memcontrol.h 28.3 KB
Newer Older
B
Balbir Singh 已提交
1 2 3 4 5
/* memcontrol.h - 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
 * 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.
 */

#ifndef _LINUX_MEMCONTROL_H
#define _LINUX_MEMCONTROL_H
22
#include <linux/cgroup.h>
23
#include <linux/vm_event_item.h>
24
#include <linux/hardirq.h>
25
#include <linux/jump_label.h>
M
Michal Hocko 已提交
26 27 28
#include <linux/page_counter.h>
#include <linux/vmpressure.h>
#include <linux/eventfd.h>
29 30
#include <linux/mm.h>
#include <linux/vmstat.h>
M
Michal Hocko 已提交
31
#include <linux/writeback.h>
32
#include <linux/page-flags.h>
33

34
struct mem_cgroup;
35 36
struct page;
struct mm_struct;
37
struct kmem_cache;
38

39 40 41 42 43 44 45 46 47
/* Cgroup-specific page state, on top of universal node page state */
enum memcg_stat_item {
	MEMCG_CACHE = NR_VM_NODE_STAT_ITEMS,
	MEMCG_RSS,
	MEMCG_RSS_HUGE,
	MEMCG_SWAP,
	MEMCG_SOCK,
	/* XXX: why are these zone and not node counters? */
	MEMCG_KERNEL_STACK_KB,
48
	MEMCG_NR_STAT,
49 50
};

51 52 53 54 55 56 57 58 59
/* Cgroup-specific events, on top of universal VM events */
enum memcg_event_item {
	MEMCG_LOW = NR_VM_EVENT_ITEMS,
	MEMCG_HIGH,
	MEMCG_MAX,
	MEMCG_OOM,
	MEMCG_NR_EVENTS,
};

60
struct mem_cgroup_reclaim_cookie {
61
	pg_data_t *pgdat;
62 63 64 65
	int priority;
	unsigned int generation;
};

66 67 68 69 70 71 72 73 74 75
#ifdef CONFIG_MEMCG

#define MEM_CGROUP_ID_SHIFT	16
#define MEM_CGROUP_ID_MAX	USHRT_MAX

struct mem_cgroup_id {
	int id;
	atomic_t ref;
};

M
Michal Hocko 已提交
76 77 78 79 80 81 82 83 84 85 86 87 88 89
/*
 * Per memcg event counter is incremented at every pagein/pageout. With THP,
 * it will be incremated by the number of pages. This counter is used for
 * for trigger some periodic events. This is straightforward and better
 * than using jiffies etc. to handle periodic memcg event.
 */
enum mem_cgroup_events_target {
	MEM_CGROUP_TARGET_THRESH,
	MEM_CGROUP_TARGET_SOFTLIMIT,
	MEM_CGROUP_TARGET_NUMAINFO,
	MEM_CGROUP_NTARGETS,
};

struct mem_cgroup_stat_cpu {
90
	long count[MEMCG_NR_STAT];
M
Michal Hocko 已提交
91 92 93 94 95 96 97 98 99 100 101
	unsigned long events[MEMCG_NR_EVENTS];
	unsigned long nr_page_events;
	unsigned long targets[MEM_CGROUP_NTARGETS];
};

struct mem_cgroup_reclaim_iter {
	struct mem_cgroup *position;
	/* scan generation, increased every round-trip */
	unsigned int generation;
};

102 103 104 105
struct lruvec_stat {
	long count[NR_VM_NODE_STAT_ITEMS];
};

M
Michal Hocko 已提交
106 107 108
/*
 * per-zone information in memory controller.
 */
109
struct mem_cgroup_per_node {
M
Michal Hocko 已提交
110
	struct lruvec		lruvec;
111
	struct lruvec_stat __percpu *lruvec_stat;
112
	unsigned long		lru_zone_size[MAX_NR_ZONES][NR_LRU_LISTS];
M
Michal Hocko 已提交
113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149

	struct mem_cgroup_reclaim_iter	iter[DEF_PRIORITY + 1];

	struct rb_node		tree_node;	/* RB tree node */
	unsigned long		usage_in_excess;/* Set to the value by which */
						/* the soft limit is exceeded*/
	bool			on_tree;
	struct mem_cgroup	*memcg;		/* Back pointer, we cannot */
						/* use container_of	   */
};

struct mem_cgroup_threshold {
	struct eventfd_ctx *eventfd;
	unsigned long threshold;
};

/* For threshold */
struct mem_cgroup_threshold_ary {
	/* An array index points to threshold just below or equal to usage. */
	int current_threshold;
	/* Size of entries[] */
	unsigned int size;
	/* Array of thresholds */
	struct mem_cgroup_threshold entries[0];
};

struct mem_cgroup_thresholds {
	/* Primary thresholds array */
	struct mem_cgroup_threshold_ary *primary;
	/*
	 * Spare threshold array.
	 * This is needed to make mem_cgroup_unregister_event() "never fail".
	 * It must be able to store at least primary->size - 1 entries.
	 */
	struct mem_cgroup_threshold_ary *spare;
};

