mmzone.h 40.6 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
L
Linus Torvalds 已提交
2 3 4 5
#ifndef _LINUX_MMZONE_H
#define _LINUX_MMZONE_H

#ifndef __ASSEMBLY__
C
Christoph Lameter 已提交
6
#ifndef __GENERATING_BOUNDS_H
L
Linus Torvalds 已提交
7 8 9 10

#include <linux/spinlock.h>
#include <linux/list.h>
#include <linux/wait.h>
11
#include <linux/bitops.h>
L
Linus Torvalds 已提交
12 13 14 15
#include <linux/cache.h>
#include <linux/threads.h>
#include <linux/numa.h>
#include <linux/init.h>
16
#include <linux/seqlock.h>
17
#include <linux/nodemask.h>
18
#include <linux/pageblock-flags.h>
19
#include <linux/page-flags-layout.h>
A
Arun Sharma 已提交
20
#include <linux/atomic.h>
21 22
#include <linux/mm_types.h>
#include <linux/page-flags.h>
R
Ralf Baechle 已提交
23
#include <asm/page.h>
L
Linus Torvalds 已提交
24

25 26
#include <linux/ali_hotfix.h>

L
Linus Torvalds 已提交
27 28 29 30 31 32
/* Free memory management - zoned buddy allocator.  */
#ifndef CONFIG_FORCE_MAX_ZONEORDER
#define MAX_ORDER 11
#else
#define MAX_ORDER CONFIG_FORCE_MAX_ZONEORDER
#endif
33
#define MAX_ORDER_NR_PAGES (1 << (MAX_ORDER - 1))
L
Linus Torvalds 已提交
34

A
Andy Whitcroft 已提交
35 36 37
/*
 * PAGE_ALLOC_COSTLY_ORDER is the order at which allocations are deemed
 * costly to service.  That is between allocation orders which should
38
 * coalesce naturally under reasonable reclaim pressure and those which
A
Andy Whitcroft 已提交
39 40 41 42
 * will not.
 */
#define PAGE_ALLOC_COSTLY_ORDER 3

43
enum migratetype {
44 45
	MIGRATE_UNMOVABLE,
	MIGRATE_MOVABLE,
46
	MIGRATE_RECLAIMABLE,
47 48
	MIGRATE_PCPTYPES,	/* the number of types on the pcp lists */
	MIGRATE_HIGHATOMIC = MIGRATE_PCPTYPES,
49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64
#ifdef CONFIG_CMA
	/*
	 * MIGRATE_CMA migration type is designed to mimic the way
	 * ZONE_MOVABLE works.  Only movable pages can be allocated
	 * from MIGRATE_CMA pageblocks and page allocator never
	 * implicitly change migration type of MIGRATE_CMA pageblock.
	 *
	 * The way to use it is to change migratetype of a range of
	 * pageblocks to MIGRATE_CMA which can be done by
	 * __free_pageblock_cma() function.  What is important though
	 * is that a range of pageblocks must be aligned to
	 * MAX_ORDER_NR_PAGES should biggest page be bigger then
	 * a single pageblock.
	 */
	MIGRATE_CMA,
#endif
65
#ifdef CONFIG_MEMORY_ISOLATION
66
	MIGRATE_ISOLATE,	/* can't allocate from here */
67
#endif
68 69 70
	MIGRATE_TYPES
};

71 72 73
/* In mm/page_alloc.c; keep in sync also with show_migration_types() there */
extern char * const migratetype_names[MIGRATE_TYPES];

74 75
#ifdef CONFIG_CMA
#  define is_migrate_cma(migratetype) unlikely((migratetype) == MIGRATE_CMA)
L
Laura Abbott 已提交
76
#  define is_migrate_cma_page(_page) (get_pageblock_migratetype(_page) == MIGRATE_CMA)
77 78
#else
#  define is_migrate_cma(migratetype) false
L
Laura Abbott 已提交
79
#  define is_migrate_cma_page(_page) false
80
#endif
81

82 83 84 85 86
static inline bool is_migrate_movable(int mt)
{
	return is_migrate_cma(mt) || mt == MIGRATE_MOVABLE;
}

87 88 89 90
#define for_each_migratetype_order(order, type) \
	for (order = 0; order < MAX_ORDER; order++) \
		for (type = 0; type < MIGRATE_TYPES; type++)

91 92
extern int page_group_by_mobility_disabled;

93 94 95
#define NR_MIGRATETYPE_BITS (PB_migrate_end - PB_migrate + 1)
#define MIGRATETYPE_MASK ((1UL << NR_MIGRATETYPE_BITS) - 1)

96 97 98 99
#define get_pageblock_migratetype(page)					\
	get_pfnblock_flags_mask(page, page_to_pfn(page),		\
			PB_migrate_end, MIGRATETYPE_MASK)

L
Linus Torvalds 已提交
100
struct free_area {
101
	struct list_head	free_list[MIGRATE_TYPES];
L
Linus Torvalds 已提交
102 103 104
	unsigned long		nr_free;
};

105 106 107 108 109 110 111 112 113 114 115 116
static inline struct page *get_page_from_free_area(struct free_area *area,
					    int migratetype)
{
	return list_first_entry_or_null(&area->free_list[migratetype],
					struct page, lru);
}

static inline bool free_area_empty(struct free_area *area, int migratetype)
{
	return list_empty(&area->free_list[migratetype]);
}

L
Linus Torvalds 已提交
117 118 119
struct pglist_data;

/*
120
 * zone->lock and the zone lru_lock are two of the hottest locks in the kernel.
L
Linus Torvalds 已提交
121 122 123 124 125 126 127
 * So add a wild amount of padding here to ensure that they fall into separate
 * cachelines.  There are very few zone structures in the machine, so space
 * consumption is not a concern here.
 */
#if defined(CONFIG_SMP)
struct zone_padding {
	char x[0];
128
} ____cacheline_internodealigned_in_smp;
L
Linus Torvalds 已提交
129 130 131 132 133
#define ZONE_PADDING(name)	struct zone_padding name;
#else
#define ZONE_PADDING(name)
#endif

134 135 136 137 138 139 140 141 142 143 144 145 146 147
#ifdef CONFIG_NUMA
enum numa_stat_item {
	NUMA_HIT,		/* allocated in intended node */
	NUMA_MISS,		/* allocated in non intended node */
	NUMA_FOREIGN,		/* was intended here, hit elsewhere */
	NUMA_INTERLEAVE_HIT,	/* interleaver preferred this zone */
	NUMA_LOCAL,		/* allocation from local node */
	NUMA_OTHER,		/* allocation from other node */
	NR_VM_NUMA_STAT_ITEMS
};
#else
#define NR_VM_NUMA_STAT_ITEMS 0
#endif

148
enum zone_stat_item {
149
	/* First 128 byte cacheline (assuming 64 bit words) */
150
	NR_FREE_PAGES,
M
Minchan Kim 已提交
151 152 153 154 155 156
	NR_ZONE_LRU_BASE, /* Used only for compaction and reclaim retry */
	NR_ZONE_INACTIVE_ANON = NR_ZONE_LRU_BASE,
	NR_ZONE_ACTIVE_ANON,
	NR_ZONE_INACTIVE_FILE,
	NR_ZONE_ACTIVE_FILE,
	NR_ZONE_UNEVICTABLE,
157
	NR_ZONE_WRITE_PENDING,	/* Count of dirty, writeback and unstable pages */
N
Nick Piggin 已提交
158
	NR_MLOCK,		/* mlock()ed pages found and moved off LRU */
159
	NR_PAGETABLE,		/* used for pagetables */
160
	NR_KERNEL_STACK_KB,	/* measured in KiB */
161
	/* Second 128 byte cacheline */
162
	NR_BOUNCE,
M
Minchan Kim 已提交
163 164
#if IS_ENABLED(CONFIG_ZSMALLOC)
	NR_ZSPAGES,		/* allocated in zsmalloc */
165
#endif
166
	NR_FREE_CMA_PAGES,
167 168
	NR_VM_ZONE_STAT_ITEMS };

