memcontrol.h 21.9 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 29 30
#include <linux/page_counter.h>
#include <linux/vmpressure.h>
#include <linux/eventfd.h>
#include <linux/mmzone.h>
#include <linux/writeback.h>
31

32
struct mem_cgroup;
33 34
struct page;
struct mm_struct;
35
struct kmem_cache;
36

S
Sha Zhengju 已提交
37 38 39 40 41 42 43 44 45 46 47 48
/*
 * The corresponding mem_cgroup_stat_names is defined in mm/memcontrol.c,
 * These two lists should keep in accord with each other.
 */
enum mem_cgroup_stat_index {
	/*
	 * For MEM_CONTAINER_TYPE_ALL, usage = pagecache + rss.
	 */
	MEM_CGROUP_STAT_CACHE,		/* # of pages charged as cache */
	MEM_CGROUP_STAT_RSS,		/* # of pages charged as anon rss */
	MEM_CGROUP_STAT_RSS_HUGE,	/* # of pages charged as anon huge */
	MEM_CGROUP_STAT_FILE_MAPPED,	/* # of pages charged as file rss */
49
	MEM_CGROUP_STAT_DIRTY,          /* # of dirty pages in page cache */
50
	MEM_CGROUP_STAT_WRITEBACK,	/* # of pages under writeback */
S
Sha Zhengju 已提交
51 52
	MEM_CGROUP_STAT_SWAP,		/* # of pages, swapped out */
	MEM_CGROUP_STAT_NSTATS,
53 54
};

55 56 57 58 59 60
struct mem_cgroup_reclaim_cookie {
	struct zone *zone;
	int priority;
	unsigned int generation;
};

61 62 63 64 65 66 67 68 69 70 71 72 73 74
enum mem_cgroup_events_index {
	MEM_CGROUP_EVENTS_PGPGIN,	/* # of pages paged in */
	MEM_CGROUP_EVENTS_PGPGOUT,	/* # of pages paged out */
	MEM_CGROUP_EVENTS_PGFAULT,	/* # of page-faults */
	MEM_CGROUP_EVENTS_PGMAJFAULT,	/* # of major page-faults */
	MEM_CGROUP_EVENTS_NSTATS,
	/* default hierarchy events */
	MEMCG_LOW = MEM_CGROUP_EVENTS_NSTATS,
	MEMCG_HIGH,
	MEMCG_MAX,
	MEMCG_OOM,
	MEMCG_NR_EVENTS,
};

M
Michal Hocko 已提交
75 76 77 78 79 80 81 82 83 84 85 86 87
/*
 * 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,
};

A
Andrew Morton 已提交
88
#ifdef CONFIG_MEMCG
M
Michal Hocko 已提交
89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
struct mem_cgroup_stat_cpu {
	long count[MEM_CGROUP_STAT_NSTATS];
	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;
};

/*
 * per-zone information in memory controller.
 */
struct mem_cgroup_per_zone {
	struct lruvec		lruvec;
	unsigned long		lru_size[NR_LRU_LISTS];

	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_per_node {
	struct mem_cgroup_per_zone zoneinfo[MAX_NR_ZONES];
};

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;
};

149 150 151 152 153 154
enum memcg_kmem_state {
	KMEM_NONE,
	KMEM_ALLOCATED,
	KMEM_ONLINE,
};

M
Michal Hocko 已提交
155 156 157 158 159 160 161 162 163 164 165
/*
 * 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;

	/* Accounted resources */
	struct page_counter memory;
166
	struct page_counter swap;
167 168

	/* Legacy consumer-oriented counters */
M
Michal Hocko 已提交
169 170
	struct page_counter memsw;
	struct page_counter kmem;
171
	struct page_counter tcpmem;
M
Michal Hocko 已提交
172 173 174 175 176

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

177 178 179
	/* Range enforcement for interrupt charges */
	struct work_struct high_work;

M
Michal Hocko 已提交
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197
	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;

