memcontrol.h 43.1 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

346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418
/*
 * 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);

	/*
	 * The lowest bit set means that memcg isn't a valid
	 * memcg pointer, but a obj_cgroups pointer.
	 * In this case the page is shared and doesn't belong
	 * to any specific memory cgroup.
	 */
	if (memcg_data & 0x1UL)
		return NULL;

	return (struct mem_cgroup *)memcg_data;
}

419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482
#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)
{
	return (struct obj_cgroup **)(READ_ONCE(page->memcg_data) & ~0x1UL);
}

/*
 * 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);

	if (memcg_data && (memcg_data & 0x1UL))
		return (struct obj_cgroup **)(memcg_data & ~0x1UL);

	return NULL;
}

/*
 * 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)
{
	return !cmpxchg(&page->memcg_data, 0, (unsigned long)objcgs | 0x1UL);
}
#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

483 484 485 486 487 488 489
static __always_inline bool memcg_stat_item_in_bytes(int idx)
{
	if (idx == MEMCG_PERCPU_B)
		return true;
	return vmstat_item_in_bytes(idx);
}

490 491 492 493 494
static inline bool mem_cgroup_is_root(struct mem_cgroup *memcg)
{
	return (memcg == root_mem_cgroup);
}

495 496 497 498 499
static inline bool mem_cgroup_disabled(void)
{
	return !cgroup_subsys_enabled(memory_cgrp_subsys);
}

500 501
static inline unsigned long mem_cgroup_protection(struct mem_cgroup *root,
						  struct mem_cgroup *memcg,
502
						  bool in_low_reclaim)
503
{
504 505 506
	if (mem_cgroup_disabled())
		return 0;

507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542
	/*
	 * 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;

543 544
	if (in_low_reclaim)
		return READ_ONCE(memcg->memory.emin);
545

546 547
	return max(READ_ONCE(memcg->memory.emin),
		   READ_ONCE(memcg->memory.elow));
548 549
}

550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579
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);
}
580

581
int mem_cgroup_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask);
582

583
void mem_cgroup_uncharge(struct page *page);
584
void mem_cgroup_uncharge_list(struct list_head *page_list);
585

586
void mem_cgroup_migrate(struct page *oldpage, struct page *newpage);
587

588 589
static struct mem_cgroup_per_node *
mem_cgroup_nodeinfo(struct mem_cgroup *memcg, int nid)
590
{
591
	return memcg->nodeinfo[nid];
592 593 594
}

/**
595
 * mem_cgroup_lruvec - get the lru list vector for a memcg & node
596 597
 * @memcg: memcg of the wanted lruvec
 *
598 599 600
 * 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.
601
 */
602 603
static inline struct lruvec *mem_cgroup_lruvec(struct mem_cgroup *memcg,
					       struct pglist_data *pgdat)
604
{
605
	struct mem_cgroup_per_node *mz;
606 607 608
	struct lruvec *lruvec;

	if (mem_cgroup_disabled()) {
609
		lruvec = &pgdat->__lruvec;
610 611 612
		goto out;
	}

613 614 615
	if (!memcg)
		memcg = root_mem_cgroup;

616
	mz = mem_cgroup_nodeinfo(memcg, pgdat->node_id);
617 618 619 620
	lruvec = &mz->lruvec;
out:
	/*
	 * Since a node can be onlined after the mem_cgroup was created,
M
Mel Gorman 已提交
621
	 * we have to be prepared to initialize lruvec->pgdat here;
622 623
	 * and if offlined then reonlined, we need to reinitialize it.
	 */
624 625
	if (unlikely(lruvec->pgdat != pgdat))
		lruvec->pgdat = pgdat;
626 627 628
	return lruvec;
}

M
Mel Gorman 已提交
629
struct lruvec *mem_cgroup_page_lruvec(struct page *, struct pglist_data *);
630

631
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
632

633 634
struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm);

635 636
struct mem_cgroup *get_mem_cgroup_from_page(struct page *page);

M
Michal Hocko 已提交
637 638 639 640 641
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
642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668
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);
}