169
enum node_stat_item {
M
Mel Gorman 已提交
170 171 172 173 174 175
	NR_LRU_BASE,
	NR_INACTIVE_ANON = NR_LRU_BASE, /* must match order of LRU_[IN]ACTIVE */
	NR_ACTIVE_ANON,		/*  "     "     "   "       "         */
	NR_INACTIVE_FILE,	/*  "     "     "   "       "         */
	NR_ACTIVE_FILE,		/*  "     "     "   "       "         */
	NR_UNEVICTABLE,		/*  "     "     "   "       "         */
176 177
	NR_SLAB_RECLAIMABLE,
	NR_SLAB_UNRECLAIMABLE,
M
Mel Gorman 已提交
178 179
	NR_ISOLATED_ANON,	/* Temporary isolated pages from anon lru */
	NR_ISOLATED_FILE,	/* Temporary isolated pages from file lru */
180 181
	WORKINGSET_REFAULT,
	WORKINGSET_ACTIVATE,
182
	WORKINGSET_RESTORE,
183
	WORKINGSET_NODERECLAIM,
184
	NR_ANON_MAPPED,	/* Mapped anonymous pages */
185 186
	NR_FILE_MAPPED,	/* pagecache pages mapped into pagetables.
			   only modified from process context */
187 188 189 190 191 192 193 194 195
	NR_FILE_PAGES,
	NR_FILE_DIRTY,
	NR_WRITEBACK,
	NR_WRITEBACK_TEMP,	/* Writeback using temporary buffers */
	NR_SHMEM,		/* shmem pages (included tmpfs/GEM pages) */
	NR_SHMEM_THPS,
	NR_SHMEM_PMDMAPPED,
	NR_ANON_THPS,
	NR_UNSTABLE_NFS,	/* NFS unstable pages */
196 197 198 199
	NR_VMSCAN_WRITE,
	NR_VMSCAN_IMMEDIATE,	/* Prioritise for reclaim when writeback ends */
	NR_DIRTIED,		/* page dirtyings since bootup */
	NR_WRITTEN,		/* page writings since bootup */
200
	NR_INDIRECTLY_RECLAIMABLE_BYTES, /* measured in bytes */
201 202 203
	NR_VM_NODE_STAT_ITEMS
};

204 205 206 207 208 209 210 211 212 213 214 215 216
/*
 * We do arithmetic on the LRU lists in various places in the code,
 * so it is important to keep the active lists LRU_ACTIVE higher in
 * the array than the corresponding inactive lists, and to keep
 * the *_FILE lists LRU_FILE higher than the corresponding _ANON lists.
 *
 * This has to be kept in sync with the statistics in zone_stat_item
 * above and the descriptions in vmstat_text in mm/vmstat.c
 */
#define LRU_BASE 0
#define LRU_ACTIVE 1
#define LRU_FILE 2

217
enum lru_list {
218 219 220 221
	LRU_INACTIVE_ANON = LRU_BASE,
	LRU_ACTIVE_ANON = LRU_BASE + LRU_ACTIVE,
	LRU_INACTIVE_FILE = LRU_BASE + LRU_FILE,
	LRU_ACTIVE_FILE = LRU_BASE + LRU_FILE + LRU_ACTIVE,
L
Lee Schermerhorn 已提交
222 223 224
	LRU_UNEVICTABLE,
	NR_LRU_LISTS
};
225

H
Hugh Dickins 已提交
226
#define for_each_lru(lru) for (lru = 0; lru < NR_LRU_LISTS; lru++)
227

H
Hugh Dickins 已提交
228
#define for_each_evictable_lru(lru) for (lru = 0; lru <= LRU_ACTIVE_FILE; lru++)
L
Lee Schermerhorn 已提交
229

H
Hugh Dickins 已提交
230
static inline int is_file_lru(enum lru_list lru)
231
{
H
Hugh Dickins 已提交
232
	return (lru == LRU_INACTIVE_FILE || lru == LRU_ACTIVE_FILE);
233 234
}

H
Hugh Dickins 已提交
235
static inline int is_active_lru(enum lru_list lru)
236
{
H
Hugh Dickins 已提交
237
	return (lru == LRU_ACTIVE_ANON || lru == LRU_ACTIVE_FILE);
238 239
}

240 241 242
struct zone_reclaim_stat {
	/*
	 * The pageout code in vmscan.c keeps track of how many of the
243
	 * mem/swap backed and file backed pages are referenced.
244 245 246 247 248 249 250 251 252
	 * The higher the rotated/scanned ratio, the more valuable
	 * that cache is.
	 *
	 * The anon LRU stats live in [0], file LRU stats in [1]
	 */
	unsigned long		recent_rotated[2];
	unsigned long		recent_scanned[2];
};

253
struct lruvec {
254 255 256 257
	struct list_head		lists[NR_LRU_LISTS];
	struct zone_reclaim_stat	reclaim_stat;
	/* Evictions & activations on the inactive file list */
	atomic_long_t			inactive_age;
258 259
	/* Refaults at the time of last reclaim cycle */
	unsigned long			refaults;
A
Andrew Morton 已提交
260
#ifdef CONFIG_MEMCG
M
Mel Gorman 已提交
261
	struct pglist_data *pgdat;
262
#endif
263 264
};

265 266 267 268 269
/* Mask used at gathering information at once (see memcontrol.c) */
#define LRU_ALL_FILE (BIT(LRU_INACTIVE_FILE) | BIT(LRU_ACTIVE_FILE))
#define LRU_ALL_ANON (BIT(LRU_INACTIVE_ANON) | BIT(LRU_ACTIVE_ANON))
#define LRU_ALL	     ((1 << NR_LRU_LISTS) - 1)

270
/* Isolate unmapped file */
271
#define ISOLATE_UNMAPPED	((__force isolate_mode_t)0x2)
272
/* Isolate for asynchronous migration */
273
#define ISOLATE_ASYNC_MIGRATE	((__force isolate_mode_t)0x4)
M
Minchan Kim 已提交
274 275
/* Isolate unevictable pages */
#define ISOLATE_UNEVICTABLE	((__force isolate_mode_t)0x8)
276 277

/* LRU Isolation modes. */
278
typedef unsigned __bitwise isolate_mode_t;
279

280 281 282 283 284 285 286
enum zone_watermarks {
	WMARK_MIN,
	WMARK_LOW,
	WMARK_HIGH,
	NR_WMARK
};

287 288 289 290
#define min_wmark_pages(z) (z->_watermark[WMARK_MIN])
#define low_wmark_pages(z) (z->_watermark[WMARK_LOW])
#define high_wmark_pages(z) (z->_watermark[WMARK_HIGH])
#define wmark_pages(z, i) (z->_watermark[i])
291

L
Linus Torvalds 已提交
292 293 294 295
struct per_cpu_pages {
	int count;		/* number of pages in the list */
	int high;		/* high watermark, emptying needed */
	int batch;		/* chunk size for buddy add/remove */
296 297 298

	/* Lists of pages, one per migrate type stored on the pcp-lists */
	struct list_head lists[MIGRATE_PCPTYPES];
L
Linus Torvalds 已提交
299 300 301
};

struct per_cpu_pageset {
302
	struct per_cpu_pages pcp;
303 304
#ifdef CONFIG_NUMA
	s8 expire;
305
	u16 vm_numa_stat_diff[NR_VM_NUMA_STAT_ITEMS];
306
#endif
307
#ifdef CONFIG_SMP
308
	s8 stat_threshold;
309 310
	s8 vm_stat_diff[NR_VM_ZONE_STAT_ITEMS];
#endif
311
};
312

