memcontrol.h 21.0 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 88 89 90
/*
 * 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 cg_proto {
	struct page_counter	memory_allocated;	/* Current allocated memory. */
	int			memory_pressure;
91
	bool			active;
M
Michal Hocko 已提交
92 93 94 95 96 97 98 99 100 101 102 103
	/*
	 * memcg field is used to find which memcg we belong directly
	 * Each memcg struct can hold more than one cg_proto, so container_of
	 * won't really cut.
	 *
	 * The elegant solution would be having an inverse function to
	 * proto_cgroup in struct proto, but that means polluting the structure
	 * for everybody, instead of just for memcg users.
	 */
	struct mem_cgroup	*memcg;
};

A
Andrew Morton 已提交
104
#ifdef CONFIG_MEMCG
M
Michal Hocko 已提交
105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203
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;
};

/*
 * 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;
	struct page_counter memsw;
	struct page_counter kmem;

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

	unsigned long soft_limit;

	/* vmpressure notifications */
	struct vmpressure vmpressure;

	/* css_online() has been completed */
	int initialized;

	/*
	 * 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;

204 205 206
	/* handle for "memory.events" */
	struct cgroup_file events_file;

M
Michal Hocko 已提交
207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
	/* 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;

#if defined(CONFIG_MEMCG_KMEM) && defined(CONFIG_INET)
	struct cg_proto tcp_mem;
#endif
#if defined(CONFIG_MEMCG_KMEM)
        /* Index in the kmem_cache->memcg_params.memcg_caches array */
	int kmemcg_id;
	bool kmem_acct_activated;
	bool kmem_acct_active;
#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 */
};
266 267

extern struct mem_cgroup *root_mem_cgroup;
T
Tejun Heo 已提交
268

M
Michal Hocko 已提交
269 270 271 272 273 274 275
/**
 * 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,
276
		       enum mem_cgroup_events_index idx,
M
Michal Hocko 已提交
277 278 279
		       unsigned int nr)
{
	this_cpu_add(memcg->stat->events[idx], nr);
280
	cgroup_file_notify(&memcg->events_file);
M
Michal Hocko 已提交
281
}
282 283 284

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

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

293
void mem_cgroup_replace_page(struct page *oldpage, struct page *newpage);
294

295 296
struct lruvec *mem_cgroup_zone_lruvec(struct zone *, struct mem_cgroup *);
struct lruvec *mem_cgroup_page_lruvec(struct page *, struct zone *);
297

298
bool task_in_mem_cgroup(struct task_struct *task, struct mem_cgroup *memcg);
299 300
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *memcg);
301

M
Michal Hocko 已提交
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
static inline
struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css){
	return css ? container_of(css, struct mem_cgroup, css) : NULL;
}

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 *);

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
321

322 323
static inline bool mm_match_cgroup(struct mm_struct *mm,
				   struct mem_cgroup *memcg)
324
{
325
	struct mem_cgroup *task_memcg;
326
	bool match = false;
327

328
	rcu_read_lock();
329
	task_memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
330
	if (task_memcg)
331
		match = mem_cgroup_is_descendant(task_memcg, memcg);
332
	rcu_read_unlock();
333
	return match;
334
}
335

336
struct cgroup_subsys_state *mem_cgroup_css_from_page(struct page *page);
337
ino_t page_cgroup_ino(struct page *page);
338

M
Michal Hocko 已提交
339 340
static inline bool mem_cgroup_disabled(void)
{
341
	return !cgroup_subsys_enabled(memory_cgrp_subsys);
M
Michal Hocko 已提交
342
}
343

344 345 346
/*
 * For memory reclaim.
 */
347
int mem_cgroup_select_victim_node(struct mem_cgroup *memcg);
M
Michal Hocko 已提交
348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374

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];
}

375
static inline bool mem_cgroup_inactive_anon_is_low(struct lruvec *lruvec)
M
Michal Hocko 已提交
376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393
{
	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;
}

394 395
void mem_cgroup_handle_over_high(void);

396 397
void mem_cgroup_print_oom_info(struct mem_cgroup *memcg,
				struct task_struct *p);
398

