memcontrol.h 22.0 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;
166 167

	/* Legacy consumer-oriented counters */
M
Michal Hocko 已提交
168 169
	struct page_counter memsw;
	struct page_counter kmem;
170
	struct page_counter tcpmem;
M
Michal Hocko 已提交
171 172 173 174 175

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

176 177 178
	/* Range enforcement for interrupt charges */
	struct work_struct high_work;

M
Michal Hocko 已提交
179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196
	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;

197 198 199
	/* handle for "memory.events" */
	struct cgroup_file events_file;

M
Michal Hocko 已提交
200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229
	/* 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;

230 231 232
	unsigned long		socket_pressure;

	/* Legacy tcp memory accounting */
233 234
	bool			tcpmem_active;
	int			tcpmem_pressure;
235

236
#ifndef CONFIG_SLOB
M
Michal Hocko 已提交
237 238
        /* Index in the kmem_cache->memcg_params.memcg_caches array */
	int kmemcg_id;
239
	enum memcg_kmem_state kmem_state;
M
Michal Hocko 已提交
240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
#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 */
};
261 262

extern struct mem_cgroup *root_mem_cgroup;
T
Tejun Heo 已提交
263

M
Michal Hocko 已提交
264 265 266 267 268 269 270
/**
 * 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,
271
		       enum mem_cgroup_events_index idx,
M
Michal Hocko 已提交
272 273 274
		       unsigned int nr)
{
	this_cpu_add(memcg->stat->events[idx], nr);
275
	cgroup_file_notify(&memcg->events_file);
M
Michal Hocko 已提交
276
}
277 278 279

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

280
int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
281 282
			  gfp_t gfp_mask, struct mem_cgroup **memcgp,
			  bool compound);
283
void mem_cgroup_commit_charge(struct page *page, struct mem_cgroup *memcg,
284 285 286
			      bool lrucare, bool compound);
void mem_cgroup_cancel_charge(struct page *page, struct mem_cgroup *memcg,
		bool compound);
287
void mem_cgroup_uncharge(struct page *page);
288
void mem_cgroup_uncharge_list(struct list_head *page_list);
289

290
void mem_cgroup_replace_page(struct page *oldpage, struct page *newpage);
291

292 293
struct lruvec *mem_cgroup_zone_lruvec(struct zone *, struct mem_cgroup *);
struct lruvec *mem_cgroup_page_lruvec(struct page *, struct zone *);
294

295
bool task_in_mem_cgroup(struct task_struct *task, struct mem_cgroup *memcg);
296
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
297

M
Michal Hocko 已提交
298 299 300 301 302
static inline
struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css){
	return css ? container_of(css, struct mem_cgroup, css) : NULL;
}

303 304 305
#define mem_cgroup_from_counter(counter, member)	\
	container_of(counter, struct mem_cgroup, member)

M
Michal Hocko 已提交
306 307 308 309 310
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 *);

311 312 313 314 315 316 317 318 319 320 321 322 323 324
/**
 * 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
325 326 327 328 329 330 331 332 333
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
334

335 336
static inline bool mm_match_cgroup(struct mm_struct *mm,
				   struct mem_cgroup *memcg)
337
{
338
	struct mem_cgroup *task_memcg;
339
	bool match = false;
340

341
	rcu_read_lock();
342
	task_memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
343
	if (task_memcg)
344
		match = mem_cgroup_is_descendant(task_memcg, memcg);
345
	rcu_read_unlock();
346
	return match;
347
}
348

349
struct cgroup_subsys_state *mem_cgroup_css_from_page(struct page *page);
350
ino_t page_cgroup_ino(struct page *page);
351

M
Michal Hocko 已提交
352 353
static inline bool mem_cgroup_disabled(void)
{
354
	return !cgroup_subsys_enabled(memory_cgrp_subsys);
M
Michal Hocko 已提交
355
}
356

357 358 359
/*
 * For memory reclaim.
 */
360
int mem_cgroup_select_victim_node(struct mem_cgroup *memcg);
M
Michal Hocko 已提交
361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387