198 199 200
	/* handle for "memory.events" */
	struct cgroup_file events_file;

M
Michal Hocko 已提交
201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
	/* 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;

231 232 233
	unsigned long		socket_pressure;

	/* Legacy tcp memory accounting */
234 235
	bool			tcpmem_active;
	int			tcpmem_pressure;
236

237
#ifndef CONFIG_SLOB
M
Michal Hocko 已提交
238 239
        /* Index in the kmem_cache->memcg_params.memcg_caches array */
	int kmemcg_id;
240
	enum memcg_kmem_state kmem_state;
M
Michal Hocko 已提交
241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261
#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 */
};
262 263

extern struct mem_cgroup *root_mem_cgroup;
T
Tejun Heo 已提交
264

M
Michal Hocko 已提交
265 266 267 268 269 270 271
/**
 * mem_cgroup_events - count memory events against a cgroup
 * @memcg: the memory cgroup
 * @idx: the event index
 * @nr: the number of events to account for
 */
static inline void mem_cgroup_events(struct mem_cgroup *memcg,
272
		       enum mem_cgroup_events_index idx,
M
Michal Hocko 已提交
273 274 275
		       unsigned int nr)
{
	this_cpu_add(memcg->stat->events[idx], nr);
276
	cgroup_file_notify(&memcg->events_file);
M
Michal Hocko 已提交
277
}
278 279 280

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

281
int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
282 283
			  gfp_t gfp_mask, struct mem_cgroup **memcgp,
			  bool compound);
284
void mem_cgroup_commit_charge(struct page *page, struct mem_cgroup *memcg,
285 286 287
			      bool lrucare, bool compound);
void mem_cgroup_cancel_charge(struct page *page, struct mem_cgroup *memcg,
		bool compound);
288
void mem_cgroup_uncharge(struct page *page);
289
void mem_cgroup_uncharge_list(struct list_head *page_list);
290

291
void mem_cgroup_replace_page(struct page *oldpage, struct page *newpage);
292

293 294
struct lruvec *mem_cgroup_zone_lruvec(struct zone *, struct mem_cgroup *);
struct lruvec *mem_cgroup_page_lruvec(struct page *, struct zone *);
295

296
bool task_in_mem_cgroup(struct task_struct *task, struct mem_cgroup *memcg);
297
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
298

M
Michal Hocko 已提交
299 300 301 302 303
static inline
struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css){
	return css ? container_of(css, struct mem_cgroup, css) : NULL;
}

304 305 306
#define mem_cgroup_from_counter(counter, member)	\
	container_of(counter, struct mem_cgroup, member)

M
Michal Hocko 已提交
307 308 309 310 311
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 *);

312 313 314 315 316 317 318 319 320 321 322 323 324 325
/**
 * 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
326 327 328 329 330 331 332 333 334
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
335

336 337
static inline bool mm_match_cgroup(struct mm_struct *mm,
				   struct mem_cgroup *memcg)
338
{
339
	struct mem_cgroup *task_memcg;
340
	bool match = false;
341

342
	rcu_read_lock();
343
	task_memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
344
	if (task_memcg)
345
		match = mem_cgroup_is_descendant(task_memcg, memcg);
346
	rcu_read_unlock();
347
	return match;
348
}
349

350
struct cgroup_subsys_state *mem_cgroup_css_from_page(struct page *page);
351
ino_t page_cgroup_ino(struct page *page);
352

M
Michal Hocko 已提交
353 354
static inline bool mem_cgroup_disabled(void)
{
355
	return !cgroup_subsys_enabled(memory_cgrp_subsys);
M
Michal Hocko 已提交
356
}
357

358 359 360 361 362 363 364
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
{
	if (mem_cgroup_disabled())
		return true;
	return !!(memcg->css.flags & CSS_ONLINE);
}