313 314 315 316 317
struct per_cpu_nodestat {
	s8 stat_threshold;
	s8 vm_node_stat_diff[NR_VM_NODE_STAT_ITEMS];
};

C
Christoph Lameter 已提交
318 319
#endif /* !__GENERATING_BOUNDS.H */

320
enum zone_type {
321
#ifdef CONFIG_ZONE_DMA
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340
	/*
	 * ZONE_DMA is used when there are devices that are not able
	 * to do DMA to all of addressable memory (ZONE_NORMAL). Then we
	 * carve out the portion of memory that is needed for these devices.
	 * The range is arch specific.
	 *
	 * Some examples
	 *
	 * Architecture		Limit
	 * ---------------------------
	 * parisc, ia64, sparc	<4G
	 * s390			<2G
	 * arm			Various
	 * alpha		Unlimited or 0-16MB.
	 *
	 * i386, x86_64 and multiple other arches
	 * 			<16M.
	 */
	ZONE_DMA,
341
#endif
342
#ifdef CONFIG_ZONE_DMA32
343 344 345 346 347 348
	/*
	 * x86_64 needs two ZONE_DMAs because it supports devices that are
	 * only able to do DMA to the lower 16M but also 32 bit devices that
	 * can only do DMA areas below 4G.
	 */
	ZONE_DMA32,
349
#endif
350 351 352 353 354 355
	/*
	 * Normal addressable memory is in ZONE_NORMAL. DMA operations can be
	 * performed on pages in ZONE_NORMAL if the DMA devices support
	 * transfers to all addressable memory.
	 */
	ZONE_NORMAL,
356
#ifdef CONFIG_HIGHMEM
357 358 359 360 361 362 363 364 365
	/*
	 * A memory area that is only addressable by the kernel through
	 * mapping portions into its own address space. This is for example
	 * used by i386 to allow the kernel to address the memory beyond
	 * 900MB. The kernel will set up special mappings (page
	 * table entries on i386) for each page that the kernel needs to
	 * access.
	 */
	ZONE_HIGHMEM,
366
#endif
M
Mel Gorman 已提交
367
	ZONE_MOVABLE,
368 369 370
#ifdef CONFIG_ZONE_DEVICE
	ZONE_DEVICE,
#endif
C
Christoph Lameter 已提交
371
	__MAX_NR_ZONES
372

373
};
L
Linus Torvalds 已提交
374

C
Christoph Lameter 已提交
375 376
#ifndef __GENERATING_BOUNDS_H

L
Linus Torvalds 已提交
377
struct zone {
378
	/* Read-mostly fields */
379 380

	/* zone watermarks, access with *_wmark_pages(zone) macros */
381
	unsigned long _watermark[NR_WMARK];
382

383 384
	unsigned long nr_reserved_highatomic;

L
Linus Torvalds 已提交
385
	/*
386 387 388 389 390 391 392
	 * We don't know if the memory that we're going to allocate will be
	 * freeable or/and it will be released eventually, so to avoid totally
	 * wasting several GB of ram we must reserve some of the lower zone
	 * memory (otherwise we risk to run OOM on the lower zones despite
	 * there being tons of freeable ram on the higher zones).  This array is
	 * recalculated at runtime if the sysctl_lowmem_reserve_ratio sysctl
	 * changes.
L
Linus Torvalds 已提交
393
	 */
394
	long lowmem_reserve[MAX_NR_ZONES];
395

396
#ifdef CONFIG_NUMA
397
	int node;
398 399
#endif
	struct pglist_data	*zone_pgdat;
400
	struct per_cpu_pageset __percpu *pageset;
401

402 403
#ifndef CONFIG_SPARSEMEM
	/*
404
	 * Flags for a pageblock_nr_pages block. See pageblock-flags.h.
405 406 407 408 409
	 * In SPARSEMEM, this map is stored in struct mem_section
	 */
	unsigned long		*pageblock_flags;
#endif /* CONFIG_SPARSEMEM */

L
Linus Torvalds 已提交
410 411 412
	/* zone_start_pfn == zone_start_paddr >> PAGE_SHIFT */
	unsigned long		zone_start_pfn;

413
	/*
414 415 416
	 * spanned_pages is the total pages spanned by the zone, including
	 * holes, which is calculated as:
	 * 	spanned_pages = zone_end_pfn - zone_start_pfn;
417
	 *
418 419
	 * present_pages is physical pages existing within the zone, which
	 * is calculated as:
420
	 *	present_pages = spanned_pages - absent_pages(pages in holes);
421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
	 *
	 * managed_pages is present pages managed by the buddy system, which
	 * is calculated as (reserved_pages includes pages allocated by the
	 * bootmem allocator):
	 *	managed_pages = present_pages - reserved_pages;
	 *
	 * So present_pages may be used by memory hotplug or memory power
	 * management logic to figure out unmanaged pages by checking
	 * (present_pages - managed_pages). And managed_pages should be used
	 * by page allocator and vm scanner to calculate all kinds of watermarks
	 * and thresholds.
	 *
	 * Locking rules:
	 *
	 * zone_start_pfn and spanned_pages are protected by span_seqlock.
	 * It is a seqlock because it has to be read outside of zone->lock,
	 * and it is done in the main allocator path.  But, it is written
	 * quite infrequently.
	 *
	 * The span_seq lock is declared along with zone->lock because it is
441 442
	 * frequently read in proximity to zone->lock.  It's good to
	 * give them a chance of being in the same cacheline.
443
	 *
444
	 * Write access to present_pages at runtime should be protected by
445 446
	 * mem_hotplug_begin/end(). Any reader who can't tolerant drift of
	 * present_pages should get_online_mems() to get a stable value.
447 448 449 450 451 452
	 *
	 * Read access to managed_pages should be safe because it's unsigned
	 * long. Write access to zone->managed_pages and totalram_pages are
	 * protected by managed_page_count_lock at runtime. Idealy only
	 * adjust_managed_page_count() should be used instead of directly
	 * touching zone->managed_pages and totalram_pages.
453
	 */
454
	unsigned long		managed_pages;
455 456
	unsigned long		spanned_pages;
	unsigned long		present_pages;
457 458

	const char		*name;
L
Linus Torvalds 已提交
459

460 461 462 463 464 465 466 467 468
#ifdef CONFIG_MEMORY_ISOLATION
	/*
	 * Number of isolated pageblock. It is used to solve incorrect
	 * freepage counting problem due to racy retrieving migratetype
	 * of pageblock. Protected by zone->lock.
	 */
	unsigned long		nr_isolate_pageblock;
#endif

469 470 471 472 473
#ifdef CONFIG_MEMORY_HOTPLUG
	/* see spanned/present_pages for more description */
	seqlock_t		span_seqlock;
#endif

474
	int initialized;
475

476
	/* Write-intensive fields used from the page allocator */
477
	ZONE_PADDING(_pad1_)
478

479 480 481 482 483 484
	/* free areas of different sizes */
	struct free_area	free_area[MAX_ORDER];

	/* zone flags, see below */
	unsigned long		flags;

485
	/* Primarily protects free_area */
486 487
	spinlock_t		lock;

488
	/* Write-intensive fields used by compaction and vmstats. */
489 490 491 492 493 494 495 496 497 498 499 500 501 502
	ZONE_PADDING(_pad2_)

	/*
	 * When free pages are below this point, additional steps are taken
	 * when reading the number of free pages to avoid per-cpu counter
	 * drift allowing watermarks to be breached
	 */
	unsigned long percpu_drift_mark;

#if defined CONFIG_COMPACTION || defined CONFIG_CMA
	/* pfn where compaction free scanner should start */
	unsigned long		compact_cached_free_pfn;
	/* pfn where async and sync compaction migration scanner should start */
	unsigned long		compact_cached_migrate_pfn[2];
503 504
	unsigned long		compact_init_migrate_pfn;
	unsigned long		compact_init_free_pfn;
505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522
#endif