150 151 152 153 154 155
enum memcg_kmem_state {
	KMEM_NONE,
	KMEM_ALLOCATED,
	KMEM_ONLINE,
};

M
Michal Hocko 已提交
156 157 158 159 160 161 162 163 164
/*
 * 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.
 */
struct mem_cgroup {
	struct cgroup_subsys_state css;

165 166 167
	/* Private memcg ID. Used to ID objects that outlive the cgroup */
	struct mem_cgroup_id id;

M
Michal Hocko 已提交
168 169
	/* Accounted resources */
	struct page_counter memory;
170
	struct page_counter swap;
171 172

	/* Legacy consumer-oriented counters */
M
Michal Hocko 已提交
173 174
	struct page_counter memsw;
	struct page_counter kmem;
175
	struct page_counter tcpmem;
M
Michal Hocko 已提交
176 177 178 179 180

	/* Normal memory consumption range */
	unsigned long low;
	unsigned long high;

181 182 183
	/* Range enforcement for interrupt charges */
	struct work_struct high_work;

M
Michal Hocko 已提交
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
	unsigned long soft_limit;

	/* vmpressure notifications */
	struct vmpressure vmpressure;

	/*
	 * Should the accounting and control be hierarchical, per subtree?
	 */
	bool use_hierarchy;

	/* protected by memcg_oom_lock */
	bool		oom_lock;
	int		under_oom;

	int	swappiness;
	/* OOM-Killer disable */
	int		oom_kill_disable;

202 203 204
	/* handle for "memory.events" */
	struct cgroup_file events_file;

M
Michal Hocko 已提交
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234
	/* protect arrays of thresholds */
	struct mutex thresholds_lock;

	/* thresholds for memory usage. RCU-protected */
	struct mem_cgroup_thresholds thresholds;

	/* thresholds for mem+swap usage. RCU-protected */
	struct mem_cgroup_thresholds memsw_thresholds;

	/* For oom notifier event fd */
	struct list_head oom_notify;

	/*
	 * 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;
	/*
	 * set > 0 if pages under this cgroup are moving to other cgroup.
	 */
	atomic_t		moving_account;
	/* taken only while moving_account > 0 */
	spinlock_t		move_lock;
	struct task_struct	*move_lock_task;
	unsigned long		move_lock_flags;
	/*
	 * percpu counter.
	 */
	struct mem_cgroup_stat_cpu __percpu *stat;

235 236 237
	unsigned long		socket_pressure;

	/* Legacy tcp memory accounting */
238 239
	bool			tcpmem_active;
	int			tcpmem_pressure;
240

241
#ifndef CONFIG_SLOB
M
Michal Hocko 已提交
242 243
        /* Index in the kmem_cache->memcg_params.memcg_caches array */
	int kmemcg_id;
244
	enum memcg_kmem_state kmem_state;
245
	struct list_head kmem_caches;
M
Michal Hocko 已提交
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266
#endif

	int last_scanned_node;
#if MAX_NUMNODES > 1
	nodemask_t	scan_nodes;
	atomic_t	numainfo_events;
	atomic_t	numainfo_updating;
#endif

#ifdef CONFIG_CGROUP_WRITEBACK
	struct list_head cgwb_list;
	struct wb_domain cgwb_domain;
#endif

	/* List of events which userspace want to receive */
	struct list_head event_list;
	spinlock_t event_list_lock;

	struct mem_cgroup_per_node *nodeinfo[0];
	/* WARNING: nodeinfo must be the last member here */
};
267 268

extern struct mem_cgroup *root_mem_cgroup;
T
Tejun Heo 已提交
269

270 271 272 273 274
static inline bool mem_cgroup_disabled(void)
{
	return !cgroup_subsys_enabled(memory_cgrp_subsys);
}

275
static inline void mem_cgroup_event(struct mem_cgroup *memcg,
276
				    enum memcg_event_item event)
M
Michal Hocko 已提交
277
{
278
	this_cpu_inc(memcg->stat->events[event]);
279
	cgroup_file_notify(&memcg->events_file);
M
Michal Hocko 已提交
280
}
281 282 283

bool mem_cgroup_low(struct mem_cgroup *root, struct mem_cgroup *memcg);

284
int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
285 286
			  gfp_t gfp_mask, struct mem_cgroup **memcgp,
			  bool compound);
287
void mem_cgroup_commit_charge(struct page *page, struct mem_cgroup *memcg,
288 289 290
			      bool lrucare, bool compound);
void mem_cgroup_cancel_charge(struct page *page, struct mem_cgroup *memcg,
		bool compound);
291
void mem_cgroup_uncharge(struct page *page);
292
void mem_cgroup_uncharge_list(struct list_head *page_list);
293

294
void mem_cgroup_migrate(struct page *oldpage, struct page *newpage);
295

296 297
static struct mem_cgroup_per_node *
mem_cgroup_nodeinfo(struct mem_cgroup *memcg, int nid)
298
{
299
	return memcg->nodeinfo[nid];
300 301 302
}

