {
	struct cgroup_subsys_state css;
	/*
	 * the counter to account for memory usage
	 */
	struct res_counter res;
195 196 197 198
	/*
	 * the counter to account for mem+swap usage.
	 */
	struct res_counter memsw;
199 200 201 202
	/*
	 * Per cgroup active and inactive list, similar to the
	 * per zone LRU lists.
	 */
203
	struct mem_cgroup_lru_info info;
204

K
KOSAKI Motohiro 已提交
205 206 207 208 209
	/*
	  protect against reclaim related member.
	*/
	spinlock_t reclaim_param_lock;

210
	int	prev_priority;	/* for recording reclaim priority */
211 212

	/*
213
	 * While reclaiming in a hierarchy, we cache the last child we
K
KAMEZAWA Hiroyuki 已提交
214
	 * reclaimed from.
215
	 */
K
KAMEZAWA Hiroyuki 已提交
216
	int last_scanned_child;
217 218 219 220
	/*
	 * Should the accounting and control be hierarchical, per subtree?
	 */
	bool use_hierarchy;
221
	unsigned long	last_oom_jiffies;
222
	atomic_t	refcnt;
223

K
KOSAKI Motohiro 已提交
224 225
	unsigned int	swappiness;

226 227 228
	/* set when res.limit == memsw.limit */
	bool		memsw_is_minimum;

229 230 231 232 233 234
	/*
	 * Should we move charges of a task when a task is moved into this
	 * mem_cgroup ? And what type of charges should we move ?
	 */
	unsigned long 	move_charge_at_immigrate;

235
	/*
236
	 * statistics. This must be placed at the end of memcg.
237 238
	 */
	struct mem_cgroup_stat stat;
B
Balbir Singh 已提交
239 240
};

241 242 243 244 245 246
/* Stuffs for move charges at task migration. */
/*
 * Types of charges to be moved. "move_charge_at_immitgrate" is treated as a
 * left-shifted bitmap of these types.
 */
enum move_type {
247
	MOVE_CHARGE_TYPE_ANON,	/* private anonymous page and swap of it */
248 249 250
	NR_MOVE_TYPE,
};

251 252 253 254 255
/* "mc" and its members are protected by cgroup_mutex */
static struct move_charge_struct {
	struct mem_cgroup *from;
	struct mem_cgroup *to;
	unsigned long precharge;
256
	unsigned long moved_charge;
257 258 259 260 261
	struct task_struct *moving_task;	/* a task moving charges */
	wait_queue_head_t waitq;		/* a waitq for other context */
} mc = {
	.waitq = __WAIT_QUEUE_HEAD_INITIALIZER(mc.waitq),
};
262

263 264 265 266 267 268 269
/*
 * Maximum loops in mem_cgroup_hierarchical_reclaim(), used for soft
 * limit reclaim to prevent infinite loops, if they ever occur.
 */
#define	MEM_CGROUP_MAX_RECLAIM_LOOPS		(100)
#define	MEM_CGROUP_MAX_SOFT_LIMIT_RECLAIM_LOOPS	(2)

270 271 272
enum charge_type {
	MEM_CGROUP_CHARGE_TYPE_CACHE = 0,
	MEM_CGROUP_CHARGE_TYPE_MAPPED,
273
	MEM_CGROUP_CHARGE_TYPE_SHMEM,	/* used by page migration of shmem */
274
	MEM_CGROUP_CHARGE_TYPE_FORCE,	/* used by force_empty */
K
KAMEZAWA Hiroyuki 已提交
275
	MEM_CGROUP_CHARGE_TYPE_SWAPOUT,	/* for accounting swapcache */
K
KAMEZAWA Hiroyuki 已提交
276
	MEM_CGROUP_CHARGE_TYPE_DROP,	/* a page was unused swap cache */
277 278 279
	NR_CHARGE_TYPE,
};

280 281 282 283
/* only for here (for easy reading.) */
#define PCGF_CACHE	(1UL << PCG_CACHE)
#define PCGF_USED	(1UL << PCG_USED)
#define PCGF_LOCK	(1UL << PCG_LOCK)
284 285
/* Not used, but added here for completeness */
#define PCGF_ACCT	(1UL << PCG_ACCT)
286

287 288 289 290 291 292 293
/* 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)

294 295 296 297 298 299 300
/*
 * Reclaim flags for mem_cgroup_hierarchical_reclaim
 */
#define MEM_CGROUP_RECLAIM_NOSWAP_BIT	0x0
#define MEM_CGROUP_RECLAIM_NOSWAP	(1 << MEM_CGROUP_RECLAIM_NOSWAP_BIT)
#define MEM_CGROUP_RECLAIM_SHRINK_BIT	0x1
#define MEM_CGROUP_RECLAIM_SHRINK	(1 << MEM_CGROUP_RECLAIM_SHRINK_BIT)
301 302
#define MEM_CGROUP_RECLAIM_SOFT_BIT	0x2
#define MEM_CGROUP_RECLAIM_SOFT		(1 << MEM_CGROUP_RECLAIM_SOFT_BIT)
303

304 305
static void mem_cgroup_get(struct mem_cgroup *mem);
static void mem_cgroup_put(struct mem_cgroup *mem);
306
static struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *mem);
307
static void drain_all_stock_async(void);
308

309 310 311 312 313 314
static struct mem_cgroup_per_zone *
mem_cgroup_zoneinfo(struct mem_cgroup *mem, int nid, int zid)
{
	return &mem->info.nodeinfo[nid]->zoneinfo[zid];
}

315 316 317 318 319
struct cgroup_subsys_state *mem_cgroup_css(struct mem_cgroup *mem)
{
	return &mem->css;
}

320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348
static struct mem_cgroup_per_zone *
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);

	if (!mem)
		return NULL;

	return mem_cgroup_zoneinfo(mem, nid, zid);
}

static struct mem_cgroup_tree_per_zone *
soft_limit_tree_node_zone(int nid, int zid)
{
	return &soft_limit_tree.rb_tree_per_node[nid]->rb_tree_per_zone[zid];
}

static struct mem_cgroup_tree_per_zone *
soft_limit_tree_from_page(struct page *page)
{
	int nid = page_to_nid(page);
	int zid = page_zonenum(page);

	return &soft_limit_tree.rb_tree_per_node[nid]->rb_tree_per_zone[zid];
}

static void
349
__mem_cgroup_insert_exceeded(struct mem_cgroup *mem,
350
				struct mem_cgroup_per_zone *mz,
351 352
				struct mem_cgroup_tree_per_zone *mctz,
				unsigned long long new_usage_in_excess)
353 354 355 356 357 358 359 360
{
	struct rb_node **p = &mctz->rb_root.rb_node;
	struct rb_node *parent = NULL;
	struct mem_cgroup_per_zone *mz_node;

	if (mz->on_tree)
		return;

361 362 363
	mz->usage_in_excess = new_usage_in_excess;
	if (!mz->usage_in_excess)
		return;
364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379
	while (*p) {
		parent = *p;
		mz_node = rb_entry(parent, struct mem_cgroup_per_zone,
					tree_node);
		if (mz->usage_in_excess < mz_node->usage_in_excess)
			p = &(*p)->rb_left;
		/*
		 * We can't avoid mem cgroups that are over their soft
		 * limit by the same amount
		 */
		else if (mz->usage_in_excess >= mz_node->usage_in_excess)
			p = &(*p)->rb_right;
	}
	rb_link_node(&mz->tree_node, parent, p);
	rb_insert_color(&mz->tree_node, &mctz->rb_root);
	mz->on_tree = true;
380 381 382 383 384 385 386 387 388 389 390 391 392
}

static void
__mem_cgroup_remove_exceeded(struct mem_cgroup *mem,
				struct mem_cgroup_per_zone *mz,
				struct mem_cgroup_tree_per_zone *mctz)
{
	if (!mz->on_tree)
		return;
	rb_erase(&mz->tree_node, &mctz->rb_root);
	mz->on_tree = false;
}

393 394 395 396 397 398
static void
mem_cgroup_remove_exceeded(struct mem_cgroup *mem,
				struct mem_cgroup_per_zone *mz,
				struct mem_cgroup_tree_per_zone *mctz)
{
	spin_lock(&mctz->lock);
399
	__mem_cgroup_remove_exceeded(mem, mz, mctz);
400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422
	spin_unlock(&mctz->lock);
}

static bool mem_cgroup_soft_limit_check(struct mem_cgroup *mem)
{
	bool ret = false;
	int cpu;
	s64 val;
	struct mem_cgroup_stat_cpu *cpustat;

	cpu = get_cpu();
	cpustat = &mem->stat.cpustat[cpu];
	val = __mem_cgroup_stat_read_local(cpustat, MEM_CGROUP_STAT_EVENTS);
	if (unlikely(val > SOFTLIMIT_EVENTS_THRESH)) {
		__mem_cgroup_stat_reset_safe(cpustat, MEM_CGROUP_STAT_EVENTS);
		ret = true;
	}
	put_cpu();
	return ret;
}

static void mem_cgroup_update_tree(struct mem_cgroup *mem, struct page *page)
{
423
	unsigned long long excess;
424 425
	struct mem_cgroup_per_zone *mz;
	struct mem_cgroup_tree_per_zone *mctz;
426 427
	int nid = page_to_nid(page);
	int zid = page_zonenum(page);
428 429 430
	mctz = soft_limit_tree_from_page(page);

	/*
431 432
	 * Necessary to update all ancestors when hierarchy is used.
	 * because their event counter is not touched.
433
	 */
434 435
	for (; mem; mem = parent_mem_cgroup(mem)) {
		mz = mem_cgroup_zoneinfo(mem, nid, zid);
436
		excess = res_counter_soft_limit_excess(&mem->res);
437 438 439 440
		/*
		 * We have to update the tree if mz is on RB-tree or
		 * mem is over its softlimit.
		 */
441
		if (excess || mz->on_tree) {
442 443 444 445 446
			spin_lock(&mctz->lock);
			/* if on-tree, remove it */
			if (mz->on_tree)
				__mem_cgroup_remove_exceeded(mem, mz, mctz);
			/*
447 448
			 * Insert again. mz->usage_in_excess will be updated.
			 * If excess is 0, no tree ops.
449
			 */
450
			__mem_cgroup_insert_exceeded(mem, mz, mctz, excess);
451 452
			spin_unlock(&mctz->lock);
		}
453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470
	}
}

static void mem_cgroup_remove_from_trees(struct mem_cgroup *mem)
{
	int node, zone;
	struct mem_cgroup_per_zone *mz;
	struct mem_cgroup_tree_per_zone *mctz;

	for_each_node_state(node, N_POSSIBLE) {
		for (zone = 0; zone < MAX_NR_ZONES; zone++) {
			mz = mem_cgroup_zoneinfo(mem, node, zone);
			mctz = soft_limit_tree_node_zone(node, zone);
			mem_cgroup_remove_exceeded(mem, mz, mctz);
		}
	}
}

471 472 473 474 475 476 477 478 479
static inline unsigned long mem_cgroup_get_excess(struct mem_cgroup *mem)
{
	return res_counter_soft_limit_excess(&mem->res) >> PAGE_SHIFT;
}

static struct mem_cgroup_per_zone *
__mem_cgroup_largest_soft_limit_node(struct mem_cgroup_tree_per_zone *mctz)
{
	struct rb_node *rightmost = NULL;
480
	struct mem_cgroup_per_zone *mz;
481 482

retry:
483
	mz = NULL;
484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512
	rightmost = rb_last(&mctz->rb_root);
	if (!rightmost)
		goto done;		/* Nothing to reclaim from */

	mz = rb_entry(rightmost, struct mem_cgroup_per_zone, tree_node);
	/*
	 * Remove the node now but someone else can add it back,
	 * we will to add it back at the end of reclaim to its correct
	 * position in the tree.
	 */
	__mem_cgroup_remove_exceeded(mz->mem, mz, mctz);
	if (!res_counter_soft_limit_excess(&mz->mem->res) ||
		!css_tryget(&mz->mem->css))
		goto retry;
done:
	return mz;
}

static struct mem_cgroup_per_zone *
mem_cgroup_largest_soft_limit_node(struct mem_cgroup_tree_per_zone *mctz)
{
	struct mem_cgroup_per_zone *mz;

	spin_lock(&mctz->lock);
	mz = __mem_cgroup_largest_soft_limit_node(mctz);
	spin_unlock(&mctz->lock);
	return mz;
}

513 514 515 516 517 518 519 520 521 522 523 524 525
static void mem_cgroup_swap_statistics(struct mem_cgroup *mem,
					 bool charge)
{
	int val = (charge) ? 1 : -1;
	struct mem_cgroup_stat *stat = &mem->stat;
	struct mem_cgroup_stat_cpu *cpustat;
	int cpu = get_cpu();

	cpustat = &stat->cpustat[cpu];
	__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_SWAPOUT, val);
	put_cpu();
}

526 527 528
static void mem_cgroup_charge_statistics(struct mem_cgroup *mem,
					 struct page_cgroup *pc,
					 bool charge)
529
{
530
	int val = (charge) ? 1 : -1;
531
	struct mem_cgroup_stat *stat = &mem->stat;
532
	struct mem_cgroup_stat_cpu *cpustat;
K
KAMEZAWA Hiroyuki 已提交
533
	int cpu = get_cpu();
534

K
KAMEZAWA Hiroyuki 已提交
535
	cpustat = &stat->cpustat[cpu];
536
	if (PageCgroupCache(pc))
537
		__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_CACHE, val);
538
	else
539
		__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_RSS, val);
540 541

	if (charge)
542
		__mem_cgroup_stat_add_safe(cpustat,
543 544
				MEM_CGROUP_STAT_PGPGIN_COUNT, 1);
	else
545
		__mem_cgroup_stat_add_safe(cpustat,
546
				MEM_CGROUP_STAT_PGPGOUT_COUNT, 1);
547
	__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_EVENTS, 1);
K
KAMEZAWA Hiroyuki 已提交
548
	put_cpu();
549 550
}