399
static inline void mem_cgroup_oom_enable(void)
400
{
T
Tejun Heo 已提交
401 402
	WARN_ON(current->memcg_may_oom);
	current->memcg_may_oom = 1;
403 404
}

405
static inline void mem_cgroup_oom_disable(void)
406
{
T
Tejun Heo 已提交
407 408
	WARN_ON(!current->memcg_may_oom);
	current->memcg_may_oom = 0;
409 410
}

411 412
static inline bool task_in_memcg_oom(struct task_struct *p)
{
T
Tejun Heo 已提交
413
	return p->memcg_in_oom;
414 415
}

416
bool mem_cgroup_oom_synchronize(bool wait);
417

A
Andrew Morton 已提交
418
#ifdef CONFIG_MEMCG_SWAP
419 420
extern int do_swap_account;
#endif
421

422 423
struct mem_cgroup *mem_cgroup_begin_page_stat(struct page *page);
void mem_cgroup_end_page_stat(struct mem_cgroup *memcg);
424

M
Michal Hocko 已提交
425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441
/**
 * 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);
}

442
static inline void mem_cgroup_inc_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
443
					    enum mem_cgroup_stat_index idx)
444
{
445
	mem_cgroup_update_page_stat(memcg, idx, 1);
446 447
}

448
static inline void mem_cgroup_dec_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
449
					    enum mem_cgroup_stat_index idx)
450
{
451
	mem_cgroup_update_page_stat(memcg, idx, -1);
452 453
}

454 455 456
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
457

458 459 460
static inline void mem_cgroup_count_vm_event(struct mm_struct *mm,
					     enum vm_event_item idx)
{
M
Michal Hocko 已提交
461 462
	struct mem_cgroup *memcg;

463 464
	if (mem_cgroup_disabled())
		return;
M
Michal Hocko 已提交
465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482

	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();
483
}
484
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
485
void mem_cgroup_split_huge_fixup(struct page *head);
486 487
#endif

A
Andrew Morton 已提交
488
#else /* CONFIG_MEMCG */
489 490
struct mem_cgroup;

491 492 493 494 495 496 497 498 499 500 501 502
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;
}

503 504 505
static inline int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
					gfp_t gfp_mask,
					struct mem_cgroup **memcgp)
506
{
507
	*memcgp = NULL;
508 509 510
	return 0;
}

511 512 513
static inline void mem_cgroup_commit_charge(struct page *page,
					    struct mem_cgroup *memcg,
					    bool lrucare)
514 515 516
{
}

517 518
static inline void mem_cgroup_cancel_charge(struct page *page,
					    struct mem_cgroup *memcg)
519 520 521
{
}

522
static inline void mem_cgroup_uncharge(struct page *page)
523 524 525
{
}

526
static inline void mem_cgroup_uncharge_list(struct list_head *page_list)
527 528 529
{
}

530
static inline void mem_cgroup_replace_page(struct page *old, struct page *new)
531 532 533
{
}

534 535
static inline struct lruvec *mem_cgroup_zone_lruvec(struct zone *zone,
						    struct mem_cgroup *memcg)
K
KAMEZAWA Hiroyuki 已提交
536
{
537
	return &zone->lruvec;
K
KAMEZAWA Hiroyuki 已提交
538 539
}

540 541
static inline struct lruvec *mem_cgroup_page_lruvec(struct page *page,
						    struct zone *zone)
542
{
543
	return &zone->lruvec;
544 545
}

546
static inline bool mm_match_cgroup(struct mm_struct *mm,
547
		struct mem_cgroup *memcg)
548
{
549
	return true;
550 551
}

552 553
static inline bool task_in_mem_cgroup(struct task_struct *task,
				      const struct mem_cgroup *memcg)
554
{
555
	return true;
556 557
}

558 559 560 561 562 563 564 565 566 567 568 569 570
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)
{
}

571 572 573 574
static inline bool mem_cgroup_disabled(void)
{
	return true;
}
575