#ifdef CONFIG_COMPACTION
	/*
	 * On compaction failure, 1<<compact_defer_shift compactions
	 * are skipped before trying again. The number attempted since
	 * last failure is tracked with compact_considered.
	 */
	unsigned int		compact_considered;
	unsigned int		compact_defer_shift;
	int			compact_order_failed;
#endif

#if defined CONFIG_COMPACTION || defined CONFIG_CMA
	/* Set to true when the PG_migrate_skip bits should be cleared */
	bool			compact_blockskip_flush;
#endif

523 524
	bool			contiguous;

525 526 527
	ZONE_PADDING(_pad3_)
	/* Zone statistics */
	atomic_long_t		vm_stat[NR_VM_ZONE_STAT_ITEMS];
528
	atomic_long_t		vm_numa_stat[NR_VM_NUMA_STAT_ITEMS];
529 530 531

	ALI_HOTFIX_RESERVE(1)
	ALI_HOTFIX_RESERVE(2)
532
} ____cacheline_internodealigned_in_smp;
L
Linus Torvalds 已提交
533

M
Mel Gorman 已提交
534 535
enum pgdat_flags {
	PGDAT_CONGESTED,		/* pgdat has many dirty pages backed by
536 537
					 * a congested BDI
					 */
M
Mel Gorman 已提交
538
	PGDAT_DIRTY,			/* reclaim scanning has recently found
539 540 541
					 * many dirty file pages at the tail
					 * of the LRU.
					 */
M
Mel Gorman 已提交
542
	PGDAT_WRITEBACK,		/* reclaim scanning has recently found
543 544
					 * many pages under writeback
					 */
545
	PGDAT_RECLAIM_LOCKED,		/* prevents concurrent reclaim */
J
Johannes Weiner 已提交
546
};
547

R
Russ Anderson 已提交
548
static inline unsigned long zone_end_pfn(const struct zone *zone)
549 550 551 552 553 554 555 556 557
{
	return zone->zone_start_pfn + zone->spanned_pages;
}

static inline bool zone_spans_pfn(const struct zone *zone, unsigned long pfn)
{
	return zone->zone_start_pfn <= pfn && pfn < zone_end_pfn(zone);
}

558 559
static inline bool zone_is_initialized(struct zone *zone)
{
560
	return zone->initialized;
561 562 563 564 565 566 567
}

static inline bool zone_is_empty(struct zone *zone)
{
	return zone->spanned_pages == 0;
}

568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583
/*
 * Return true if [start_pfn, start_pfn + nr_pages) range has a non-empty
 * intersection with the given zone
 */
static inline bool zone_intersects(struct zone *zone,
		unsigned long start_pfn, unsigned long nr_pages)
{
	if (zone_is_empty(zone))
		return false;
	if (start_pfn >= zone_end_pfn(zone) ||
	    start_pfn + nr_pages <= zone->zone_start_pfn)
		return false;

	return true;
}

L
Linus Torvalds 已提交
584 585 586 587 588 589 590
/*
 * The "priority" of VM scanning is how much of the queues we will scan in one
 * go. A value of 12 for DEF_PRIORITY implies that we will scan 1/4096th of the
 * queues ("queue_length >> 12") during an aging round.
 */
#define DEF_PRIORITY 12

591 592 593
/* Maximum number of zones on a zonelist */
#define MAX_ZONES_PER_ZONELIST (MAX_NUMNODES * MAX_NR_ZONES)

594 595
enum {
	ZONELIST_FALLBACK,	/* zonelist with fallback */
596
#ifdef CONFIG_NUMA
597 598 599 600 601
	/*
	 * The NUMA zonelists are doubled because we need zonelists that
	 * restrict the allocations to a single node for __GFP_THISNODE.
	 */
	ZONELIST_NOFALLBACK,	/* zonelist without fallback (__GFP_THISNODE) */
602
#endif
603 604
	MAX_ZONELISTS
};
605

606 607 608 609 610 611 612 613 614
/*
 * This struct contains information about a zone in a zonelist. It is stored
 * here to avoid dereferences into large structures and lookups of tables
 */
struct zoneref {
	struct zone *zone;	/* Pointer to actual zone */
	int zone_idx;		/* zone_idx(zoneref->zone) */
};

L
Linus Torvalds 已提交
615 616 617 618 619 620
/*
 * One allocation request operates on a zonelist. A zonelist
 * is a list of zones, the first one is the 'goal' of the
 * allocation, the other zones are fallback zones, in decreasing
 * priority.
 *
621 622 623 624 625 626 627
 * To speed the reading of the zonelist, the zonerefs contain the zone index
 * of the entry being read. Helper functions to access information given
 * a struct zoneref are
 *
 * zonelist_zone()	- Return the struct zone * for an entry in _zonerefs
 * zonelist_zone_idx()	- Return the index of the zone for an entry
 * zonelist_node_idx()	- Return the index of the node for an entry
L
Linus Torvalds 已提交
628 629
 */
struct zonelist {
630
	struct zoneref _zonerefs[MAX_ZONES_PER_ZONELIST + 1];
L
Linus Torvalds 已提交
631 632
};

633 634 635 636 637
#ifndef CONFIG_DISCONTIGMEM
/* The array of struct pages - for discontigmem use pgdat->lmem_map */
extern struct page *mem_map;
#endif

638 639 640 641 642 643 644 645
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
struct deferred_split {
	spinlock_t split_queue_lock;
	struct list_head split_queue;
	unsigned long split_queue_len;
};
#endif

L
Linus Torvalds 已提交
646 647
/*
 * On NUMA machines, each NUMA node would have a pg_data_t to describe
648 649
 * it's memory layout. On UMA machines there is a single pglist_data which
 * describes the whole memory.
L
Linus Torvalds 已提交
650 651 652 653 654 655 656
 *
 * Memory statistics and page replacement data structures are maintained on a
 * per-zone basis.
 */
struct bootmem_data;
typedef struct pglist_data {
	struct zone node_zones[MAX_NR_ZONES];
657
	struct zonelist node_zonelists[MAX_ZONELISTS];
L
Linus Torvalds 已提交
658
	int nr_zones;
659
#ifdef CONFIG_FLAT_NODE_MEM_MAP	/* means !SPARSEMEM */
L
Linus Torvalds 已提交
660
	struct page *node_mem_map;
661 662 663
#ifdef CONFIG_PAGE_EXTENSION
	struct page_ext *node_page_ext;
#endif
A
Andy Whitcroft 已提交
664
#endif
665
#ifndef CONFIG_NO_BOOTMEM
L
Linus Torvalds 已提交
666
	struct bootmem_data *bdata;
667
#endif
668
#if defined(CONFIG_MEMORY_HOTPLUG) || defined(CONFIG_DEFERRED_STRUCT_PAGE_INIT)
669 670 671 672 673
	/*
	 * Must be held any time you expect node_start_pfn, node_present_pages
	 * or node_spanned_pages stay constant.  Holding this will also
	 * guarantee that any pfn_valid() stays that way.
	 *
674
	 * pgdat_resize_lock() and pgdat_resize_unlock() are provided to
675 676
	 * manipulate node_size_lock without checking for CONFIG_MEMORY_HOTPLUG
	 * or CONFIG_DEFERRED_STRUCT_PAGE_INIT.
677
	 *
678
	 * Nests above zone->lock and zone->span_seqlock
679 680 681
	 */
	spinlock_t node_size_lock;
#endif
L
Linus Torvalds 已提交
682 683 684 685
	unsigned long node_start_pfn;
	unsigned long node_present_pages; /* total number of physical pages */
	unsigned long node_spanned_pages; /* total size of physical page
					     range, including holes */
G
Gavin Shan 已提交
686 687 688 689 690
#ifdef CONFIG_KIDLED
	unsigned long node_idle_scan_pfn;
	u8 *node_page_age;
#endif