/**
303 304
 * mem_cgroup_lruvec - get the lru list vector for a node or a memcg zone
 * @node: node of the wanted lruvec
305 306
 * @memcg: memcg of the wanted lruvec
 *
307 308
 * Returns the lru list vector holding pages for a given @node or a given
 * @memcg and @zone. This can be the node lruvec, if the memory controller
309 310
 * is disabled.
 */
311
static inline struct lruvec *mem_cgroup_lruvec(struct pglist_data *pgdat,
312
				struct mem_cgroup *memcg)
313
{
314
	struct mem_cgroup_per_node *mz;
315 316 317
	struct lruvec *lruvec;

	if (mem_cgroup_disabled()) {
318
		lruvec = node_lruvec(pgdat);
319 320 321
		goto out;
	}

322
	mz = mem_cgroup_nodeinfo(memcg, pgdat->node_id);
323 324 325 326
	lruvec = &mz->lruvec;
out:
	/*
	 * Since a node can be onlined after the mem_cgroup was created,
M
Mel Gorman 已提交
327
	 * we have to be prepared to initialize lruvec->pgdat here;
328 329
	 * and if offlined then reonlined, we need to reinitialize it.
	 */
330 331
	if (unlikely(lruvec->pgdat != pgdat))
		lruvec->pgdat = pgdat;
332 333 334
	return lruvec;
}

M
Mel Gorman 已提交
335
struct lruvec *mem_cgroup_page_lruvec(struct page *, struct pglist_data *);
336

337
bool task_in_mem_cgroup(struct task_struct *task, struct mem_cgroup *memcg);
338
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
339

M
Michal Hocko 已提交
340 341 342 343 344
static inline
struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css){
	return css ? container_of(css, struct mem_cgroup, css) : NULL;
}

345 346 347
#define mem_cgroup_from_counter(counter, member)	\
	container_of(counter, struct mem_cgroup, member)

M
Michal Hocko 已提交
348 349 350 351
struct mem_cgroup *mem_cgroup_iter(struct mem_cgroup *,
				   struct mem_cgroup *,
				   struct mem_cgroup_reclaim_cookie *);
void mem_cgroup_iter_break(struct mem_cgroup *, struct mem_cgroup *);
352 353
int mem_cgroup_scan_tasks(struct mem_cgroup *,
			  int (*)(struct task_struct *, void *), void *);
M
Michal Hocko 已提交
354

355 356 357 358 359
static inline unsigned short mem_cgroup_id(struct mem_cgroup *memcg)
{
	if (mem_cgroup_disabled())
		return 0;

360
	return memcg->id.id;
361
}
362
struct mem_cgroup *mem_cgroup_from_id(unsigned short id);
363

364 365 366 367 368 369 370 371 372 373 374
static inline struct mem_cgroup *lruvec_memcg(struct lruvec *lruvec)
{
	struct mem_cgroup_per_node *mz;

	if (mem_cgroup_disabled())
		return NULL;

	mz = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
	return mz->memcg;
}

375 376 377 378 379 380 381 382 383 384 385 386 387 388
/**
 * parent_mem_cgroup - find the accounting parent of a memcg
 * @memcg: memcg whose parent to find
 *
 * Returns the parent memcg, or NULL if this is the root or the memory
 * controller is in legacy no-hierarchy mode.
 */
static inline struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *memcg)
{
	if (!memcg->memory.parent)
		return NULL;
	return mem_cgroup_from_counter(memcg->memory.parent, memory);
}

M
Michal Hocko 已提交
389 390 391 392 393 394 395 396 397
static inline bool mem_cgroup_is_descendant(struct mem_cgroup *memcg,
			      struct mem_cgroup *root)
{
	if (root == memcg)
		return true;
	if (!root->use_hierarchy)
		return false;
	return cgroup_is_descendant(memcg->css.cgroup, root->css.cgroup);
}
G
Glauber Costa 已提交
398

399 400
static inline bool mm_match_cgroup(struct mm_struct *mm,
				   struct mem_cgroup *memcg)
401
{
402
	struct mem_cgroup *task_memcg;
403
	bool match = false;
404

405
	rcu_read_lock();
406
	task_memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
407
	if (task_memcg)
408
		match = mem_cgroup_is_descendant(task_memcg, memcg);
409
	rcu_read_unlock();
410
	return match;
411
}
412

413
struct cgroup_subsys_state *mem_cgroup_css_from_page(struct page *page);
414
ino_t page_cgroup_ino(struct page *page);
415

416 417 418 419 420 421 422
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
{
	if (mem_cgroup_disabled())
		return true;
	return !!(memcg->css.flags & CSS_ONLINE);
}

423 424 425
/*
 * For memory reclaim.
 */
426
int mem_cgroup_select_victim_node(struct mem_cgroup *memcg);
M
Michal Hocko 已提交
427 428

void mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
429
		int zid, int nr_pages);
M
Michal Hocko 已提交
430