669 670
static inline void mem_cgroup_put(struct mem_cgroup *memcg)
{
671 672
	if (memcg)
		css_put(&memcg->css);
673 674
}

675 676 677
#define mem_cgroup_from_counter(counter, member)	\
	container_of(counter, struct mem_cgroup, member)

M
Michal Hocko 已提交
678 679 680 681
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 *);
682 683
int mem_cgroup_scan_tasks(struct mem_cgroup *,
			  int (*)(struct task_struct *, void *), void *);
M
Michal Hocko 已提交
684

685 686 687 688 689
static inline unsigned short mem_cgroup_id(struct mem_cgroup *memcg)
{
	if (mem_cgroup_disabled())
		return 0;

690
	return memcg->id.id;
691
}
692
struct mem_cgroup *mem_cgroup_from_id(unsigned short id);
693

694 695 696 697 698
static inline struct mem_cgroup *mem_cgroup_from_seq(struct seq_file *m)
{
	return mem_cgroup_from_css(seq_css(m));
}

699 700 701 702 703 704 705 706 707 708 709
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;
}

710 711 712 713 714 715 716 717 718 719 720 721 722 723
/**
 * 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
724 725 726 727 728 729 730 731 732
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
733

734 735
static inline bool mm_match_cgroup(struct mm_struct *mm,
				   struct mem_cgroup *memcg)
736
{
737
	struct mem_cgroup *task_memcg;
738
	bool match = false;
739

740
	rcu_read_lock();
741
	task_memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
742
	if (task_memcg)
743
		match = mem_cgroup_is_descendant(task_memcg, memcg);
744
	rcu_read_unlock();
745
	return match;
746
}
747

748
struct cgroup_subsys_state *mem_cgroup_css_from_page(struct page *page);
749
ino_t page_cgroup_ino(struct page *page);
750

751 752 753 754 755 756 757
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
{
	if (mem_cgroup_disabled())
		return true;
	return !!(memcg->css.flags & CSS_ONLINE);
}

758 759 760
/*
 * For memory reclaim.
 */
761
int mem_cgroup_select_victim_node(struct mem_cgroup *memcg);
M
Michal Hocko 已提交
762 763

void mem_cgroup_update_lru_size(struct lruvec *lruvec, enum lru_list lru,
764
		int zid, int nr_pages);
M
Michal Hocko 已提交
765

766 767 768 769 770 771 772
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);
773
	return READ_ONCE(mz->lru_zone_size[zone_idx][lru]);
M
Michal Hocko 已提交
774 775
}

776 777
void mem_cgroup_handle_over_high(void);

778
unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg);
779

780 781
unsigned long mem_cgroup_size(struct mem_cgroup *memcg);

782
void mem_cgroup_print_oom_context(struct mem_cgroup *memcg,
783
				struct task_struct *p);
784

785 786
void mem_cgroup_print_oom_meminfo(struct mem_cgroup *memcg);

787
static inline void mem_cgroup_enter_user_fault(void)
788
{
789 790
	WARN_ON(current->in_user_fault);
	current->in_user_fault = 1;
791 792
}

793
static inline void mem_cgroup_exit_user_fault(void)
794
{
795 796
	WARN_ON(!current->in_user_fault);
	current->in_user_fault = 0;
797 798
}

799 800
static inline bool task_in_memcg_oom(struct task_struct *p)
{
T
Tejun Heo 已提交
801
	return p->memcg_in_oom;
802 803
}

804
bool mem_cgroup_oom_synchronize(bool wait);
805 806 807
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);
808

A
Andrew Morton 已提交
809
#ifdef CONFIG_MEMCG_SWAP
810
extern bool cgroup_memory_noswap;
811
#endif
812

813 814
struct mem_cgroup *lock_page_memcg(struct page *page);
void __unlock_page_memcg(struct mem_cgroup *memcg);
J
Johannes Weiner 已提交
815
void unlock_page_memcg(struct page *page);
816

817 818 819 820 821 822 823 824 825 826 827 828 829 830
/*
 * 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;
}

