memcontrol.h 43.4 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0-or-later */
B
Balbir Singh 已提交
2 3 4 5 6
/* memcontrol.h - Memory Controller
 *
 * Copyright IBM Corporation, 2007
 * Author Balbir Singh <balbir@linux.vnet.ibm.com>
 *
7 8
 * Copyright 2007 OpenVZ SWsoft Inc
 * Author: Pavel Emelianov <xemul@openvz.org>
B
Balbir Singh 已提交
9 10 11 12
 */

#ifndef _LINUX_MEMCONTROL_H
#define _LINUX_MEMCONTROL_H
13
#include <linux/cgroup.h>
14
#include <linux/vm_event_item.h>
15
#include <linux/hardirq.h>
16
#include <linux/jump_label.h>
M
Michal Hocko 已提交
17 18 19
#include <linux/page_counter.h>
#include <linux/vmpressure.h>
#include <linux/eventfd.h>
20 21
#include <linux/mm.h>
#include <linux/vmstat.h>
M
Michal Hocko 已提交
22
#include <linux/writeback.h>
23
#include <linux/page-flags.h>
24

25
struct mem_cgroup;
R
Roman Gushchin 已提交
26
struct obj_cgroup;
27 28
struct page;
struct mm_struct;
29
struct kmem_cache;
30

31 32
/* Cgroup-specific page state, on top of universal node page state */
enum memcg_stat_item {
33
	MEMCG_SWAP = NR_VM_NODE_STAT_ITEMS,
34
	MEMCG_SOCK,
35
	MEMCG_PERCPU_B,
36
	MEMCG_NR_STAT,
37 38
};

39 40
enum memcg_memory_event {
	MEMCG_LOW,
41 42 43
	MEMCG_HIGH,
	MEMCG_MAX,
	MEMCG_OOM,
R
Roman Gushchin 已提交
44
	MEMCG_OOM_KILL,
45
	MEMCG_SWAP_HIGH,
46 47
	MEMCG_SWAP_MAX,
	MEMCG_SWAP_FAIL,
48
	MEMCG_NR_MEMORY_EVENTS,
49 50
};

51
struct mem_cgroup_reclaim_cookie {
52
	pg_data_t *pgdat;
53 54 55
	unsigned int generation;
};

56 57 58 59 60 61 62
#ifdef CONFIG_MEMCG

#define MEM_CGROUP_ID_SHIFT	16
#define MEM_CGROUP_ID_MAX	USHRT_MAX

struct mem_cgroup_id {
	int id;
63
	refcount_t ref;
64 65
};

M
Michal Hocko 已提交
66 67
/*
 * Per memcg event counter is incremented at every pagein/pageout. With THP,
68 69
 * it will be incremented by the number of pages. This counter is used
 * to trigger some periodic events. This is straightforward and better
M
Michal Hocko 已提交
70 71 72 73 74 75 76 77
 * than using jiffies etc. to handle periodic memcg event.
 */
enum mem_cgroup_events_target {
	MEM_CGROUP_TARGET_THRESH,
	MEM_CGROUP_TARGET_SOFTLIMIT,
	MEM_CGROUP_NTARGETS,
};

78 79
struct memcg_vmstats_percpu {
	long stat[MEMCG_NR_STAT];
80
	unsigned long events[NR_VM_EVENT_ITEMS];
M
Michal Hocko 已提交
81 82 83 84 85 86 87 88 89 90
	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;
};

91 92 93 94
struct lruvec_stat {
	long count[NR_VM_NODE_STAT_ITEMS];
};

95 96 97 98 99 100
/*
 * Bitmap of shrinker::id corresponding to memcg-aware shrinkers,
 * which have elements charged to this memcg.
 */
struct memcg_shrinker_map {
	struct rcu_head rcu;
101
	unsigned long map[];
102 103
};

M
Michal Hocko 已提交
104
/*
105
 * per-node information in memory controller.
M
Michal Hocko 已提交
106
 */
107
struct mem_cgroup_per_node {
M
Michal Hocko 已提交
108
	struct lruvec		lruvec;
109

110 111 112 113
	/* Legacy local VM stats */
	struct lruvec_stat __percpu *lruvec_stat_local;

	/* Subtree VM stats (batched updates) */
114 115 116
	struct lruvec_stat __percpu *lruvec_stat_cpu;
	atomic_long_t		lruvec_stat[NR_VM_NODE_STAT_ITEMS];

117
	unsigned long		lru_zone_size[MAX_NR_ZONES][NR_LRU_LISTS];
M
Michal Hocko 已提交
118

119
	struct mem_cgroup_reclaim_iter	iter;
M
Michal Hocko 已提交
120

121
	struct memcg_shrinker_map __rcu	*shrinker_map;
122

M
Michal Hocko 已提交
123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142
	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 */
143
	struct mem_cgroup_threshold entries[];
M
Michal Hocko 已提交
144 145 146 147 148 149 150 151 152 153 154 155 156
};

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;
};

157 158 159 160 161 162
enum memcg_kmem_state {
	KMEM_NONE,
	KMEM_ALLOCATED,
	KMEM_ONLINE,
};

163 164 165 166 167 168 169 170 171
#if defined(CONFIG_SMP)
struct memcg_padding {
	char x[0];
} ____cacheline_internodealigned_in_smp;
#define MEMCG_PADDING(name)      struct memcg_padding name;
#else
#define MEMCG_PADDING(name)
#endif

172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188
/*
 * Remember four most recent foreign writebacks with dirty pages in this
 * cgroup.  Inode sharing is expected to be uncommon and, even if we miss
 * one in a given round, we're likely to catch it later if it keeps
 * foreign-dirtying, so a fairly low count should be enough.
 *
 * See mem_cgroup_track_foreign_dirty_slowpath() for details.
 */
#define MEMCG_CGWB_FRN_CNT	4

struct memcg_cgwb_frn {
	u64 bdi_id;			/* bdi->id of the foreign inode */
	int memcg_id;			/* memcg->css.id of foreign inode */
	u64 at;				/* jiffies_64 at the time of dirtying */
	struct wb_completion done;	/* tracks in-flight foreign writebacks */
};

R
Roman Gushchin 已提交
189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204
/*
 * Bucket for arbitrarily byte-sized objects charged to a memory
 * cgroup. The bucket can be reparented in one piece when the cgroup
 * is destroyed, without having to round up the individual references
 * of all live memory objects in the wild.
 */
struct obj_cgroup {
	struct percpu_ref refcnt;
	struct mem_cgroup *memcg;
	atomic_t nr_charged_bytes;
	union {
		struct list_head list;
		struct rcu_head rcu;
	};
};

M
Michal Hocko 已提交
205 206 207 208 209 210 211 212 213
/*
 * 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;

214 215 216
	/* Private memcg ID. Used to ID objects that outlive the cgroup */
	struct mem_cgroup_id id;