K
KAMEZAWA Hiroyuki 已提交
551
static unsigned long mem_cgroup_get_local_zonestat(struct mem_cgroup *mem,
552
					enum lru_list idx)
553 554 555 556 557 558 559 560 561 562 563
{
	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;
564 565
}

566
static struct mem_cgroup *mem_cgroup_from_cont(struct cgroup *cont)
B
Balbir Singh 已提交
567 568 569 570 571 572
{
	return container_of(cgroup_subsys_state(cont,
				mem_cgroup_subsys_id), struct mem_cgroup,
				css);
}

573
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p)
574
{
575 576 577 578 579 580 581 582
	/*
	 * 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;

583 584 585 586
	return container_of(task_subsys_state(p, mem_cgroup_subsys_id),
				struct mem_cgroup, css);
}

587 588 589
static struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm)
{
	struct mem_cgroup *mem = NULL;
590 591 592

	if (!mm)
		return NULL;
593 594 595 596 597 598 599 600 601 602 603 604 605 606 607
	/*
	 * 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;
}

K
KAMEZAWA Hiroyuki 已提交
608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
/*
 * 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;
}

643 644 645 646 647
static inline bool mem_cgroup_is_root(struct mem_cgroup *mem)
{
	return (mem == root_mem_cgroup);
}

K
KAMEZAWA Hiroyuki 已提交
648 649 650 651 652 653 654 655 656 657 658 659 660
/*
 * 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.
 */
661

K
KAMEZAWA Hiroyuki 已提交
662 663 664 665
void mem_cgroup_del_lru_list(struct page *page, enum lru_list lru)
{
	struct page_cgroup *pc;
	struct mem_cgroup_per_zone *mz;
666

667
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
668 669 670
		return;
	pc = lookup_page_cgroup(page);
	/* can happen while we handle swapcache. */
671
	if (!TestClearPageCgroupAcctLRU(pc))
K
KAMEZAWA Hiroyuki 已提交
672
		return;
673
	VM_BUG_ON(!pc->mem_cgroup);
674 675 676 677
	/*
	 * 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
678
	mz = page_cgroup_zoneinfo(pc);
679
	MEM_CGROUP_ZSTAT(mz, lru) -= 1;
680 681 682
	if (mem_cgroup_is_root(pc->mem_cgroup))
		return;
	VM_BUG_ON(list_empty(&pc->lru));
K
KAMEZAWA Hiroyuki 已提交
683 684
	list_del_init(&pc->lru);
	return;
685 686
}

K
KAMEZAWA Hiroyuki 已提交
687
void mem_cgroup_del_lru(struct page *page)
688
{
K
KAMEZAWA Hiroyuki 已提交
689 690
	mem_cgroup_del_lru_list(page, page_lru(page));
}
691

K
KAMEZAWA Hiroyuki 已提交
692 693 694 695
void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru)
{
	struct mem_cgroup_per_zone *mz;
	struct page_cgroup *pc;
696

697
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
698
		return;
699

K
KAMEZAWA Hiroyuki 已提交
700
	pc = lookup_page_cgroup(page);
701 702 703 704
	/*
	 * 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
705
	smp_rmb();
706 707
	/* unused or root page is not rotated. */
	if (!PageCgroupUsed(pc) || mem_cgroup_is_root(pc->mem_cgroup))
K
KAMEZAWA Hiroyuki 已提交
708 709 710
		return;
	mz = page_cgroup_zoneinfo(pc);
	list_move(&pc->lru, &mz->lists[lru]);
711 712
}

K
KAMEZAWA Hiroyuki 已提交
713
void mem_cgroup_add_lru_list(struct page *page, enum lru_list lru)
714
{
K
KAMEZAWA Hiroyuki 已提交
715 716
	struct page_cgroup *pc;
	struct mem_cgroup_per_zone *mz;
717

718
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
719 720
		return;
	pc = lookup_page_cgroup(page);
721
	VM_BUG_ON(PageCgroupAcctLRU(pc));
722 723 724 725
	/*
	 * 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
726 727
	smp_rmb();
	if (!PageCgroupUsed(pc))
L
Lee Schermerhorn 已提交
728
		return;
729

K
KAMEZAWA Hiroyuki 已提交
730
	mz = page_cgroup_zoneinfo(pc);
731
	MEM_CGROUP_ZSTAT(mz, lru) += 1;
732 733 734
	SetPageCgroupAcctLRU(pc);
	if (mem_cgroup_is_root(pc->mem_cgroup))
		return;
K
KAMEZAWA Hiroyuki 已提交
735 736
	list_add(&pc->lru, &mz->lists[lru]);
}
737

K
KAMEZAWA Hiroyuki 已提交
738
/*
739 740 741 742 743
 * 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
744
 */
745
static void mem_cgroup_lru_del_before_commit_swapcache(struct page *page)
K
KAMEZAWA Hiroyuki 已提交
746
{
747 748 749 750 751 752 753 754 755 756 757 758
	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
759 760
}

761 762 763 764 765 766 767 768
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 */
769
	if (PageLRU(page) && !PageCgroupAcctLRU(pc))
770 771 772 773 774
		mem_cgroup_add_lru_list(page, page_lru(page));
	spin_unlock_irqrestore(&zone->lru_lock, flags);
}


K
KAMEZAWA Hiroyuki 已提交
775 776 777
void mem_cgroup_move_lists(struct page *page,
			   enum lru_list from, enum lru_list to)
{
778
	if (mem_cgroup_disabled())
K
KAMEZAWA Hiroyuki 已提交
779 780 781
		return;
	mem_cgroup_del_lru_list(page, from);
	mem_cgroup_add_lru_list(page, to);
782 783
}

784 785 786
int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem)
{
	int ret;
787
	struct mem_cgroup *curr = NULL;
788 789

	task_lock(task);
790 791 792
	rcu_read_lock();
	curr = try_get_mem_cgroup_from_mm(task->mm);
	rcu_read_unlock();
793
	task_unlock(task);
794 795
	if (!curr)
		return 0;
796 797 798 799 800 801 802
	/*
	 * We should check use_hierarchy of "mem" not "curr". Because checking
	 * use_hierarchy of "curr" here make this function true if hierarchy is
	 * enabled in "curr" and "curr" is a child of "mem" in *cgroup*
	 * hierarchy(even if use_hierarchy is disabled in "mem").
	 */
	if (mem->use_hierarchy)
803 804 805 806
		ret = css_is_ancestor(&curr->css, &mem->css);
	else
		ret = (curr == mem);
	css_put(&curr->css);
807 808 809
	return ret;
}

810 811 812 813 814
/*
 * 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
815 816 817 818 819 820 821
	int prev_priority;

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

	return prev_priority;
822 823 824 825
}

void mem_cgroup_note_reclaim_priority(struct mem_cgroup *mem, int priority)
{
K
KOSAKI Motohiro 已提交
826
	spin_lock(&mem->reclaim_param_lock);
827 828
	if (priority < mem->prev_priority)
		mem->prev_priority = priority;
K
KOSAKI Motohiro 已提交
829
	spin_unlock(&mem->reclaim_param_lock);
830 831 832 833
}

void mem_cgroup_record_reclaim_priority(struct mem_cgroup *mem, int priority)
{
K
KOSAKI Motohiro 已提交
834
	spin_lock(&mem->reclaim_param_lock);
835
	mem->prev_priority = priority;
K
KOSAKI Motohiro 已提交
836
	spin_unlock(&mem->reclaim_param_lock);
837 838
}

839
static int calc_inactive_ratio(struct mem_cgroup *memcg, unsigned long *present_pages)
840 841 842
{
	unsigned long active;
	unsigned long inactive;
843 844
	unsigned long gb;
	unsigned long inactive_ratio;
845

K
KAMEZAWA Hiroyuki 已提交
846 847
	inactive = mem_cgroup_get_local_zonestat(memcg, LRU_INACTIVE_ANON);
	active = mem_cgroup_get_local_zonestat(memcg, LRU_ACTIVE_ANON);
848

849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875
	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)
876 877 878 879 880
		return 1;

	return 0;
}

881 882 883 884 885 886 887 888 889 890 891
int mem_cgroup_inactive_file_is_low(struct mem_cgroup *memcg)
{
	unsigned long active;
	unsigned long inactive;

	inactive = mem_cgroup_get_local_zonestat(memcg, LRU_INACTIVE_FILE);
	active = mem_cgroup_get_local_zonestat(memcg, LRU_ACTIVE_FILE);

	return (active > inactive);
}

892 893 894 895 896 897 898 899 900 901 902
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
903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922
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);
923 924 925 926 927 928 929 930
	/*
	 * 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
931 932 933 934 935 936 937
	mz = page_cgroup_zoneinfo(pc);
	if (!mz)
		return NULL;

	return &mz->reclaim_stat;
}

938 939 940 941 942
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,
943
					int active, int file)
944 945 946 947 948 949
{
	unsigned long nr_taken = 0;
	struct page *page;
	unsigned long scan;
	LIST_HEAD(pc_list);
	struct list_head *src;
950
	struct page_cgroup *pc, *tmp;
951 952 953
	int nid = z->zone_pgdat->node_id;
	int zid = zone_idx(z);
	struct mem_cgroup_per_zone *mz;
954
	int lru = LRU_FILE * file + active;
955
	int ret;
956

957
	BUG_ON(!mem_cont);
958
	mz = mem_cgroup_zoneinfo(mem_cont, nid, zid);
959
	src = &mz->lists[lru];
960

961 962
	scan = 0;
	list_for_each_entry_safe_reverse(pc, tmp, src, lru) {
H
Hugh Dickins 已提交
963
		if (scan >= nr_to_scan)
964
			break;
K
KAMEZAWA Hiroyuki 已提交
965 966

		page = pc->page;
967 968
		if (unlikely(!PageCgroupUsed(pc)))
			continue;
H
Hugh Dickins 已提交
969
		if (unlikely(!PageLRU(page)))
970 971
			continue;

H
Hugh Dickins 已提交
972
		scan++;
973 974 975
		ret = __isolate_lru_page(page, mode, file);
		switch (ret) {
		case 0:
976
			list_move(&page->lru, dst);
977
			mem_cgroup_del_lru(page);
978
			nr_taken++;
979 980 981 982 983 984 985
			break;
		case -EBUSY:
			/* we don't affect global LRU but rotate in our LRU */
			mem_cgroup_rotate_lru_list(page, page_lru(page));
			break;
		default:
			break;
986 987 988 989 990 991 992
		}
	}

	*scanned = scan;
	return nr_taken;
}

993 994 995
#define mem_cgroup_from_res_counter(counter, member)	\
	container_of(counter, struct mem_cgroup, member)

996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
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
1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023
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;
}

1024 1025 1026 1027 1028 1029
static int mem_cgroup_count_children_cb(struct mem_cgroup *mem, void *data)
{
	int *val = data;
	(*val)++;
	return 0;
}
1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050

/**
 * 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;

1051
	if (!memcg || !p)
1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
		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));
}

1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108
/*
 * 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;
}

1109
/*
K
KAMEZAWA Hiroyuki 已提交
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
 * 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.
1152 1153
 *
 * root_mem is the original ancestor that we've been reclaim from.
K
KAMEZAWA Hiroyuki 已提交
1154 1155 1156
 *
 * We give up and return to the caller when we visit root_mem twice.
 * (other groups can be removed while we're walking....)
1157 1158
 *
 * If shrink==true, for avoiding to free too much, this returns immedieately.
1159 1160
 */
static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem,
1161
						struct zone *zone,
1162 1163
						gfp_t gfp_mask,
						unsigned long reclaim_options)
1164
{
K
KAMEZAWA Hiroyuki 已提交
1165 1166 1167
	struct mem_cgroup *victim;
	int ret, total = 0;
	int loop = 0;
1168 1169
	bool noswap = reclaim_options & MEM_CGROUP_RECLAIM_NOSWAP;
	bool shrink = reclaim_options & MEM_CGROUP_RECLAIM_SHRINK;
1170 1171
	bool check_soft = reclaim_options & MEM_CGROUP_RECLAIM_SOFT;
	unsigned long excess = mem_cgroup_get_excess(root_mem);
K
KAMEZAWA Hiroyuki 已提交
1172

1173 1174 1175 1176
	/* If memsw_is_minimum==1, swap-out is of-no-use. */
	if (root_mem->memsw_is_minimum)
		noswap = true;

1177
	while (1) {
K
KAMEZAWA Hiroyuki 已提交
1178
		victim = mem_cgroup_select_victim(root_mem);
1179
		if (victim == root_mem) {
K
KAMEZAWA Hiroyuki 已提交
1180
			loop++;
1181 1182
			if (loop >= 1)
				drain_all_stock_async();
1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205
			if (loop >= 2) {
				/*
				 * If we have not been able to reclaim
				 * anything, it might because there are
				 * no reclaimable pages under this hierarchy
				 */
				if (!check_soft || !total) {
					css_put(&victim->css);
					break;
				}
				/*
				 * We want to do more targetted reclaim.
				 * excess >> 2 is not to excessive so as to
				 * reclaim too much, nor too less that we keep
				 * coming back to reclaim from this cgroup
				 */
				if (total >= (excess >> 2) ||
					(loop > MEM_CGROUP_MAX_RECLAIM_LOOPS)) {
					css_put(&victim->css);
					break;
				}
			}
		}
K
KAMEZAWA Hiroyuki 已提交
1206 1207 1208
		if (!mem_cgroup_local_usage(&victim->stat)) {
			/* this cgroup's local usage == 0 */
			css_put(&victim->css);
1209 1210
			continue;
		}
K
KAMEZAWA Hiroyuki 已提交
1211
		/* we use swappiness of local cgroup */
1212 1213 1214 1215 1216 1217 1218
		if (check_soft)
			ret = mem_cgroup_shrink_node_zone(victim, gfp_mask,
				noswap, get_swappiness(victim), zone,
				zone->zone_pgdat->node_id);
		else
			ret = try_to_free_mem_cgroup_pages(victim, gfp_mask,
						noswap, get_swappiness(victim));
K
KAMEZAWA Hiroyuki 已提交
1219
		css_put(&victim->css);
1220 1221 1222 1223 1224 1225 1226
		/*
		 * 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
1227
		total += ret;
1228 1229 1230 1231
		if (check_soft) {
			if (res_counter_check_under_soft_limit(&root_mem->res))
				return total;
		} else if (mem_cgroup_check_under_limit(root_mem))
K
KAMEZAWA Hiroyuki 已提交
1232
			return 1 + total;
1233
	}
K
KAMEZAWA Hiroyuki 已提交
1234
	return total;
1235 1236
}