831 832 833 834
/*
 * idx can be of type enum memcg_stat_item or node_stat_item.
 * Keep in sync with memcg_exact_page_state().
 */
835 836
static inline unsigned long memcg_page_state_local(struct mem_cgroup *memcg,
						   int idx)
837
{
838 839 840 841 842
	long x = 0;
	int cpu;

	for_each_possible_cpu(cpu)
		x += per_cpu(memcg->vmstats_local->stat[idx], cpu);
843 844 845 846 847
#ifdef CONFIG_SMP
	if (x < 0)
		x = 0;
#endif
	return x;
848 849
}

850
void __mod_memcg_state(struct mem_cgroup *memcg, int idx, int val);
851

852
/* idx can be of type enum memcg_stat_item or node_stat_item */
853
static inline void mod_memcg_state(struct mem_cgroup *memcg,
854
				   int idx, int val)
855
{
856 857 858
	unsigned long flags;

	local_irq_save(flags);
859
	__mod_memcg_state(memcg, idx, val);
860
	local_irq_restore(flags);
861 862
}

M
Michal Hocko 已提交
863
/**
864
 * mod_memcg_page_state - update page state statistics
J
Johannes Weiner 已提交
865
 * @page: the page
M
Michal Hocko 已提交
866 867 868
 * @idx: page state item to account
 * @val: number of pages (positive or negative)
 *
869 870 871
 * 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:
872
 *
873
 *   lock_page(page) or lock_page_memcg(page)
874
 *   if (TestClearPageState(page))
875
 *     mod_memcg_page_state(page, state, -1);
876
 *   unlock_page(page) or unlock_page_memcg(page)
877 878
 *
 * Kernel pages are an exception to this, since they'll never move.
M
Michal Hocko 已提交
879
 */
880
static inline void __mod_memcg_page_state(struct page *page,
881
					  int idx, int val)
882
{
883 884 885 886
	struct mem_cgroup *memcg = page_memcg(page);

	if (memcg)
		__mod_memcg_state(memcg, idx, val);
887 888
}

889
static inline void mod_memcg_page_state(struct page *page,
890
					int idx, int val)
M
Michal Hocko 已提交
891
{
892 893 894 895
	struct mem_cgroup *memcg = page_memcg(page);

	if (memcg)
		mod_memcg_state(memcg, idx, val);
M
Michal Hocko 已提交
896 897
}

898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915
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;
}

916 917
static inline unsigned long lruvec_page_state_local(struct lruvec *lruvec,
						    enum node_stat_item idx)
918
{
919
	struct mem_cgroup_per_node *pn;
920 921
	long x = 0;
	int cpu;
922 923 924 925 926

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

	pn = container_of(lruvec, struct mem_cgroup_per_node, lruvec);
927 928
	for_each_possible_cpu(cpu)
		x += per_cpu(pn->lruvec_stat_local->count[idx], cpu);
929 930 931 932 933
#ifdef CONFIG_SMP
	if (x < 0)
		x = 0;
#endif
	return x;
934 935
}

936 937
void __mod_memcg_lruvec_state(struct lruvec *lruvec, enum node_stat_item idx,
			      int val);
938 939
void __mod_lruvec_state(struct lruvec *lruvec, enum node_stat_item idx,
			int val);
940
void __mod_lruvec_slab_state(void *p, enum node_stat_item idx, int val);
941

942
void mod_memcg_obj_state(void *p, int idx, int val);
943

944 945 946 947 948 949 950 951 952 953
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);
}

954 955 956 957 958 959 960 961 962 963
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);
}

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

	local_irq_save(flags);
970
	__mod_lruvec_state(lruvec, idx, val);
971
	local_irq_restore(flags);
972 973 974 975 976
}

static inline void __mod_lruvec_page_state(struct page *page,
					   enum node_stat_item idx, int val)
{
977
	struct page *head = compound_head(page); /* rmap on tail pages */
978
	struct mem_cgroup *memcg = page_memcg(head);
979 980
	pg_data_t *pgdat = page_pgdat(page);
	struct lruvec *lruvec;
981