431 432 433
unsigned long mem_cgroup_node_nr_lru_pages(struct mem_cgroup *memcg,
					   int nid, unsigned int lru_mask);

M
Michal Hocko 已提交
434 435 436
static inline
unsigned long mem_cgroup_get_lru_size(struct lruvec *lruvec, enum lru_list lru)
{
437
	struct mem_cgroup_per_node *mz;
438 439
	unsigned long nr_pages = 0;
	int zid;
M
Michal Hocko 已提交
440

441
	mz = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
442 443 444 445 446 447 448 449 450 451 452 453 454
	for (zid = 0; zid < MAX_NR_ZONES; zid++)
		nr_pages += mz->lru_zone_size[zid][lru];
	return nr_pages;
}

static inline
unsigned long mem_cgroup_get_zone_lru_size(struct lruvec *lruvec,
		enum lru_list lru, int zone_idx)
{
	struct mem_cgroup_per_node *mz;

	mz = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
	return mz->lru_zone_size[zone_idx][lru];
M
Michal Hocko 已提交
455 456
}

457 458
void mem_cgroup_handle_over_high(void);

459 460
unsigned long mem_cgroup_get_limit(struct mem_cgroup *memcg);

461 462
void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
				struct task_struct *p);
463

464
static inline void mem_cgroup_oom_enable(void)
465
{
T
Tejun Heo 已提交
466 467
	WARN_ON(current->memcg_may_oom);
	current->memcg_may_oom = 1;
468 469
}

470
static inline void mem_cgroup_oom_disable(void)
471
{
T
Tejun Heo 已提交
472 473
	WARN_ON(!current->memcg_may_oom);
	current->memcg_may_oom = 0;
474 475
}

476 477
static inline bool task_in_memcg_oom(struct task_struct *p)
{
T
Tejun Heo 已提交
478
	return p->memcg_in_oom;
479 480
}

481
bool mem_cgroup_oom_synchronize(bool wait);
482

A
Andrew Morton 已提交
483
#ifdef CONFIG_MEMCG_SWAP
484 485
extern int do_swap_account;
#endif
486

487 488
struct mem_cgroup *lock_page_memcg(struct page *page);
void __unlock_page_memcg(struct mem_cgroup *memcg);
J
Johannes Weiner 已提交
489
void unlock_page_memcg(struct page *page);
490

491
/* idx can be of type enum memcg_stat_item or node_stat_item */
492
static inline unsigned long memcg_page_state(struct mem_cgroup *memcg,
493
					     int idx)
494 495 496 497 498 499 500 501 502 503 504 505 506
{
	long val = 0;
	int cpu;

	for_each_possible_cpu(cpu)
		val += per_cpu(memcg->stat->count[idx], cpu);

	if (val < 0)
		val = 0;

	return val;
}

507
/* idx can be of type enum memcg_stat_item or node_stat_item */
508
static inline void __mod_memcg_state(struct mem_cgroup *memcg,
509
				     int idx, int val)
510 511
{
	if (!mem_cgroup_disabled())
512
		__this_cpu_add(memcg->stat->count[idx], val);
513 514
}

515
/* idx can be of type enum memcg_stat_item or node_stat_item */
516
static inline void mod_memcg_state(struct mem_cgroup *memcg,
517
				   int idx, int val)
518
{
519 520
	if (!mem_cgroup_disabled())
		this_cpu_add(memcg->stat->count[idx], val);
521 522
}

M
Michal Hocko 已提交
523
/**
524
 * mod_memcg_page_state - update page state statistics
J
Johannes Weiner 已提交
525
 * @page: the page
M
Michal Hocko 已提交
526 527 528
 * @idx: page state item to account
 * @val: number of pages (positive or negative)
 *
529 530 531
 * The @page must be locked or the caller must use lock_page_memcg()
 * to prevent double accounting when the page is concurrently being
 * moved to another memcg:
532
 *
533
 *   lock_page(page) or lock_page_memcg(page)
534
 *   if (TestClearPageState(page))
535
 *     mod_memcg_page_state(page, state, -1);
536
 *   unlock_page(page) or unlock_page_memcg(page)
537 538
 *
 * Kernel pages are an exception to this, since they'll never move.
M
Michal Hocko 已提交
539
 */
540
static inline void __mod_memcg_page_state(struct page *page,
541
					  int idx, int val)
542 543 544 545 546
{
	if (page->mem_cgroup)
		__mod_memcg_state(page->mem_cgroup, idx, val);
}

547
static inline void mod_memcg_page_state(struct page *page,
548
					int idx, int val)
M
Michal Hocko 已提交
549
{
J
Johannes Weiner 已提交
550
	if (page->mem_cgroup)
551
		mod_memcg_state(page->mem_cgroup, idx, val);
M
Michal Hocko 已提交
552 553
}

554 555
static inline unsigned long lruvec_page_state(struct lruvec *lruvec,
					      enum node_stat_item idx)
556
{
557 558 559 560 561 562 563 564 565 566 567 568 569 570 571
	struct mem_cgroup_per_node *pn;
	long val = 0;
	int cpu;

	if (mem_cgroup_disabled())
		return node_page_state(lruvec_pgdat(lruvec), idx);