M
Michal Hocko 已提交
217
	/* Accounted resources */
218 219 220 221 222 223
	struct page_counter memory;		/* Both v1 & v2 */

	union {
		struct page_counter swap;	/* v2 only */
		struct page_counter memsw;	/* v1 only */
	};
224 225

	/* Legacy consumer-oriented counters */
226 227
	struct page_counter kmem;		/* v1 only */
	struct page_counter tcpmem;		/* v1 only */
M
Michal Hocko 已提交
228

229 230 231
	/* Range enforcement for interrupt charges */
	struct work_struct high_work;

M
Michal Hocko 已提交
232 233 234 235 236 237 238 239 240 241
	unsigned long soft_limit;

	/* vmpressure notifications */
	struct vmpressure vmpressure;

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

242 243 244 245 246
	/*
	 * Should the OOM killer kill all belonging tasks, had it kill one?
	 */
	bool oom_group;

M
Michal Hocko 已提交
247 248 249 250 251 252 253 254
	/* protected by memcg_oom_lock */
	bool		oom_lock;
	int		under_oom;

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

255
	/* memory.events and memory.events.local */
256
	struct cgroup_file events_file;
257
	struct cgroup_file events_local_file;
258

259 260 261
	/* handle for "memory.swap.events" */
	struct cgroup_file swap_events_file;

M
Michal Hocko 已提交
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278
	/* 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;
279 280 281 282 283 284
	/* taken only while moving_account > 0 */
	spinlock_t		move_lock;
	unsigned long		move_lock_flags;

	MEMCG_PADDING(_pad1_);

M
Michal Hocko 已提交
285 286 287 288 289
	/*
	 * set > 0 if pages under this cgroup are moving to other cgroup.
	 */
	atomic_t		moving_account;
	struct task_struct	*move_lock_task;
290

291 292 293 294
	/* Legacy local VM stats and events */
	struct memcg_vmstats_percpu __percpu *vmstats_local;

	/* Subtree VM stats and events (batched updates) */
295
	struct memcg_vmstats_percpu __percpu *vmstats_percpu;
296 297 298

	MEMCG_PADDING(_pad2_);

299 300
	atomic_long_t		vmstats[MEMCG_NR_STAT];
	atomic_long_t		vmevents[NR_VM_EVENT_ITEMS];
301

302
	/* memory.events */
303
	atomic_long_t		memory_events[MEMCG_NR_MEMORY_EVENTS];
304
	atomic_long_t		memory_events_local[MEMCG_NR_MEMORY_EVENTS];
M
Michal Hocko 已提交
305

306 307 308
	unsigned long		socket_pressure;

	/* Legacy tcp memory accounting */
309 310
	bool			tcpmem_active;
	int			tcpmem_pressure;
311

312
#ifdef CONFIG_MEMCG_KMEM
M
Michal Hocko 已提交
313 314
        /* Index in the kmem_cache->memcg_params.memcg_caches array */
	int kmemcg_id;
315
	enum memcg_kmem_state kmem_state;
R
Roman Gushchin 已提交
316 317
	struct obj_cgroup __rcu *objcg;
	struct list_head objcg_list; /* list of inherited objcgs */
M
Michal Hocko 已提交
318 319 320 321 322
#endif

#ifdef CONFIG_CGROUP_WRITEBACK
	struct list_head cgwb_list;
	struct wb_domain cgwb_domain;
323
	struct memcg_cgwb_frn cgwb_frn[MEMCG_CGWB_FRN_CNT];
M
Michal Hocko 已提交
324 325 326 327 328 329
#endif

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

330 331 332 333
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	struct deferred_split deferred_split_queue;
#endif

M
Michal Hocko 已提交
334 335 336
	struct mem_cgroup_per_node *nodeinfo[0];
	/* WARNING: nodeinfo must be the last member here */
};
337

338 339 340 341 342 343
/*
 * size of first charge trial. "32" comes from vmscan.c's magic value.
 * TODO: maybe necessary to use big numbers in big irons.
 */
#define MEMCG_CHARGE_BATCH 32U

344
extern struct mem_cgroup *root_mem_cgroup;
T
Tejun Heo 已提交
345

R
Roman Gushchin 已提交
346 347 348 349 350 351 352 353 354
enum page_memcg_data_flags {
	/* page->memcg_data is a pointer to an objcgs vector */
	MEMCG_DATA_OBJCGS = (1UL << 0),
	/* the next bit after the last actual flag */
	__NR_MEMCG_DATA_FLAGS  = (1UL << 1),
};

#define MEMCG_DATA_FLAGS_MASK (__NR_MEMCG_DATA_FLAGS - 1)

355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415
/*
 * page_memcg - get the memory cgroup associated with a page
 * @page: a pointer to the page struct
 *
 * Returns a pointer to the memory cgroup associated with the page,
 * or NULL. This function assumes that the page is known to have a
 * proper memory cgroup pointer. It's not safe to call this function
 * against some type of pages, e.g. slab pages or ex-slab pages.
 *
 * Any of the following ensures page and memcg binding stability:
 * - the page lock
 * - LRU isolation
 * - lock_page_memcg()
 * - exclusive reference
 */
static inline struct mem_cgroup *page_memcg(struct page *page)
{
	VM_BUG_ON_PAGE(PageSlab(page), page);
	return (struct mem_cgroup *)page->memcg_data;
}

/*
 * page_memcg_rcu - locklessly get the memory cgroup associated with a page
 * @page: a pointer to the page struct
 *
 * Returns a pointer to the memory cgroup associated with the page,
 * or NULL. This function assumes that the page is known to have a
 * proper memory cgroup pointer. It's not safe to call this function
 * against some type of pages, e.g. slab pages or ex-slab pages.
 */
static inline struct mem_cgroup *page_memcg_rcu(struct page *page)
{
	VM_BUG_ON_PAGE(PageSlab(page), page);
	WARN_ON_ONCE(!rcu_read_lock_held());

	return (struct mem_cgroup *)READ_ONCE(page->memcg_data);
}

/*
 * page_memcg_check - get the memory cgroup associated with a page
 * @page: a pointer to the page struct
 *
 * Returns a pointer to the memory cgroup associated with the page,
 * or NULL. This function unlike page_memcg() can take any  page
 * as an argument. It has to be used in cases when it's not known if a page
 * has an associated memory cgroup pointer or an object cgroups vector.
 *
 * Any of the following ensures page and memcg binding stability:
 * - the page lock
 * - LRU isolation
 * - lock_page_memcg()
 * - exclusive reference
 */
static inline struct mem_cgroup *page_memcg_check(struct page *page)
{
	/*
	 * Because page->memcg_data might be changed asynchronously
	 * for slab pages, READ_ONCE() should be used here.
	 */
	unsigned long memcg_data = READ_ONCE(page->memcg_data);