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

static inline bool mem_cgroup_lruvec_online(struct lruvec *lruvec)
{
	struct mem_cgroup_per_zone *mz;
	struct mem_cgroup *memcg;

	if (mem_cgroup_disabled())
		return true;

	mz = container_of(lruvec, struct mem_cgroup_per_zone, lruvec);
	memcg = mz->memcg;

	return !!(memcg->css.flags & CSS_ONLINE);
}

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];
}

388
static inline bool mem_cgroup_inactive_anon_is_low(struct lruvec *lruvec)
M
Michal Hocko 已提交
389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
{
	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;
}

407 408
void mem_cgroup_handle_over_high(void);

409 410
void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
				struct task_struct *p);
411

412
static inline void mem_cgroup_oom_enable(void)
413
{
T
Tejun Heo 已提交
414 415
	WARN_ON(current->memcg_may_oom);
	current->memcg_may_oom = 1;
416 417
}

418
static inline void mem_cgroup_oom_disable(void)
419
{
T
Tejun Heo 已提交
420 421
	WARN_ON(!current->memcg_may_oom);
	current->memcg_may_oom = 0;
422 423
}

424 425
static inline bool task_in_memcg_oom(struct task_struct *p)
{
T
Tejun Heo 已提交
426
	return p->memcg_in_oom;
427 428
}

429
bool mem_cgroup_oom_synchronize(bool wait);
430

A
Andrew Morton 已提交
431
#ifdef CONFIG_MEMCG_SWAP
432 433
extern int do_swap_account;
#endif
434

435 436
struct mem_cgroup *mem_cgroup_begin_page_stat(struct page *page);
void mem_cgroup_end_page_stat(struct mem_cgroup *memcg);
437

M
Michal Hocko 已提交
438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454
/**
 * 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);
}

455
static inline void mem_cgroup_inc_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
static inline void mem_cgroup_dec_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
462
					    enum mem_cgroup_stat_index idx)
463
{
464
	mem_cgroup_update_page_stat(memcg, idx, -1);
465 466
}

467 468 469
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
470

471 472 473
static inline void mem_cgroup_count_vm_event(struct mm_struct *mm,
					     enum vm_event_item idx)
{
M
Michal Hocko 已提交
474 475
	struct mem_cgroup *memcg;

476 477
	if (mem_cgroup_disabled())
		return;
M
Michal Hocko 已提交
478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495

	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();
496
}
497
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
498
void mem_cgroup_split_huge_fixup(struct page *head);
499 500
#endif

A
Andrew Morton 已提交
501
#else /* CONFIG_MEMCG */
502 503
struct mem_cgroup;

504 505 506 507 508 509 510 511 512 513 514 515
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;
}

516 517
static inline int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
					gfp_t gfp_mask,
518 519
					struct mem_cgroup **memcgp,
					bool compound)
520
{
521
	*memcgp = NULL;
522 523 524
	return 0;
}

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

531
static inline void mem_cgroup_cancel_charge(struct page *page,
532 533
					    struct mem_cgroup *memcg,
					    bool compound)
534 535 536
{
}

537
static inline void mem_cgroup_uncharge(struct page *page)
538 539 540
{
}

541
static inline void mem_cgroup_uncharge_list(struct list_head *page_list)
542 543 544
{
}

545
static inline void mem_cgroup_replace_page(struct page *old, struct page *new)
546 547 548
{
}

549 550
static inline struct lruvec *mem_cgroup_zone_lruvec(struct zone *zone,
						    struct mem_cgroup *memcg)
K
KAMEZAWA Hiroyuki 已提交
551
{
552
	return &zone->lruvec;
K
KAMEZAWA Hiroyuki 已提交
553 554
}

555 556
static inline struct lruvec *mem_cgroup_page_lruvec(struct page *page,
						    struct zone *zone)
557
{
558
	return &zone->lruvec;
559 560
}

561
static inline bool mm_match_cgroup(struct mm_struct *mm,
562
		struct mem_cgroup *memcg)
563
{
564
	return true;
565 566
}

567 568
static inline bool task_in_mem_cgroup(struct task_struct *task,
				      const struct mem_cgroup *memcg)
569
{
570
	return true;
571 572
}

573 574 575 576 577 578 579 580 581 582 583 584 585
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)
{
}

586 587 588 589
static inline bool mem_cgroup_disabled(void)
{
	return true;
}
590

591
static inline bool
592
mem_cgroup_inactive_anon_is_low(struct lruvec *lruvec)
593
{
594
	return true;
595 596
}