982
	/* Untracked pages have no memcg, no lruvec. Update only the node */
983
	if (!memcg) {
984
		__mod_node_page_state(pgdat, idx, val);
985
		return;
986 987
	}

988
	lruvec = mem_cgroup_lruvec(memcg, pgdat);
989
	__mod_lruvec_state(lruvec, idx, val);
990 991 992 993 994
}

static inline void mod_lruvec_page_state(struct page *page,
					 enum node_stat_item idx, int val)
{
995 996 997
	unsigned long flags;

	local_irq_save(flags);
998
	__mod_lruvec_page_state(page, idx, val);
999
	local_irq_restore(flags);
1000 1001
}

1002
unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order,
1003 1004
						gfp_t gfp_mask,
						unsigned long *total_scanned);
D
David Rientjes 已提交
1005

1006 1007
void __count_memcg_events(struct mem_cgroup *memcg, enum vm_event_item idx,
			  unsigned long count);
1008

1009
static inline void count_memcg_events(struct mem_cgroup *memcg,
1010 1011
				      enum vm_event_item idx,
				      unsigned long count)
1012
{
1013 1014 1015
	unsigned long flags;

	local_irq_save(flags);
1016
	__count_memcg_events(memcg, idx, count);
1017
	local_irq_restore(flags);
1018 1019 1020
}

static inline void count_memcg_page_event(struct page *page,
1021
					  enum vm_event_item idx)
1022
{
1023 1024 1025 1026
	struct mem_cgroup *memcg = page_memcg(page);

	if (memcg)
		count_memcg_events(memcg, idx, 1);
1027 1028 1029 1030
}

static inline void count_memcg_event_mm(struct mm_struct *mm,
					enum vm_event_item idx)
1031
{
M
Michal Hocko 已提交
1032 1033
	struct mem_cgroup *memcg;

1034 1035
	if (mem_cgroup_disabled())
		return;
M
Michal Hocko 已提交
1036 1037 1038

	rcu_read_lock();
	memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
R
Roman Gushchin 已提交
1039
	if (likely(memcg))
1040
		count_memcg_events(memcg, idx, 1);
M
Michal Hocko 已提交
1041
	rcu_read_unlock();
1042
}
1043

1044 1045
static inline void memcg_memory_event(struct mem_cgroup *memcg,
				      enum memcg_memory_event event)
1046
{
1047 1048 1049
	atomic_long_inc(&memcg->memory_events_local[event]);
	cgroup_file_notify(&memcg->events_local_file);

1050 1051 1052 1053
	do {
		atomic_long_inc(&memcg->memory_events[event]);
		cgroup_file_notify(&memcg->events_file);

1054 1055
		if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
			break;
1056 1057 1058 1059
		if (cgrp_dfl_root.flags & CGRP_ROOT_MEMORY_LOCAL_EVENTS)
			break;
	} while ((memcg = parent_mem_cgroup(memcg)) &&
		 !mem_cgroup_is_root(memcg));
1060 1061
}

R
Roman Gushchin 已提交
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076
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();
}

1077
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
1078
void mem_cgroup_split_huge_fixup(struct page *head);
1079 1080
#endif

A
Andrew Morton 已提交
1081
#else /* CONFIG_MEMCG */
1082 1083 1084 1085

#define MEM_CGROUP_ID_SHIFT	0
#define MEM_CGROUP_ID_MAX	0

1086 1087
struct mem_cgroup;

1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103
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;
}

1104 1105 1106 1107 1108
static inline bool mem_cgroup_is_root(struct mem_cgroup *memcg)
{
	return true;
}

1109 1110 1111 1112 1113
static inline bool mem_cgroup_disabled(void)
{
	return true;
}

1114 1115
static inline void memcg_memory_event(struct mem_cgroup *memcg,
				      enum memcg_memory_event event)
1116 1117 1118
{
}

R
Roman Gushchin 已提交
1119 1120 1121 1122 1123
static inline void memcg_memory_event_mm(struct mm_struct *mm,
					 enum memcg_memory_event event)
{
}