R
Roman Gushchin 已提交
416
	if (memcg_data & MEMCG_DATA_OBJCGS)
417 418 419 420 421
		return NULL;

	return (struct mem_cgroup *)memcg_data;
}

422 423 424 425 426 427 428 429 430 431 432 433 434
#ifdef CONFIG_MEMCG_KMEM
/*
 * page_objcgs - get the object cgroups vector associated with a page
 * @page: a pointer to the page struct
 *
 * Returns a pointer to the object cgroups vector associated with the page,
 * or NULL. This function assumes that the page is known to have an
 * associated object cgroups vector. It's not safe to call this function
 * against pages, which might have an associated memory cgroup: e.g.
 * kernel stack pages.
 */
static inline struct obj_cgroup **page_objcgs(struct page *page)
{
R
Roman Gushchin 已提交
435 436 437 438 439
	unsigned long memcg_data = READ_ONCE(page->memcg_data);

	VM_BUG_ON_PAGE(memcg_data && !(memcg_data & MEMCG_DATA_OBJCGS), page);

	return (struct obj_cgroup **)(memcg_data & ~MEMCG_DATA_FLAGS_MASK);
440 441 442 443 444 445 446 447 448 449 450 451 452 453
}

/*
 * page_objcgs_check - get the object cgroups vector associated with a page
 * @page: a pointer to the page struct
 *
 * Returns a pointer to the object cgroups vector associated with the page,
 * or NULL. This function is safe to use if the page can be directly associated
 * with a memory cgroup.
 */
static inline struct obj_cgroup **page_objcgs_check(struct page *page)
{
	unsigned long memcg_data = READ_ONCE(page->memcg_data);

R
Roman Gushchin 已提交
454 455
	if (!memcg_data || !(memcg_data & MEMCG_DATA_OBJCGS))
		return NULL;
456

R
Roman Gushchin 已提交
457
	return (struct obj_cgroup **)(memcg_data & ~MEMCG_DATA_FLAGS_MASK);
458 459 460 461 462 463 464 465 466 467 468 469
}

/*
 * set_page_objcgs - associate a page with a object cgroups vector
 * @page: a pointer to the page struct
 * @objcgs: a pointer to the object cgroups vector
 *
 * Atomically associates a page with a vector of object cgroups.
 */
static inline bool set_page_objcgs(struct page *page,
					struct obj_cgroup **objcgs)
{
R
Roman Gushchin 已提交
470 471
	return !cmpxchg(&page->memcg_data, 0, (unsigned long)objcgs |
			MEMCG_DATA_OBJCGS);
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490
}
#else
static inline struct obj_cgroup **page_objcgs(struct page *page)
{
	return NULL;
}

static inline struct obj_cgroup **page_objcgs_check(struct page *page)
{
	return NULL;
}

static inline bool set_page_objcgs(struct page *page,
					struct obj_cgroup **objcgs)
{
	return true;
}
#endif

491 492 493 494 495 496 497
static __always_inline bool memcg_stat_item_in_bytes(int idx)
{
	if (idx == MEMCG_PERCPU_B)
		return true;
	return vmstat_item_in_bytes(idx);
}

498 499 500 501 502
static inline bool mem_cgroup_is_root(struct mem_cgroup *memcg)
{
	return (memcg == root_mem_cgroup);
}

503 504 505 506 507
static inline bool mem_cgroup_disabled(void)
{
	return !cgroup_subsys_enabled(memory_cgrp_subsys);
}

508 509
static inline unsigned long mem_cgroup_protection(struct mem_cgroup *root,
						  struct mem_cgroup *memcg,
510
						  bool in_low_reclaim)
511
{
512 513 514
	if (mem_cgroup_disabled())
		return 0;

515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
	/*
	 * There is no reclaim protection applied to a targeted reclaim.
	 * We are special casing this specific case here because
	 * mem_cgroup_protected calculation is not robust enough to keep
	 * the protection invariant for calculated effective values for
	 * parallel reclaimers with different reclaim target. This is
	 * especially a problem for tail memcgs (as they have pages on LRU)
	 * which would want to have effective values 0 for targeted reclaim
	 * but a different value for external reclaim.
	 *
	 * Example
	 * Let's have global and A's reclaim in parallel:
	 *  |
	 *  A (low=2G, usage = 3G, max = 3G, children_low_usage = 1.5G)
	 *  |\
	 *  | C (low = 1G, usage = 2.5G)
	 *  B (low = 1G, usage = 0.5G)
	 *
	 * For the global reclaim
	 * A.elow = A.low
	 * B.elow = min(B.usage, B.low) because children_low_usage <= A.elow
	 * C.elow = min(C.usage, C.low)
	 *
	 * With the effective values resetting we have A reclaim
	 * A.elow = 0
	 * B.elow = B.low
	 * C.elow = C.low
	 *
	 * If the global reclaim races with A's reclaim then
	 * B.elow = C.elow = 0 because children_low_usage > A.elow)
	 * is possible and reclaiming B would be violating the protection.
	 *
	 */
	if (root == memcg)
		return 0;

551 552
	if (in_low_reclaim)
		return READ_ONCE(memcg->memory.emin);
553

554 555
	return max(READ_ONCE(memcg->memory.emin),
		   READ_ONCE(memcg->memory.elow));
556 557
}

558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587
void mem_cgroup_calculate_protection(struct mem_cgroup *root,
				     struct mem_cgroup *memcg);

static inline bool mem_cgroup_supports_protection(struct mem_cgroup *memcg)
{
	/*
	 * The root memcg doesn't account charges, and doesn't support
	 * protection.
	 */
	return !mem_cgroup_disabled() && !mem_cgroup_is_root(memcg);

}

static inline bool mem_cgroup_below_low(struct mem_cgroup *memcg)
{
	if (!mem_cgroup_supports_protection(memcg))
		return false;

	return READ_ONCE(memcg->memory.elow) >=
		page_counter_read(&memcg->memory);
}

static inline bool mem_cgroup_below_min(struct mem_cgroup *memcg)
{
	if (!mem_cgroup_supports_protection(memcg))
		return false;

	return READ_ONCE(memcg->memory.emin) >=
		page_counter_read(&memcg->memory);
}
588

589
int mem_cgroup_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask);
590

591
void mem_cgroup_uncharge(struct page *page);
592
void mem_cgroup_uncharge_list(struct list_head *page_list);
593

594
void mem_cgroup_migrate(struct page *oldpage, struct page *newpage);
595

596 597
static struct mem_cgroup_per_node *
mem_cgroup_nodeinfo(struct mem_cgroup *memcg, int nid)
598
{
599
	return memcg->nodeinfo[nid];
600 601 602
}

/**
603
 * mem_cgroup_lruvec - get the lru list vector for a memcg & node
604 605
 * @memcg: memcg of the wanted lruvec
 *
606 607 608
 * Returns the lru list vector holding pages for a given @memcg &
 * @node combination. This can be the node lruvec, if the memory
 * controller is disabled.
609
 */
610 611
static inline struct lruvec *mem_cgroup_lruvec(struct mem_cgroup *memcg,
					       struct pglist_data *pgdat)
612
{
613
	struct mem_cgroup_per_node *mz;
614 615 616
	struct lruvec *lruvec;

	if (mem_cgroup_disabled()) {
617
		lruvec = &pgdat->__lruvec;
618 619 620
		goto out;
	}