L
Linus Torvalds 已提交
691 692
	int node_id;
	wait_queue_head_t kswapd_wait;
693
	wait_queue_head_t pfmemalloc_wait;
694 695
	struct task_struct *kswapd;	/* Protected by
					   mem_hotplug_begin/end() */
696 697 698
	int kswapd_order;
	enum zone_type kswapd_classzone_idx;

699 700
	int kswapd_failures;		/* Number of 'reclaimed == 0' runs */

701 702 703 704 705
#ifdef CONFIG_COMPACTION
	int kcompactd_max_order;
	enum zone_type kcompactd_classzone_idx;
	wait_queue_head_t kcompactd_wait;
	struct task_struct *kcompactd;
706
#endif
707 708 709 710 711 712
	/*
	 * This is a per-node reserve of pages that are not available
	 * to userspace allocations.
	 */
	unsigned long		totalreserve_pages;

713 714 715 716 717 718 719 720
#ifdef CONFIG_NUMA
	/*
	 * zone reclaim becomes active if more unmapped pages exist.
	 */
	unsigned long		min_unmapped_pages;
	unsigned long		min_slab_pages;
#endif /* CONFIG_NUMA */

721 722 723
	/* Write-intensive fields used by page reclaim */
	ZONE_PADDING(_pad1_)
	spinlock_t		lru_lock;
724 725 726 727 728 729 730

#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT
	/*
	 * If memory initialisation on large machines is deferred then this
	 * is the first PFN that needs to be initialised.
	 */
	unsigned long first_deferred_pfn;
731 732
	/* Number of non-deferred pages */
	unsigned long static_init_pgcnt;
733
#endif /* CONFIG_DEFERRED_STRUCT_PAGE_INIT */
734 735

#ifdef CONFIG_TRANSPARENT_HUGEPAGE
736
	struct deferred_split deferred_split_queue;
737
#endif
738

M
Mel Gorman 已提交
739 740 741 742 743 744 745
	/* Fields commonly accessed by the page reclaim scanner */
	struct lruvec		lruvec;

	unsigned long		flags;

	ZONE_PADDING(_pad2_)

746 747 748
	/* Per-node vmstats */
	struct per_cpu_nodestat __percpu *per_cpu_nodestats;
	atomic_long_t		vm_stat[NR_VM_NODE_STAT_ITEMS];
749 750 751

	ALI_HOTFIX_RESERVE(1)
	ALI_HOTFIX_RESERVE(2)
L
Linus Torvalds 已提交
752 753 754 755
} pg_data_t;

#define node_present_pages(nid)	(NODE_DATA(nid)->node_present_pages)
#define node_spanned_pages(nid)	(NODE_DATA(nid)->node_spanned_pages)
A
Andy Whitcroft 已提交
756
#ifdef CONFIG_FLAT_NODE_MEM_MAP
757
#define pgdat_page_nr(pgdat, pagenr)	((pgdat)->node_mem_map + (pagenr))
A
Andy Whitcroft 已提交
758 759 760
#else
#define pgdat_page_nr(pgdat, pagenr)	pfn_to_page((pgdat)->node_start_pfn + (pagenr))
#endif
761
#define nid_page_nr(nid, pagenr) 	pgdat_page_nr(NODE_DATA(nid),(pagenr))
L
Linus Torvalds 已提交
762

763
#define node_start_pfn(nid)	(NODE_DATA(nid)->node_start_pfn)
764
#define node_end_pfn(nid) pgdat_end_pfn(NODE_DATA(nid))
765 766 767 768
static inline spinlock_t *zone_lru_lock(struct zone *zone)
{
	return &zone->zone_pgdat->lru_lock;
}
769

770
static inline struct lruvec *node_lruvec(struct pglist_data *pgdat)
M
Mel Gorman 已提交
771
{
772
	return &pgdat->lruvec;
M
Mel Gorman 已提交
773 774
}

775 776 777 778 779 780 781 782 783
static inline unsigned long pgdat_end_pfn(pg_data_t *pgdat)
{
	return pgdat->node_start_pfn + pgdat->node_spanned_pages;
}

static inline bool pgdat_is_empty(pg_data_t *pgdat)
{
	return !pgdat->node_start_pfn && !pgdat->node_spanned_pages;
}
784

785 786
#include <linux/memory_hotplug.h>

787
void build_all_zonelists(pg_data_t *pgdat);
788 789
void wakeup_kswapd(struct zone *zone, gfp_t gfp_mask, int order,
		   enum zone_type classzone_idx);
790 791 792
bool __zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark,
			 int classzone_idx, unsigned int alloc_flags,
			 long free_pages);
793
bool zone_watermark_ok(struct zone *z, unsigned int order,
794 795
		unsigned long mark, int classzone_idx,
		unsigned int alloc_flags);
796
bool zone_watermark_ok_safe(struct zone *z, unsigned int order,
797
		unsigned long mark, int classzone_idx);
D
Dave Hansen 已提交
798 799 800 801
enum memmap_context {
	MEMMAP_EARLY,
	MEMMAP_HOTPLUG,
};
802
extern void init_currently_empty_zone(struct zone *zone, unsigned long start_pfn,
803
				     unsigned long size);
804

805
extern void lruvec_init(struct lruvec *lruvec);
806

M
Mel Gorman 已提交
807
static inline struct pglist_data *lruvec_pgdat(struct lruvec *lruvec)
808
{
A
Andrew Morton 已提交
809
#ifdef CONFIG_MEMCG
M
Mel Gorman 已提交
810
	return lruvec->pgdat;
811
#else
M
Mel Gorman 已提交
812
	return container_of(lruvec, struct pglist_data, lruvec);
813 814 815
#endif
}

816
extern unsigned long lruvec_lru_size(struct lruvec *lruvec, enum lru_list lru, int zone_idx);
817

L
Linus Torvalds 已提交
818 819 820 821 822 823
#ifdef CONFIG_HAVE_MEMORY_PRESENT
void memory_present(int nid, unsigned long start, unsigned long end);
#else
static inline void memory_present(int nid, unsigned long start, unsigned long end) {}
#endif

824 825 826 827 828 829
#ifdef CONFIG_HAVE_MEMORYLESS_NODES
int local_memory_node(int node_id);
#else
static inline int local_memory_node(int node_id) { return node_id; };
#endif

L
Linus Torvalds 已提交
830 831 832 833 834
/*
 * zone_idx() returns 0 for the ZONE_DMA zone, 1 for the ZONE_NORMAL zone, etc.
 */
#define zone_idx(zone)		((zone) - (zone)->zone_pgdat->node_zones)

835 836 837 838 839 840 841 842 843 844 845 846
#ifdef CONFIG_ZONE_DEVICE
static inline bool is_dev_zone(const struct zone *zone)
{
	return zone_idx(zone) == ZONE_DEVICE;
}
#else
static inline bool is_dev_zone(const struct zone *zone)
{
	return false;
}
#endif

847 848 849 850 851 852 853 854 855 856 857 858 859
/*
 * Returns true if a zone has pages managed by the buddy allocator.
 * All the reclaim decisions have to use this function rather than
 * populated_zone(). If the whole zone is reserved then we can easily
 * end up with populated_zone() && !managed_zone().
 */
static inline bool managed_zone(struct zone *zone)
{
	return zone->managed_pages;
}

/* Returns true if a zone has memory */
static inline bool populated_zone(struct zone *zone)
860
{
861
	return zone->present_pages;
862 863
}