1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252
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;
}
1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264

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);
}

1265 1266 1267 1268
/*
 * Currently used to update mapped file statistics, but the routine can be
 * generalized to update other statistics as well.
 */
1269
void mem_cgroup_update_file_mapped(struct page *page, int val)
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
{
	struct mem_cgroup *mem;
	struct mem_cgroup_stat *stat;
	struct mem_cgroup_stat_cpu *cpustat;
	int cpu;
	struct page_cgroup *pc;

	pc = lookup_page_cgroup(page);
	if (unlikely(!pc))
		return;

	lock_page_cgroup(pc);
	mem = pc->mem_cgroup;
	if (!mem)
		goto done;

	if (!PageCgroupUsed(pc))
		goto done;

	/*
	 * Preemption is already disabled, we don't need get_cpu()
	 */
	cpu = smp_processor_id();
	stat = &mem->stat;
	cpustat = &stat->cpustat[cpu];

1296
	__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_FILE_MAPPED, val);
1297 1298 1299
done:
	unlock_page_cgroup(pc);
}
1300

1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427
/*
 * size of first charge trial. "32" comes from vmscan.c's magic value.
 * TODO: maybe necessary to use big numbers in big irons.
 */
#define CHARGE_SIZE	(32 * PAGE_SIZE)
struct memcg_stock_pcp {
	struct mem_cgroup *cached; /* this never be root cgroup */
	int charge;
	struct work_struct work;
};
static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock);
static atomic_t memcg_drain_count;

/*
 * Try to consume stocked charge on this cpu. If success, PAGE_SIZE is consumed
 * from local stock and true is returned. If the stock is 0 or charges from a
 * cgroup which is not current target, returns false. This stock will be
 * refilled.
 */
static bool consume_stock(struct mem_cgroup *mem)
{
	struct memcg_stock_pcp *stock;
	bool ret = true;

	stock = &get_cpu_var(memcg_stock);
	if (mem == stock->cached && stock->charge)
		stock->charge -= PAGE_SIZE;
	else /* need to call res_counter_charge */
		ret = false;
	put_cpu_var(memcg_stock);
	return ret;
}

/*
 * Returns stocks cached in percpu to res_counter and reset cached information.
 */
static void drain_stock(struct memcg_stock_pcp *stock)
{
	struct mem_cgroup *old = stock->cached;

	if (stock->charge) {
		res_counter_uncharge(&old->res, stock->charge);
		if (do_swap_account)
			res_counter_uncharge(&old->memsw, stock->charge);
	}
	stock->cached = NULL;
	stock->charge = 0;
}

/*
 * This must be called under preempt disabled or must be called by
 * a thread which is pinned to local cpu.
 */
static void drain_local_stock(struct work_struct *dummy)
{
	struct memcg_stock_pcp *stock = &__get_cpu_var(memcg_stock);
	drain_stock(stock);
}

/*
 * Cache charges(val) which is from res_counter, to local per_cpu area.
 * This will be consumed by consumt_stock() function, later.
 */
static void refill_stock(struct mem_cgroup *mem, int val)
{
	struct memcg_stock_pcp *stock = &get_cpu_var(memcg_stock);

	if (stock->cached != mem) { /* reset if necessary */
		drain_stock(stock);
		stock->cached = mem;
	}
	stock->charge += val;
	put_cpu_var(memcg_stock);
}

/*
 * Tries to drain stocked charges in other cpus. This function is asynchronous
 * and just put a work per cpu for draining localy on each cpu. Caller can
 * expects some charges will be back to res_counter later but cannot wait for
 * it.
 */
static void drain_all_stock_async(void)
{
	int cpu;
	/* This function is for scheduling "drain" in asynchronous way.
	 * The result of "drain" is not directly handled by callers. Then,
	 * if someone is calling drain, we don't have to call drain more.
	 * Anyway, WORK_STRUCT_PENDING check in queue_work_on() will catch if
	 * there is a race. We just do loose check here.
	 */
	if (atomic_read(&memcg_drain_count))
		return;
	/* Notify other cpus that system-wide "drain" is running */
	atomic_inc(&memcg_drain_count);
	get_online_cpus();
	for_each_online_cpu(cpu) {
		struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu);
		schedule_work_on(cpu, &stock->work);
	}
 	put_online_cpus();
	atomic_dec(&memcg_drain_count);
	/* We don't wait for flush_work */
}

/* This is a synchronous drain interface. */
static void drain_all_stock_sync(void)
{
	/* called when force_empty is called */
	atomic_inc(&memcg_drain_count);
	schedule_on_each_cpu(drain_local_stock);
	atomic_dec(&memcg_drain_count);
}

static int __cpuinit memcg_stock_cpu_callback(struct notifier_block *nb,
					unsigned long action,
					void *hcpu)
{
	int cpu = (unsigned long)hcpu;
	struct memcg_stock_pcp *stock;

	if (action != CPU_DEAD)
		return NOTIFY_OK;
	stock = &per_cpu(memcg_stock, cpu);
	drain_stock(stock);
	return NOTIFY_OK;
}

1428 1429 1430
/*
 * Unlike exported interface, "oom" parameter is added. if oom==true,
 * oom-killer can be invoked.
1431
 */
1432
static int __mem_cgroup_try_charge(struct mm_struct *mm,
1433
			gfp_t gfp_mask, struct mem_cgroup **memcg,
1434
			bool oom, struct page *page)
1435
{
1436
	struct mem_cgroup *mem, *mem_over_limit;
1437
	int nr_retries = MEM_CGROUP_RECLAIM_RETRIES;
1438
	struct res_counter *fail_res;
1439
	int csize = CHARGE_SIZE;
1440 1441 1442 1443 1444 1445 1446

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

1447
	/*
1448 1449
	 * We always charge the cgroup the mm_struct belongs to.
	 * The mm_struct's mem_cgroup changes on task migration if the
1450 1451 1452
	 * thread group leader migrates. It's possible that mm is not
	 * set, if so charge the init_mm (happens for pagecache usage).
	 */
1453 1454 1455
	mem = *memcg;
	if (likely(!mem)) {
		mem = try_get_mem_cgroup_from_mm(mm);
1456
		*memcg = mem;
1457
	} else {
1458
		css_get(&mem->css);
1459
	}
1460 1461 1462
	if (unlikely(!mem))
		return 0;

1463
	VM_BUG_ON(css_is_removed(&mem->css));
1464 1465
	if (mem_cgroup_is_root(mem))
		goto done;
1466

1467
	while (1) {
1468
		int ret = 0;
1469
		unsigned long flags = 0;
1470

1471 1472 1473 1474
		if (consume_stock(mem))
			goto charged;

		ret = res_counter_charge(&mem->res, csize, &fail_res);
1475 1476 1477
		if (likely(!ret)) {
			if (!do_swap_account)
				break;
1478
			ret = res_counter_charge(&mem->memsw, csize, &fail_res);
1479 1480 1481
			if (likely(!ret))
				break;
			/* mem+swap counter fails */
1482
			res_counter_uncharge(&mem->res, csize);
1483
			flags |= MEM_CGROUP_RECLAIM_NOSWAP;
1484 1485 1486 1487 1488 1489 1490
			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);

1491 1492 1493 1494 1495
		/* reduce request size and retry */
		if (csize > PAGE_SIZE) {
			csize = PAGE_SIZE;
			continue;
		}
1496
		if (!(gfp_mask & __GFP_WAIT))
1497
			goto nomem;
1498

1499 1500
		ret = mem_cgroup_hierarchical_reclaim(mem_over_limit, NULL,
						gfp_mask, flags);
1501 1502
		if (ret)
			continue;
1503 1504

		/*
1505 1506 1507 1508 1509
		 * 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
1510
		 *
1511
		 */
1512 1513
		if (mem_cgroup_check_under_limit(mem_over_limit))
			continue;
1514

1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556
		/* try to avoid oom while someone is moving charge */
		if (mc.moving_task && current != mc.moving_task) {
			struct mem_cgroup *from, *to;
			bool do_continue = false;
			/*
			 * There is a small race that "from" or "to" can be
			 * freed by rmdir, so we use css_tryget().
			 */
			rcu_read_lock();
			from = mc.from;
			to = mc.to;
			if (from && css_tryget(&from->css)) {
				if (mem_over_limit->use_hierarchy)
					do_continue = css_is_ancestor(
							&from->css,
							&mem_over_limit->css);
				else
					do_continue = (from == mem_over_limit);
				css_put(&from->css);
			}
			if (!do_continue && to && css_tryget(&to->css)) {
				if (mem_over_limit->use_hierarchy)
					do_continue = css_is_ancestor(
							&to->css,
							&mem_over_limit->css);
				else
					do_continue = (to == mem_over_limit);
				css_put(&to->css);
			}
			rcu_read_unlock();
			if (do_continue) {
				DEFINE_WAIT(wait);
				prepare_to_wait(&mc.waitq, &wait,
							TASK_INTERRUPTIBLE);
				/* moving charge context might have finished. */
				if (mc.moving_task)
					schedule();
				finish_wait(&mc.waitq, &wait);
				continue;
			}
		}

1557
		if (!nr_retries--) {
1558
			if (oom) {
1559
				mem_cgroup_out_of_memory(mem_over_limit, gfp_mask);
1560
				record_last_oom(mem_over_limit);
1561
			}
1562
			goto nomem;
1563
		}
1564
	}
1565 1566 1567
	if (csize > PAGE_SIZE)
		refill_stock(mem, csize - PAGE_SIZE);
charged:
1568
	/*
1569 1570
	 * Insert ancestor (and ancestor's ancestors), to softlimit RB-tree.
	 * if they exceeds softlimit.
1571
	 */
1572
	if (page && mem_cgroup_soft_limit_check(mem))
1573
		mem_cgroup_update_tree(mem, page);
1574
done:
1575 1576 1577 1578 1579
	return 0;
nomem:
	css_put(&mem->css);
	return -ENOMEM;
}
1580

1581 1582 1583 1584 1585
/*
 * Somemtimes we have to undo a charge we got by try_charge().
 * This function is for that and do uncharge, put css's refcnt.
 * gotten by try_charge().
 */
1586 1587
static void __mem_cgroup_cancel_charge(struct mem_cgroup *mem,
							unsigned long count)
1588 1589
{
	if (!mem_cgroup_is_root(mem)) {
1590
		res_counter_uncharge(&mem->res, PAGE_SIZE * count);
1591
		if (do_swap_account)
1592 1593 1594 1595
			res_counter_uncharge(&mem->memsw, PAGE_SIZE * count);
		VM_BUG_ON(test_bit(CSS_ROOT, &mem->css.flags));
		WARN_ON_ONCE(count > INT_MAX);
		__css_put(&mem->css, (int)count);
1596
	}
1597 1598 1599 1600 1601 1602
	/* we don't need css_put for root */
}

static void mem_cgroup_cancel_charge(struct mem_cgroup *mem)
{
	__mem_cgroup_cancel_charge(mem, 1);
1603 1604
}

1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623
/*
 * 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);
}

1624
struct mem_cgroup *try_get_mem_cgroup_from_page(struct page *page)
1625
{
1626
	struct mem_cgroup *mem = NULL;
1627
	struct page_cgroup *pc;
1628
	unsigned short id;
1629 1630
	swp_entry_t ent;

1631 1632 1633
	VM_BUG_ON(!PageLocked(page));

	pc = lookup_page_cgroup(page);
1634
	lock_page_cgroup(pc);
1635
	if (PageCgroupUsed(pc)) {
1636
		mem = pc->mem_cgroup;
1637 1638
		if (mem && !css_tryget(&mem->css))
			mem = NULL;
1639
	} else if (PageSwapCache(page)) {
1640
		ent.val = page_private(page);
1641 1642 1643 1644 1645 1646
		id = lookup_swap_cgroup(ent);
		rcu_read_lock();
		mem = mem_cgroup_lookup(id);
		if (mem && !css_tryget(&mem->css))
			mem = NULL;
		rcu_read_unlock();
1647
	}
1648
	unlock_page_cgroup(pc);
1649 1650 1651
	return mem;
}

1652
/*
1653
 * commit a charge got by __mem_cgroup_try_charge() and makes page_cgroup to be
1654 1655 1656 1657 1658 1659 1660 1661 1662 1663
 * 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;
1664 1665 1666 1667

	lock_page_cgroup(pc);
	if (unlikely(PageCgroupUsed(pc))) {
		unlock_page_cgroup(pc);
1668
		mem_cgroup_cancel_charge(mem);
1669
		return;
1670
	}
1671