621 622 623
	if (!memcg)
		memcg = root_mem_cgroup;

624
	mz = mem_cgroup_nodeinfo(memcg, pgdat->node_id);
625 626 627 628
	lruvec = &mz->lruvec;
out:
	/*
	 * Since a node can be onlined after the mem_cgroup was created,
M
Mel Gorman 已提交
629
	 * we have to be prepared to initialize lruvec->pgdat here;
630 631
	 * and if offlined then reonlined, we need to reinitialize it.
	 */
632 633
	if (unlikely(lruvec->pgdat != pgdat))
		lruvec->pgdat = pgdat;
634 635 636
	return lruvec;
}

M
Mel Gorman 已提交
637
struct lruvec *mem_cgroup_page_lruvec(struct page *, struct pglist_data *);
638

639
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
640

641 642
struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm);

643 644
struct mem_cgroup *get_mem_cgroup_from_page(struct page *page);

M
Michal Hocko 已提交
645 646 647 648 649
static inline
struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css){
	return css ? container_of(css, struct mem_cgroup, css) : NULL;
}

R
Roman Gushchin 已提交
650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676
static inline bool obj_cgroup_tryget(struct obj_cgroup *objcg)
{
	return percpu_ref_tryget(&objcg->refcnt);
}

static inline void obj_cgroup_get(struct obj_cgroup *objcg)
{
	percpu_ref_get(&objcg->refcnt);
}

static inline void obj_cgroup_put(struct obj_cgroup *objcg)
{
	percpu_ref_put(&objcg->refcnt);
}

/*
 * After the initialization objcg->memcg is always pointing at
 * a valid memcg, but can be atomically swapped to the parent memcg.
 *
 * The caller must ensure that the returned memcg won't be released:
 * e.g. acquire the rcu_read_lock or css_set_lock.
 */
static inline struct mem_cgroup *obj_cgroup_memcg(struct obj_cgroup *objcg)
{
	return READ_ONCE(objcg->memcg);
}

677 678
static inline void mem_cgroup_put(struct mem_cgroup *memcg)
{
679 680
	if (memcg)
		css_put(&memcg->css);
681 682
}

683 684 685
#define mem_cgroup_from_counter(counter, member)	\
	container_of(counter, struct mem_cgroup, member)

M
Michal Hocko 已提交
686 687 688 689
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 *);
690 691
int mem_cgroup_scan_tasks(struct mem_cgroup *,
			  int (*)(struct task_struct *, void *), void *);
M
Michal Hocko 已提交
692

693 694 695 696 697
static inline unsigned short mem_cgroup_id(struct mem_cgroup *memcg)
{
	if (mem_cgroup_disabled())
		return 0;

698
	return memcg->id.id;
699
}
700
struct mem_cgroup *mem_cgroup_from_id(unsigned short id);
701

702 703 704 705 706
static inline struct mem_cgroup *mem_cgroup_from_seq(struct seq_file *m)
{
	return mem_cgroup_from_css(seq_css(m));
}

707 708 709 710 711 712 713 714 715 716 717
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;
}

718 719 720 721 722 723 724 725 726 727 728 729 730 731
/**
 * 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
732 733 734 735 736 737 738 739 740
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
741

742 743
static inline bool mm_match_cgroup(struct mm_struct *mm,
				   struct mem_cgroup *memcg)
744
{
745
	struct mem_cgroup *task_memcg;
746
	bool match = false;
747

748
	rcu_read_lock();
749
	task_memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
750
	if (task_memcg)
751
		match = mem_cgroup_is_descendant(task_memcg, memcg);
752
	rcu_read_unlock();
753
	return match;
754
}
755

756
struct cgroup_subsys_state *mem_cgroup_css_from_page(struct page *page);
757
ino_t page_cgroup_ino(struct page *page);
758

759 760 761 762 763 764 765
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
{
	if (mem_cgroup_disabled())
		return true;
	return !!(memcg->css.flags & CSS_ONLINE);
}

766 767 768
/*
 * For memory reclaim.
 */
769
int mem_cgroup_select_victim_node(struct mem_cgroup *memcg);
M
Michal Hocko 已提交
770 771

void mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
772
		int zid, int nr_pages);
M
Michal Hocko 已提交
773

774 775 776 777 778 779 780
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);
781
	return READ_ONCE(mz->lru_zone_size[zone_idx][lru]);
M
Michal Hocko 已提交
782 783
}

784 785
void mem_cgroup_handle_over_high(void);

786
unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg);
787

788 789
unsigned long mem_cgroup_size(struct mem_cgroup *memcg);

790
void mem_cgroup_print_oom_context(struct mem_cgroup *memcg,
791
				struct task_struct *p);
792

793 794
void mem_cgroup_print_oom_meminfo(struct mem_cgroup *memcg);

795
static inline void mem_cgroup_enter_user_fault(void)
796
{
797 798
	WARN_ON(current->in_user_fault);
	current->in_user_fault = 1;
799 800
}

801
static inline void mem_cgroup_exit_user_fault(void)
802
{
803 804
	WARN_ON(!current->in_user_fault);
	current->in_user_fault = 0;
805 806
}

807 808
static inline bool task_in_memcg_oom(struct task_struct *p)
{
T
Tejun Heo 已提交
809
	return p->memcg_in_oom;
810 811
}

812
bool mem_cgroup_oom_synchronize(bool wait);
813 814 815
struct mem_cgroup *mem_cgroup_get_oom_group(struct task_struct *victim,
					    struct mem_cgroup *oom_domain);
void mem_cgroup_print_oom_group(struct mem_cgroup *memcg);
816

A
Andrew Morton 已提交
817
#ifdef CONFIG_MEMCG_SWAP
818
extern bool cgroup_memory_noswap;
819
#endif
820

821 822
struct mem_cgroup *lock_page_memcg(struct page *page);
void __unlock_page_memcg(struct mem_cgroup *memcg);
J
Johannes Weiner 已提交
823
void unlock_page_memcg(struct page *page);
824

825 826 827 828 829 830 831 832 833 834 835 836 837 838
/*
 * idx can be of type enum memcg_stat_item or node_stat_item.
 * Keep in sync with memcg_exact_page_state().
 */
static inline unsigned long memcg_page_state(struct mem_cgroup *memcg, int idx)
{
	long x = atomic_long_read(&memcg->vmstats[idx]);
#ifdef CONFIG_SMP
	if (x < 0)
		x = 0;
#endif
	return x;
}

839 840 841 842
/*
 * idx can be of type enum memcg_stat_item or node_stat_item.
 * Keep in sync with memcg_exact_page_state().
 */
843 844
static inline unsigned long memcg_page_state_local(struct mem_cgroup *memcg,
						   int idx)
845
{
846 847 848 849 850
	long x = 0;
	int cpu;

	for_each_possible_cpu(cpu)
		x += per_cpu(memcg->vmstats_local->stat[idx], cpu);
851 852 853 854 855
#ifdef CONFIG_SMP
	if (x < 0)
		x = 0;
#endif
	return x;
856 857
}