365 366 367
/*
 * For memory reclaim.
 */
368
int mem_cgroup_select_victim_node(struct mem_cgroup *memcg);
M
Michal Hocko 已提交
369 370 371 372 373 374 375 376 377 378 379 380 381

void mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
		int nr_pages);

static inline
unsigned long mem_cgroup_get_lru_size(struct lruvec *lruvec, enum lru_list lru)
{
	struct mem_cgroup_per_zone *mz;

	mz = container_of(lruvec, struct mem_cgroup_per_zone, lruvec);
	return mz->lru_size[lru];
}

382
static inline bool mem_cgroup_inactive_anon_is_low(struct lruvec *lruvec)
M
Michal Hocko 已提交
383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
{
	unsigned long inactive_ratio;
	unsigned long inactive;
	unsigned long active;
	unsigned long gb;

	inactive = mem_cgroup_get_lru_size(lruvec, LRU_INACTIVE_ANON);
	active = mem_cgroup_get_lru_size(lruvec, LRU_ACTIVE_ANON);

	gb = (inactive + active) >> (30 - PAGE_SHIFT);
	if (gb)
		inactive_ratio = int_sqrt(10 * gb);
	else
		inactive_ratio = 1;

	return inactive * inactive_ratio < active;
}

401 402
void mem_cgroup_handle_over_high(void);

403 404
void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
				struct task_struct *p);
405

406
static inline void mem_cgroup_oom_enable(void)
407
{
T
Tejun Heo 已提交
408 409
	WARN_ON(current->memcg_may_oom);
	current->memcg_may_oom = 1;
410 411
}

412
static inline void mem_cgroup_oom_disable(void)
413
{
T
Tejun Heo 已提交
414 415
	WARN_ON(!current->memcg_may_oom);
	current->memcg_may_oom = 0;
416 417
}

418 419
static inline bool task_in_memcg_oom(struct task_struct *p)
{
T
Tejun Heo 已提交
420
	return p->memcg_in_oom;
421 422
}

423
bool mem_cgroup_oom_synchronize(bool wait);
424

A
Andrew Morton 已提交
425
#ifdef CONFIG_MEMCG_SWAP
426 427
extern int do_swap_account;
#endif
428

429 430
struct mem_cgroup *mem_cgroup_begin_page_stat(struct page *page);
void mem_cgroup_end_page_stat(struct mem_cgroup *memcg);
431

M
Michal Hocko 已提交
432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448
/**
 * mem_cgroup_update_page_stat - update page state statistics
 * @memcg: memcg to account against
 * @idx: page state item to account
 * @val: number of pages (positive or negative)
 *
 * See mem_cgroup_begin_page_stat() for locking requirements.
 */
static inline void mem_cgroup_update_page_stat(struct mem_cgroup *memcg,
				 enum mem_cgroup_stat_index idx, int val)
{
	VM_BUG_ON(!rcu_read_lock_held());

	if (memcg)
		this_cpu_add(memcg->stat->count[idx], val);
}

449
static inline void mem_cgroup_inc_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
450
					    enum mem_cgroup_stat_index idx)
451
{
452
	mem_cgroup_update_page_stat(memcg, idx, 1);
453 454
}

455
static inline void mem_cgroup_dec_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
456
					    enum mem_cgroup_stat_index idx)
457
{
458
	mem_cgroup_update_page_stat(memcg, idx, -1);
459 460
}

461 462 463
unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
						gfp_t gfp_mask,
						unsigned long *total_scanned);
D
David Rientjes 已提交
464

465 466 467
static inline void mem_cgroup_count_vm_event(struct mm_struct *mm,
					     enum vm_event_item idx)
{
M
Michal Hocko 已提交
468 469
	struct mem_cgroup *memcg;