1672
	pc->mem_cgroup = mem;
1673 1674 1675 1676 1677 1678 1679
	/*
	 * We access a page_cgroup asynchronously without lock_page_cgroup().
	 * Especially when a page_cgroup is taken from a page, pc->mem_cgroup
	 * is accessed after testing USED bit. To make pc->mem_cgroup visible
	 * before USED bit, we need memory barrier here.
	 * See mem_cgroup_add_lru_list(), etc.
 	 */
K
KAMEZAWA Hiroyuki 已提交
1680
	smp_wmb();
1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693
	switch (ctype) {
	case MEM_CGROUP_CHARGE_TYPE_CACHE:
	case MEM_CGROUP_CHARGE_TYPE_SHMEM:
		SetPageCgroupCache(pc);
		SetPageCgroupUsed(pc);
		break;
	case MEM_CGROUP_CHARGE_TYPE_MAPPED:
		ClearPageCgroupCache(pc);
		SetPageCgroupUsed(pc);
		break;
	default:
		break;
	}
1694

K
KAMEZAWA Hiroyuki 已提交
1695
	mem_cgroup_charge_statistics(mem, pc, true);
1696 1697

	unlock_page_cgroup(pc);
1698
}
1699

1700
/**
1701
 * __mem_cgroup_move_account - move account of the page
1702 1703 1704
 * @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.
1705
 * @uncharge: whether we should call uncharge and css_put against @from.
1706 1707
 *
 * The caller must confirm following.
K
KAMEZAWA Hiroyuki 已提交
1708
 * - page is not on LRU (isolate_page() is useful.)
1709
 * - the pc is locked, used, and ->mem_cgroup points to @from.
1710
 *
1711 1712 1713 1714
 * This function doesn't do "charge" nor css_get to new cgroup. It should be
 * done by a caller(__mem_cgroup_try_charge would be usefull). If @uncharge is
 * true, this function does "uncharge" from old cgroup, but it doesn't if
 * @uncharge is false, so a caller should do "uncharge".
1715 1716
 */

1717
static void __mem_cgroup_move_account(struct page_cgroup *pc,
1718
	struct mem_cgroup *from, struct mem_cgroup *to, bool uncharge)
1719
{
1720 1721 1722 1723
	struct page *page;
	int cpu;
	struct mem_cgroup_stat *stat;
	struct mem_cgroup_stat_cpu *cpustat;
1724 1725

	VM_BUG_ON(from == to);
K
KAMEZAWA Hiroyuki 已提交
1726
	VM_BUG_ON(PageLRU(pc->page));
1727 1728 1729
	VM_BUG_ON(!PageCgroupLocked(pc));
	VM_BUG_ON(!PageCgroupUsed(pc));
	VM_BUG_ON(pc->mem_cgroup != from);
1730

1731
	page = pc->page;
1732
	if (page_mapped(page) && !PageAnon(page)) {
1733 1734 1735 1736
		cpu = smp_processor_id();
		/* Update mapped_file data for mem_cgroup "from" */
		stat = &from->stat;
		cpustat = &stat->cpustat[cpu];
1737
		__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_FILE_MAPPED,
1738 1739 1740 1741 1742
						-1);

		/* Update mapped_file data for mem_cgroup "to" */
		stat = &to->stat;
		cpustat = &stat->cpustat[cpu];
1743
		__mem_cgroup_stat_add_safe(cpustat, MEM_CGROUP_STAT_FILE_MAPPED,
1744 1745
						1);
	}
1746 1747 1748 1749
	mem_cgroup_charge_statistics(from, pc, false);
	if (uncharge)
		/* This is not "cancel", but cancel_charge does all we need. */
		mem_cgroup_cancel_charge(from);
1750

1751
	/* caller should have done css_get */
K
KAMEZAWA Hiroyuki 已提交
1752 1753
	pc->mem_cgroup = to;
	mem_cgroup_charge_statistics(to, pc, true);
1754 1755 1756
	/*
	 * We charges against "to" which may not have any tasks. Then, "to"
	 * can be under rmdir(). But in current implementation, caller of
1757 1758 1759
	 * this function is just force_empty() and move charge, so it's
	 * garanteed that "to" is never removed. So, we don't check rmdir
	 * status here.
1760
	 */
1761 1762 1763 1764 1765 1766 1767
}

/*
 * check whether the @pc is valid for moving account and call
 * __mem_cgroup_move_account()
 */
static int mem_cgroup_move_account(struct page_cgroup *pc,
1768
		struct mem_cgroup *from, struct mem_cgroup *to, bool uncharge)
1769 1770 1771 1772
{
	int ret = -EINVAL;
	lock_page_cgroup(pc);
	if (PageCgroupUsed(pc) && pc->mem_cgroup == from) {
1773
		__mem_cgroup_move_account(pc, from, to, uncharge);
1774 1775 1776
		ret = 0;
	}
	unlock_page_cgroup(pc);
1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787
	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
1788
	struct page *page = pc->page;
1789 1790 1791 1792 1793 1794 1795 1796 1797
	struct cgroup *cg = child->css.cgroup;
	struct cgroup *pcg = cg->parent;
	struct mem_cgroup *parent;
	int ret;

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

1798 1799 1800 1801 1802
	ret = -EBUSY;
	if (!get_page_unless_zero(page))
		goto out;
	if (isolate_lru_page(page))
		goto put;
K
KAMEZAWA Hiroyuki 已提交
1803

1804
	parent = mem_cgroup_from_cont(pcg);
1805
	ret = __mem_cgroup_try_charge(NULL, gfp_mask, &parent, false, page);
1806
	if (ret || !parent)
1807
		goto put_back;
1808

1809 1810 1811
	ret = mem_cgroup_move_account(pc, child, parent, true);
	if (ret)
		mem_cgroup_cancel_charge(parent);
1812
put_back:
K
KAMEZAWA Hiroyuki 已提交
1813
	putback_lru_page(page);
1814
put:
1815
	put_page(page);
1816
out:
1817 1818 1819
	return ret;
}

1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840
/*
 * 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;
1841
	ret = __mem_cgroup_try_charge(mm, gfp_mask, &mem, true, page);
1842
	if (ret || !mem)
1843 1844 1845
		return ret;

	__mem_cgroup_commit_charge(mem, pc, ctype);
1846 1847 1848
	return 0;
}

1849 1850
int mem_cgroup_newpage_charge(struct page *page,
			      struct mm_struct *mm, gfp_t gfp_mask)
1851
{
1852
	if (mem_cgroup_disabled())
1853
		return 0;
1854 1855
	if (PageCompound(page))
		return 0;
1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866
	/*
	 * 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;
1867
	return mem_cgroup_charge_common(page, mm, gfp_mask,
1868
				MEM_CGROUP_CHARGE_TYPE_MAPPED, NULL);
1869 1870
}

D
Daisuke Nishimura 已提交
1871 1872 1873 1874
static void
__mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr,
					enum charge_type ctype);

1875 1876
int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
				gfp_t gfp_mask)
1877
{
1878 1879 1880
	struct mem_cgroup *mem = NULL;
	int ret;

1881
	if (mem_cgroup_disabled())
1882
		return 0;
1883 1884
	if (PageCompound(page))
		return 0;
1885 1886 1887 1888 1889 1890 1891 1892
	/*
	 * 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.)
1893 1894
	 * And when the page is SwapCache, it should take swap information
	 * into account. This is under lock_page() now.
1895 1896 1897 1898
	 */
	if (!(gfp_mask & __GFP_WAIT)) {
		struct page_cgroup *pc;

1899 1900 1901 1902 1903 1904 1905

		pc = lookup_page_cgroup(page);
		if (!pc)
			return 0;
		lock_page_cgroup(pc);
		if (PageCgroupUsed(pc)) {
			unlock_page_cgroup(pc);
1906 1907
			return 0;
		}
1908
		unlock_page_cgroup(pc);
1909 1910
	}

1911
	if (unlikely(!mm && !mem))
1912
		mm = &init_mm;
1913

1914 1915
	if (page_is_file_cache(page))
		return mem_cgroup_charge_common(page, mm, gfp_mask,
1916
				MEM_CGROUP_CHARGE_TYPE_CACHE, NULL);
1917

D
Daisuke Nishimura 已提交
1918 1919 1920 1921 1922 1923 1924 1925 1926
	/* 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);
1927 1928

	return ret;
1929 1930
}

1931 1932 1933
/*
 * While swap-in, try_charge -> commit or cancel, the page is locked.
 * And when try_charge() successfully returns, one refcnt to memcg without
1934
 * struct page_cgroup is acquired. This refcnt will be consumed by
1935 1936
 * "commit()" or removed by "cancel()"
 */
1937 1938 1939 1940 1941
int mem_cgroup_try_charge_swapin(struct mm_struct *mm,
				 struct page *page,
				 gfp_t mask, struct mem_cgroup **ptr)
{
	struct mem_cgroup *mem;
1942
	int ret;
1943

1944
	if (mem_cgroup_disabled())
1945 1946 1947 1948 1949 1950
		return 0;

	if (!do_swap_account)
		goto charge_cur_mm;
	/*
	 * A racing thread's fault, or swapoff, may have already updated
H
Hugh Dickins 已提交
1951 1952 1953
	 * the pte, and even removed page from swap cache: in those cases
	 * do_swap_page()'s pte_same() test will fail; but there's also a
	 * KSM case which does need to charge the page.
1954 1955
	 */
	if (!PageSwapCache(page))
H
Hugh Dickins 已提交
1956
		goto charge_cur_mm;
1957
	mem = try_get_mem_cgroup_from_page(page);
1958 1959
	if (!mem)
		goto charge_cur_mm;
1960
	*ptr = mem;
1961
	ret = __mem_cgroup_try_charge(NULL, mask, ptr, true, page);
1962 1963 1964
	/* drop extra refcnt from tryget */
	css_put(&mem->css);
	return ret;
1965 1966 1967
charge_cur_mm:
	if (unlikely(!mm))
		mm = &init_mm;
1968
	return __mem_cgroup_try_charge(mm, mask, ptr, true, page);
1969 1970
}

D
Daisuke Nishimura 已提交
1971 1972 1973
static void
__mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr,
					enum charge_type ctype)
1974 1975 1976
{
	struct page_cgroup *pc;

1977
	if (mem_cgroup_disabled())
1978 1979 1980
		return;
	if (!ptr)
		return;
1981
	cgroup_exclude_rmdir(&ptr->css);
1982
	pc = lookup_page_cgroup(page);
1983
	mem_cgroup_lru_del_before_commit_swapcache(page);
D
Daisuke Nishimura 已提交
1984
	__mem_cgroup_commit_charge(ptr, pc, ctype);
1985
	mem_cgroup_lru_add_after_commit_swapcache(page);
1986 1987 1988
	/*
	 * Now swap is on-memory. This means this page may be
	 * counted both as mem and swap....double count.
1989 1990 1991
	 * 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.
1992
	 */
1993
	if (do_swap_account && PageSwapCache(page)) {
1994
		swp_entry_t ent = {.val = page_private(page)};
1995
		unsigned short id;
1996
		struct mem_cgroup *memcg;
1997 1998 1999 2000

		id = swap_cgroup_record(ent, 0);
		rcu_read_lock();
		memcg = mem_cgroup_lookup(id);
2001
		if (memcg) {
2002 2003 2004 2005
			/*
			 * This recorded memcg can be obsolete one. So, avoid
			 * calling css_tryget
			 */
2006
			if (!mem_cgroup_is_root(memcg))
2007
				res_counter_uncharge(&memcg->memsw, PAGE_SIZE);
2008
			mem_cgroup_swap_statistics(memcg, false);
2009 2010
			mem_cgroup_put(memcg);
		}
2011
		rcu_read_unlock();
2012
	}
2013 2014 2015 2016 2017 2018
	/*
	 * At swapin, we may charge account against cgroup which has no tasks.
	 * So, rmdir()->pre_destroy() can be called while we do this charge.
	 * In that case, we need to call pre_destroy() again. check it here.
	 */
	cgroup_release_and_wakeup_rmdir(&ptr->css);
2019 2020
}

D
Daisuke Nishimura 已提交
2021 2022 2023 2024 2025 2026
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);
}

2027 2028
void mem_cgroup_cancel_charge_swapin(struct mem_cgroup *mem)
{
2029
	if (mem_cgroup_disabled())
2030 2031 2032
		return;
	if (!mem)
		return;
2033
	mem_cgroup_cancel_charge(mem);
2034 2035
}

2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
static void
__do_uncharge(struct mem_cgroup *mem, const enum charge_type ctype)
{
	struct memcg_batch_info *batch = NULL;
	bool uncharge_memsw = true;
	/* If swapout, usage of swap doesn't decrease */
	if (!do_swap_account || ctype == MEM_CGROUP_CHARGE_TYPE_SWAPOUT)
		uncharge_memsw = false;
	/*
	 * do_batch > 0 when unmapping pages or inode invalidate/truncate.
	 * In those cases, all pages freed continously can be expected to be in
	 * the same cgroup and we have chance to coalesce uncharges.
	 * But we do uncharge one by one if this is killed by OOM(TIF_MEMDIE)
	 * because we want to do uncharge as soon as possible.
	 */
	if (!current->memcg_batch.do_batch || test_thread_flag(TIF_MEMDIE))
		goto direct_uncharge;

	batch = &current->memcg_batch;
	/*
	 * In usual, we do css_get() when we remember memcg pointer.
	 * But in this case, we keep res->usage until end of a series of
	 * uncharges. Then, it's ok to ignore memcg's refcnt.
	 */
	if (!batch->memcg)
		batch->memcg = mem;
	/*
	 * In typical case, batch->memcg == mem. This means we can
	 * merge a series of uncharges to an uncharge of res_counter.
	 * If not, we uncharge res_counter ony by one.
	 */
	if (batch->memcg != mem)
		goto direct_uncharge;
	/* remember freed charge and uncharge it later */
	batch->bytes += PAGE_SIZE;
	if (uncharge_memsw)
		batch->memsw_bytes += PAGE_SIZE;
	return;
direct_uncharge:
	res_counter_uncharge(&mem->res, PAGE_SIZE);
	if (uncharge_memsw)
		res_counter_uncharge(&mem->memsw, PAGE_SIZE);
	return;
}
2080