597 598 599 600 601
static inline bool mem_cgroup_lruvec_online(struct lruvec *lruvec)
{
	return true;
}

602
static inline unsigned long
603
mem_cgroup_get_lru_size(struct lruvec *lruvec, enum lru_list lru)
604 605 606 607
{
	return 0;
}

608 609 610
static inline void
mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
			      int increment)
K
KOSAKI Motohiro 已提交
611 612 613
{
}

614 615 616 617 618
static inline void
mem_cgroup_print_oom_info(struct mem_cgroup *memcg, struct task_struct *p)
{
}

619
static inline struct mem_cgroup *mem_cgroup_begin_page_stat(struct page *page)
620
{
621
	return NULL;
622 623
}

624
static inline void mem_cgroup_end_page_stat(struct mem_cgroup *memcg)
625 626 627
{
}

628 629 630 631
static inline void mem_cgroup_handle_over_high(void)
{
}

632
static inline void mem_cgroup_oom_enable(void)
633 634 635
{
}

636
static inline void mem_cgroup_oom_disable(void)
637 638 639
{
}

640 641 642 643 644
static inline bool task_in_memcg_oom(struct task_struct *p)
{
	return false;
}

645
static inline bool mem_cgroup_oom_synchronize(bool wait)
646 647 648 649
{
	return false;
}

650
static inline void mem_cgroup_inc_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
651
					    enum mem_cgroup_stat_index idx)
652 653 654
{
}

655
static inline void mem_cgroup_dec_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
656
					    enum mem_cgroup_stat_index idx)
657 658 659
{
}

660
static inline
661 662 663
unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
					    gfp_t gfp_mask,
					    unsigned long *total_scanned)
664
{
665
	return 0;
666 667
}

668
static inline void mem_cgroup_split_huge_fixup(struct page *head)
669 670 671
{
}

672 673 674 675
static inline
void mem_cgroup_count_vm_event(struct mm_struct *mm, enum vm_event_item idx)
{
}
A
Andrew Morton 已提交
676
#endif /* CONFIG_MEMCG */
677

678
#ifdef CONFIG_CGROUP_WRITEBACK
T
Tejun Heo 已提交
679

680
struct list_head *mem_cgroup_cgwb_list(struct mem_cgroup *memcg);
T
Tejun Heo 已提交
681
struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
682 683 684
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
685 686 687 688 689 690 691 692

#else	/* CONFIG_CGROUP_WRITEBACK */

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

693
static inline void mem_cgroup_wb_stats(struct bdi_writeback *wb,
694 695
				       unsigned long *pfilepages,
				       unsigned long *pheadroom,
696 697 698 699 700
				       unsigned long *pdirty,
				       unsigned long *pwriteback)
{
}

T
Tejun Heo 已提交
701
#endif	/* CONFIG_CGROUP_WRITEBACK */
702

G
Glauber Costa 已提交
703 704 705
struct sock;
void sock_update_memcg(struct sock *sk);
void sock_release_memcg(struct sock *sk);
706 707
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);
708
#ifdef CONFIG_MEMCG
709 710
extern struct static_key_false memcg_sockets_enabled_key;
#define mem_cgroup_sockets_enabled static_branch_unlikely(&memcg_sockets_enabled_key)
711
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
712
{
713
	if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) && memcg->tcpmem_pressure)
714 715 716 717 718 719
		return true;
	do {
		if (time_before(jiffies, memcg->socket_pressure))
			return true;
	} while ((memcg = parent_mem_cgroup(memcg)));
	return false;
720 721
}
#else
722
#define mem_cgroup_sockets_enabled 0
723
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
724 725 726 727
{
	return false;
}
#endif
728

729
#if defined(CONFIG_MEMCG) && !defined(CONFIG_SLOB)
730
extern struct static_key_false memcg_kmem_enabled_key;
731

732
extern int memcg_nr_cache_ids;
733 734
void memcg_get_cache_ids(void);
void memcg_put_cache_ids(void);
735 736 737 738 739 740

/*
 * 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
 */
741
#define for_each_memcg_cache_index(_idx)	\
742
	for ((_idx) = 0; (_idx) < memcg_nr_cache_ids; (_idx)++)
743

