swap.h 14.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7
#ifndef _LINUX_SWAP_H
#define _LINUX_SWAP_H

#include <linux/spinlock.h>
#include <linux/linkage.h>
#include <linux/mmzone.h>
#include <linux/list.h>
8
#include <linux/memcontrol.h>
L
Linus Torvalds 已提交
9
#include <linux/sched.h>
10
#include <linux/node.h>
A
Andrew Morton 已提交
11

A
Arun Sharma 已提交
12
#include <linux/atomic.h>
L
Linus Torvalds 已提交
13 14
#include <asm/page.h>

15 16
struct notifier_block;

A
Andrew Morton 已提交
17 18
struct bio;

L
Linus Torvalds 已提交
19 20 21
#define SWAP_FLAG_PREFER	0x8000	/* set if swap priority specified */
#define SWAP_FLAG_PRIO_MASK	0x7fff
#define SWAP_FLAG_PRIO_SHIFT	0
22
#define SWAP_FLAG_DISCARD	0x10000 /* discard swap cluster after use */
L
Linus Torvalds 已提交
23

24 25 26
#define SWAP_FLAGS_VALID	(SWAP_FLAG_PRIO_MASK | SWAP_FLAG_PREFER | \
				 SWAP_FLAG_DISCARD)

L
Linus Torvalds 已提交
27 28 29 30 31 32 33 34 35 36 37 38 39 40
static inline int current_is_kswapd(void)
{
	return current->flags & PF_KSWAPD;
}

/*
 * MAX_SWAPFILES defines the maximum number of swaptypes: things which can
 * be swapped to.  The swap type and the offset into that swap type are
 * encoded into pte's and into pgoff_t's in the swapcache.  Using five bits
 * for the type means that the maximum number of swapcache pages is 27 bits
 * on 32-bit-pgoff_t architectures.  And that assumes that the architecture packs
 * the type/offset into the pte as 5/27 as well.
 */
#define MAX_SWAPFILES_SHIFT	5
41 42 43 44 45 46 47 48 49 50 51 52 53 54

/*
 * Use some of the swap files numbers for other purposes. This
 * is a convenient way to hook into the VM to trigger special
 * actions on faults.
 */

/*
 * NUMA node memory migration support
 */
#ifdef CONFIG_MIGRATION
#define SWP_MIGRATION_NUM 2
#define SWP_MIGRATION_READ	(MAX_SWAPFILES + SWP_HWPOISON_NUM)
#define SWP_MIGRATION_WRITE	(MAX_SWAPFILES + SWP_HWPOISON_NUM + 1)
55
#else
56
#define SWP_MIGRATION_NUM 0
57
#endif
L
Linus Torvalds 已提交
58

59 60 61 62 63 64 65 66 67 68 69 70 71
/*
 * Handling of hardware poisoned pages with memory corruption.
 */
#ifdef CONFIG_MEMORY_FAILURE
#define SWP_HWPOISON_NUM 1
#define SWP_HWPOISON		MAX_SWAPFILES
#else
#define SWP_HWPOISON_NUM 0
#endif

#define MAX_SWAPFILES \
	((1 << MAX_SWAPFILES_SHIFT) - SWP_MIGRATION_NUM - SWP_HWPOISON_NUM)

L
Linus Torvalds 已提交
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89
/*
 * Magic header for a swap area. The first part of the union is
 * what the swap magic looks like for the old (limited to 128MB)
 * swap area format, the second part of the union adds - in the
 * old reserved area - some extra information. Note that the first
 * kilobyte is reserved for boot loader or disk label stuff...
 *
 * Having the magic at the end of the PAGE_SIZE makes detecting swap
 * areas somewhat tricky on machines that support multiple page sizes.
 * For 2.5 we'll probably want to move the magic to just beyond the
 * bootbits...
 */
union swap_header {
	struct {
		char reserved[PAGE_SIZE - 10];
		char magic[10];			/* SWAP-SPACE or SWAPSPACE2 */
	} magic;
	struct {
90 91 92 93 94 95 96 97
		char		bootbits[1024];	/* Space for disklabel etc. */
		__u32		version;
		__u32		last_page;
		__u32		nr_badpages;
		unsigned char	sws_uuid[16];
		unsigned char	sws_volume[16];
		__u32		padding[117];
		__u32		badpages[1];
L
Linus Torvalds 已提交
98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
	} info;
};