2081
/*
2082
 * uncharge if !page_mapped(page)
2083
 */
2084
static struct mem_cgroup *
2085
__mem_cgroup_uncharge_common(struct page *page, enum charge_type ctype)
2086
{
H
Hugh Dickins 已提交
2087
	struct page_cgroup *pc;
2088
	struct mem_cgroup *mem = NULL;
2089
	struct mem_cgroup_per_zone *mz;
2090

2091
	if (mem_cgroup_disabled())
2092
		return NULL;
2093

K
KAMEZAWA Hiroyuki 已提交
2094
	if (PageSwapCache(page))
2095
		return NULL;
K
KAMEZAWA Hiroyuki 已提交
2096

2097
	/*
2098
	 * Check if our page_cgroup is valid
2099
	 */
2100 2101
	pc = lookup_page_cgroup(page);
	if (unlikely(!pc || !PageCgroupUsed(pc)))
2102
		return NULL;
2103

2104
	lock_page_cgroup(pc);
K
KAMEZAWA Hiroyuki 已提交
2105

2106 2107
	mem = pc->mem_cgroup;

K
KAMEZAWA Hiroyuki 已提交
2108 2109 2110 2111 2112
	if (!PageCgroupUsed(pc))
		goto unlock_out;

	switch (ctype) {
	case MEM_CGROUP_CHARGE_TYPE_MAPPED:
K
KAMEZAWA Hiroyuki 已提交
2113
	case MEM_CGROUP_CHARGE_TYPE_DROP:
K
KAMEZAWA Hiroyuki 已提交
2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125
		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;
2126
	}
K
KAMEZAWA Hiroyuki 已提交
2127

2128 2129
	if (!mem_cgroup_is_root(mem))
		__do_uncharge(mem, ctype);
2130 2131
	if (ctype == MEM_CGROUP_CHARGE_TYPE_SWAPOUT)
		mem_cgroup_swap_statistics(mem, true);
K
KAMEZAWA Hiroyuki 已提交
2132
	mem_cgroup_charge_statistics(mem, pc, false);
K
KAMEZAWA Hiroyuki 已提交
2133

2134
	ClearPageCgroupUsed(pc);
2135 2136 2137 2138 2139 2140
	/*
	 * 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.
	 */
2141

2142
	mz = page_cgroup_zoneinfo(pc);
2143
	unlock_page_cgroup(pc);
H
Hugh Dickins 已提交
2144

2145
	if (mem_cgroup_soft_limit_check(mem))
2146
		mem_cgroup_update_tree(mem, page);
K
KAMEZAWA Hiroyuki 已提交
2147 2148 2149
	/* at swapout, this memcg will be accessed to record to swap */
	if (ctype != MEM_CGROUP_CHARGE_TYPE_SWAPOUT)
		css_put(&mem->css);
2150

2151
	return mem;
K
KAMEZAWA Hiroyuki 已提交
2152 2153 2154

unlock_out:
	unlock_page_cgroup(pc);
2155
	return NULL;
2156 2157
}

2158 2159
void mem_cgroup_uncharge_page(struct page *page)
{
2160 2161 2162 2163 2164
	/* early check. */
	if (page_mapped(page))
		return;
	if (page->mapping && !PageAnon(page))
		return;
2165 2166 2167 2168 2169 2170
	__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));
2171
	VM_BUG_ON(page->mapping);
2172 2173 2174
	__mem_cgroup_uncharge_common(page, MEM_CGROUP_CHARGE_TYPE_CACHE);
}

2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218
/*
 * Batch_start/batch_end is called in unmap_page_range/invlidate/trucate.
 * In that cases, pages are freed continuously and we can expect pages
 * are in the same memcg. All these calls itself limits the number of
 * pages freed at once, then uncharge_start/end() is called properly.
 * This may be called prural(2) times in a context,
 */

void mem_cgroup_uncharge_start(void)
{
	current->memcg_batch.do_batch++;
	/* We can do nest. */
	if (current->memcg_batch.do_batch == 1) {
		current->memcg_batch.memcg = NULL;
		current->memcg_batch.bytes = 0;
		current->memcg_batch.memsw_bytes = 0;
	}
}

void mem_cgroup_uncharge_end(void)
{
	struct memcg_batch_info *batch = &current->memcg_batch;

	if (!batch->do_batch)
		return;

	batch->do_batch--;
	if (batch->do_batch) /* If stacked, do nothing. */
		return;

	if (!batch->memcg)
		return;
	/*
	 * This "batch->memcg" is valid without any css_get/put etc...
	 * bacause we hide charges behind us.
	 */
	if (batch->bytes)
		res_counter_uncharge(&batch->memcg->res, batch->bytes);
	if (batch->memsw_bytes)
		res_counter_uncharge(&batch->memcg->memsw, batch->memsw_bytes);
	/* forget this pointer (for sanity check) */
	batch->memcg = NULL;
}

2219
#ifdef CONFIG_SWAP
2220
/*
2221
 * called after __delete_from_swap_cache() and drop "page" account.
2222 2223
 * memcg information is recorded to swap_cgroup of "ent"
 */
K
KAMEZAWA Hiroyuki 已提交
2224 2225
void
mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
2226 2227
{
	struct mem_cgroup *memcg;
K
KAMEZAWA Hiroyuki 已提交
2228 2229 2230 2231 2232 2233
	int ctype = MEM_CGROUP_CHARGE_TYPE_SWAPOUT;

	if (!swapout) /* this was a swap cache but the swap is unused ! */
		ctype = MEM_CGROUP_CHARGE_TYPE_DROP;

	memcg = __mem_cgroup_uncharge_common(page, ctype);
2234 2235

	/* record memcg information */
K
KAMEZAWA Hiroyuki 已提交
2236
	if (do_swap_account && swapout && memcg) {
2237
		swap_cgroup_record(ent, css_id(&memcg->css));
2238 2239
		mem_cgroup_get(memcg);
	}
K
KAMEZAWA Hiroyuki 已提交
2240
	if (swapout && memcg)
K
KAMEZAWA Hiroyuki 已提交
2241
		css_put(&memcg->css);
2242
}
2243
#endif
2244 2245 2246 2247 2248 2249 2250

#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
2251
{
2252
	struct mem_cgroup *memcg;
2253
	unsigned short id;
2254 2255 2256 2257

	if (!do_swap_account)
		return;

2258 2259 2260
	id = swap_cgroup_record(ent, 0);
	rcu_read_lock();
	memcg = mem_cgroup_lookup(id);
2261
	if (memcg) {
2262 2263 2264 2265
		/*
		 * We uncharge this because swap is freed.
		 * This memcg can be obsolete one. We avoid calling css_tryget
		 */
2266
		if (!mem_cgroup_is_root(memcg))
2267
			res_counter_uncharge(&memcg->memsw, PAGE_SIZE);
2268
		mem_cgroup_swap_statistics(memcg, false);
2269 2270
		mem_cgroup_put(memcg);
	}
2271
	rcu_read_unlock();
K
KAMEZAWA Hiroyuki 已提交
2272
}
2273
#endif
K
KAMEZAWA Hiroyuki 已提交
2274

2275
/*
2276 2277
 * Before starting migration, account PAGE_SIZE to mem_cgroup that the old
 * page belongs to.
2278
 */
2279
int mem_cgroup_prepare_migration(struct page *page, struct mem_cgroup **ptr)
2280 2281
{
	struct page_cgroup *pc;
2282 2283
	struct mem_cgroup *mem = NULL;
	int ret = 0;
2284

2285
	if (mem_cgroup_disabled())
2286 2287
		return 0;

2288 2289 2290
	pc = lookup_page_cgroup(page);
	lock_page_cgroup(pc);
	if (PageCgroupUsed(pc)) {
2291 2292 2293
		mem = pc->mem_cgroup;
		css_get(&mem->css);
	}
2294
	unlock_page_cgroup(pc);
2295

2296
	if (mem) {
2297 2298
		ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, &mem, false,
						page);
2299 2300
		css_put(&mem->css);
	}
2301
	*ptr = mem;
2302
	return ret;
2303
}
2304

2305
/* remove redundant charge if migration failed*/
2306 2307
void mem_cgroup_end_migration(struct mem_cgroup *mem,
		struct page *oldpage, struct page *newpage)
2308
{
2309 2310 2311 2312 2313 2314
	struct page *target, *unused;
	struct page_cgroup *pc;
	enum charge_type ctype;

	if (!mem)
		return;
2315
	cgroup_exclude_rmdir(&mem->css);
2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332
	/* 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
2333
	if (unused)
2334 2335 2336
		__mem_cgroup_uncharge_common(unused, ctype);

	pc = lookup_page_cgroup(target);
2337
	/*
2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351
	 * __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.
2352
	 */
2353 2354
	if (ctype == MEM_CGROUP_CHARGE_TYPE_MAPPED)
		mem_cgroup_uncharge_page(target);
2355 2356 2357 2358 2359 2360
	/*
	 * At migration, we may charge account against cgroup which has no tasks
	 * So, rmdir()->pre_destroy() can be called while we do this charge.
	 * In that case, we need to call pre_destroy() again. check it here.
	 */
	cgroup_release_and_wakeup_rmdir(&mem->css);
2361
}
2362

2363
/*
2364 2365 2366 2367 2368 2369
 * 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.
2370
 */
2371
int mem_cgroup_shmem_charge_fallback(struct page *page,
2372 2373
			    struct mm_struct *mm,
			    gfp_t gfp_mask)
2374
{
2375
	struct mem_cgroup *mem = NULL;
2376
	int ret;
2377

2378
	if (mem_cgroup_disabled())
2379
		return 0;
2380

2381 2382 2383
	ret = mem_cgroup_try_charge_swapin(mm, page, gfp_mask, &mem);
	if (!ret)
		mem_cgroup_cancel_charge_swapin(mem); /* it does !mem check */
2384

2385
	return ret;
2386 2387
}

2388 2389
static DEFINE_MUTEX(set_limit_mutex);

2390
static int mem_cgroup_resize_limit(struct mem_cgroup *memcg,
2391
				unsigned long long val)
2392
{
2393
	int retry_count;
2394
	u64 memswlimit;
2395
	int ret = 0;
2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406
	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);
2407

2408
	while (retry_count) {
2409 2410 2411 2412
		if (signal_pending(current)) {
			ret = -EINTR;
			break;
		}
2413 2414 2415 2416 2417 2418 2419 2420 2421 2422
		/*
		 * 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);
2423 2424
			break;
		}
2425
		ret = res_counter_set_limit(&memcg->res, val);
2426 2427 2428 2429 2430 2431
		if (!ret) {
			if (memswlimit == val)
				memcg->memsw_is_minimum = true;
			else
				memcg->memsw_is_minimum = false;
		}
2432 2433 2434 2435 2436
		mutex_unlock(&set_limit_mutex);

		if (!ret)
			break;

2437
		mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL,
2438
						MEM_CGROUP_RECLAIM_SHRINK);
2439 2440 2441 2442 2443 2444
		curusage = res_counter_read_u64(&memcg->res, RES_USAGE);
		/* Usage is reduced ? */
  		if (curusage >= oldusage)
			retry_count--;
		else
			oldusage = curusage;
2445
	}
2446

2447 2448 2449
	return ret;
}

L
Li Zefan 已提交
2450 2451
static int mem_cgroup_resize_memsw_limit(struct mem_cgroup *memcg,
					unsigned long long val)
2452
{
2453
	int retry_count;
2454
	u64 memlimit, oldusage, curusage;
2455 2456
	int children = mem_cgroup_count_children(memcg);
	int ret = -EBUSY;
2457

2458 2459 2460
	/* see mem_cgroup_resize_res_limit */
 	retry_count = children * MEM_CGROUP_RECLAIM_RETRIES;
	oldusage = res_counter_read_u64(&memcg->memsw, RES_USAGE);
2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478
	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);
2479 2480 2481 2482 2483 2484
		if (!ret) {
			if (memlimit == val)
				memcg->memsw_is_minimum = true;
			else
				memcg->memsw_is_minimum = false;
		}
2485 2486 2487 2488 2489
		mutex_unlock(&set_limit_mutex);

		if (!ret)
			break;

2490
		mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL,
2491 2492
						MEM_CGROUP_RECLAIM_NOSWAP |
						MEM_CGROUP_RECLAIM_SHRINK);
2493
		curusage = res_counter_read_u64(&memcg->memsw, RES_USAGE);
2494
		/* Usage is reduced ? */
2495
		if (curusage >= oldusage)
2496
			retry_count--;
2497 2498
		else
			oldusage = curusage;
2499 2500 2501 2502
	}
	return ret;
}

2503 2504 2505 2506 2507 2508 2509 2510 2511
unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
						gfp_t gfp_mask, int nid,
						int zid)
{
	unsigned long nr_reclaimed = 0;
	struct mem_cgroup_per_zone *mz, *next_mz = NULL;
	unsigned long reclaimed;
	int loop = 0;
	struct mem_cgroup_tree_per_zone *mctz;
2512
	unsigned long long excess;
2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564

	if (order > 0)
		return 0;

	mctz = soft_limit_tree_node_zone(nid, zid);
	/*
	 * This loop can run a while, specially if mem_cgroup's continuously
	 * keep exceeding their soft limit and putting the system under
	 * pressure
	 */
	do {
		if (next_mz)
			mz = next_mz;
		else
			mz = mem_cgroup_largest_soft_limit_node(mctz);
		if (!mz)
			break;

		reclaimed = mem_cgroup_hierarchical_reclaim(mz->mem, zone,
						gfp_mask,
						MEM_CGROUP_RECLAIM_SOFT);
		nr_reclaimed += reclaimed;
		spin_lock(&mctz->lock);