864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882
#ifdef CONFIG_NUMA
static inline int zone_to_nid(struct zone *zone)
{
	return zone->node;
}

static inline void zone_set_nid(struct zone *zone, int nid)
{
	zone->node = nid;
}
#else
static inline int zone_to_nid(struct zone *zone)
{
	return 0;
}

static inline void zone_set_nid(struct zone *zone, int nid) {}
#endif

M
Mel Gorman 已提交
883 884
extern int movable_zone;

885
#ifdef CONFIG_HIGHMEM
M
Mel Gorman 已提交
886 887
static inline int zone_movable_is_highmem(void)
{
888
#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP
M
Mel Gorman 已提交
889 890
	return movable_zone == ZONE_HIGHMEM;
#else
891
	return (ZONE_MOVABLE - 1) == ZONE_HIGHMEM;
M
Mel Gorman 已提交
892 893
#endif
}
894
#endif
M
Mel Gorman 已提交
895

896
static inline int is_highmem_idx(enum zone_type idx)
L
Linus Torvalds 已提交
897
{
898
#ifdef CONFIG_HIGHMEM
M
Mel Gorman 已提交
899 900
	return (idx == ZONE_HIGHMEM ||
		(idx == ZONE_MOVABLE && zone_movable_is_highmem()));
901 902 903
#else
	return 0;
#endif
L
Linus Torvalds 已提交
904 905 906 907 908 909 910 911 912 913
}

/**
 * is_highmem - helper function to quickly check if a struct zone is a 
 *              highmem zone or not.  This is an attempt to keep references
 *              to ZONE_{DMA/NORMAL/HIGHMEM/etc} in general code to a minimum.
 * @zone - pointer to struct zone variable
 */
static inline int is_highmem(struct zone *zone)
{
914
#ifdef CONFIG_HIGHMEM
C
Chanho Min 已提交
915
	return is_highmem_idx(zone_idx(zone));
916 917 918
#else
	return 0;
#endif
L
Linus Torvalds 已提交
919 920 921 922
}

/* These two functions are used to setup the per zone pages min values */
struct ctl_table;
923
int min_free_kbytes_sysctl_handler(struct ctl_table *, int,
L
Linus Torvalds 已提交
924
					void __user *, size_t *, loff_t *);
925 926
int watermark_scale_factor_sysctl_handler(struct ctl_table *, int,
					void __user *, size_t *, loff_t *);
927
extern int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES];
928
int lowmem_reserve_ratio_sysctl_handler(struct ctl_table *, int,
L
Linus Torvalds 已提交
929
					void __user *, size_t *, loff_t *);
930
int percpu_pagelist_fraction_sysctl_handler(struct ctl_table *, int,
931
					void __user *, size_t *, loff_t *);
932
int sysctl_min_unmapped_ratio_sysctl_handler(struct ctl_table *, int,
933
			void __user *, size_t *, loff_t *);
934
int sysctl_min_slab_ratio_sysctl_handler(struct ctl_table *, int,
935
			void __user *, size_t *, loff_t *);
L
Linus Torvalds 已提交
936

937
extern int numa_zonelist_order_handler(struct ctl_table *, int,
938
			void __user *, size_t *, loff_t *);
939
extern char numa_zonelist_order[];
940
#define NUMA_ZONELIST_ORDER_LEN	16
941

942
#ifndef CONFIG_NEED_MULTIPLE_NODES
L
Linus Torvalds 已提交
943 944 945 946 947

extern struct pglist_data contig_page_data;
#define NODE_DATA(nid)		(&contig_page_data)
#define NODE_MEM_MAP(nid)	mem_map

948
#else /* CONFIG_NEED_MULTIPLE_NODES */
L
Linus Torvalds 已提交
949 950 951

#include <asm/mmzone.h>

952
#endif /* !CONFIG_NEED_MULTIPLE_NODES */
953

954 955 956
extern struct pglist_data *first_online_pgdat(void);
extern struct pglist_data *next_online_pgdat(struct pglist_data *pgdat);
extern struct zone *next_zone(struct zone *zone);
957 958

/**
959
 * for_each_online_pgdat - helper macro to iterate over all online nodes
960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977
 * @pgdat - pointer to a pg_data_t variable
 */
#define for_each_online_pgdat(pgdat)			\
	for (pgdat = first_online_pgdat();		\
	     pgdat;					\
	     pgdat = next_online_pgdat(pgdat))
/**
 * for_each_zone - helper macro to iterate over all memory zones
 * @zone - pointer to struct zone variable
 *
 * The user only needs to declare the zone variable, for_each_zone
 * fills it in.
 */
#define for_each_zone(zone)			        \
	for (zone = (first_online_pgdat())->node_zones; \
	     zone;					\
	     zone = next_zone(zone))

978 979 980 981 982 983 984 985
#define for_each_populated_zone(zone)		        \
	for (zone = (first_online_pgdat())->node_zones; \
	     zone;					\
	     zone = next_zone(zone))			\
		if (!populated_zone(zone))		\
			; /* do nothing */		\
		else

986 987 988 989 990 991 992 993 994 995 996 997
static inline struct zone *zonelist_zone(struct zoneref *zoneref)
{
	return zoneref->zone;
}

static inline int zonelist_zone_idx(struct zoneref *zoneref)
{
	return zoneref->zone_idx;
}

static inline int zonelist_node_idx(struct zoneref *zoneref)
{
998
	return zone_to_nid(zoneref->zone);
999 1000
}

1001 1002 1003 1004
struct zoneref *__next_zones_zonelist(struct zoneref *z,
					enum zone_type highest_zoneidx,
					nodemask_t *nodes);

1005 1006 1007 1008 1009 1010 1011 1012
/**
 * next_zones_zonelist - Returns the next zone at or below highest_zoneidx within the allowed nodemask using a cursor within a zonelist as a starting point
 * @z - The cursor used as a starting point for the search
 * @highest_zoneidx - The zone index of the highest zone to return
 * @nodes - An optional nodemask to filter the zonelist with
 *
 * This function returns the next zone at or below a given zone index that is
 * within the allowed nodemask using a cursor as the starting point for the
1013 1014 1015
 * search. The zoneref returned is a cursor that represents the current zone
 * being examined. It should be advanced by one before calling
 * next_zones_zonelist again.
1016
 */
1017
static __always_inline struct zoneref *next_zones_zonelist(struct zoneref *z,
1018
					enum zone_type highest_zoneidx,
1019 1020 1021 1022 1023 1024
					nodemask_t *nodes)
{
	if (likely(!nodes && zonelist_zone_idx(z) <= highest_zoneidx))
		return z;
	return __next_zones_zonelist(z, highest_zoneidx, nodes);
}
1025

1026 1027 1028 1029 1030
/**
 * first_zones_zonelist - Returns the first zone at or below highest_zoneidx within the allowed nodemask in a zonelist
 * @zonelist - The zonelist to search for a suitable zone
 * @highest_zoneidx - The zone index of the highest zone to return
 * @nodes - An optional nodemask to filter the zonelist with
1031
 * @return - Zoneref pointer for the first suitable zone found (see below)
1032 1033 1034
 *
 * This function returns the first zone at or below a given zone index that is
 * within the allowed nodemask. The zoneref returned is a cursor that can be
1035 1036
 * used to iterate the zonelist with next_zones_zonelist by advancing it by
 * one before calling.
1037 1038 1039 1040
 *
 * When no eligible zone is found, zoneref->zone is NULL (zoneref itself is
 * never NULL). This may happen either genuinely, or due to concurrent nodemask
 * update due to cpuset modification.
1041
 */
1042
static inline struct zoneref *first_zones_zonelist(struct zonelist *zonelist,
1043
					enum zone_type highest_zoneidx,
1044
					nodemask_t *nodes)
1045
{
1046
	return next_zones_zonelist(zonelist->_zonerefs,
1047
							highest_zoneidx, nodes);
1048 1049
}