1124 1125
static inline unsigned long mem_cgroup_protection(struct mem_cgroup *root,
						  struct mem_cgroup *memcg,
1126
						  bool in_low_reclaim)
1127
{
1128
	return 0;
1129 1130
}

1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141
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)
1142
{
1143
	return false;
1144 1145
}

1146
static inline int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
1147
				    gfp_t gfp_mask)
1148 1149 1150 1151
{
	return 0;
}

1152
static inline void mem_cgroup_uncharge(struct page *page)
1153 1154 1155
{
}

1156
static inline void mem_cgroup_uncharge_list(struct list_head *page_list)
1157 1158 1159
{
}

1160
static inline void mem_cgroup_migrate(struct page *old, struct page *new)
1161 1162 1163
{
}

1164 1165
static inline struct lruvec *mem_cgroup_lruvec(struct mem_cgroup *memcg,
					       struct pglist_data *pgdat)
K
KAMEZAWA Hiroyuki 已提交
1166
{
1167
	return &pgdat->__lruvec;
K
KAMEZAWA Hiroyuki 已提交
1168 1169
}

1170
static inline struct lruvec *mem_cgroup_page_lruvec(struct page *page,
M
Mel Gorman 已提交
1171
						    struct pglist_data *pgdat)
1172
{
1173
	return &pgdat->__lruvec;
1174 1175
}

1176 1177 1178 1179 1180
static inline struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *memcg)
{
	return NULL;
}

1181
static inline bool mm_match_cgroup(struct mm_struct *mm,
1182
		struct mem_cgroup *memcg)
1183
{
1184
	return true;
1185 1186
}

1187 1188 1189 1190 1191
static inline struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm)
{
	return NULL;
}

1192 1193 1194 1195 1196
static inline struct mem_cgroup *get_mem_cgroup_from_page(struct page *page)
{
	return NULL;
}

1197 1198 1199 1200
static inline void mem_cgroup_put(struct mem_cgroup *memcg)
{
}

1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213
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)
{
}

1214 1215 1216 1217 1218 1219
static inline int mem_cgroup_scan_tasks(struct mem_cgroup *memcg,
		int (*fn)(struct task_struct *, void *), void *arg)
{
	return 0;
}

1220
static inline unsigned short mem_cgroup_id(struct mem_cgroup *memcg)
1221
{
1222 1223 1224 1225 1226 1227 1228 1229
	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;
1230
}
1231

1232 1233 1234 1235 1236
static inline struct mem_cgroup *mem_cgroup_from_seq(struct seq_file *m)
{
	return NULL;
}

1237 1238 1239 1240 1241
static inline struct mem_cgroup *lruvec_memcg(struct lruvec *lruvec)
{
	return NULL;
}

1242
static inline bool mem_cgroup_online(struct mem_cgroup *memcg)
1243
{
1244
	return true;
1245 1246
}

1247 1248 1249 1250 1251 1252
static inline
unsigned long mem_cgroup_get_zone_lru_size(struct lruvec *lruvec,
		enum lru_list lru, int zone_idx)
{
	return 0;
}
1253

1254
static inline unsigned long mem_cgroup_get_max(struct mem_cgroup *memcg)
1255 1256 1257 1258
{
	return 0;
}

1259 1260 1261 1262 1263
static inline unsigned long mem_cgroup_size(struct mem_cgroup *memcg)
{
	return 0;
}

1264
static inline void
1265 1266 1267 1268 1269 1270
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)
1271 1272 1273
{
}

1274 1275 1276 1277 1278 1279
static inline struct mem_cgroup *lock_page_memcg(struct page *page)
{
	return NULL;
}

static inline void __unlock_page_memcg(struct mem_cgroup *memcg)
1280 1281 1282
{
}

J
Johannes Weiner 已提交
1283
static inline void unlock_page_memcg(struct page *page)
1284 1285 1286
{
}