		/*
		 * If we failed to reclaim anything from this memory cgroup
		 * it is time to move on to the next cgroup
		 */
		next_mz = NULL;
		if (!reclaimed) {
			do {
				/*
				 * Loop until we find yet another one.
				 *
				 * By the time we get the soft_limit lock
				 * again, someone might have aded the
				 * group back on the RB tree. Iterate to
				 * make sure we get a different mem.
				 * mem_cgroup_largest_soft_limit_node returns
				 * NULL if no other cgroup is present on
				 * the tree
				 */
				next_mz =
				__mem_cgroup_largest_soft_limit_node(mctz);
				if (next_mz == mz) {
					css_put(&next_mz->mem->css);
					next_mz = NULL;
				} else /* next_mz == NULL or other memcg */
					break;
			} while (1);
		}
		__mem_cgroup_remove_exceeded(mz->mem, mz, mctz);
2565
		excess = res_counter_soft_limit_excess(&mz->mem->res);
2566 2567 2568 2569 2570 2571 2572 2573
		/*
		 * One school of thought says that we should not add
		 * back the node to the tree if reclaim returns 0.
		 * But our reclaim could return 0, simply because due
		 * to priority we are exposing a smaller subset of
		 * memory to reclaim from. Consider this as a longer
		 * term TODO.
		 */
2574 2575
		/* If excess == 0, no tree ops */
		__mem_cgroup_insert_exceeded(mz->mem, mz, mctz, excess);
2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593
		spin_unlock(&mctz->lock);
		css_put(&mz->mem->css);
		loop++;
		/*
		 * Could not reclaim anything and there are no more
		 * mem cgroups to try or we seem to be looping without
		 * reclaiming anything.
		 */
		if (!nr_reclaimed &&
			(next_mz == NULL ||
			loop > MEM_CGROUP_MAX_SOFT_LIMIT_RECLAIM_LOOPS))
			break;
	} while (!nr_reclaimed);
	if (next_mz)
		css_put(&next_mz->mem->css);
	return nr_reclaimed;
}

2594 2595 2596 2597
/*
 * This routine traverse page_cgroup in given list and drop them all.
 * *And* this routine doesn't reclaim page itself, just removes page_cgroup.
 */
2598
static int mem_cgroup_force_empty_list(struct mem_cgroup *mem,
K
KAMEZAWA Hiroyuki 已提交
2599
				int node, int zid, enum lru_list lru)
2600
{
K
KAMEZAWA Hiroyuki 已提交
2601 2602
	struct zone *zone;
	struct mem_cgroup_per_zone *mz;
2603
	struct page_cgroup *pc, *busy;
K
KAMEZAWA Hiroyuki 已提交
2604
	unsigned long flags, loop;
2605
	struct list_head *list;
2606
	int ret = 0;
2607

K
KAMEZAWA Hiroyuki 已提交
2608 2609
	zone = &NODE_DATA(node)->node_zones[zid];
	mz = mem_cgroup_zoneinfo(mem, node, zid);
2610
	list = &mz->lists[lru];
2611

2612 2613 2614 2615 2616 2617
	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
2618
		spin_lock_irqsave(&zone->lru_lock, flags);
2619
		if (list_empty(list)) {
K
KAMEZAWA Hiroyuki 已提交
2620
			spin_unlock_irqrestore(&zone->lru_lock, flags);
2621
			break;
2622 2623 2624 2625
		}
		pc = list_entry(list->prev, struct page_cgroup, lru);
		if (busy == pc) {
			list_move(&pc->lru, list);
2626
			busy = NULL;
K
KAMEZAWA Hiroyuki 已提交
2627
			spin_unlock_irqrestore(&zone->lru_lock, flags);
2628 2629
			continue;
		}
K
KAMEZAWA Hiroyuki 已提交
2630
		spin_unlock_irqrestore(&zone->lru_lock, flags);
2631

K
KAMEZAWA Hiroyuki 已提交
2632
		ret = mem_cgroup_move_parent(pc, mem, GFP_KERNEL);
2633
		if (ret == -ENOMEM)
2634
			break;
2635 2636 2637 2638 2639 2640 2641

		if (ret == -EBUSY || ret == -EINVAL) {
			/* found lock contention or "pc" is obsolete. */
			busy = pc;
			cond_resched();
		} else
			busy = NULL;
2642
	}
K
KAMEZAWA Hiroyuki 已提交
2643

2644 2645 2646
	if (!ret && !list_empty(list))
		return -EBUSY;
	return ret;
2647 2648 2649 2650 2651 2652
}

/*
 * make mem_cgroup's charge to be 0 if there is no task.
 * This enables deleting this mem_cgroup.
 */
2653
static int mem_cgroup_force_empty(struct mem_cgroup *mem, bool free_all)
2654
{
2655 2656 2657
	int ret;
	int node, zid, shrink;
	int nr_retries = MEM_CGROUP_RECLAIM_RETRIES;
2658
	struct cgroup *cgrp = mem->css.cgroup;
2659

2660
	css_get(&mem->css);
2661 2662

	shrink = 0;
2663 2664 2665
	/* should free all ? */
	if (free_all)
		goto try_to_free;
2666
move_account:
2667
	do {
2668
		ret = -EBUSY;
2669 2670 2671 2672
		if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children))
			goto out;
		ret = -EINTR;
		if (signal_pending(current))
2673
			goto out;
2674 2675
		/* This is for making all *used* pages to be on LRU. */
		lru_add_drain_all();
2676
		drain_all_stock_sync();
2677
		ret = 0;
2678
		for_each_node_state(node, N_HIGH_MEMORY) {
2679
			for (zid = 0; !ret && zid < MAX_NR_ZONES; zid++) {
2680
				enum lru_list l;
2681 2682
				for_each_lru(l) {
					ret = mem_cgroup_force_empty_list(mem,
K
KAMEZAWA Hiroyuki 已提交
2683
							node, zid, l);
2684 2685 2686
					if (ret)
						break;
				}
2687
			}
2688 2689 2690 2691 2692 2693
			if (ret)
				break;
		}
		/* it seems parent cgroup doesn't have enough mem */
		if (ret == -ENOMEM)
			goto try_to_free;
2694
		cond_resched();
2695 2696
	/* "ret" should also be checked to ensure all lists are empty. */
	} while (mem->res.usage > 0 || ret);
2697 2698 2699
out:
	css_put(&mem->css);
	return ret;
2700 2701

try_to_free:
2702 2703
	/* returns EBUSY if there is a task or if we come here twice. */
	if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children) || shrink) {
2704 2705 2706
		ret = -EBUSY;
		goto out;
	}
2707 2708
	/* we call try-to-free pages for make this cgroup empty */
	lru_add_drain_all();
2709 2710 2711 2712
	/* try to free all pages in this cgroup */
	shrink = 1;
	while (nr_retries && mem->res.usage > 0) {
		int progress;
2713 2714 2715 2716 2717

		if (signal_pending(current)) {
			ret = -EINTR;
			goto out;
		}
K
KOSAKI Motohiro 已提交
2718 2719
		progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL,
						false, get_swappiness(mem));
2720
		if (!progress) {
2721
			nr_retries--;
2722
			/* maybe some writeback is necessary */
2723
			congestion_wait(BLK_RW_ASYNC, HZ/10);
2724
		}
2725 2726

	}
K
KAMEZAWA Hiroyuki 已提交
2727
	lru_add_drain();
2728
	/* try move_account...there may be some *locked* pages. */
2729
	goto move_account;
2730 2731
}

2732 2733 2734 2735 2736 2737
int mem_cgroup_force_empty_write(struct cgroup *cont, unsigned int event)
{
	return mem_cgroup_force_empty(mem_cgroup_from_cont(cont), true);
}


2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755
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();
	/*
2756
	 * If parent's use_hierarchy is set, we can't make any modifications
2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775
	 * 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;
}

2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799
struct mem_cgroup_idx_data {
	s64 val;
	enum mem_cgroup_stat_index idx;
};

static int
mem_cgroup_get_idx_stat(struct mem_cgroup *mem, void *data)
{
	struct mem_cgroup_idx_data *d = data;
	d->val += mem_cgroup_read_stat(&mem->stat, d->idx);
	return 0;
}

static void
mem_cgroup_get_recursive_idx_stat(struct mem_cgroup *mem,
				enum mem_cgroup_stat_index idx, s64 *val)
{
	struct mem_cgroup_idx_data d;
	d.idx = idx;
	d.val = 0;
	mem_cgroup_walk_tree(mem, &d, mem_cgroup_get_idx_stat);
	*val = d.val;
}

2800
static u64 mem_cgroup_read(struct cgroup *cont, struct cftype *cft)
B
Balbir Singh 已提交
2801
{
2802
	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);
2803
	u64 idx_val, val;
2804 2805 2806 2807 2808 2809
	int type, name;

	type = MEMFILE_TYPE(cft->private);
	name = MEMFILE_ATTR(cft->private);
	switch (type) {
	case _MEM:
2810 2811 2812 2813 2814 2815 2816 2817 2818 2819
		if (name == RES_USAGE && mem_cgroup_is_root(mem)) {
			mem_cgroup_get_recursive_idx_stat(mem,
				MEM_CGROUP_STAT_CACHE, &idx_val);
			val = idx_val;
			mem_cgroup_get_recursive_idx_stat(mem,
				MEM_CGROUP_STAT_RSS, &idx_val);
			val += idx_val;
			val <<= PAGE_SHIFT;
		} else
			val = res_counter_read_u64(&mem->res, name);
2820 2821
		break;
	case _MEMSWAP:
2822 2823 2824 2825 2826 2827 2828 2829 2830
		if (name == RES_USAGE && mem_cgroup_is_root(mem)) {
			mem_cgroup_get_recursive_idx_stat(mem,
				MEM_CGROUP_STAT_CACHE, &idx_val);
			val = idx_val;
			mem_cgroup_get_recursive_idx_stat(mem,
				MEM_CGROUP_STAT_RSS, &idx_val);
			val += idx_val;
			mem_cgroup_get_recursive_idx_stat(mem,
				MEM_CGROUP_STAT_SWAPOUT, &idx_val);
2831
			val += idx_val;
2832 2833 2834
			val <<= PAGE_SHIFT;
		} else
			val = res_counter_read_u64(&mem->memsw, name);
2835 2836 2837 2838 2839 2840
		break;
	default:
		BUG();
		break;
	}
	return val;
B
Balbir Singh 已提交
2841
}
2842 2843 2844 2845
/*
 * The user of this function is...
 * RES_LIMIT.
 */
2846 2847
static int mem_cgroup_write(struct cgroup *cont, struct cftype *cft,
			    const char *buffer)
B
Balbir Singh 已提交
2848
{
2849
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cont);
2850
	int type, name;
2851 2852 2853
	unsigned long long val;
	int ret;

2854 2855 2856
	type = MEMFILE_TYPE(cft->private);
	name = MEMFILE_ATTR(cft->private);
	switch (name) {
2857
	case RES_LIMIT:
2858 2859 2860 2861
		if (mem_cgroup_is_root(memcg)) { /* Can't set limit on root */
			ret = -EINVAL;
			break;
		}
2862 2863
		/* This function does all necessary parse...reuse it */
		ret = res_counter_memparse_write_strategy(buffer, &val);
2864 2865 2866
		if (ret)
			break;
		if (type == _MEM)
2867
			ret = mem_cgroup_resize_limit(memcg, val);
2868 2869
		else
			ret = mem_cgroup_resize_memsw_limit(memcg, val);
2870
		break;
2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884
	case RES_SOFT_LIMIT:
		ret = res_counter_memparse_write_strategy(buffer, &val);
		if (ret)
			break;
		/*
		 * For memsw, soft limits are hard to implement in terms
		 * of semantics, for now, we support soft limits for
		 * control without swap
		 */
		if (type == _MEM)
			ret = res_counter_set_soft_limit(&memcg->res, val);
		else
			ret = -EINVAL;
		break;
2885 2886 2887 2888 2889
	default:
		ret = -EINVAL; /* should be BUG() ? */
		break;
	}
	return ret;
B
Balbir Singh 已提交
2890 2891
}

2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919
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;
}

2920
static int mem_cgroup_reset(struct cgroup *cont, unsigned int event)
2921 2922
{
	struct mem_cgroup *mem;
2923
	int type, name;
2924 2925

	mem = mem_cgroup_from_cont(cont);
2926 2927 2928
	type = MEMFILE_TYPE(event);
	name = MEMFILE_ATTR(event);
	switch (name) {
2929
	case RES_MAX_USAGE:
2930 2931 2932 2933
		if (type == _MEM)
			res_counter_reset_max(&mem->res);
		else
			res_counter_reset_max(&mem->memsw);
2934 2935
		break;
	case RES_FAILCNT:
2936 2937 2938 2939
		if (type == _MEM)
			res_counter_reset_failcnt(&mem->res);
		else
			res_counter_reset_failcnt(&mem->memsw);
2940 2941
		break;
	}
2942