1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
/**
 * for_each_zone_zonelist_nodemask - helper macro to iterate over valid zones in a zonelist at or below a given zone index and within a nodemask
 * @zone - The current zone in the iterator
 * @z - The current pointer within zonelist->zones being iterated
 * @zlist - The zonelist being iterated
 * @highidx - The zone index of the highest zone to return
 * @nodemask - Nodemask allowed by the allocator
 *
 * This iterator iterates though all zones at or below a given zone index and
 * within a given nodemask
 */
#define for_each_zone_zonelist_nodemask(zone, z, zlist, highidx, nodemask) \
1062
	for (z = first_zones_zonelist(zlist, highidx, nodemask), zone = zonelist_zone(z);	\
1063
		zone;							\
1064
		z = next_zones_zonelist(++z, highidx, nodemask),	\
1065 1066 1067 1068 1069 1070 1071 1072
			zone = zonelist_zone(z))

#define for_next_zone_zonelist_nodemask(zone, z, zlist, highidx, nodemask) \
	for (zone = z->zone;	\
		zone;							\
		z = next_zones_zonelist(++z, highidx, nodemask),	\
			zone = zonelist_zone(z))

1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083

/**
 * for_each_zone_zonelist - helper macro to iterate over valid zones in a zonelist at or below a given zone index
 * @zone - The current zone in the iterator
 * @z - The current pointer within zonelist->zones being iterated
 * @zlist - The zonelist being iterated
 * @highidx - The zone index of the highest zone to return
 *
 * This iterator iterates though all zones at or below a given zone index.
 */
#define for_each_zone_zonelist(zone, z, zlist, highidx) \
1084
	for_each_zone_zonelist_nodemask(zone, z, zlist, highidx, NULL)
1085

A
Andy Whitcroft 已提交
1086 1087 1088 1089
#ifdef CONFIG_SPARSEMEM
#include <asm/sparsemem.h>
#endif

1090
#if !defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID) && \
T
Tejun Heo 已提交
1091
	!defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP)
1092 1093
static inline unsigned long early_pfn_to_nid(unsigned long pfn)
{
1094
	BUILD_BUG_ON(IS_ENABLED(CONFIG_NUMA));
1095 1096
	return 0;
}
1097 1098
#endif

1099 1100 1101 1102
#ifdef CONFIG_FLATMEM
#define pfn_to_nid(pfn)		(0)
#endif

A
Andy Whitcroft 已提交
1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118
#ifdef CONFIG_SPARSEMEM

/*
 * SECTION_SHIFT    		#bits space required to store a section #
 *
 * PA_SECTION_SHIFT		physical address to/from section number
 * PFN_SECTION_SHIFT		pfn to/from section number
 */
#define PA_SECTION_SHIFT	(SECTION_SIZE_BITS)
#define PFN_SECTION_SHIFT	(SECTION_SIZE_BITS - PAGE_SHIFT)

#define NR_MEM_SECTIONS		(1UL << SECTIONS_SHIFT)

#define PAGES_PER_SECTION       (1UL << PFN_SECTION_SHIFT)
#define PAGE_SECTION_MASK	(~(PAGES_PER_SECTION-1))

1119
#define SECTION_BLOCKFLAGS_BITS \
1120
	((1UL << (PFN_SECTION_SHIFT - pageblock_order)) * NR_PAGEBLOCK_BITS)
1121

A
Andy Whitcroft 已提交
1122 1123 1124 1125
#if (MAX_ORDER - 1 + PAGE_SHIFT) > SECTION_SIZE_BITS
#error Allocator MAX_ORDER exceeds SECTION_SIZE
#endif

1126 1127 1128 1129 1130 1131 1132 1133
static inline unsigned long pfn_to_section_nr(unsigned long pfn)
{
	return pfn >> PFN_SECTION_SHIFT;
}
static inline unsigned long section_nr_to_pfn(unsigned long sec)
{
	return sec << PFN_SECTION_SHIFT;
}
1134

1135 1136 1137
#define SECTION_ALIGN_UP(pfn)	(((pfn) + PAGES_PER_SECTION - 1) & PAGE_SECTION_MASK)
#define SECTION_ALIGN_DOWN(pfn)	((pfn) & PAGE_SECTION_MASK)

A
Andy Whitcroft 已提交
1138
struct page;
1139
struct page_ext;
A
Andy Whitcroft 已提交
1140
struct mem_section {
A
Andy Whitcroft 已提交
1141 1142 1143 1144 1145
	/*
	 * This is, logically, a pointer to an array of struct
	 * pages.  However, it is stored with some other magic.
	 * (see sparse.c::sparse_init_one_section())
	 *
1146 1147 1148 1149
	 * Additionally during early boot we encode node id of
	 * the location of the section here to guide allocation.
	 * (see sparse.c::memory_present())
	 *
A
Andy Whitcroft 已提交
1150 1151 1152 1153
	 * Making it a UL at least makes someone do a cast
	 * before using it wrong.
	 */
	unsigned long section_mem_map;
1154 1155 1156

	/* See declaration of similar field in struct zone */
	unsigned long *pageblock_flags;
1157 1158
#ifdef CONFIG_PAGE_EXTENSION
	/*
1159
	 * If SPARSEMEM, pgdat doesn't have page_ext pointer. We use
1160 1161 1162 1163 1164
	 * section. (see page_ext.h about this.)
	 */
	struct page_ext *page_ext;
	unsigned long pad;
#endif
1165 1166 1167 1168
	/*
	 * WARNING: mem_section must be a power-of-2 in size for the
	 * calculation and use of SECTION_ROOT_MASK to make sense.
	 */
A
Andy Whitcroft 已提交
1169 1170
};

1171 1172 1173 1174 1175
#ifdef CONFIG_SPARSEMEM_EXTREME
#define SECTIONS_PER_ROOT       (PAGE_SIZE / sizeof (struct mem_section))
#else
#define SECTIONS_PER_ROOT	1
#endif
B
Bob Picco 已提交
1176

1177
#define SECTION_NR_TO_ROOT(sec)	((sec) / SECTIONS_PER_ROOT)
1178
#define NR_SECTION_ROOTS	DIV_ROUND_UP(NR_MEM_SECTIONS, SECTIONS_PER_ROOT)
1179
#define SECTION_ROOT_MASK	(SECTIONS_PER_ROOT - 1)
B
Bob Picco 已提交
1180

1181
#ifdef CONFIG_SPARSEMEM_EXTREME
1182
extern struct mem_section **mem_section;
B
Bob Picco 已提交
1183
#else
1184 1185
extern struct mem_section mem_section[NR_SECTION_ROOTS][SECTIONS_PER_ROOT];
#endif
A
Andy Whitcroft 已提交
1186

A
Andy Whitcroft 已提交
1187 1188
static inline struct mem_section *__nr_to_section(unsigned long nr)
{
1189 1190 1191 1192
#ifdef CONFIG_SPARSEMEM_EXTREME
	if (!mem_section)
		return NULL;
#endif
1193 1194 1195
	if (!mem_section[SECTION_NR_TO_ROOT(nr)])
		return NULL;
	return &mem_section[SECTION_NR_TO_ROOT(nr)][nr & SECTION_ROOT_MASK];
A
Andy Whitcroft 已提交
1196
}
1197
extern int __section_nr(struct mem_section* ms);
1198
extern unsigned long usemap_size(void);
A
Andy Whitcroft 已提交
1199 1200 1201