858
void __mod_memcg_state(struct mem_cgroup *memcg, int idx, int val);
859

860
/* idx can be of type enum memcg_stat_item or node_stat_item */
861
static inline void mod_memcg_state(struct mem_cgroup *memcg,
862
				   int idx, int val)
863
{
864 865 866
	unsigned long flags;

	local_irq_save(flags);
867
	__mod_memcg_state(memcg, idx, val);
868
	local_irq_restore(flags);
869 870
}

M
Michal Hocko 已提交
871
/**
872
 * mod_memcg_page_state - update page state statistics
J
Johannes Weiner 已提交
873
 * @page: the page
M
Michal Hocko 已提交
874 875 876
 * @idx: page state item to account
 * @val: number of pages (positive or negative)
 *
877 878 879
 * 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:
880
 *
881
 *   lock_page(page) or lock_page_memcg(page)
882
 *   if (TestClearPageState(page))
883
 *     mod_memcg_page_state(page, state, -1);
884
 *   unlock_page(page) or unlock_page_memcg(page)
885 886
 *
 * Kernel pages are an exception to this, since they'll never move.
M
Michal Hocko 已提交
887
 */
888
static inline void __mod_memcg_page_state(struct page *page,
889
					  int idx, int val)
890
{
891 892 893 894
	struct mem_cgroup *memcg = page_memcg(page);

	if (memcg)
		__mod_memcg_state(memcg, idx, val);
895 896
}

897
static inline void mod_memcg_page_state(struct page *page,
898
					int idx, int val)
M
Michal Hocko 已提交
899
{
900 901 902 903
	struct mem_cgroup *memcg = page_memcg(page);

	if (memcg)
		mod_memcg_state(memcg, idx, val);
M
Michal Hocko 已提交
904 905
}

906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923
static inline unsigned long lruvec_page_state(struct lruvec *lruvec,
					      enum node_stat_item idx)
{
	struct mem_cgroup_per_node *pn;
	long x;

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

	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
	x = atomic_long_read(&pn->lruvec_stat[idx]);
#ifdef CONFIG_SMP
	if (x < 0)
		x = 0;
#endif
	return x;
}

924 925
static inline unsigned long lruvec_page_state_local(struct lruvec *lruvec,
						    enum node_stat_item idx)
926
{
927
	struct mem_cgroup_per_node *pn;
928 929
	long x = 0;
	int cpu;
930 931 932 933 934

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

	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
935 936
	for_each_possible_cpu(cpu)
		x += per_cpu(pn->lruvec_stat_local->count[idx], cpu);
937 938 939 940 941
#ifdef CONFIG_SMP
	if (x < 0)
		x = 0;
#endif
	return x;
942 943
}

944 945
void __mod_memcg_lruvec_state(struct lruvec *lruvec, enum node_stat_item idx,
			      int val);
946 947
void __mod_lruvec_state(struct lruvec *lruvec, enum node_stat_item idx,
			int val);
948
void __mod_lruvec_slab_state(void *p, enum node_stat_item idx, int val);
949

950
void mod_memcg_obj_state(void *p, int idx, int val);
951

952 953 954 955 956 957 958 959 960 961
static inline void mod_lruvec_slab_state(void *p, enum node_stat_item idx,
					 int val)
{
	unsigned long flags;

	local_irq_save(flags);
	__mod_lruvec_slab_state(p, idx, val);
	local_irq_restore(flags);
}

962 963 964 965 966 967 968 969 970 971
static inline void mod_memcg_lruvec_state(struct lruvec *lruvec,
					  enum node_stat_item idx, int val)
{
	unsigned long flags;

	local_irq_save(flags);
	__mod_memcg_lruvec_state(lruvec, idx, val);
	local_irq_restore(flags);
}

972 973 974
static inline void mod_lruvec_state(struct lruvec *lruvec,
				    enum node_stat_item idx, int val)
{
975 976 977
	unsigned long flags;

	local_irq_save(flags);
978
	__mod_lruvec_state(lruvec, idx, val);
979
	local_irq_restore(flags);
980 981 982 983 984
}

static inline void __mod_lruvec_page_state(struct page *page,
					   enum node_stat_item idx, int val)
{
985
	struct page *head = compound_head(page); /* rmap on tail pages */
986
	struct mem_cgroup *memcg = page_memcg(head);
987 988
	pg_data_t *pgdat = page_pgdat(page);
	struct lruvec *lruvec;
989

990
	/* Untracked pages have no memcg, no lruvec. Update only the node */
991
	if (!memcg) {
992
		__mod_node_page_state(pgdat, idx, val);
993
		return;
994 995
	}

996
	lruvec = mem_cgroup_lruvec(memcg, pgdat);
997
	__mod_lruvec_state(lruvec, idx, val);
998 999 1000 1001 1002
}

static inline void mod_lruvec_page_state(struct page *page,
					 enum node_stat_item idx, int val)
{
1003 1004 1005
	unsigned long flags;

	local_irq_save(flags);
1006
	__mod_lruvec_page_state(page, idx, val);
1007
	local_irq_restore(flags);
1008 1009
}

1010
unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order,
1011 1012
						gfp_t gfp_mask,
						unsigned long *total_scanned);
D
David Rientjes 已提交
1013

1014 1015
void __count_memcg_events(struct mem_cgroup *memcg, enum vm_event_item idx,
			  unsigned long count);
1016

1017
static inline void count_memcg_events(struct mem_cgroup *memcg,
1018 1019
				      enum vm_event_item idx,
				      unsigned long count)
1020
{
1021 1022 1023
	unsigned long flags;

	local_irq_save(flags);
1024
	__count_memcg_events(memcg, idx, count);
1025
	local_irq_restore(flags);
1026 1027 1028
}

static inline void count_memcg_page_event(struct page *page,
1029
					  enum vm_event_item idx)
1030
{
1031 1032 1033 1034
	struct mem_cgroup *memcg = page_memcg(page);

	if (memcg)
		count_memcg_events(memcg, idx, 1);
1035 1036 1037 1038
}

static inline void count_memcg_event_mm(struct mm_struct *mm,
					enum vm_event_item idx)
1039
{
M
Michal Hocko 已提交
1040 1041
	struct mem_cgroup *memcg;

1042 1043
	if (mem_cgroup_disabled())
		return;
M
Michal Hocko 已提交
1044 1045 1046

	rcu_read_lock();
	memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
R
Roman Gushchin 已提交
1047
	if (likely(memcg))
1048
		count_memcg_events(memcg, idx, 1);
M
Michal Hocko 已提交
1049
	rcu_read_unlock();
1050
}
1051

1052 1053
static inline void memcg_memory_event(struct mem_cgroup *memcg,
				      enum memcg_memory_event event)
1054
{
1055 1056 1057
	atomic_long_inc(&memcg->memory_events_local[event]);
	cgroup_file_notify(&memcg->events_local_file);