	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
	for_each_possible_cpu(cpu)
		val += per_cpu(pn->lruvec_stat->count[idx], cpu);

	if (val < 0)
		val = 0;

	return val;
572 573
}

574 575
static inline void __mod_lruvec_state(struct lruvec *lruvec,
				      enum node_stat_item idx, int val)
576
{
577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623
	struct mem_cgroup_per_node *pn;

	__mod_node_page_state(lruvec_pgdat(lruvec), idx, val);
	if (mem_cgroup_disabled())
		return;
	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
	__mod_memcg_state(pn->memcg, idx, val);
	__this_cpu_add(pn->lruvec_stat->count[idx], val);
}

static inline void mod_lruvec_state(struct lruvec *lruvec,
				    enum node_stat_item idx, int val)
{
	struct mem_cgroup_per_node *pn;

	mod_node_page_state(lruvec_pgdat(lruvec), idx, val);
	if (mem_cgroup_disabled())
		return;
	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
	mod_memcg_state(pn->memcg, idx, val);
	this_cpu_add(pn->lruvec_stat->count[idx], val);
}

static inline void __mod_lruvec_page_state(struct page *page,
					   enum node_stat_item idx, int val)
{
	struct mem_cgroup_per_node *pn;

	__mod_node_page_state(page_pgdat(page), idx, val);
	if (mem_cgroup_disabled() || !page->mem_cgroup)
		return;
	__mod_memcg_state(page->mem_cgroup, idx, val);
	pn = page->mem_cgroup->nodeinfo[page_to_nid(page)];
	__this_cpu_add(pn->lruvec_stat->count[idx], val);
}

static inline void mod_lruvec_page_state(struct page *page,
					 enum node_stat_item idx, int val)
{
	struct mem_cgroup_per_node *pn;

	mod_node_page_state(page_pgdat(page), idx, val);
	if (mem_cgroup_disabled() || !page->mem_cgroup)
		return;
	mod_memcg_state(page->mem_cgroup, idx, val);
	pn = page->mem_cgroup->nodeinfo[page_to_nid(page)];
	this_cpu_add(pn->lruvec_stat->count[idx], val);
624 625
}

626
unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order,
627 628
						gfp_t gfp_mask,
						unsigned long *total_scanned);
D
David Rientjes 已提交
629

630 631 632 633 634 635 636 637
static inline void count_memcg_events(struct mem_cgroup *memcg,
				      enum vm_event_item idx,
				      unsigned long count)
{
	if (!mem_cgroup_disabled())
		this_cpu_add(memcg->stat->events[idx], count);
}

638
/* idx can be of type enum memcg_stat_item or node_stat_item */
639
static inline void count_memcg_page_event(struct page *page,
640
					  int idx)
641 642 643 644 645 646 647
{
	if (page->mem_cgroup)
		count_memcg_events(page->mem_cgroup, idx, 1);
}

static inline void count_memcg_event_mm(struct mm_struct *mm,
					enum vm_event_item idx)
648
{
M
Michal Hocko 已提交
649 650
	struct mem_cgroup *memcg;

651 652
	if (mem_cgroup_disabled())
		return;
M
Michal Hocko 已提交
653 654 655

	rcu_read_lock();
	memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
656
	if (likely(memcg)) {
657
		this_cpu_inc(memcg->stat->events[idx]);
658 659 660
		if (idx == OOM_KILL)
			cgroup_file_notify(&memcg->events_file);
	}
M
Michal Hocko 已提交
661
	rcu_read_unlock();
662
}
663
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
664
void mem_cgroup_split_huge_fixup(struct page *head);
665 666
#endif

A
Andrew Morton 已提交
667
#else /* CONFIG_MEMCG */
668 669 670 671

#define MEM_CGROUP_ID_SHIFT	0
#define MEM_CGROUP_ID_MAX	0

672 673
struct mem_cgroup;

674 675 676 677 678
static inline bool mem_cgroup_disabled(void)
{
	return true;
}

679
static inline void mem_cgroup_event(struct mem_cgroup *memcg,
680
				    enum memcg_event_item event)
681 682 683 684 685 686 687 688 689
{
}

static inline bool mem_cgroup_low(struct mem_cgroup *root,
				  struct mem_cgroup *memcg)
{
	return false;
}

690 691
static inline int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
					gfp_t gfp_mask,
692 693
					struct mem_cgroup **memcgp,
					bool compound)
694
{
695
	*memcgp = NULL;
696 697 698
	return 0;
}

699 700
static inline void mem_cgroup_commit_charge(struct page *page,
					    struct mem_cgroup *memcg,
701
					    bool lrucare, bool compound)
702 703 704
{
}

705
static inline void mem_cgroup_cancel_charge(struct page *page,
706 707
					    struct mem_cgroup *memcg,
					    bool compound)
708 709 710
{
}