 /* A swap entry has to fit into a "unsigned long", as
  * the entry is hidden in the "index" field of the
  * swapper address space.
  */
typedef struct {
	unsigned long val;
} swp_entry_t;

/*
 * current->reclaim_state points to one of these when a task is running
 * memory reclaim
 */
struct reclaim_state {
	unsigned long reclaimed_slab;
};

#ifdef __KERNEL__

struct address_space;
struct sysinfo;
struct writeback_control;
struct zone;

/*
 * A swap extent maps a range of a swapfile's PAGE_SIZE pages onto a range of
 * disk blocks.  A list of swap extents maps the entire swapfile.  (Where the
 * term `swapfile' refers to either a blockdevice or an IS_REG file.  Apart
 * from setup, they're handled identically.
 *
 * We always assume that blocks are of size PAGE_SIZE.
 */
struct swap_extent {
	struct list_head list;
	pgoff_t start_page;
	pgoff_t nr_pages;
	sector_t start_block;
};

/*
 * Max bad pages in the new format..
 */
#define __swapoffset(x) ((unsigned long)&((union swap_header *)0)->x)
#define MAX_SWAP_BADPAGES \
	((__swapoffset(magic.magic) - __swapoffset(info.badpages)) / sizeof(int))

enum {
	SWP_USED	= (1 << 0),	/* is slot in swap_info[] used? */
	SWP_WRITEOK	= (1 << 1),	/* ok to write to this swap?	*/
149
	SWP_DISCARDABLE = (1 << 2),	/* swapon+blkdev support discard */
150
	SWP_DISCARDING	= (1 << 3),	/* now discarding a free cluster */
151
	SWP_SOLIDSTATE	= (1 << 4),	/* blkdev seeks are cheap */
H
Hugh Dickins 已提交
152
	SWP_CONTINUED	= (1 << 5),	/* swap_map has count continuation */
153
	SWP_BLKDEV	= (1 << 6),	/* its a block device */
154
	SWP_FILE	= (1 << 7),	/* set after swap_activate success */
155 156
					/* add others here before... */
	SWP_SCANNING	= (1 << 8),	/* refcount in scan_swap_map */
L
Linus Torvalds 已提交
157 158 159
};

#define SWAP_CLUSTER_MAX 32
160
#define COMPACT_CLUSTER_MAX SWAP_CLUSTER_MAX
L
Linus Torvalds 已提交
161

162 163 164 165 166 167 168 169 170
/*
 * Ratio between the present memory in the zone and the "gap" that
 * we're allowing kswapd to shrink in addition to the per-zone high
 * wmark, even for zones that already have the high wmark satisfied,
 * in order to provide better per-zone lru behavior. We are ok to
 * spend not more than 1% of the memory for this zone balancing "gap".
 */
#define KSWAPD_ZONE_BALANCE_GAP_RATIO 100

H
Hugh Dickins 已提交
171 172 173 174 175
#define SWAP_MAP_MAX	0x3e	/* Max duplication count, in first swap_map */
#define SWAP_MAP_BAD	0x3f	/* Note pageblock is bad, in first swap_map */
#define SWAP_HAS_CACHE	0x40	/* Flag page is cached, in first swap_map */
#define SWAP_CONT_MAX	0x7f	/* Max count, in each swap_map continuation */
#define COUNT_CONTINUED	0x80	/* See swap_map continuation for full count */
H
Hugh Dickins 已提交
176
#define SWAP_MAP_SHMEM	0xbf	/* Owned by shmem/tmpfs, in first swap_map */
H
Hugh Dickins 已提交
177