1058 1059 1060 1061
	do {
		atomic_long_inc(&memcg->memory_events[event]);
		cgroup_file_notify(&memcg->events_file);

1062 1063
		if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
			break;
1064 1065 1066 1067
		if (cgrp_dfl_root.flags & CGRP_ROOT_MEMORY_LOCAL_EVENTS)
			break;
	} while ((memcg = parent_mem_cgroup(memcg)) &&
		 !mem_cgroup_is_root(memcg));
1068 1069
}

R
Roman Gushchin 已提交
1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084
static inline void memcg_memory_event_mm(struct mm_struct *mm,
					 enum memcg_memory_event event)
{
	struct mem_cgroup *memcg;

	if (mem_cgroup_disabled())
		return;

	rcu_read_lock();
	memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
	if (likely(memcg))
		memcg_memory_event(memcg, event);
	rcu_read_unlock();
}

1085
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
1086
void mem_cgroup_split_huge_fixup(struct page *head);
1087 1088
#endif

A
Andrew Morton 已提交
1089
#else /* CONFIG_MEMCG */
1090 1091 1092 1093

#define MEM_CGROUP_ID_SHIFT	0
#define MEM_CGROUP_ID_MAX	0

1094 1095
struct mem_cgroup;

1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111
static inline struct mem_cgroup *page_memcg(struct page *page)
{
	return NULL;
}

static inline struct mem_cgroup *page_memcg_rcu(struct page *page)
{
	WARN_ON_ONCE(!rcu_read_lock_held());
	return NULL;
}

static inline struct mem_cgroup *page_memcg_check(struct page *page)
{
	return NULL;
}

1112 1113 1114 1115 1116
static inline bool mem_cgroup_is_root(struct mem_cgroup *memcg)
{
	return true;
}

1117 1118 1119 1120 1121
static inline bool mem_cgroup_disabled(void)
{
	return true;
}

1122 1123
static inline void memcg_memory_event(struct mem_cgroup *memcg,
				      enum memcg_memory_event event)
1124 1125 1126
{
}

R
Roman Gushchin 已提交
1127 1128 1129 1130 1131
static inline void memcg_memory_event_mm(struct mm_struct *mm,
					 enum memcg_memory_event event)
{
}

1132 1133
static inline unsigned long mem_cgroup_protection(struct mem_cgroup *root,
						  struct mem_cgroup *memcg,
1134
						  bool in_low_reclaim)
1135
{
1136
	return 0;
1137 1138
}

1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149
static inline void mem_cgroup_calculate_protection(struct mem_cgroup *root,
						   struct mem_cgroup *memcg)
{
}

static inline bool mem_cgroup_below_low(struct mem_cgroup *memcg)
{
	return false;
}

static inline bool mem_cgroup_below_min(struct mem_cgroup *memcg)
1150
{
1151
	return false;
1152 1153
}

1154
static inline int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
1155
				    gfp_t gfp_mask)
1156 1157 1158 1159
{
	return 0;
}

1160
static inline void mem_cgroup_uncharge(struct page *page)
1161 1162 1163
{
}

1164
static inline void mem_cgroup_uncharge_list(struct list_head *page_list)
1165 1166 1167
{
}

1168
static inline void mem_cgroup_migrate(struct page *old, struct page *new)
1169 1170 1171
{
}

1172 1173
static inline struct lruvec *mem_cgroup_lruvec(struct mem_cgroup *memcg,
					       struct pglist_data *pgdat)
K
KAMEZAWA Hiroyuki 已提交
1174
{
1175
	return &pgdat->__lruvec;
K
KAMEZAWA Hiroyuki 已提交
1176 1177
}

1178
static inline struct lruvec *mem_cgroup_page_lruvec(struct page *page,
M
Mel Gorman 已提交
1179
						    struct pglist_data *pgdat)
1180
{
1181
	return &pgdat->__lruvec;
1182 1183
}

1184 1185 1186 1187 1188
static inline struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *memcg)
{
	return NULL;
}

1189
static inline bool mm_match_cgroup(struct mm_struct *mm,
1190
		struct mem_cgroup *memcg)
1191
{
1192
	return true;
1193 1194
}

1195 1196 1197 1198 1199
static inline struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm)
{
	return NULL;
}

1200 1201 1202 1203 1204
static inline struct mem_cgroup *get_mem_cgroup_from_page(struct page *page)
{
	return NULL;
}

1205 1206 1207 1208
static inline void mem_cgroup_put(struct mem_cgroup *memcg)
{
}

1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221
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)
{
}

1222 1223 1224 1225 1226 1227
static inline int mem_cgroup_scan_tasks(struct mem_cgroup *memcg,
		int (*fn)(struct task_struct *, void *), void *arg)
{
	return 0;
}

1228
static inline unsigned short mem_cgroup_id(struct mem_cgroup *memcg)
1229
{
1230 1231 1232 1233 1234 1235 1236 1237
	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;
1238
}
1239

1240 1241 1242 1243 1244
static inline struct mem_cgroup *mem_cgroup_from_seq(struct seq_file *m)
{
	return NULL;
}

1245 1246 1247 1248 1249
static inline struct mem_cgroup *lruvec_memcg(struct lruvec *lruvec)
{
	return NULL;
}

1250
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
1251
{
1252
	return true;
1253 1254
}

1255 1256 1257 1258 1259 1260
static inline
unsigned long mem_cgroup_get_zone_lru_size(struct lruvec *lruvec,
		enum lru_list lru, int zone_idx)
{
	return 0;
}
1261

1262
static inline unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg)
1263 1264 1265 1266
{
	return 0;
}

1267 1268 1269 1270 1271
static inline unsigned long mem_cgroup_size(struct mem_cgroup *memcg)
{
	return 0;
}

1272
static inline void
1273 1274 1275 1276 1277 1278
mem_cgroup_print_oom_context(struct mem_cgroup *memcg, struct task_struct *p)
{
}

static inline void
mem_cgroup_print_oom_meminfo(struct mem_cgroup *memcg)
1279 1280 1281
{
}

1282 1283 1284 1285 1286 1287
static inline struct mem_cgroup *lock_page_memcg(struct page *page)
{
	return NULL;
}

static inline void __unlock_page_memcg(struct mem_cgroup *memcg)
1288 1289 1290
{
}

J
Johannes Weiner 已提交
1291
static inline void unlock_page_memcg(struct page *page)
1292 1293 1294
{
}

1295 1296 1297 1298
static inline void mem_cgroup_handle_over_high(void)
{
}

1299
static inline void mem_cgroup_enter_user_fault(void)
1300 1301 1302
{
}

1303
static inline void mem_cgroup_exit_user_fault(void)
1304 1305 1306
{
}

1307 1308 1309 1310 1311
static inline bool task_in_memcg_oom(struct task_struct *p)
{
	return false;
}

1312
static inline bool mem_cgroup_oom_synchronize(bool wait)
1313 1314 1315 1316
{
	return false;
}