2943
	return 0;
2944 2945
}

2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970
static u64 mem_cgroup_move_charge_read(struct cgroup *cgrp,
					struct cftype *cft)
{
	return mem_cgroup_from_cont(cgrp)->move_charge_at_immigrate;
}

static int mem_cgroup_move_charge_write(struct cgroup *cgrp,
					struct cftype *cft, u64 val)
{
	struct mem_cgroup *mem = mem_cgroup_from_cont(cgrp);

	if (val >= (1 << NR_MOVE_TYPE))
		return -EINVAL;
	/*
	 * We check this value several times in both in can_attach() and
	 * attach(), so we need cgroup lock to prevent this value from being
	 * inconsistent.
	 */
	cgroup_lock();
	mem->move_charge_at_immigrate = val;
	cgroup_unlock();

	return 0;
}

K
KAMEZAWA Hiroyuki 已提交
2971 2972 2973 2974 2975

/* For read statistics */
enum {
	MCS_CACHE,
	MCS_RSS,
2976
	MCS_FILE_MAPPED,
K
KAMEZAWA Hiroyuki 已提交
2977 2978
	MCS_PGPGIN,
	MCS_PGPGOUT,
2979
	MCS_SWAP,
K
KAMEZAWA Hiroyuki 已提交
2980 2981 2982 2983 2984 2985 2986 2987 2988 2989
	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];
2990 2991
};

K
KAMEZAWA Hiroyuki 已提交
2992 2993 2994 2995 2996 2997
struct {
	char *local_name;
	char *total_name;
} memcg_stat_strings[NR_MCS_STAT] = {
	{"cache", "total_cache"},
	{"rss", "total_rss"},
2998
	{"mapped_file", "total_mapped_file"},
K
KAMEZAWA Hiroyuki 已提交
2999 3000
	{"pgpgin", "total_pgpgin"},
	{"pgpgout", "total_pgpgout"},
3001
	{"swap", "total_swap"},
K
KAMEZAWA Hiroyuki 已提交
3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019
	{"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;
3020 3021
	val = mem_cgroup_read_stat(&mem->stat, MEM_CGROUP_STAT_FILE_MAPPED);
	s->stat[MCS_FILE_MAPPED] += val * PAGE_SIZE;
K
KAMEZAWA Hiroyuki 已提交
3022 3023 3024 3025
	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;
3026 3027 3028 3029
	if (do_swap_account) {
		val = mem_cgroup_read_stat(&mem->stat, MEM_CGROUP_STAT_SWAPOUT);
		s->stat[MCS_SWAP] += val * PAGE_SIZE;
	}
K
KAMEZAWA Hiroyuki 已提交
3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050

	/* 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);
}

3051 3052
static int mem_control_stat_show(struct cgroup *cont, struct cftype *cft,
				 struct cgroup_map_cb *cb)
3053 3054
{
	struct mem_cgroup *mem_cont = mem_cgroup_from_cont(cont);
K
KAMEZAWA Hiroyuki 已提交
3055
	struct mcs_total_stat mystat;
3056 3057
	int i;

K
KAMEZAWA Hiroyuki 已提交
3058 3059
	memset(&mystat, 0, sizeof(mystat));
	mem_cgroup_get_local_stat(mem_cont, &mystat);
3060

3061 3062 3063
	for (i = 0; i < NR_MCS_STAT; i++) {
		if (i == MCS_SWAP && !do_swap_account)
			continue;
K
KAMEZAWA Hiroyuki 已提交
3064
		cb->fill(cb, memcg_stat_strings[i].local_name, mystat.stat[i]);
3065
	}
L
Lee Schermerhorn 已提交
3066

K
KAMEZAWA Hiroyuki 已提交
3067
	/* Hierarchical information */
3068 3069 3070 3071 3072 3073 3074
	{
		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
3075

K
KAMEZAWA Hiroyuki 已提交
3076 3077
	memset(&mystat, 0, sizeof(mystat));
	mem_cgroup_get_total_stat(mem_cont, &mystat);
3078 3079 3080
	for (i = 0; i < NR_MCS_STAT; i++) {
		if (i == MCS_SWAP && !do_swap_account)
			continue;
K
KAMEZAWA Hiroyuki 已提交
3081
		cb->fill(cb, memcg_stat_strings[i].total_name, mystat.stat[i]);
3082
	}
K
KAMEZAWA Hiroyuki 已提交
3083

K
KOSAKI Motohiro 已提交
3084
#ifdef CONFIG_DEBUG_VM
3085
	cb->fill(cb, "inactive_ratio", calc_inactive_ratio(mem_cont, NULL));
K
KOSAKI Motohiro 已提交
3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112

	{
		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

3113 3114 3115
	return 0;
}

K
KOSAKI Motohiro 已提交
3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127
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;
3128

K
KOSAKI Motohiro 已提交
3129 3130 3131 3132 3133 3134 3135
	if (val > 100)
		return -EINVAL;

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

	parent = mem_cgroup_from_cont(cgrp->parent);
3136 3137 3138

	cgroup_lock();

K
KOSAKI Motohiro 已提交
3139 3140
	/* If under hierarchy, only empty-root can set this value */
	if ((parent->use_hierarchy) ||
3141 3142
	    (memcg->use_hierarchy && !list_empty(&cgrp->children))) {
		cgroup_unlock();
K
KOSAKI Motohiro 已提交
3143
		return -EINVAL;
3144
	}
K
KOSAKI Motohiro 已提交
3145 3146 3147 3148 3149

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

3150 3151
	cgroup_unlock();

K
KOSAKI Motohiro 已提交
3152 3153 3154
	return 0;
}

3155

B
Balbir Singh 已提交
3156 3157
static struct cftype mem_cgroup_files[] = {
	{
3158
		.name = "usage_in_bytes",
3159
		.private = MEMFILE_PRIVATE(_MEM, RES_USAGE),
3160
		.read_u64 = mem_cgroup_read,
B
Balbir Singh 已提交
3161
	},
3162 3163
	{
		.name = "max_usage_in_bytes",
3164
		.private = MEMFILE_PRIVATE(_MEM, RES_MAX_USAGE),
3165
		.trigger = mem_cgroup_reset,
3166 3167
		.read_u64 = mem_cgroup_read,
	},
B
Balbir Singh 已提交
3168
	{
3169
		.name = "limit_in_bytes",
3170
		.private = MEMFILE_PRIVATE(_MEM, RES_LIMIT),
3171
		.write_string = mem_cgroup_write,
3172
		.read_u64 = mem_cgroup_read,
B
Balbir Singh 已提交
3173
	},
3174 3175 3176 3177 3178 3179
	{
		.name = "soft_limit_in_bytes",
		.private = MEMFILE_PRIVATE(_MEM, RES_SOFT_LIMIT),
		.write_string = mem_cgroup_write,
		.read_u64 = mem_cgroup_read,
	},
B
Balbir Singh 已提交
3180 3181
	{
		.name = "failcnt",
3182
		.private = MEMFILE_PRIVATE(_MEM, RES_FAILCNT),
3183
		.trigger = mem_cgroup_reset,
3184
		.read_u64 = mem_cgroup_read,
B
Balbir Singh 已提交
3185
	},
3186 3187
	{
		.name = "stat",
3188
		.read_map = mem_control_stat_show,
3189
	},
3190 3191 3192 3193
	{
		.name = "force_empty",
		.trigger = mem_cgroup_force_empty_write,
	},
3194 3195 3196 3197 3198
	{
		.name = "use_hierarchy",
		.write_u64 = mem_cgroup_hierarchy_write,
		.read_u64 = mem_cgroup_hierarchy_read,
	},
K
KOSAKI Motohiro 已提交
3199 3200 3201 3202 3203
	{
		.name = "swappiness",
		.read_u64 = mem_cgroup_swappiness_read,
		.write_u64 = mem_cgroup_swappiness_write,
	},
3204 3205 3206 3207 3208
	{
		.name = "move_charge_at_immigrate",
		.read_u64 = mem_cgroup_move_charge_read,
		.write_u64 = mem_cgroup_move_charge_write,
	},
B
Balbir Singh 已提交
3209 3210
};

3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251
#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

3252 3253 3254
static int alloc_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node)
{
	struct mem_cgroup_per_node *pn;
3255
	struct mem_cgroup_per_zone *mz;
3256
	enum lru_list l;
3257
	int zone, tmp = node;
3258 3259 3260 3261 3262 3263 3264 3265
	/*
	 * 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.
	 */
3266 3267 3268
	if (!node_state(node, N_NORMAL_MEMORY))
		tmp = -1;
	pn = kmalloc_node(sizeof(*pn), GFP_KERNEL, tmp);
3269 3270
	if (!pn)
		return 1;
3271

3272 3273
	mem->info.nodeinfo[node] = pn;
	memset(pn, 0, sizeof(*pn));
3274 3275 3276

	for (zone = 0; zone < MAX_NR_ZONES; zone++) {
		mz = &pn->zoneinfo[zone];
3277 3278
		for_each_lru(l)
			INIT_LIST_HEAD(&mz->lists[l]);
3279
		mz->usage_in_excess = 0;
3280 3281
		mz->on_tree = false;
		mz->mem = mem;
3282
	}
3283 3284 3285
	return 0;
}

3286 3287 3288 3289 3290
static void free_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node)
{
	kfree(mem->info.nodeinfo[node]);
}

3291 3292 3293 3294 3295 3296
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;
}

3297 3298 3299
static struct mem_cgroup *mem_cgroup_alloc(void)
{
	struct mem_cgroup *mem;
3300
	int size = mem_cgroup_size();
3301

3302 3303
	if (size < PAGE_SIZE)
		mem = kmalloc(size, GFP_KERNEL);
3304
	else
3305
		mem = vmalloc(size);
3306 3307

	if (mem)
3308
		memset(mem, 0, size);
3309 3310 3311
	return mem;
}

3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322
/*
 * 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.
 */

3323
static void __mem_cgroup_free(struct mem_cgroup *mem)
3324
{
K
KAMEZAWA Hiroyuki 已提交
3325 3326
	int node;

3327
	mem_cgroup_remove_from_trees(mem);
K
KAMEZAWA Hiroyuki 已提交
3328 3329
	free_css_id(&mem_cgroup_subsys, &mem->css);

K
KAMEZAWA Hiroyuki 已提交
3330 3331 3332
	for_each_node_state(node, N_POSSIBLE)
		free_mem_cgroup_per_zone_info(mem, node);

3333
	if (mem_cgroup_size() < PAGE_SIZE)
3334 3335 3336 3337 3338
		kfree(mem);
	else
		vfree(mem);
}

3339 3340 3341 3342 3343 3344 3345
static void mem_cgroup_get(struct mem_cgroup *mem)
{
	atomic_inc(&mem->refcnt);
}

static void mem_cgroup_put(struct mem_cgroup *mem)
{
3346 3347
	if (atomic_dec_and_test(&mem->refcnt)) {
		struct mem_cgroup *parent = parent_mem_cgroup(mem);
3348
		__mem_cgroup_free(mem);
3349 3350 3351
		if (parent)
			mem_cgroup_put(parent);
	}
3352 3353
}

3354 3355 3356 3357 3358 3359 3360 3361 3362
/*
 * 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);
}
3363

3364 3365 3366
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
static void __init enable_swap_cgroup(void)
{
3367
	if (!mem_cgroup_disabled() && really_do_swap_account)
3368 3369 3370 3371 3372 3373 3374 3375
		do_swap_account = 1;
}
#else
static void __init enable_swap_cgroup(void)
{
}
#endif

3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400
static int mem_cgroup_soft_limit_tree_init(void)
{
	struct mem_cgroup_tree_per_node *rtpn;
	struct mem_cgroup_tree_per_zone *rtpz;
	int tmp, node, zone;

	for_each_node_state(node, N_POSSIBLE) {
		tmp = node;
		if (!node_state(node, N_NORMAL_MEMORY))
			tmp = -1;
		rtpn = kzalloc_node(sizeof(*rtpn), GFP_KERNEL, tmp);
		if (!rtpn)
			return 1;

		soft_limit_tree.rb_tree_per_node[node] = rtpn;

		for (zone = 0; zone < MAX_NR_ZONES; zone++) {
			rtpz = &rtpn->rb_tree_per_zone[zone];
			rtpz->rb_root = RB_ROOT;
			spin_lock_init(&rtpz->lock);
		}
	}
	return 0;
}

L
Li Zefan 已提交
3401
static struct cgroup_subsys_state * __ref
B
Balbir Singh 已提交
3402 3403
mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont)
{
3404
	struct mem_cgroup *mem, *parent;
K
KAMEZAWA Hiroyuki 已提交
3405
	long error = -ENOMEM;
3406
	int node;
B
Balbir Singh 已提交
3407

3408 3409
	mem = mem_cgroup_alloc();
	if (!mem)
K
KAMEZAWA Hiroyuki 已提交
3410
		return ERR_PTR(error);
3411

3412 3413 3414
	for_each_node_state(node, N_POSSIBLE)
		if (alloc_mem_cgroup_per_zone_info(mem, node))
			goto free_out;
3415

3416
	/* root ? */
3417
	if (cont->parent == NULL) {
3418
		int cpu;
3419
		enable_swap_cgroup();
3420
		parent = NULL;
3421
		root_mem_cgroup = mem;
3422 3423
		if (mem_cgroup_soft_limit_tree_init())
			goto free_out;
3424 3425 3426 3427 3428 3429
		for_each_possible_cpu(cpu) {
			struct memcg_stock_pcp *stock =
						&per_cpu(memcg_stock, cpu);
			INIT_WORK(&stock->work, drain_local_stock);
		}
		hotcpu_notifier(memcg_stock_cpu_callback, 0);
3430
	} else {
3431
		parent = mem_cgroup_from_cont(cont->parent);
3432 3433
		mem->use_hierarchy = parent->use_hierarchy;
	}
3434

3435 3436 3437
	if (parent && parent->use_hierarchy) {
		res_counter_init(&mem->res, &parent->res);
		res_counter_init(&mem->memsw, &parent->memsw);
3438 3439 3440 3441 3442 3443 3444
		/*
		 * 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);
3445 3446 3447 3448
	} else {
		res_counter_init(&mem->res, NULL);
		res_counter_init(&mem->memsw, NULL);
	}
K
KAMEZAWA Hiroyuki 已提交
3449
	mem->last_scanned_child = 0;
K
KOSAKI Motohiro 已提交
3450
	spin_lock_init(&mem->reclaim_param_lock);
3451