1287 1288 1289 1290
static inline void mem_cgroup_handle_over_high(void)
{
}

1291
static inline void mem_cgroup_enter_user_fault(void)
1292 1293 1294
{
}

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

1299 1300 1301 1302 1303
static inline bool task_in_memcg_oom(struct task_struct *p)
{
	return false;
}

1304
static inline bool mem_cgroup_oom_synchronize(bool wait)
1305 1306 1307 1308
{
	return false;
}

1309 1310 1311 1312 1313 1314 1315 1316 1317 1318
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)
{
}

1319 1320 1321 1322 1323
static inline unsigned long memcg_page_state(struct mem_cgroup *memcg, int idx)
{
	return 0;
}

1324 1325
static inline unsigned long memcg_page_state_local(struct mem_cgroup *memcg,
						   int idx)
1326 1327 1328 1329
{
	return 0;
}

1330
static inline void __mod_memcg_state(struct mem_cgroup *memcg,
1331
				     int idx,
1332
				     int nr)
1333 1334 1335
{
}

1336
static inline void mod_memcg_state(struct mem_cgroup *memcg,
1337
				   int idx,
1338
				   int nr)
1339 1340 1341
{
}

1342
static inline void __mod_memcg_page_state(struct page *page,
1343
					  int idx,
1344
					  int nr)
1345 1346 1347
{
}

1348
static inline void mod_memcg_page_state(struct page *page,
1349
					int idx,
1350
					int nr)
1351 1352 1353
{
}

1354 1355 1356 1357 1358 1359
static inline unsigned long lruvec_page_state(struct lruvec *lruvec,
					      enum node_stat_item idx)
{
	return node_page_state(lruvec_pgdat(lruvec), idx);
}

1360 1361
static inline unsigned long lruvec_page_state_local(struct lruvec *lruvec,
						    enum node_stat_item idx)
1362
{
1363
	return node_page_state(lruvec_pgdat(lruvec), idx);
1364 1365
}

1366 1367 1368 1369 1370
static inline void __mod_memcg_lruvec_state(struct lruvec *lruvec,
					    enum node_stat_item idx, int val)
{
}

1371 1372
static inline void __mod_lruvec_state(struct lruvec *lruvec,
				      enum node_stat_item idx, int val)
1373
{
1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392
	__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);
1393 1394
}

1395 1396 1397 1398 1399 1400 1401 1402
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);
}

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
static inline void mod_memcg_obj_state(void *p, int idx, int val)
{
}

1415
static inline
1416
unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order,
1417 1418
					    gfp_t gfp_mask,
					    unsigned long *total_scanned)
1419
{
1420
	return 0;
1421 1422
}

1423
static inline void mem_cgroup_split_huge_fixup(struct page *head)
1424 1425 1426
{
}

1427 1428 1429 1430 1431 1432
static inline void count_memcg_events(struct mem_cgroup *memcg,
				      enum vm_event_item idx,
				      unsigned long count)
{
}

1433 1434 1435 1436 1437 1438
static inline void __count_memcg_events(struct mem_cgroup *memcg,
					enum vm_event_item idx,
					unsigned long count)
{
}

1439
static inline void count_memcg_page_event(struct page *page,
1440
					  int idx)
1441 1442 1443
{
}

1444
static inline
1445
void count_memcg_event_mm(struct mm_struct *mm, enum vm_event_item idx)
1446 1447
{
}
A
Andrew Morton 已提交
1448
#endif /* CONFIG_MEMCG */
1449

1450
/* idx can be of type enum memcg_stat_item or node_stat_item */
1451
static inline void __inc_memcg_state(struct mem_cgroup *memcg,
1452
				     int idx)
1453 1454 1455 1456
{
	__mod_memcg_state(memcg, idx, 1);
}

1457
/* idx can be of type enum memcg_stat_item or node_stat_item */
1458
static inline void __dec_memcg_state(struct mem_cgroup *memcg,
1459
				     int idx)
1460 1461 1462 1463
{
	__mod_memcg_state(memcg, idx, -1);
}

1464
/* idx can be of type enum memcg_stat_item or node_stat_item */
1465
static inline void __inc_memcg_page_state(struct page *page,
1466
					  int idx)
1467 1468 1469 1470
{
	__mod_memcg_page_state(page, idx, 1);
}