711
static inline void mem_cgroup_uncharge(struct page *page)
712 713 714
{
}

715
static inline void mem_cgroup_uncharge_list(struct list_head *page_list)
716 717 718
{
}

719
static inline void mem_cgroup_migrate(struct page *old, struct page *new)
720 721 722
{
}

723
static inline struct lruvec *mem_cgroup_lruvec(struct pglist_data *pgdat,
724
				struct mem_cgroup *memcg)
K
KAMEZAWA Hiroyuki 已提交
725
{
726
	return node_lruvec(pgdat);
K
KAMEZAWA Hiroyuki 已提交
727 728
}

729
static inline struct lruvec *mem_cgroup_page_lruvec(struct page *page,
M
Mel Gorman 已提交
730
						    struct pglist_data *pgdat)
731
{
M
Mel Gorman 已提交
732
	return &pgdat->lruvec;
733 734
}

735
static inline bool mm_match_cgroup(struct mm_struct *mm,
736
		struct mem_cgroup *memcg)
737
{
738
	return true;
739 740
}

741 742
static inline bool task_in_mem_cgroup(struct task_struct *task,
				      const struct mem_cgroup *memcg)
743
{
744
	return true;
745 746
}

747 748 749 750 751 752 753 754 755 756 757 758 759
static inline struct mem_cgroup *
mem_cgroup_iter(struct mem_cgroup *root,
		struct mem_cgroup *prev,
		struct mem_cgroup_reclaim_cookie *reclaim)
{
	return NULL;
}

static inline void mem_cgroup_iter_break(struct mem_cgroup *root,
					 struct mem_cgroup *prev)
{
}

760 761 762 763 764 765
static inline int mem_cgroup_scan_tasks(struct mem_cgroup *memcg,
		int (*fn)(struct task_struct *, void *), void *arg)
{
	return 0;
}

766
static inline unsigned short mem_cgroup_id(struct mem_cgroup *memcg)
767
{
768 769 770 771 772 773 774 775
	return 0;
}

static inline struct mem_cgroup *mem_cgroup_from_id(unsigned short id)
{
	WARN_ON_ONCE(id);
	/* XXX: This should always return root_mem_cgroup */
	return NULL;
776
}
777

778 779 780 781 782
static inline struct mem_cgroup *lruvec_memcg(struct lruvec *lruvec)
{
	return NULL;
}

783
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
784
{
785
	return true;
786 787
}

788
static inline unsigned long
789
mem_cgroup_get_lru_size(struct lruvec *lruvec, enum lru_list lru)
790 791 792
{
	return 0;
}
793 794 795 796 797 798
static inline
unsigned long mem_cgroup_get_zone_lru_size(struct lruvec *lruvec,
		enum lru_list lru, int zone_idx)
{
	return 0;
}
799

800 801 802 803 804 805 806
static inline unsigned long
mem_cgroup_node_nr_lru_pages(struct mem_cgroup *memcg,
			     int nid, unsigned int lru_mask)
{
	return 0;
}

807 808 809 810 811
static inline unsigned long mem_cgroup_get_limit(struct mem_cgroup *memcg)
{
	return 0;
}

812 813 814 815 816
static inline void
mem_cgroup_print_oom_info(struct mem_cgroup *memcg, struct task_struct *p)
{
}

817 818 819 820 821 822
static inline struct mem_cgroup *lock_page_memcg(struct page *page)
{
	return NULL;
}

static inline void __unlock_page_memcg(struct mem_cgroup *memcg)
823 824 825
{
}

J
Johannes Weiner 已提交
826
static inline void unlock_page_memcg(struct page *page)
827 828 829
{
}

830 831 832 833
static inline void mem_cgroup_handle_over_high(void)
{
}

834
static inline void mem_cgroup_oom_enable(void)
835 836 837
{
}

838
static inline void mem_cgroup_oom_disable(void)
839 840 841
{
}

842 843 844 845 846
static inline bool task_in_memcg_oom(struct task_struct *p)
{
	return false;
}

847
static inline bool mem_cgroup_oom_synchronize(bool wait)
848 849 850 851
{
	return false;
}

852
static inline unsigned long memcg_page_state(struct mem_cgroup *memcg,
853
					     int idx)
854 855 856 857
{
	return 0;
}

858
static inline void __mod_memcg_state(struct mem_cgroup *memcg,
859
				     int idx,
860
				     int nr)
861 862 863
{
}

864
static inline void mod_memcg_state(struct mem_cgroup *memcg,
865
				   int idx,
866
				   int nr)
867 868 869
{
}

870
static inline void __mod_memcg_page_state(struct page *page,
871
					  int idx,
872
					  int nr)
873 874 875
{
}

876
static inline void mod_memcg_page_state(struct page *page,
877
					int idx,
878
					int nr)
879 880 881
{
}

882 883
static inline unsigned long lruvec_page_state(struct lruvec *lruvec,
					      enum node_stat_item idx)
884
{
885
	return node_page_state(lruvec_pgdat(lruvec), idx);
886 887
}