K
KOSAKI Motohiro 已提交
3452 3453
	if (parent)
		mem->swappiness = get_swappiness(parent);
3454
	atomic_set(&mem->refcnt, 1);
3455
	mem->move_charge_at_immigrate = 0;
B
Balbir Singh 已提交
3456
	return &mem->css;
3457
free_out:
3458
	__mem_cgroup_free(mem);
3459
	root_mem_cgroup = NULL;
K
KAMEZAWA Hiroyuki 已提交
3460
	return ERR_PTR(error);
B
Balbir Singh 已提交
3461 3462
}

3463
static int mem_cgroup_pre_destroy(struct cgroup_subsys *ss,
3464 3465 3466
					struct cgroup *cont)
{
	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);
3467 3468

	return mem_cgroup_force_empty(mem, false);
3469 3470
}

B
Balbir Singh 已提交
3471 3472 3473
static void mem_cgroup_destroy(struct cgroup_subsys *ss,
				struct cgroup *cont)
{
3474 3475 3476
	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);

	mem_cgroup_put(mem);
B
Balbir Singh 已提交
3477 3478 3479 3480 3481
}

static int mem_cgroup_populate(struct cgroup_subsys *ss,
				struct cgroup *cont)
{
3482 3483 3484 3485 3486 3487 3488 3489
	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
3490 3491
}

3492
/* Handlers for move charge at task migration. */
3493 3494
#define PRECHARGE_COUNT_AT_ONCE	256
static int mem_cgroup_do_precharge(unsigned long count)
3495
{
3496 3497
	int ret = 0;
	int batch_count = PRECHARGE_COUNT_AT_ONCE;
3498 3499
	struct mem_cgroup *mem = mc.to;

3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544
	if (mem_cgroup_is_root(mem)) {
		mc.precharge += count;
		/* we don't need css_get for root */
		return ret;
	}
	/* try to charge at once */
	if (count > 1) {
		struct res_counter *dummy;
		/*
		 * "mem" cannot be under rmdir() because we've already checked
		 * by cgroup_lock_live_cgroup() that it is not removed and we
		 * are still under the same cgroup_mutex. So we can postpone
		 * css_get().
		 */
		if (res_counter_charge(&mem->res, PAGE_SIZE * count, &dummy))
			goto one_by_one;
		if (do_swap_account && res_counter_charge(&mem->memsw,
						PAGE_SIZE * count, &dummy)) {
			res_counter_uncharge(&mem->res, PAGE_SIZE * count);
			goto one_by_one;
		}
		mc.precharge += count;
		VM_BUG_ON(test_bit(CSS_ROOT, &mem->css.flags));
		WARN_ON_ONCE(count > INT_MAX);
		__css_get(&mem->css, (int)count);
		return ret;
	}
one_by_one:
	/* fall back to one by one charge */
	while (count--) {
		if (signal_pending(current)) {
			ret = -EINTR;
			break;
		}
		if (!batch_count--) {
			batch_count = PRECHARGE_COUNT_AT_ONCE;
			cond_resched();
		}
		ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, &mem,
								false, NULL);
		if (ret || !mem)
			/* mem_cgroup_clear_mc() will do uncharge later */
			return -ENOMEM;
		mc.precharge++;
	}
3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635
	return ret;
}

/**
 * is_target_pte_for_mc - check a pte whether it is valid for move charge
 * @vma: the vma the pte to be checked belongs
 * @addr: the address corresponding to the pte to be checked
 * @ptent: the pte to be checked
 * @target: the pointer the target page will be stored(can be NULL)
 *
 * Returns
 *   0(MC_TARGET_NONE): if the pte is not a target for move charge.
 *   1(MC_TARGET_PAGE): if the page corresponding to this pte is a target for
 *     move charge. if @target is not NULL, the page is stored in target->page
 *     with extra refcnt got(Callers should handle it).
 *
 * Called with pte lock held.
 */
/* We add a new member later. */
union mc_target {
	struct page	*page;
};

/* We add a new type later. */
enum mc_target_type {
	MC_TARGET_NONE,	/* not used */
	MC_TARGET_PAGE,
};

static int is_target_pte_for_mc(struct vm_area_struct *vma,
		unsigned long addr, pte_t ptent, union mc_target *target)
{
	struct page *page;
	struct page_cgroup *pc;
	int ret = 0;
	bool move_anon = test_bit(MOVE_CHARGE_TYPE_ANON,
					&mc.to->move_charge_at_immigrate);

	if (!pte_present(ptent))
		return 0;

	page = vm_normal_page(vma, addr, ptent);
	if (!page || !page_mapped(page))
		return 0;
	/*
	 * TODO: We don't move charges of file(including shmem/tmpfs) pages for
	 * now.
	 */
	if (!move_anon || !PageAnon(page))
		return 0;
	/*
	 * TODO: We don't move charges of shared(used by multiple processes)
	 * pages for now.
	 */
	if (page_mapcount(page) > 1)
		return 0;
	if (!get_page_unless_zero(page))
		return 0;

	pc = lookup_page_cgroup(page);
	/*
	 * Do only loose check w/o page_cgroup lock. mem_cgroup_move_account()
	 * checks the pc is valid or not under the lock.
	 */
	if (PageCgroupUsed(pc) && pc->mem_cgroup == mc.from) {
		ret = MC_TARGET_PAGE;
		if (target)
			target->page = page;
	}

	if (!ret || !target)
		put_page(page);

	return ret;
}

static int mem_cgroup_count_precharge_pte_range(pmd_t *pmd,
					unsigned long addr, unsigned long end,
					struct mm_walk *walk)
{
	struct vm_area_struct *vma = walk->private;
	pte_t *pte;
	spinlock_t *ptl;

	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
	for (; addr != end; pte++, addr += PAGE_SIZE)
		if (is_target_pte_for_mc(vma, addr, *pte, NULL))
			mc.precharge++;	/* increment precharge temporarily */
	pte_unmap_unlock(pte - 1, ptl);
	cond_resched();

3636 3637 3638
	return 0;
}

3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668
static unsigned long mem_cgroup_count_precharge(struct mm_struct *mm)
{
	unsigned long precharge;
	struct vm_area_struct *vma;

	down_read(&mm->mmap_sem);
	for (vma = mm->mmap; vma; vma = vma->vm_next) {
		struct mm_walk mem_cgroup_count_precharge_walk = {
			.pmd_entry = mem_cgroup_count_precharge_pte_range,
			.mm = mm,
			.private = vma,
		};
		if (is_vm_hugetlb_page(vma))
			continue;
		/* TODO: We don't move charges of shmem/tmpfs pages for now. */
		if (vma->vm_flags & VM_SHARED)
			continue;
		walk_page_range(vma->vm_start, vma->vm_end,
					&mem_cgroup_count_precharge_walk);
	}
	up_read(&mm->mmap_sem);

	precharge = mc.precharge;
	mc.precharge = 0;

	return precharge;
}

static int mem_cgroup_precharge_mc(struct mm_struct *mm)
{
3669
	return mem_cgroup_do_precharge(mem_cgroup_count_precharge(mm));
3670 3671 3672 3673 3674
}

static void mem_cgroup_clear_mc(void)
{
	/* we must uncharge all the leftover precharges from mc.to */
3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685
	if (mc.precharge) {
		__mem_cgroup_cancel_charge(mc.to, mc.precharge);
		mc.precharge = 0;
	}
	/*
	 * we didn't uncharge from mc.from at mem_cgroup_move_account(), so
	 * we must uncharge here.
	 */
	if (mc.moved_charge) {
		__mem_cgroup_cancel_charge(mc.from, mc.moved_charge);
		mc.moved_charge = 0;
3686 3687 3688
	}
	mc.from = NULL;
	mc.to = NULL;
3689 3690
	mc.moving_task = NULL;
	wake_up_all(&mc.waitq);
3691 3692
}

3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710
static int mem_cgroup_can_attach(struct cgroup_subsys *ss,
				struct cgroup *cgroup,
				struct task_struct *p,
				bool threadgroup)
{
	int ret = 0;
	struct mem_cgroup *mem = mem_cgroup_from_cont(cgroup);

	if (mem->move_charge_at_immigrate) {
		struct mm_struct *mm;
		struct mem_cgroup *from = mem_cgroup_from_task(p);

		VM_BUG_ON(from == mem);

		mm = get_task_mm(p);
		if (!mm)
			return 0;
		/* We move charges only when we move a owner of the mm */
3711 3712 3713 3714
		if (mm->owner == p) {
			VM_BUG_ON(mc.from);
			VM_BUG_ON(mc.to);
			VM_BUG_ON(mc.precharge);
3715
			VM_BUG_ON(mc.moved_charge);
3716
			VM_BUG_ON(mc.moving_task);
3717 3718 3719
			mc.from = from;
			mc.to = mem;
			mc.precharge = 0;
3720
			mc.moved_charge = 0;
3721
			mc.moving_task = current;
3722 3723 3724 3725 3726

			ret = mem_cgroup_precharge_mc(mm);
			if (ret)
				mem_cgroup_clear_mc();
		}
3727 3728 3729 3730 3731 3732 3733 3734 3735 3736
		mmput(mm);
	}
	return ret;
}

static void mem_cgroup_cancel_attach(struct cgroup_subsys *ss,
				struct cgroup *cgroup,
				struct task_struct *p,
				bool threadgroup)
{
3737
	mem_cgroup_clear_mc();
3738 3739
}

3740 3741 3742
static int mem_cgroup_move_charge_pte_range(pmd_t *pmd,
				unsigned long addr, unsigned long end,
				struct mm_walk *walk)
3743
{
3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767
	int ret = 0;
	struct vm_area_struct *vma = walk->private;
	pte_t *pte;
	spinlock_t *ptl;

retry:
	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
	for (; addr != end; addr += PAGE_SIZE) {
		pte_t ptent = *(pte++);
		union mc_target target;
		int type;
		struct page *page;
		struct page_cgroup *pc;

		if (!mc.precharge)
			break;

		type = is_target_pte_for_mc(vma, addr, ptent, &target);
		switch (type) {
		case MC_TARGET_PAGE:
			page = target.page;
			if (isolate_lru_page(page))
				goto put;
			pc = lookup_page_cgroup(page);
3768 3769
			if (!mem_cgroup_move_account(pc,
						mc.from, mc.to, false)) {
3770
				mc.precharge--;
3771 3772
				/* we uncharge from mc.from later. */
				mc.moved_charge++;
3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791
			}
			putback_lru_page(page);
put:			/* is_target_pte_for_mc() gets the page */
			put_page(page);
			break;
		default:
			break;
		}
	}
	pte_unmap_unlock(pte - 1, ptl);
	cond_resched();

	if (addr != end) {
		/*
		 * We have consumed all precharges we got in can_attach().
		 * We try charge one by one, but don't do any additional
		 * charges to mc.to if we have failed in charge once in attach()
		 * phase.
		 */
3792
		ret = mem_cgroup_do_precharge(1);
3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827
		if (!ret)
			goto retry;
	}

	return ret;
}

static void mem_cgroup_move_charge(struct mm_struct *mm)
{
	struct vm_area_struct *vma;

	lru_add_drain_all();
	down_read(&mm->mmap_sem);
	for (vma = mm->mmap; vma; vma = vma->vm_next) {
		int ret;
		struct mm_walk mem_cgroup_move_charge_walk = {
			.pmd_entry = mem_cgroup_move_charge_pte_range,
			.mm = mm,
			.private = vma,
		};
		if (is_vm_hugetlb_page(vma))
			continue;
		/* TODO: We don't move charges of shmem/tmpfs pages for now. */
		if (vma->vm_flags & VM_SHARED)
			continue;
		ret = walk_page_range(vma->vm_start, vma->vm_end,
						&mem_cgroup_move_charge_walk);
		if (ret)
			/*
			 * means we have consumed all precharges and failed in
			 * doing additional charge. Just abandon here.
			 */
			break;
	}
	up_read(&mm->mmap_sem);
3828 3829
}

B
Balbir Singh 已提交
3830 3831 3832
static void mem_cgroup_move_task(struct cgroup_subsys *ss,
				struct cgroup *cont,
				struct cgroup *old_cont,
3833 3834
				struct task_struct *p,
				bool threadgroup)
B
Balbir Singh 已提交
3835
{
3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847
	struct mm_struct *mm;

	if (!mc.to)
		/* no need to move charge */
		return;

	mm = get_task_mm(p);
	if (mm) {
		mem_cgroup_move_charge(mm);
		mmput(mm);
	}
	mem_cgroup_clear_mc();
B
Balbir Singh 已提交
3848 3849
}

B
Balbir Singh 已提交
3850 3851 3852 3853
struct cgroup_subsys mem_cgroup_subsys = {
	.name = "memory",
	.subsys_id = mem_cgroup_subsys_id,
	.create = mem_cgroup_create,
3854
	.pre_destroy = mem_cgroup_pre_destroy,
B
Balbir Singh 已提交
3855 3856
	.destroy = mem_cgroup_destroy,
	.populate = mem_cgroup_populate,
3857 3858
	.can_attach = mem_cgroup_can_attach,
	.cancel_attach = mem_cgroup_cancel_attach,
B
Balbir Singh 已提交
3859
	.attach = mem_cgroup_move_task,
3860
	.early_init = 0,
K
KAMEZAWA Hiroyuki 已提交
3861
	.use_id = 1,
B
Balbir Singh 已提交
3862
};
3863 3864 3865 3866 3867 3868 3869 3870 3871 3872

#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