1317 1318 1319 1320 1321 1322 1323 1324 1325 1326
static inline struct mem_cgroup *mem_cgroup_get_oom_group(
	struct task_struct *victim, struct mem_cgroup *oom_domain)
{
	return NULL;
}

static inline void mem_cgroup_print_oom_group(struct mem_cgroup *memcg)
{
}

1327 1328 1329 1330 1331
static inline unsigned long memcg_page_state(struct mem_cgroup *memcg, int idx)
{
	return 0;
}

1332 1333
static inline unsigned long memcg_page_state_local(struct mem_cgroup *memcg,
						   int idx)
1334 1335 1336 1337
{
	return 0;
}

1338
static inline void __mod_memcg_state(struct mem_cgroup *memcg,
1339
				     int idx,
1340
				     int nr)
1341 1342 1343
{
}

1344
static inline void mod_memcg_state(struct mem_cgroup *memcg,
1345
				   int idx,
1346
				   int nr)
1347 1348 1349
{
}

1350
static inline void __mod_memcg_page_state(struct page *page,
1351
					  int idx,
1352
					  int nr)
1353 1354 1355
{
}

1356
static inline void mod_memcg_page_state(struct page *page,
1357
					int idx,
1358
					int nr)
1359 1360 1361
{
}

1362 1363 1364 1365 1366 1367
static inline unsigned long lruvec_page_state(struct lruvec *lruvec,
					      enum node_stat_item idx)
{
	return node_page_state(lruvec_pgdat(lruvec), idx);
}

1368 1369
static inline unsigned long lruvec_page_state_local(struct lruvec *lruvec,
						    enum node_stat_item idx)
1370
{
1371
	return node_page_state(lruvec_pgdat(lruvec), idx);
1372 1373
}

1374 1375 1376 1377 1378
static inline void __mod_memcg_lruvec_state(struct lruvec *lruvec,
					    enum node_stat_item idx, int val)
{
}

1379 1380
static inline void __mod_lruvec_state(struct lruvec *lruvec,
				      enum node_stat_item idx, int val)
1381
{
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400
	__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);
1401 1402
}

1403 1404 1405 1406 1407 1408 1409 1410
static inline void __mod_lruvec_slab_state(void *p, enum node_stat_item idx,
					   int val)
{
	struct page *page = virt_to_head_page(p);

	__mod_node_page_state(page_pgdat(page), idx, val);
}

1411 1412 1413 1414 1415 1416 1417 1418
static inline void mod_lruvec_slab_state(void *p, enum node_stat_item idx,
					 int val)
{
	struct page *page = virt_to_head_page(p);

	mod_node_page_state(page_pgdat(page), idx, val);
}

1419 1420 1421 1422
static inline void mod_memcg_obj_state(void *p, int idx, int val)
{
}

1423
static inline
1424
unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order,
1425 1426
					    gfp_t gfp_mask,
					    unsigned long *total_scanned)
1427
{
1428
	return 0;
1429 1430
}

1431
static inline void mem_cgroup_split_huge_fixup(struct page *head)
1432 1433 1434
{
}

1435 1436 1437 1438 1439 1440
static inline void count_memcg_events(struct mem_cgroup *memcg,
				      enum vm_event_item idx,
				      unsigned long count)
{
}

1441 1442 1443 1444 1445 1446
static inline void __count_memcg_events(struct mem_cgroup *memcg,
					enum vm_event_item idx,
					unsigned long count)
{
}

1447
static inline void count_memcg_page_event(struct page *page,
1448
					  int idx)
1449 1450 1451
{
}

1452
static inline
1453
void count_memcg_event_mm(struct mm_struct *mm, enum vm_event_item idx)
1454 1455
{
}
A
Andrew Morton 已提交
1456
#endif /* CONFIG_MEMCG */
1457

1458
/* idx can be of type enum memcg_stat_item or node_stat_item */
1459
static inline void __inc_memcg_state(struct mem_cgroup *memcg,
1460
				     int idx)
1461 1462 1463 1464
{
	__mod_memcg_state(memcg, idx, 1);
}

1465
/* idx can be of type enum memcg_stat_item or node_stat_item */
1466
static inline void __dec_memcg_state(struct mem_cgroup *memcg,
1467
				     int idx)
1468 1469 1470 1471
{
	__mod_memcg_state(memcg, idx, -1);
}

1472
/* idx can be of type enum memcg_stat_item or node_stat_item */
1473
static inline void __inc_memcg_page_state(struct page *page,
1474
					  int idx)
1475 1476 1477 1478
{
	__mod_memcg_page_state(page, idx, 1);
}

1479
/* idx can be of type enum memcg_stat_item or node_stat_item */
1480
static inline void __dec_memcg_page_state(struct page *page,
1481
					  int idx)
1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509
{
	__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);
}

1510 1511 1512 1513 1514 1515 1516 1517 1518 1519
static inline void __inc_lruvec_slab_state(void *p, enum node_stat_item idx)
{
	__mod_lruvec_slab_state(p, idx, 1);
}

static inline void __dec_lruvec_slab_state(void *p, enum node_stat_item idx)
{
	__mod_lruvec_slab_state(p, idx, -1);
}

1520
/* idx can be of type enum memcg_stat_item or node_stat_item */
1521
static inline void inc_memcg_state(struct mem_cgroup *memcg,
1522
				   int idx)
1523 1524 1525 1526
{
	mod_memcg_state(memcg, idx, 1);
}

1527
/* idx can be of type enum memcg_stat_item or node_stat_item */
1528
static inline void dec_memcg_state(struct mem_cgroup *memcg,
1529
				   int idx)
1530 1531 1532 1533
{
	mod_memcg_state(memcg, idx, -1);
}

1534
/* idx can be of type enum memcg_stat_item or node_stat_item */
1535
static inline void inc_memcg_page_state(struct page *page,
1536
					int idx)
1537 1538 1539 1540
{
	mod_memcg_page_state(page, idx, 1);
}

1541
/* idx can be of type enum memcg_stat_item or node_stat_item */
1542
static inline void dec_memcg_page_state(struct page *page,
1543
					int idx)
1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571
{
	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);
}

1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584
static inline struct lruvec *parent_lruvec(struct lruvec *lruvec)
{
	struct mem_cgroup *memcg;

	memcg = lruvec_memcg(lruvec);
	if (!memcg)
		return NULL;
	memcg = parent_mem_cgroup(memcg);
	if (!memcg)
		return NULL;
	return mem_cgroup_lruvec(memcg, lruvec_pgdat(lruvec));
}

1585
#ifdef CONFIG_CGROUP_WRITEBACK
T
Tejun Heo 已提交
1586 1587

struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
1588 1589 1590
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
1591

1592 1593 1594 1595 1596 1597
void mem_cgroup_track_foreign_dirty_slowpath(struct page *page,
					     struct bdi_writeback *wb);

static inline void mem_cgroup_track_foreign_dirty(struct page *page,
						  struct bdi_writeback *wb)
{
1598 1599 1600
	if (mem_cgroup_disabled())
		return;