L
Linus Torvalds 已提交
178 179 180 181
/*
 * The in-memory structure used to track swap areas.
 */
struct swap_info_struct {
182 183 184 185
	unsigned long	flags;		/* SWP_USED etc: see above */
	signed short	prio;		/* swap priority of this type */
	signed char	type;		/* strange name for an index */
	signed char	next;		/* next type on the swap list */
H
Hugh Dickins 已提交
186 187 188 189 190 191 192 193
	unsigned int	max;		/* extent of the swap_map */
	unsigned char *swap_map;	/* vmalloc'ed array of usage counts */
	unsigned int lowest_bit;	/* index of first free in swap_map */
	unsigned int highest_bit;	/* index of last free in swap_map */
	unsigned int pages;		/* total of usable pages of swap */
	unsigned int inuse_pages;	/* number of those currently in use */
	unsigned int cluster_next;	/* likely index for next allocation */
	unsigned int cluster_nr;	/* countdown to next cluster search */
194 195
	unsigned int lowest_alloc;	/* while preparing discard cluster */
	unsigned int highest_alloc;	/* while preparing discard cluster */
H
Hugh Dickins 已提交
196 197 198 199 200
	struct swap_extent *curr_swap_extent;
	struct swap_extent first_swap_extent;
	struct block_device *bdev;	/* swap device or bdev of swap file */
	struct file *swap_file;		/* seldom referenced */
	unsigned int old_block_size;	/* seldom referenced */
201 202 203 204
#ifdef CONFIG_FRONTSWAP
	unsigned long *frontswap_map;	/* frontswap in-use, one bit per page */
	atomic_t frontswap_pages;	/* frontswap pages in-use counter */
#endif
L
Linus Torvalds 已提交
205 206 207 208 209 210 211 212 213 214 215 216
};

struct swap_list_t {
	int head;	/* head of priority-ordered swapfile list */
	int next;	/* swapfile to be used next */
};

/* Swap 50% full? Release swapcache more aggressively.. */
#define vm_swap_full() (nr_swap_pages*2 < total_swap_pages)

/* linux/mm/page_alloc.c */
extern unsigned long totalram_pages;
217
extern unsigned long totalreserve_pages;
218
extern unsigned long dirty_balance_reserve;
L
Linus Torvalds 已提交
219 220 221
extern unsigned int nr_free_buffer_pages(void);
extern unsigned int nr_free_pagecache_pages(void);

C
Christoph Lameter 已提交
222 223 224 225
/* Definition of global_page_state not available yet */
#define nr_free_pages() global_page_state(NR_FREE_PAGES)


L
Linus Torvalds 已提交
226
/* linux/mm/swap.c */
227 228
extern void __lru_cache_add(struct page *, enum lru_list lru);
extern void lru_cache_add_lru(struct page *, enum lru_list lru);
229 230
extern void lru_add_page_tail(struct page *page, struct page *page_tail,
			      struct lruvec *lruvec);
231 232
extern void activate_page(struct page *);
extern void mark_page_accessed(struct page *);
L
Linus Torvalds 已提交
233
extern void lru_add_drain(void);
234
extern void lru_add_drain_cpu(int cpu);
235
extern int lru_add_drain_all(void);
236
extern void rotate_reclaimable_page(struct page *page);
M
Minchan Kim 已提交
237
extern void deactivate_page(struct page *page);
L
Linus Torvalds 已提交
238 239
extern void swap_setup(void);

L
Lee Schermerhorn 已提交
240 241
extern void add_page_to_unevictable_list(struct page *page);

242 243 244 245
/**
 * lru_cache_add: add a page to the page lists
 * @page: the page to add
 */
246
static inline void lru_cache_add_anon(struct page *page)
247
{
248
	__lru_cache_add(page, LRU_INACTIVE_ANON);
249 250
}

251 252 253 254 255
static inline void lru_cache_add_file(struct page *page)
{
	__lru_cache_add(page, LRU_INACTIVE_FILE);
}

L
Linus Torvalds 已提交
256
/* linux/mm/vmscan.c */
257
extern unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
258
					gfp_t gfp_mask, nodemask_t *mask);
259
extern int __isolate_lru_page(struct page *page, isolate_mode_t mode);
260 261 262 263 264 265
extern unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem,
						  gfp_t gfp_mask, bool noswap);
extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
						gfp_t gfp_mask, bool noswap,
						struct zone *zone,
						unsigned long *nr_scanned);
266
extern unsigned long shrink_all_memory(unsigned long nr_pages);
L
Linus Torvalds 已提交
267
extern int vm_swappiness;
C
Christoph Lameter 已提交
268
extern int remove_mapping(struct address_space *mapping, struct page *page);
269
extern long vm_total_pages;
C
Christoph Lameter 已提交
270

271 272
#ifdef CONFIG_NUMA
extern int zone_reclaim_mode;
273
extern int sysctl_min_unmapped_ratio;
274
extern int sysctl_min_slab_ratio;
275 276 277 278 279 280 281 282 283
extern int zone_reclaim(struct zone *, gfp_t, unsigned int);
#else
#define zone_reclaim_mode 0
static inline int zone_reclaim(struct zone *z, gfp_t mask, unsigned int order)
{
	return 0;
}
#endif