576
static inline bool
577
mem_cgroup_inactive_anon_is_low(struct lruvec *lruvec)
578
{
579
	return true;
580 581
}

582 583 584 585 586
static inline bool mem_cgroup_lruvec_online(struct lruvec *lruvec)
{
	return true;
}

587
static inline unsigned long
588
mem_cgroup_get_lru_size(struct lruvec *lruvec, enum lru_list lru)
589 590 591 592
{
	return 0;
}

593 594 595
static inline void
mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
			      int increment)
K
KOSAKI Motohiro 已提交
596 597 598
{
}

599 600 601 602 603
static inline void
mem_cgroup_print_oom_info(struct mem_cgroup *memcg, struct task_struct *p)
{
}

604
static inline struct mem_cgroup *mem_cgroup_begin_page_stat(struct page *page)
605
{
606
	return NULL;
607 608
}

609
static inline void mem_cgroup_end_page_stat(struct mem_cgroup *memcg)
610 611 612
{
}

613 614 615 616
static inline void mem_cgroup_handle_over_high(void)
{
}

617
static inline void mem_cgroup_oom_enable(void)
618 619 620
{
}

621
static inline void mem_cgroup_oom_disable(void)
622 623 624
{
}

625 626 627 628 629
static inline bool task_in_memcg_oom(struct task_struct *p)
{
	return false;
}

630
static inline bool mem_cgroup_oom_synchronize(bool wait)
631 632 633 634
{
	return false;
}

635
static inline void mem_cgroup_inc_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
636
					    enum mem_cgroup_stat_index idx)
637 638 639
{
}

640
static inline void mem_cgroup_dec_page_stat(struct mem_cgroup *memcg,
S
Sha Zhengju 已提交
641
					    enum mem_cgroup_stat_index idx)
642 643 644
{
}

645
static inline
646 647 648
unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
					    gfp_t gfp_mask,
					    unsigned long *total_scanned)
649
{
650
	return 0;
651 652
}

653
static inline void mem_cgroup_split_huge_fixup(struct page *head)
654 655 656
{
}

657 658 659 660
static inline
void mem_cgroup_count_vm_event(struct mm_struct *mm, enum vm_event_item idx)
{
}
A
Andrew Morton 已提交
661
#endif /* CONFIG_MEMCG */
662

G
Glauber Costa 已提交
663 664 665 666 667 668
enum {
	UNDER_LIMIT,
	SOFT_LIMIT,
	OVER_LIMIT,
};

669
#ifdef CONFIG_CGROUP_WRITEBACK
T
Tejun Heo 已提交
670

671
struct list_head *mem_cgroup_cgwb_list(struct mem_cgroup *memcg);
T
Tejun Heo 已提交
672
struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
673 674 675
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
676 677 678 679 680 681 682 683

#else	/* CONFIG_CGROUP_WRITEBACK */

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

684
static inline void mem_cgroup_wb_stats(struct bdi_writeback *wb,
685 686
				       unsigned long *pfilepages,
				       unsigned long *pheadroom,
687 688 689 690 691
				       unsigned long *pdirty,
				       unsigned long *pwriteback)
{
}

T
Tejun Heo 已提交
692
#endif	/* CONFIG_CGROUP_WRITEBACK */
693

G
Glauber Costa 已提交
694 695 696
struct sock;
void sock_update_memcg(struct sock *sk);
void sock_release_memcg(struct sock *sk);
697 698

#ifdef CONFIG_MEMCG_KMEM
699
extern struct static_key memcg_kmem_enabled_key;
700

701
extern int memcg_nr_cache_ids;
702 703
void memcg_get_cache_ids(void);
void memcg_put_cache_ids(void);
704 705 706 707 708 709

/*
 * 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
 */
710
#define for_each_memcg_cache_index(_idx)	\
711
	for ((_idx) = 0; (_idx) < memcg_nr_cache_ids; (_idx)++)
712

713 714
static inline bool memcg_kmem_enabled(void)
{
715
	return static_key_false(&memcg_kmem_enabled_key);
716 717
}