/*
 * We use the lower bits of the mem_map pointer to store
1202 1203 1204 1205 1206 1207 1208 1209 1210 1211
 * a little bit of information.  The pointer is calculated
 * as mem_map - section_nr_to_pfn(pnum).  The result is
 * aligned to the minimum alignment of the two values:
 *   1. All mem_map arrays are page-aligned.
 *   2. section_nr_to_pfn() always clears PFN_SECTION_SHIFT
 *      lowest bits.  PFN_SECTION_SHIFT is arch-specific
 *      (equal SECTION_SIZE_BITS - PAGE_SHIFT), and the
 *      worst combination is powerpc with 256k pages,
 *      which results in PFN_SECTION_SHIFT equal 6.
 * To sum it up, at least 6 bits are available.
A
Andy Whitcroft 已提交
1212 1213 1214
 */
#define	SECTION_MARKED_PRESENT	(1UL<<0)
#define SECTION_HAS_MEM_MAP	(1UL<<1)
1215 1216
#define SECTION_IS_ONLINE	(1UL<<2)
#define SECTION_MAP_LAST_BIT	(1UL<<3)
A
Andy Whitcroft 已提交
1217
#define SECTION_MAP_MASK	(~(SECTION_MAP_LAST_BIT-1))
1218
#define SECTION_NID_SHIFT	3
A
Andy Whitcroft 已提交
1219 1220 1221 1222 1223 1224 1225 1226

static inline struct page *__section_mem_map_addr(struct mem_section *section)
{
	unsigned long map = section->section_mem_map;
	map &= SECTION_MAP_MASK;
	return (struct page *)map;
}

1227
static inline int present_section(struct mem_section *section)
A
Andy Whitcroft 已提交
1228
{
B
Bob Picco 已提交
1229
	return (section && (section->section_mem_map & SECTION_MARKED_PRESENT));
A
Andy Whitcroft 已提交
1230 1231
}

1232 1233 1234 1235 1236 1237
static inline int present_section_nr(unsigned long nr)
{
	return present_section(__nr_to_section(nr));
}

static inline int valid_section(struct mem_section *section)
A
Andy Whitcroft 已提交
1238
{
B
Bob Picco 已提交
1239
	return (section && (section->section_mem_map & SECTION_HAS_MEM_MAP));
A
Andy Whitcroft 已提交
1240 1241 1242 1243 1244 1245 1246
}

static inline int valid_section_nr(unsigned long nr)
{
	return valid_section(__nr_to_section(nr));
}

1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263
static inline int online_section(struct mem_section *section)
{
	return (section && (section->section_mem_map & SECTION_IS_ONLINE));
}

static inline int online_section_nr(unsigned long nr)
{
	return online_section(__nr_to_section(nr));
}

#ifdef CONFIG_MEMORY_HOTPLUG
void online_mem_sections(unsigned long start_pfn, unsigned long end_pfn);
#ifdef CONFIG_MEMORY_HOTREMOVE
void offline_mem_sections(unsigned long start_pfn, unsigned long end_pfn);
#endif
#endif

A
Andy Whitcroft 已提交
1264 1265
static inline struct mem_section *__pfn_to_section(unsigned long pfn)
{
A
Andy Whitcroft 已提交
1266
	return __nr_to_section(pfn_to_section_nr(pfn));
A
Andy Whitcroft 已提交
1267 1268
}

1269 1270
extern int __highest_present_section_nr;

1271
#ifndef CONFIG_HAVE_ARCH_PFN_VALID
A
Andy Whitcroft 已提交
1272 1273 1274 1275
static inline int pfn_valid(unsigned long pfn)
{
	if (pfn_to_section_nr(pfn) >= NR_MEM_SECTIONS)
		return 0;
A
Andy Whitcroft 已提交
1276
	return valid_section(__nr_to_section(pfn_to_section_nr(pfn)));
A
Andy Whitcroft 已提交
1277
}
1278
#endif
A
Andy Whitcroft 已提交
1279

1280 1281 1282 1283 1284 1285 1286
static inline int pfn_present(unsigned long pfn)
{
	if (pfn_to_section_nr(pfn) >= NR_MEM_SECTIONS)
		return 0;
	return present_section(__nr_to_section(pfn_to_section_nr(pfn)));
}

A
Andy Whitcroft 已提交
1287 1288 1289 1290 1291 1292
/*
 * These are _only_ used during initialisation, therefore they
 * can use __initdata ...  They could have names to indicate
 * this restriction.
 */
#ifdef CONFIG_NUMA
1293 1294 1295 1296 1297
#define pfn_to_nid(pfn)							\
({									\
	unsigned long __pfn_to_nid_pfn = (pfn);				\
	page_to_nid(pfn_to_page(__pfn_to_nid_pfn));			\
})
1298 1299
#else
#define pfn_to_nid(pfn)		(0)
A
Andy Whitcroft 已提交
1300 1301 1302 1303 1304 1305
#endif

#define early_pfn_valid(pfn)	pfn_valid(pfn)
void sparse_init(void);
#else
#define sparse_init()	do {} while (0)
1306
#define sparse_index_init(_sec, _nid)  do {} while (0)
A
Andy Whitcroft 已提交
1307 1308
#endif /* CONFIG_SPARSEMEM */

1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319
/*
 * During memory init memblocks map pfns to nids. The search is expensive and
 * this caches recent lookups. The implementation of __early_pfn_to_nid
 * may treat start/end as pfns or sections.
 */
struct mminit_pfnnid_cache {
	unsigned long last_start;
	unsigned long last_end;
	int last_nid;
};

A
Andy Whitcroft 已提交
1320 1321 1322 1323 1324 1325
#ifndef early_pfn_valid
#define early_pfn_valid(pfn)	(1)
#endif

void memory_present(int nid, unsigned long start, unsigned long end);

1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337
/*
 * If it is possible to have holes within a MAX_ORDER_NR_PAGES, then we
 * need to check pfn validility within that MAX_ORDER_NR_PAGES block.
 * pfn_valid_within() should be used in this case; we optimise this away
 * when we have no holes within a MAX_ORDER_NR_PAGES block.
 */
#ifdef CONFIG_HOLES_IN_ZONE
#define pfn_valid_within(pfn) pfn_valid(pfn)
#else
#define pfn_valid_within(pfn) (1)
#endif

1338 1339 1340
#ifdef CONFIG_ARCH_HAS_HOLES_MEMORYMODEL
/*
 * pfn_valid() is meant to be able to tell if a given PFN has valid memmap
1341 1342 1343 1344 1345 1346 1347 1348 1349
 * associated with it or not. This means that a struct page exists for this
 * pfn. The caller cannot assume the page is fully initialized in general.
 * Hotplugable pages might not have been onlined yet. pfn_to_online_page()
 * will ensure the struct page is fully online and initialized. Special pages
 * (e.g. ZONE_DEVICE) are never onlined and should be treated accordingly.
 *
 * In FLATMEM, it is expected that holes always have valid memmap as long as
 * there is valid PFNs either side of the hole. In SPARSEMEM, it is assumed
 * that a valid section has a memmap for the entire section.
1350 1351 1352 1353 1354 1355 1356 1357 1358
 *
 * However, an ARM, and maybe other embedded architectures in the future
 * free memmap backing holes to save memory on the assumption the memmap is
 * never used. The page_zone linkages are then broken even though pfn_valid()
 * returns true. A walker of the full memmap must then do this additional
 * check to ensure the memmap they are looking at is sane by making sure
 * the zone and PFN linkages are still valid. This is expensive, but walkers
 * of the full memmap are extremely rare.
 */
1359
bool memmap_valid_within(unsigned long pfn,
1360 1361
					struct page *page, struct zone *zone);
#else
1362
static inline bool memmap_valid_within(unsigned long pfn,
1363 1364
					struct page *page, struct zone *zone)
{
1365
	return true;
1366 1367 1368
}
#endif /* CONFIG_ARCH_HAS_HOLES_MEMORYMODEL */

C
Christoph Lameter 已提交
1369
#endif /* !__GENERATING_BOUNDS.H */
L
Linus Torvalds 已提交
1370 1371
#endif /* !__ASSEMBLY__ */
#endif /* _LINUX_MMZONE_H */