L
Lee Schermerhorn 已提交
284
extern int page_evictable(struct page *page, struct vm_area_struct *vma);
285
extern void check_move_unevictable_pages(struct page **, int nr_pages);
286 287

extern unsigned long scan_unevictable_pages;
288
extern int scan_unevictable_handler(struct ctl_table *, int,
289
					void __user *, size_t *, loff_t *);
290
#ifdef CONFIG_NUMA
291 292
extern int scan_unevictable_register_node(struct node *node);
extern void scan_unevictable_unregister_node(struct node *node);
293 294 295 296 297 298 299 300 301
#else
static inline int scan_unevictable_register_node(struct node *node)
{
	return 0;
}
static inline void scan_unevictable_unregister_node(struct node *node)
{
}
#endif
L
Lee Schermerhorn 已提交
302

303
extern int kswapd_run(int nid);
304
extern void kswapd_stop(int nid);
A
Andrew Morton 已提交
305
#ifdef CONFIG_MEMCG
306 307 308 309 310 311 312
extern int mem_cgroup_swappiness(struct mem_cgroup *mem);
#else
static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
{
	return vm_swappiness;
}
#endif
A
Andrew Morton 已提交
313
#ifdef CONFIG_MEMCG_SWAP
314 315 316 317 318 319
extern void mem_cgroup_uncharge_swap(swp_entry_t ent);
#else
static inline void mem_cgroup_uncharge_swap(swp_entry_t ent)
{
}
#endif
L
Linus Torvalds 已提交
320 321
#ifdef CONFIG_SWAP
/* linux/mm/page_io.c */
322
extern int swap_readpage(struct page *);
L
Linus Torvalds 已提交
323
extern int swap_writepage(struct page *page, struct writeback_control *wbc);
324
extern int swap_set_page_dirty(struct page *page);
325
extern void end_swap_bio_read(struct bio *bio, int err);
L
Linus Torvalds 已提交
326

327 328 329 330 331
int add_swap_extent(struct swap_info_struct *sis, unsigned long start_page,
		unsigned long nr_pages, sector_t start_block);
int generic_swapfile_activate(struct swap_info_struct *, struct file *,
		sector_t *);

L
Linus Torvalds 已提交
332 333 334 335
/* linux/mm/swap_state.c */
extern struct address_space swapper_space;
#define total_swapcache_pages  swapper_space.nrpages
extern void show_swap_cache_info(void);
336
extern int add_to_swap(struct page *);
337
extern int add_to_swap_cache(struct page *, swp_entry_t, gfp_t);
L
Linus Torvalds 已提交
338 339 340 341
extern void __delete_from_swap_cache(struct page *);
extern void delete_from_swap_cache(struct page *);
extern void free_page_and_swap_cache(struct page *);
extern void free_pages_and_swap_cache(struct page **, int);
342
extern struct page *lookup_swap_cache(swp_entry_t);
343
extern struct page *read_swap_cache_async(swp_entry_t, gfp_t,
344
			struct vm_area_struct *vma, unsigned long addr);
345
extern struct page *swapin_readahead(swp_entry_t, gfp_t,
346 347
			struct vm_area_struct *vma, unsigned long addr);

L
Linus Torvalds 已提交
348
/* linux/mm/swapfile.c */
349
extern long nr_swap_pages;
L
Linus Torvalds 已提交
350 351 352
extern long total_swap_pages;
extern void si_swapinfo(struct sysinfo *);
extern swp_entry_t get_swap_page(void);
353
extern swp_entry_t get_swap_page_of_type(int);
H
Hugh Dickins 已提交
354
extern int add_swap_count_continuation(swp_entry_t, gfp_t);
H
Hugh Dickins 已提交
355
extern void swap_shmem_alloc(swp_entry_t);
H
Hugh Dickins 已提交
356 357
extern int swap_duplicate(swp_entry_t);
extern int swapcache_prepare(swp_entry_t);
L
Linus Torvalds 已提交
358
extern void swap_free(swp_entry_t);
359
extern void swapcache_free(swp_entry_t, struct page *page);
360
extern int free_swap_and_cache(swp_entry_t);
361
extern int swap_type_of(dev_t, sector_t, struct block_device **);
362
extern unsigned int count_swap_pages(int, int);
363
extern sector_t map_swap_page(struct page *, struct block_device **);
364
extern sector_t swapdev_block(int, pgoff_t);
365
extern int page_swapcount(struct page *);
366
extern struct swap_info_struct *page_swap_info(struct page *);
367
extern int reuse_swap_page(struct page *);
368
extern int try_to_free_swap(struct page *);
L
Linus Torvalds 已提交
369 370
struct backing_dev_info;