1601
	if (unlikely(&page_memcg(page)->css != wb->memcg_css))
1602 1603 1604 1605 1606
		mem_cgroup_track_foreign_dirty_slowpath(page, wb);
}

void mem_cgroup_flush_foreign(struct bdi_writeback *wb);

T
Tejun Heo 已提交
1607 1608 1609 1610 1611 1612 1613
#else	/* CONFIG_CGROUP_WRITEBACK */

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

1614
static inline void mem_cgroup_wb_stats(struct bdi_writeback *wb,
1615 1616
				       unsigned long *pfilepages,
				       unsigned long *pheadroom,
1617 1618 1619 1620 1621
				       unsigned long *pdirty,
				       unsigned long *pwriteback)
{
}

1622 1623 1624 1625 1626 1627 1628 1629 1630
static inline void mem_cgroup_track_foreign_dirty(struct page *page,
						  struct bdi_writeback *wb)
{
}

static inline void mem_cgroup_flush_foreign(struct bdi_writeback *wb)
{
}

T
Tejun Heo 已提交
1631
#endif	/* CONFIG_CGROUP_WRITEBACK */
1632

G
Glauber Costa 已提交
1633
struct sock;
1634 1635
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);
1636
#ifdef CONFIG_MEMCG
1637 1638
extern struct static_key_false memcg_sockets_enabled_key;
#define mem_cgroup_sockets_enabled static_branch_unlikely(&memcg_sockets_enabled_key)
1639 1640
void mem_cgroup_sk_alloc(struct sock *sk);
void mem_cgroup_sk_free(struct sock *sk);
1641
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
1642
{
1643
	if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) && memcg->tcpmem_pressure)
1644 1645 1646 1647 1648 1649
		return true;
	do {
		if (time_before(jiffies, memcg->socket_pressure))
			return true;
	} while ((memcg = parent_mem_cgroup(memcg)));
	return false;
1650
}
1651 1652 1653 1654 1655

extern int memcg_expand_shrinker_maps(int new_id);

extern void memcg_set_shrinker_bit(struct mem_cgroup *memcg,
				   int nid, int shrinker_id);
1656
#else
1657
#define mem_cgroup_sockets_enabled 0
1658 1659
static inline void mem_cgroup_sk_alloc(struct sock *sk) { };
static inline void mem_cgroup_sk_free(struct sock *sk) { };
1660
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
1661 1662 1663
{
	return false;
}
1664 1665 1666 1667 1668

static inline void memcg_set_shrinker_bit(struct mem_cgroup *memcg,
					  int nid, int shrinker_id)
{
}
1669
#endif
1670

1671
#ifdef CONFIG_MEMCG_KMEM
1672 1673 1674
int __memcg_kmem_charge(struct mem_cgroup *memcg, gfp_t gfp,
			unsigned int nr_pages);
void __memcg_kmem_uncharge(struct mem_cgroup *memcg, unsigned int nr_pages);
1675 1676
int __memcg_kmem_charge_page(struct page *page, gfp_t gfp, int order);
void __memcg_kmem_uncharge_page(struct page *page, int order);
1677

R
Roman Gushchin 已提交
1678 1679 1680 1681 1682
struct obj_cgroup *get_obj_cgroup_from_current(void);

int obj_cgroup_charge(struct obj_cgroup *objcg, gfp_t gfp, size_t size);
void obj_cgroup_uncharge(struct obj_cgroup *objcg, size_t size);

1683
extern struct static_key_false memcg_kmem_enabled_key;
1684

1685
extern int memcg_nr_cache_ids;
1686 1687
void memcg_get_cache_ids(void);
void memcg_put_cache_ids(void);
1688 1689 1690 1691 1692 1693

/*
 * 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
 */
1694
#define for_each_memcg_cache_index(_idx)	\
1695
	for ((_idx) = 0; (_idx) < memcg_nr_cache_ids; (_idx)++)
1696

1697 1698
static inline bool memcg_kmem_enabled(void)
{
1699
	return static_branch_likely(&memcg_kmem_enabled_key);
1700 1701
}

1702 1703
static inline int memcg_kmem_charge_page(struct page *page, gfp_t gfp,
					 int order)
1704 1705
{
	if (memcg_kmem_enabled())
1706
		return __memcg_kmem_charge_page(page, gfp, order);
1707 1708 1709
	return 0;
}

1710
static inline void memcg_kmem_uncharge_page(struct page *page, int order)
1711 1712
{
	if (memcg_kmem_enabled())
1713
		__memcg_kmem_uncharge_page(page, order);
1714 1715
}

1716 1717
static inline int memcg_kmem_charge(struct mem_cgroup *memcg, gfp_t gfp,
				    unsigned int nr_pages)
1718 1719
{
	if (memcg_kmem_enabled())
1720
		return __memcg_kmem_charge(memcg, gfp, nr_pages);
1721 1722
	return 0;
}
1723

1724 1725
static inline void memcg_kmem_uncharge(struct mem_cgroup *memcg,
				       unsigned int nr_pages)
1726 1727
{
	if (memcg_kmem_enabled())
1728
		__memcg_kmem_uncharge(memcg, nr_pages);
1729 1730
}

M
Michal Hocko 已提交
1731
/*
J
Jesper Dangaard Brouer 已提交
1732
 * helper for accessing a memcg's index. It will be used as an index in the
M
Michal Hocko 已提交
1733 1734 1735 1736 1737 1738 1739
 * 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;
}
1740

1741 1742
struct mem_cgroup *mem_cgroup_from_obj(void *p);

1743
#else
1744

1745 1746
static inline int memcg_kmem_charge_page(struct page *page, gfp_t gfp,
					 int order)
1747 1748 1749 1750
{
	return 0;
}

1751
static inline void memcg_kmem_uncharge_page(struct page *page, int order)
1752 1753 1754
{
}

1755 1756
static inline int __memcg_kmem_charge_page(struct page *page, gfp_t gfp,
					   int order)
1757 1758 1759 1760
{
	return 0;
}

1761
static inline void __memcg_kmem_uncharge_page(struct page *page, int order)
1762 1763 1764
{
}

1765 1766 1767
#define for_each_memcg_cache_index(_idx)	\
	for (; NULL; )

1768 1769 1770 1771 1772
static inline bool memcg_kmem_enabled(void)
{
	return false;
}

1773 1774 1775 1776 1777
static inline int memcg_cache_id(struct mem_cgroup *memcg)
{
	return -1;
}

1778 1779 1780 1781 1782 1783 1784 1785
static inline void memcg_get_cache_ids(void)
{
}

static inline void memcg_put_cache_ids(void)
{
}

1786 1787 1788 1789 1790
static inline struct mem_cgroup *mem_cgroup_from_obj(void *p)
{
       return NULL;
}

1791
#endif /* CONFIG_MEMCG_KMEM */
1792

B
Balbir Singh 已提交
1793
#endif /* _LINUX_MEMCONTROL_H */