744 745
static inline bool memcg_kmem_enabled(void)
{
746
	return static_branch_unlikely(&memcg_kmem_enabled_key);
747 748
}

749
static inline bool memcg_kmem_online(struct mem_cgroup *memcg)
M
Michal Hocko 已提交
750
{
751
	return memcg->kmem_state == KMEM_ONLINE;
M
Michal Hocko 已提交
752
}
753

754 755 756 757 758 759 760 761 762 763 764
/*
 * 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.
 */
765 766
int __memcg_kmem_charge_memcg(struct page *page, gfp_t gfp, int order,
			      struct mem_cgroup *memcg);
767 768
int __memcg_kmem_charge(struct page *page, gfp_t gfp, int order);
void __memcg_kmem_uncharge(struct page *page, int order);
769

M
Michal Hocko 已提交
770 771 772 773 774 775 776 777 778
/*
 * 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;
}
779

V
Vladimir Davydov 已提交
780
struct kmem_cache *__memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp);
781
void __memcg_kmem_put_cache(struct kmem_cache *cachep);
782

V
Vladimir Davydov 已提交
783
static inline bool __memcg_kmem_bypass(void)
784 785 786 787 788
{
	if (!memcg_kmem_enabled())
		return true;
	if (in_interrupt() || (!current->mm) || (current->flags & PF_KTHREAD))
		return true;
789
	return false;
790 791 792
}

/**
793 794 795 796
 * memcg_kmem_charge: charge a kmem page
 * @page: page to charge
 * @gfp: reclaim mode
 * @order: allocation order
797
 *
798
 * Returns 0 on success, an error code on failure.
799
 */
800 801
static __always_inline int memcg_kmem_charge(struct page *page,
					     gfp_t gfp, int order)
802
{
V
Vladimir Davydov 已提交
803 804 805
	if (__memcg_kmem_bypass())
		return 0;
	if (!(gfp & __GFP_ACCOUNT))
806 807
		return 0;
	return __memcg_kmem_charge(page, gfp, order);
808 809 810
}

/**
811 812 813
 * memcg_kmem_uncharge: uncharge a kmem page
 * @page: page to uncharge
 * @order: allocation order
814
 */
815
static __always_inline void memcg_kmem_uncharge(struct page *page, int order)
816 817
{
	if (memcg_kmem_enabled())
818
		__memcg_kmem_uncharge(page, order);
819 820
}

821 822 823 824
/**
 * memcg_kmem_get_cache: selects the correct per-memcg cache for allocation
 * @cachep: the original global kmem cache
 *
825
 * All memory allocated from a per-memcg cache is charged to the owner memcg.
826 827 828 829
 */
static __always_inline struct kmem_cache *
memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp)
{
V
Vladimir Davydov 已提交
830
	if (__memcg_kmem_bypass())
831
		return cachep;
V
Vladimir Davydov 已提交
832
	return __memcg_kmem_get_cache(cachep, gfp);
833
}
834 835 836 837 838 839

static __always_inline void memcg_kmem_put_cache(struct kmem_cache *cachep)
{
	if (memcg_kmem_enabled())
		__memcg_kmem_put_cache(cachep);
}
840
#else
841 842 843
#define for_each_memcg_cache_index(_idx)	\
	for (; NULL; )

844 845 846 847 848
static inline bool memcg_kmem_enabled(void)
{
	return false;
}

849
static inline bool memcg_kmem_online(struct mem_cgroup *memcg)
850 851 852 853
{
	return false;
}

854
static inline int memcg_kmem_charge(struct page *page, gfp_t gfp, int order)
855
{
856
	return 0;
857 858
}

859
static inline void memcg_kmem_uncharge(struct page *page, int order)
860 861
{
}
862 863 864 865 866 867

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

868 869 870 871 872 873 874 875
static inline void memcg_get_cache_ids(void)
{
}

static inline void memcg_put_cache_ids(void)
{
}

876 877 878 879 880
static inline struct kmem_cache *
memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp)
{
	return cachep;
}
881 882 883 884

static inline void memcg_kmem_put_cache(struct kmem_cache *cachep)
{
}
885 886
#endif /* CONFIG_MEMCG && !CONFIG_SLOB */

B
Balbir Singh 已提交
887
#endif /* _LINUX_MEMCONTROL_H */