A
Andrew Morton 已提交
371
#ifdef CONFIG_MEMCG
K
KAMEZAWA Hiroyuki 已提交
372 373
extern void
mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout);
K
KAMEZAWA Hiroyuki 已提交
374
#else
375
static inline void
K
KAMEZAWA Hiroyuki 已提交
376
mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
377 378 379
{
}
#endif
K
KAMEZAWA Hiroyuki 已提交
380

L
Linus Torvalds 已提交
381 382
#else /* CONFIG_SWAP */

383 384
#define nr_swap_pages				0L
#define total_swap_pages			0L
L
Linus Torvalds 已提交
385 386 387 388
#define total_swapcache_pages			0UL

#define si_swapinfo(val) \
	do { (val)->freeswap = (val)->totalswap = 0; } while (0)
389 390
/* only sparc can not include linux/pagemap.h in this file
 * so leave page_cache_release and release_pages undeclared... */
L
Linus Torvalds 已提交
391 392 393 394 395
#define free_page_and_swap_cache(page) \
	page_cache_release(page)
#define free_pages_and_swap_cache(pages, nr) \
	release_pages((pages), (nr), 0);

396 397 398 399
static inline void show_swap_cache_info(void)
{
}

400
#define free_swap_and_cache(swp)	is_migration_entry(swp)
401
#define swapcache_prepare(swp)		is_migration_entry(swp)
402

H
Hugh Dickins 已提交
403
static inline int add_swap_count_continuation(swp_entry_t swp, gfp_t gfp_mask)
404
{
H
Hugh Dickins 已提交
405 406 407
	return 0;
}

H
Hugh Dickins 已提交
408 409 410 411
static inline void swap_shmem_alloc(swp_entry_t swp)
{
}

H
Hugh Dickins 已提交
412 413 414
static inline int swap_duplicate(swp_entry_t swp)
{
	return 0;
415 416
}

417 418 419 420
static inline void swap_free(swp_entry_t swp)
{
}

421 422 423 424
static inline void swapcache_free(swp_entry_t swp, struct page *page)
{
}

425
static inline struct page *swapin_readahead(swp_entry_t swp, gfp_t gfp_mask,
426 427 428 429 430
			struct vm_area_struct *vma, unsigned long addr)
{
	return NULL;
}

H
Hugh Dickins 已提交
431 432 433 434 435
static inline int swap_writepage(struct page *p, struct writeback_control *wbc)
{
	return 0;
}

436 437 438 439 440
static inline struct page *lookup_swap_cache(swp_entry_t swp)
{
	return NULL;
}

H
Hugh Dickins 已提交
441 442 443 444 445
static inline int add_to_swap(struct page *page)
{
	return 0;
}

446 447
static inline int add_to_swap_cache(struct page *page, swp_entry_t entry,
							gfp_t gfp_mask)
448
{
449
	return -1;
450 451 452 453 454 455 456 457 458 459
}

static inline void __delete_from_swap_cache(struct page *page)
{
}

static inline void delete_from_swap_cache(struct page *page)
{
}

460 461 462 463 464
static inline int page_swapcount(struct page *page)
{
	return 0;
}

465
#define reuse_swap_page(page)	(page_mapcount(page) == 1)
L
Linus Torvalds 已提交
466

467
static inline int try_to_free_swap(struct page *page)
468 469 470 471
{
	return 0;
}

L
Linus Torvalds 已提交
472 473 474 475 476 477 478
static inline swp_entry_t get_swap_page(void)
{
	swp_entry_t entry;
	entry.val = 0;
	return entry;
}

479 480 481 482 483
static inline void
mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent)
{
}

L
Linus Torvalds 已提交
484 485 486
#endif /* CONFIG_SWAP */
#endif /* __KERNEL__*/
#endif /* _LINUX_SWAP_H */