888 889
static inline void __mod_lruvec_state(struct lruvec *lruvec,
				      enum node_stat_item idx, int val)
890
{
891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909
	__mod_node_page_state(lruvec_pgdat(lruvec), idx, val);
}

static inline void mod_lruvec_state(struct lruvec *lruvec,
				    enum node_stat_item idx, int val)
{
	mod_node_page_state(lruvec_pgdat(lruvec), idx, val);
}

static inline void __mod_lruvec_page_state(struct page *page,
					   enum node_stat_item idx, int val)
{
	__mod_node_page_state(page_pgdat(page), idx, val);
}

static inline void mod_lruvec_page_state(struct page *page,
					 enum node_stat_item idx, int val)
{
	mod_node_page_state(page_pgdat(page), idx, val);
910 911
}

912
static inline
913
unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order,
914 915
					    gfp_t gfp_mask,
					    unsigned long *total_scanned)
916
{
917
	return 0;
918 919
}

920
static inline void mem_cgroup_split_huge_fixup(struct page *head)
921 922 923
{
}

924 925 926 927 928 929 930
static inline void count_memcg_events(struct mem_cgroup *memcg,
				      enum vm_event_item idx,
				      unsigned long count)
{
}

static inline void count_memcg_page_event(struct page *page,
931
					  int idx)
932 933 934
{
}

935
static inline
936
void count_memcg_event_mm(struct mm_struct *mm, enum vm_event_item idx)
937 938
{
}
A
Andrew Morton 已提交
939
#endif /* CONFIG_MEMCG */
940

941
/* idx can be of type enum memcg_stat_item or node_stat_item */
942
static inline void __inc_memcg_state(struct mem_cgroup *memcg,
943
				     int idx)
944 945 946 947
{
	__mod_memcg_state(memcg, idx, 1);
}

948
/* idx can be of type enum memcg_stat_item or node_stat_item */
949
static inline void __dec_memcg_state(struct mem_cgroup *memcg,
950
				     int idx)
951 952 953 954
{
	__mod_memcg_state(memcg, idx, -1);
}

955
/* idx can be of type enum memcg_stat_item or node_stat_item */
956
static inline void __inc_memcg_page_state(struct page *page,
957
					  int idx)
958 959 960 961
{
	__mod_memcg_page_state(page, idx, 1);
}

962
/* idx can be of type enum memcg_stat_item or node_stat_item */
963
static inline void __dec_memcg_page_state(struct page *page,
964
					  int idx)
965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992
{
	__mod_memcg_page_state(page, idx, -1);
}

static inline void __inc_lruvec_state(struct lruvec *lruvec,
				      enum node_stat_item idx)
{
	__mod_lruvec_state(lruvec, idx, 1);
}

static inline void __dec_lruvec_state(struct lruvec *lruvec,
				      enum node_stat_item idx)
{
	__mod_lruvec_state(lruvec, idx, -1);
}

static inline void __inc_lruvec_page_state(struct page *page,
					   enum node_stat_item idx)
{
	__mod_lruvec_page_state(page, idx, 1);
}

static inline void __dec_lruvec_page_state(struct page *page,
					   enum node_stat_item idx)
{
	__mod_lruvec_page_state(page, idx, -1);
}

993
/* idx can be of type enum memcg_stat_item or node_stat_item */
994
static inline void inc_memcg_state(struct mem_cgroup *memcg,
995
				   int idx)
996 997 998 999
{
	mod_memcg_state(memcg, idx, 1);
}

1000
/* idx can be of type enum memcg_stat_item or node_stat_item */
1001
static inline void dec_memcg_state(struct mem_cgroup *memcg,
1002
				   int idx)
1003 1004 1005 1006
{
	mod_memcg_state(memcg, idx, -1);
}

1007
/* idx can be of type enum memcg_stat_item or node_stat_item */
1008
static inline void inc_memcg_page_state(struct page *page,
1009
					int idx)
1010 1011 1012 1013
{
	mod_memcg_page_state(page, idx, 1);
}

1014
/* idx can be of type enum memcg_stat_item or node_stat_item */
1015
static inline void dec_memcg_page_state(struct page *page,
1016
					int idx)
1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044
{
	mod_memcg_page_state(page, idx, -1);
}

static inline void inc_lruvec_state(struct lruvec *lruvec,
				    enum node_stat_item idx)
{
	mod_lruvec_state(lruvec, idx, 1);
}

static inline void dec_lruvec_state(struct lruvec *lruvec,
				    enum node_stat_item idx)
{
	mod_lruvec_state(lruvec, idx, -1);
}

static inline void inc_lruvec_page_state(struct page *page,
					 enum node_stat_item idx)
{
	mod_lruvec_page_state(page, idx, 1);
}

static inline void dec_lruvec_page_state(struct page *page,
					 enum node_stat_item idx)
{
	mod_lruvec_page_state(page, idx, -1);
}