470 471
	if (mem_cgroup_disabled())
		return;
M
Michal Hocko 已提交
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489

	rcu_read_lock();
	memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
	if (unlikely(!memcg))
		goto out;

	switch (idx) {
	case PGFAULT:
		this_cpu_inc(memcg->stat->events[MEM_CGROUP_EVENTS_PGFAULT]);
		break;
	case PGMAJFAULT:
		this_cpu_inc(memcg->stat->events[MEM_CGROUP_EVENTS_PGMAJFAULT]);
		break;
	default:
		BUG();
	}
out:
	rcu_read_unlock();
490
}
491
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
492
void mem_cgroup_split_huge_fixup(struct page *head);
493 494
#endif

A
Andrew Morton 已提交
495
#else /* CONFIG_MEMCG */
496 497
struct mem_cgroup;

498 499 500 501 502 503 504 505 506 507 508 509
static inline void mem_cgroup_events(struct mem_cgroup *memcg,
				     enum mem_cgroup_events_index idx,
				     unsigned int nr)
{
}

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

510 511
static inline int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
					gfp_t gfp_mask,
512 513
					struct mem_cgroup **memcgp,
					bool compound)
514
{
515
	*memcgp = NULL;
516 517 518
	return 0;
}

519 520
static inline void mem_cgroup_commit_charge(struct page *page,
					    struct mem_cgroup *memcg,
521
					    bool lrucare, bool compound)
522 523 524
{
}

525
static inline void mem_cgroup_cancel_charge(struct page *page,
526 527
					    struct mem_cgroup *memcg,
					    bool compound)
528 529 530
{
}

531
static inline void mem_cgroup_uncharge(struct page *page)
532 533 534
{
}

535
static inline void mem_cgroup_uncharge_list(struct list_head *page_list)
536 537 538
{
}

539
static inline void mem_cgroup_replace_page(struct page *old, struct page *new)
540 541 542
{
}

543 544
static inline struct lruvec *mem_cgroup_zone_lruvec(struct zone *zone,
						    struct mem_cgroup *memcg)
K
KAMEZAWA Hiroyuki 已提交
545
{
546
	return &zone->lruvec;
K
KAMEZAWA Hiroyuki 已提交
547 548
}

549 550
static inline struct lruvec *mem_cgroup_page_lruvec(struct page *page,
						    struct zone *zone)
551
{
552
	return &zone->lruvec;
553 554
}

555
static inline bool mm_match_cgroup(struct mm_struct *mm,
556
		struct mem_cgroup *memcg)
557
{
558
	return true;
559 560
}

561 562
static inline bool task_in_mem_cgroup(struct task_struct *task,
				      const struct mem_cgroup *memcg)
563
{
564
	return true;
565 566
}

567 568 569 570 571 572 573 574 575 576 577 578 579
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)
{
}

580 581 582 583
static inline bool mem_cgroup_disabled(void)
{
	return true;
}
584

585
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
586
{
587
	return true;
588 589
}

590 591
static inline bool
mem_cgroup_inactive_anon_is_low(struct lruvec *lruvec)
592 593 594 595
{
	return true;
}

596
static inline unsigned long
597
mem_cgroup_get_lru_size(struct lruvec *lruvec, enum lru_list lru)
598 599 600 601
{
	return 0;
}

602 603 604
static inline void
mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
			      int increment)
K
KOSAKI Motohiro 已提交
605 606 607
{
}

608 609 610 611 612
static inline void
mem_cgroup_print_oom_info(struct mem_cgroup *memcg, struct task_struct *p)
{
}

613
static inline struct mem_cgroup *mem_cgroup_begin_page_stat(struct page *page)
614
{
615
	return NULL;
616 617
}

618
static inline void mem_cgroup_end_page_stat(struct mem_cgroup *memcg)
619 620 621
{
}

622 623 624 625
static inline void mem_cgroup_handle_over_high(void)
{
}

626
static inline void mem_cgroup_oom_enable(void)
627 628 629
{
}

630
static inline void mem_cgroup_oom_disable(void)
631 632 633
{
}

634 635 636 637 638
static inline bool task_in_memcg_oom(struct task_struct *p)
{
	return false;
}