1471
/* idx can be of type enum memcg_stat_item or node_stat_item */
1472
static inline void __dec_memcg_page_state(struct page *page,
1473
					  int idx)
1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501
{
	__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);
}

1502 1503 1504 1505 1506 1507 1508 1509 1510 1511
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);
}

1512
/* idx can be of type enum memcg_stat_item or node_stat_item */
1513
static inline void inc_memcg_state(struct mem_cgroup *memcg,
1514
				   int idx)
1515 1516 1517 1518
{
	mod_memcg_state(memcg, idx, 1);
}

1519
/* idx can be of type enum memcg_stat_item or node_stat_item */
1520
static inline void dec_memcg_state(struct mem_cgroup *memcg,
1521
				   int idx)
1522 1523 1524 1525
{
	mod_memcg_state(memcg, idx, -1);
}

1526
/* idx can be of type enum memcg_stat_item or node_stat_item */
1527
static inline void inc_memcg_page_state(struct page *page,
1528
					int idx)
1529 1530 1531 1532
{
	mod_memcg_page_state(page, idx, 1);
}

1533
/* idx can be of type enum memcg_stat_item or node_stat_item */
1534
static inline void dec_memcg_page_state(struct page *page,
1535
					int idx)
1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563
{
	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);
}

1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576
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));
}

1577
#ifdef CONFIG_CGROUP_WRITEBACK
T
Tejun Heo 已提交
1578 1579

struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
1580 1581 1582
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
1583

1584 1585 1586 1587 1588 1589
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)
{
1590 1591 1592
	if (mem_cgroup_disabled())
		return;

1593
	if (unlikely(&page_memcg(page)->css != wb->memcg_css))
1594 1595 1596 1597 1598
		mem_cgroup_track_foreign_dirty_slowpath(page, wb);
}

void mem_cgroup_flush_foreign(struct bdi_writeback *wb);

T
Tejun Heo 已提交
1599 1600 1601 1602 1603 1604 1605
#else	/* CONFIG_CGROUP_WRITEBACK */

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

1606
static inline void mem_cgroup_wb_stats(struct bdi_writeback *wb,
1607 1608
				       unsigned long *pfilepages,
				       unsigned long *pheadroom,
1609 1610 1611 1612 1613
				       unsigned long *pdirty,
				       unsigned long *pwriteback)
{
}

1614 1615 1616 1617 1618 1619 1620 1621 1622
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
1623
#endif	/* CONFIG_CGROUP_WRITEBACK */
1624

G
Glauber Costa 已提交
1625
struct sock;
1626 1627
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);
1628
#ifdef CONFIG_MEMCG
1629 1630
extern struct static_key_false memcg_sockets_enabled_key;
#define mem_cgroup_sockets_enabled static_branch_unlikely(&memcg_sockets_enabled_key)
1631 1632
void mem_cgroup_sk_alloc(struct sock *sk);
void mem_cgroup_sk_free(struct sock *sk);
1633
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
1634
{
1635
	if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) && memcg->tcpmem_pressure)
1636 1637 1638 1639 1640 1641
		return true;
	do {
		if (time_before(jiffies, memcg->socket_pressure))
			return true;
	} while ((memcg = parent_mem_cgroup(memcg)));
	return false;
1642
}
1643 1644 1645 1646 1647

extern int memcg_expand_shrinker_maps(int new_id);

extern void memcg_set_shrinker_bit(struct mem_cgroup *memcg,
				   int nid, int shrinker_id);
1648
#else
1649
#define mem_cgroup_sockets_enabled 0
1650 1651
static inline void mem_cgroup_sk_alloc(struct sock *sk) { };
static inline void mem_cgroup_sk_free(struct sock *sk) { };
1652
static inline bool mem_cgroup_under_socket_pressure(struct mem_cgroup *memcg)
1653 1654 1655
{
	return false;
}
1656 1657 1658 1659 1660

static inline void memcg_set_shrinker_bit(struct mem_cgroup *memcg,
					  int nid, int shrinker_id)
{
}
1661
#endif
1662