M
Michal Hocko 已提交
718 719 720 721
static inline bool memcg_kmem_is_active(struct mem_cgroup *memcg)
{
	return memcg->kmem_acct_active;
}
722

723 724 725 726 727 728 729 730 731 732 733
/*
 * 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.
 */
734 735
int __memcg_kmem_charge_memcg(struct page *page, gfp_t gfp, int order,
			      struct mem_cgroup *memcg);
736 737
int __memcg_kmem_charge(struct page *page, gfp_t gfp, int order);
void __memcg_kmem_uncharge(struct page *page, int order);
738

M
Michal Hocko 已提交
739 740 741 742 743 744 745 746 747
/*
 * 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;
}
748

V
Vladimir Davydov 已提交
749
struct kmem_cache *__memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp);
750
void __memcg_kmem_put_cache(struct kmem_cache *cachep);
751

V
Vladimir Davydov 已提交
752
static inline bool __memcg_kmem_bypass(void)
753 754 755 756 757
{
	if (!memcg_kmem_enabled())
		return true;
	if (in_interrupt() || (!current->mm) || (current->flags & PF_KTHREAD))
		return true;
758
	return false;
759 760 761
}

/**
762 763 764 765
 * memcg_kmem_charge: charge a kmem page
 * @page: page to charge
 * @gfp: reclaim mode
 * @order: allocation order
766
 *
767
 * Returns 0 on success, an error code on failure.
768
 */
769 770
static __always_inline int memcg_kmem_charge(struct page *page,
					     gfp_t gfp, int order)
771
{
V
Vladimir Davydov 已提交
772 773 774
	if (__memcg_kmem_bypass())
		return 0;
	if (!(gfp & __GFP_ACCOUNT))
775 776
		return 0;
	return __memcg_kmem_charge(page, gfp, order);
777 778 779
}

/**
780 781 782
 * memcg_kmem_uncharge: uncharge a kmem page
 * @page: page to uncharge
 * @order: allocation order
783
 */
784
static __always_inline void memcg_kmem_uncharge(struct page *page, int order)
785 786
{
	if (memcg_kmem_enabled())
787
		__memcg_kmem_uncharge(page, order);
788 789
}

790 791 792 793
/**
 * memcg_kmem_get_cache: selects the correct per-memcg cache for allocation
 * @cachep: the original global kmem cache
 *
794
 * All memory allocated from a per-memcg cache is charged to the owner memcg.
795 796 797 798
 */
static __always_inline struct kmem_cache *
memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp)
{
V
Vladimir Davydov 已提交
799
	if (__memcg_kmem_bypass())
800
		return cachep;
V
Vladimir Davydov 已提交
801
	return __memcg_kmem_get_cache(cachep, gfp);
802
}
803 804 805 806 807 808

static __always_inline void memcg_kmem_put_cache(struct kmem_cache *cachep)
{
	if (memcg_kmem_enabled())
		__memcg_kmem_put_cache(cachep);
}
809
#else
810 811 812
#define for_each_memcg_cache_index(_idx)	\
	for (; NULL; )

813 814 815 816 817
static inline bool memcg_kmem_enabled(void)
{
	return false;
}

818 819 820 821 822
static inline bool memcg_kmem_is_active(struct mem_cgroup *memcg)
{
	return false;
}

823
static inline int memcg_kmem_charge(struct page *page, gfp_t gfp, int order)
824
{
825
	return 0;
826 827
}

828
static inline void memcg_kmem_uncharge(struct page *page, int order)
829 830
{
}
831 832 833 834 835 836

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

837 838 839 840 841 842 843 844
static inline void memcg_get_cache_ids(void)
{
}

static inline void memcg_put_cache_ids(void)
{
}

845 846 847 848 849
static inline struct kmem_cache *
memcg_kmem_get_cache(struct kmem_cache *cachep, gfp_t gfp)
{
	return cachep;
}
850 851 852 853

static inline void memcg_kmem_put_cache(struct kmem_cache *cachep)
{
}
854
#endif /* CONFIG_MEMCG_KMEM */
B
Balbir Singh 已提交
855
#endif /* _LINUX_MEMCONTROL_H */