639
static inline bool mem_cgroup_oom_synchronize(bool wait)
640 641 642 643
{
	return false;
}

644
static inline void mem_cgroup_inc_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
645
					    enum mem_cgroup_stat_index idx)
646 647 648
{
}

649
static inline void mem_cgroup_dec_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
650
					    enum mem_cgroup_stat_index idx)
651 652 653
{
}

654
static inline
655 656 657
unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
					    gfp_t gfp_mask,
					    unsigned long *total_scanned)
658
{
659
	return 0;
660 661
}

662
static inline void mem_cgroup_split_huge_fixup(struct page *head)
663 664 665
{
}

666 667 668 669
static inline
void mem_cgroup_count_vm_event(struct mm_struct *mm, enum vm_event_item idx)
{
}
A
Andrew Morton 已提交
670
#endif /* CONFIG_MEMCG */
671

672
#ifdef CONFIG_CGROUP_WRITEBACK
T
Tejun Heo 已提交
673

674
struct list_head *mem_cgroup_cgwb_list(struct mem_cgroup *memcg);
T
Tejun Heo 已提交
675
struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
676 677 678
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
679 680 681 682 683 684 685 686

#else	/* CONFIG_CGROUP_WRITEBACK */

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

687
static inline void mem_cgroup_wb_stats(struct bdi_writeback *wb,
688 689
				       unsigned long *pfilepages,
				       unsigned long *pheadroom,
690 691 692 693 694
				       unsigned long *pdirty,
				       unsigned long *pwriteback)
{
}

T
Tejun Heo 已提交
695
#endif	/* CONFIG_CGROUP_WRITEBACK */
696

G
Glauber Costa 已提交
697 698 699
struct sock;
void sock_update_memcg(struct sock *sk);
void sock_release_memcg(struct sock *sk);
700 701
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);
702
#ifdef CONFIG_MEMCG
703 704
extern struct static_key_false memcg_sockets_enabled_key;
#define mem_cgroup_sockets_enabled static_branch_unlikely(&memcg_sockets_enabled_key)
705
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
706
{
707
	if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) && memcg->tcpmem_pressure)
708 709 710 711 712 713
		return true;
	do {
		if (time_before(jiffies, memcg->socket_pressure))
			return true;
	} while ((memcg = parent_mem_cgroup(memcg)));
	return false;
714 715
}
#else
716
#define mem_cgroup_sockets_enabled 0
717
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
718 719 720 721
{
	return false;
}
#endif
722

723
#if defined(CONFIG_MEMCG) && !defined(CONFIG_SLOB)
724
extern struct static_key_false memcg_kmem_enabled_key;
725

726
extern int memcg_nr_cache_ids;
727 728
void memcg_get_cache_ids(void);
void memcg_put_cache_ids(void);
729 730 731 732 733 734

/*
 * 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
 */
735
#define for_each_memcg_cache_index(_idx)	\
736
	for ((_idx) = 0; (_idx) < memcg_nr_cache_ids; (_idx)++)
737

738 739
static inline bool memcg_kmem_enabled(void)
{
740
	return static_branch_unlikely(&memcg_kmem_enabled_key);
741 742
}

743
static inline bool memcg_kmem_online(struct mem_cgroup *memcg)
M
Michal Hocko 已提交
744
{
745
	return memcg->kmem_state == KMEM_ONLINE;
M
Michal Hocko 已提交
746
}
747

748 749 750 751 752 753 754 755 756 757 758
/*
 * In general, we'll do everything in our power to not incur in any overhead
 * for non-memcg users for the kmem functions. Not even a function call, if we
 * can avoid it.
 *
 * Therefore, we'll inline all those functions so that in the best case, we'll
 * see that kmemcg is off for everybody and proceed quickly.  If it is on,
 * we'll still do most of the flag checking inline. We check a lot of
 * conditions, but because they are pretty simple, they are expected to be
 * fast.
 */
759 760
int __memcg_kmem_charge_memcg(struct page *page, gfp_t gfp, int order,
			      struct mem_cgroup *memcg);
761 762
int __memcg_kmem_charge(struct page *page, gfp_t gfp, int order);
void __memcg_kmem_uncharge(struct page *page, int order);
763