1045
#ifdef CONFIG_CGROUP_WRITEBACK
T
Tejun Heo 已提交
1046

1047
struct list_head *mem_cgroup_cgwb_list(struct mem_cgroup *memcg);
T
Tejun Heo 已提交
1048
struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
1049 1050 1051
void mem_cgroup_wb_stats(struct bdi_writeback *wb, unsigned long *pfilepages,
			 unsigned long *pheadroom, unsigned long *pdirty,
			 unsigned long *pwriteback);
T
Tejun Heo 已提交
1052 1053 1054 1055 1056 1057 1058 1059

#else	/* CONFIG_CGROUP_WRITEBACK */

static inline struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb)
{
	return NULL;
}

1060
static inline void mem_cgroup_wb_stats(struct bdi_writeback *wb,
1061 1062
				       unsigned long *pfilepages,
				       unsigned long *pheadroom,
1063 1064 1065 1066 1067
				       unsigned long *pdirty,
				       unsigned long *pwriteback)
{
}

T
Tejun Heo 已提交
1068
#endif	/* CONFIG_CGROUP_WRITEBACK */
1069

G
Glauber Costa 已提交
1070
struct sock;
1071 1072
bool mem_cgroup_charge_skmem(struct mem_cgroup *memcg, unsigned int nr_pages);
void mem_cgroup_uncharge_skmem(struct mem_cgroup *memcg, unsigned int nr_pages);
1073
#ifdef CONFIG_MEMCG
1074 1075
extern struct static_key_false memcg_sockets_enabled_key;
#define mem_cgroup_sockets_enabled static_branch_unlikely(&memcg_sockets_enabled_key)
1076 1077
void mem_cgroup_sk_alloc(struct sock *sk);
void mem_cgroup_sk_free(struct sock *sk);
1078
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
1079
{
1080
	if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) && memcg->tcpmem_pressure)
1081 1082 1083 1084 1085 1086
		return true;
	do {
		if (time_before(jiffies, memcg->socket_pressure))
			return true;
	} while ((memcg = parent_mem_cgroup(memcg)));
	return false;
1087 1088
}
#else
1089
#define mem_cgroup_sockets_enabled 0
1090 1091
static inline void mem_cgroup_sk_alloc(struct sock *sk) { };
static inline void mem_cgroup_sk_free(struct sock *sk) { };
1092
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
1093 1094 1095 1096
{
	return false;
}
#endif
1097

1098 1099 1100 1101 1102 1103 1104
struct kmem_cache *memcg_kmem_get_cache(struct kmem_cache *cachep);
void memcg_kmem_put_cache(struct kmem_cache *cachep);
int memcg_kmem_charge_memcg(struct page *page, gfp_t gfp, int order,
			    struct mem_cgroup *memcg);
int memcg_kmem_charge(struct page *page, gfp_t gfp, int order);
void memcg_kmem_uncharge(struct page *page, int order);

1105
#if defined(CONFIG_MEMCG) && !defined(CONFIG_SLOB)
1106
extern struct static_key_false memcg_kmem_enabled_key;
1107
extern struct workqueue_struct *memcg_kmem_cache_wq;
1108

1109
extern int memcg_nr_cache_ids;
1110 1111
void memcg_get_cache_ids(void);
void memcg_put_cache_ids(void);
1112 1113 1114 1115 1116 1117

/*
 * Helper macro to loop through all memcg-specific caches. Callers must still
 * check if the cache is valid (it is either valid or NULL).
 * the slab_mutex must be held when looping through those caches
 */
1118
#define for_each_memcg_cache_index(_idx)	\
1119
	for ((_idx) = 0; (_idx) < memcg_nr_cache_ids; (_idx)++)
1120

1121 1122
static inline bool memcg_kmem_enabled(void)
{
1123
	return static_branch_unlikely(&memcg_kmem_enabled_key);
1124 1125
}

M
Michal Hocko 已提交
1126
/*
J
Jesper Dangaard Brouer 已提交
1127
 * helper for accessing a memcg's index. It will be used as an index in the
M
Michal Hocko 已提交
1128 1129 1130 1131 1132 1133 1134
 * child cache array in kmem_cache, and also to derive its name. This function
 * will return -1 when this is not a kmem-limited memcg.
 */
static inline int memcg_cache_id(struct mem_cgroup *memcg)
{
	return memcg ? memcg->kmemcg_id : -1;
}
1135

1136
#else
1137 1138 1139
#define for_each_memcg_cache_index(_idx)	\
	for (; NULL; )

1140 1141 1142 1143 1144
static inline bool memcg_kmem_enabled(void)
{
	return false;
}

1145 1146 1147 1148 1149
static inline int memcg_cache_id(struct mem_cgroup *memcg)
{
	return -1;
}

1150 1151 1152 1153 1154 1155 1156 1157
static inline void memcg_get_cache_ids(void)
{
}

static inline void memcg_put_cache_ids(void)
{
}

1158 1159
#endif /* CONFIG_MEMCG && !CONFIG_SLOB */

B
Balbir Singh 已提交
1160
#endif /* _LINUX_MEMCONTROL_H */