1663
#ifdef CONFIG_MEMCG_KMEM
1664 1665 1666
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);
1667 1668
int __memcg_kmem_charge_page(struct page *page, gfp_t gfp, int order);
void __memcg_kmem_uncharge_page(struct page *page, int order);
1669

R
Roman Gushchin 已提交
1670 1671 1672 1673 1674
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);

1675
extern struct static_key_false memcg_kmem_enabled_key;
1676

1677
extern int memcg_nr_cache_ids;
1678 1679
void memcg_get_cache_ids(void);
void memcg_put_cache_ids(void);
1680 1681 1682 1683 1684 1685

/*
 * 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
 */
1686
#define for_each_memcg_cache_index(_idx)	\
1687
	for ((_idx) = 0; (_idx) < memcg_nr_cache_ids; (_idx)++)
1688

1689 1690
static inline bool memcg_kmem_enabled(void)
{
1691
	return static_branch_likely(&memcg_kmem_enabled_key);
1692 1693
}

1694 1695
static inline int memcg_kmem_charge_page(struct page *page, gfp_t gfp,
					 int order)
1696 1697
{
	if (memcg_kmem_enabled())
1698
		return __memcg_kmem_charge_page(page, gfp, order);
1699 1700 1701
	return 0;
}

1702
static inline void memcg_kmem_uncharge_page(struct page *page, int order)
1703 1704
{
	if (memcg_kmem_enabled())
1705
		__memcg_kmem_uncharge_page(page, order);
1706 1707
}

1708 1709
static inline int memcg_kmem_charge(struct mem_cgroup *memcg, gfp_t gfp,
				    unsigned int nr_pages)
1710 1711
{
	if (memcg_kmem_enabled())
1712
		return __memcg_kmem_charge(memcg, gfp, nr_pages);
1713 1714
	return 0;
}
1715

1716 1717
static inline void memcg_kmem_uncharge(struct mem_cgroup *memcg,
				       unsigned int nr_pages)
1718 1719
{
	if (memcg_kmem_enabled())
1720
		__memcg_kmem_uncharge(memcg, nr_pages);
1721 1722
}

M
Michal Hocko 已提交
1723
/*
J
Jesper Dangaard Brouer 已提交
1724
 * helper for accessing a memcg's index. It will be used as an index in the
M
Michal Hocko 已提交
1725 1726 1727 1728 1729 1730 1731
 * 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;
}
1732

1733 1734
struct mem_cgroup *mem_cgroup_from_obj(void *p);

1735
#else
1736

1737 1738
static inline int memcg_kmem_charge_page(struct page *page, gfp_t gfp,
					 int order)
1739 1740 1741 1742
{
	return 0;
}

1743
static inline void memcg_kmem_uncharge_page(struct page *page, int order)
1744 1745 1746
{
}

1747 1748
static inline int __memcg_kmem_charge_page(struct page *page, gfp_t gfp,
					   int order)
1749 1750 1751 1752
{
	return 0;
}

1753
static inline void __memcg_kmem_uncharge_page(struct page *page, int order)
1754 1755 1756
{
}

1757 1758 1759
#define for_each_memcg_cache_index(_idx)	\
	for (; NULL; )

1760 1761 1762 1763 1764
static inline bool memcg_kmem_enabled(void)
{
	return false;
}

1765 1766 1767 1768 1769
static inline int memcg_cache_id(struct mem_cgroup *memcg)
{
	return -1;
}

1770 1771 1772 1773 1774 1775 1776 1777
static inline void memcg_get_cache_ids(void)
{
}

static inline void memcg_put_cache_ids(void)
{
}

1778 1779 1780 1781 1782
static inline struct mem_cgroup *mem_cgroup_from_obj(void *p)
{
       return NULL;
}

1783
#endif /* CONFIG_MEMCG_KMEM */
1784

B
Balbir Singh 已提交
1785
#endif /* _LINUX_MEMCONTROL_H */