M
Michal Hocko 已提交
764 765 766 767 768 769 770 771 772
/*
 * helper for acessing a memcg's index. It will be used as an index in the
 * 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;
}
773

V
Vladimir Davydov 已提交
774
struct kmem_cache *__memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp);
775
void __memcg_kmem_put_cache(struct kmem_cache *cachep);
776

V
Vladimir Davydov 已提交
777
static inline bool __memcg_kmem_bypass(void)
778 779 780 781 782
{
	if (!memcg_kmem_enabled())
		return true;
	if (in_interrupt() || (!current->mm) || (current->flags & PF_KTHREAD))
		return true;
783
	return false;
784 785 786
}

/**
787 788 789 790
 * memcg_kmem_charge: charge a kmem page
 * @page: page to charge
 * @gfp: reclaim mode
 * @order: allocation order
791
 *
792
 * Returns 0 on success, an error code on failure.
793
 */
794 795
static __always_inline int memcg_kmem_charge(struct page *page,
					     gfp_t gfp, int order)
796
{
V
Vladimir Davydov 已提交
797 798 799
	if (__memcg_kmem_bypass())
		return 0;
	if (!(gfp & __GFP_ACCOUNT))
800 801
		return 0;
	return __memcg_kmem_charge(page, gfp, order);
802 803 804
}

/**
805 806 807
 * memcg_kmem_uncharge: uncharge a kmem page
 * @page: page to uncharge
 * @order: allocation order
808
 */
809
static __always_inline void memcg_kmem_uncharge(struct page *page, int order)
810 811
{
	if (memcg_kmem_enabled())
812
		__memcg_kmem_uncharge(page, order);
813 814
}

815 816 817 818
/**
 * memcg_kmem_get_cache: selects the correct per-memcg cache for allocation
 * @cachep: the original global kmem cache
 *
819
 * All memory allocated from a per-memcg cache is charged to the owner memcg.
820 821 822 823
 */
static __always_inline struct kmem_cache *
memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp)
{
V
Vladimir Davydov 已提交
824
	if (__memcg_kmem_bypass())
825
		return cachep;
V
Vladimir Davydov 已提交
826
	return __memcg_kmem_get_cache(cachep, gfp);
827
}
828 829 830 831 832 833

static __always_inline void memcg_kmem_put_cache(struct kmem_cache *cachep)
{
	if (memcg_kmem_enabled())
		__memcg_kmem_put_cache(cachep);
}
834
#else
835 836 837
#define for_each_memcg_cache_index(_idx)	\
	for (; NULL; )

838 839 840 841 842
static inline bool memcg_kmem_enabled(void)
{
	return false;
}

843
static inline bool memcg_kmem_online(struct mem_cgroup *memcg)
844 845 846 847
{
	return false;
}

848
static inline int memcg_kmem_charge(struct page *page, gfp_t gfp, int order)
849
{
850
	return 0;
851 852
}

853
static inline void memcg_kmem_uncharge(struct page *page, int order)
854 855
{
}
856 857 858 859 860 861

static inline int memcg_cache_id(struct mem_cgroup *memcg)
{
	return -1;
}

862 863 864 865 866 867 868 869
static inline void memcg_get_cache_ids(void)
{
}

static inline void memcg_put_cache_ids(void)
{
}

870 871 872 873 874
static inline struct kmem_cache *
memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp)
{
	return cachep;
}
875 876 877 878

static inline void memcg_kmem_put_cache(struct kmem_cache *cachep)
{
}
879 880
#endif /* CONFIG_MEMCG && !CONFIG_SLOB */

B
Balbir Singh 已提交
881
#endif /* _LINUX_MEMCONTROL_H */