Blame view
mm/swap.c
28.1 KB
1da177e4c
|
1 2 3 4 5 6 7 |
/* * linux/mm/swap.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds */ /* |
183ff22bb
|
8 |
* This file contains the default values for the operation of the |
1da177e4c
|
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 |
* Linux VM subsystem. Fine-tuning documentation can be found in * Documentation/sysctl/vm.txt. * Started 18.12.91 * Swap aging added 23.2.95, Stephen Tweedie. * Buffermem limits added 12.3.98, Rik van Riel. */ #include <linux/mm.h> #include <linux/sched.h> #include <linux/kernel_stat.h> #include <linux/swap.h> #include <linux/mman.h> #include <linux/pagemap.h> #include <linux/pagevec.h> #include <linux/init.h> |
b95f1b31b
|
24 |
#include <linux/export.h> |
1da177e4c
|
25 |
#include <linux/mm_inline.h> |
1da177e4c
|
26 |
#include <linux/percpu_counter.h> |
3565fce3a
|
27 |
#include <linux/memremap.h> |
1da177e4c
|
28 29 30 |
#include <linux/percpu.h> #include <linux/cpu.h> #include <linux/notifier.h> |
e0bf68dde
|
31 |
#include <linux/backing-dev.h> |
66e1707bc
|
32 |
#include <linux/memcontrol.h> |
5a0e3ad6a
|
33 |
#include <linux/gfp.h> |
a27bb332c
|
34 |
#include <linux/uio.h> |
822fc6136
|
35 |
#include <linux/hugetlb.h> |
33c3fc71c
|
36 |
#include <linux/page_idle.h> |
1da177e4c
|
37 |
|
64d6519dd
|
38 |
#include "internal.h" |
c6286c983
|
39 40 |
#define CREATE_TRACE_POINTS #include <trace/events/pagemap.h> |
1da177e4c
|
41 42 |
/* How many pages do we try to swap or page in/out together? */ int page_cluster; |
13f7f7898
|
43 |
static DEFINE_PER_CPU(struct pagevec, lru_add_pvec); |
f84f9504b
|
44 |
static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs); |
cc5993bd7
|
45 |
static DEFINE_PER_CPU(struct pagevec, lru_deactivate_file_pvecs); |
f7ad2a6cb
|
46 |
static DEFINE_PER_CPU(struct pagevec, lru_lazyfree_pvecs); |
a4a921aa5
|
47 48 49 |
#ifdef CONFIG_SMP static DEFINE_PER_CPU(struct pagevec, activate_page_pvecs); #endif |
902aaed0d
|
50 |
|
b221385bc
|
51 52 53 54 |
/* * This path almost never happens for VM activity - pages are normally * freed via pagevecs. But it gets used by networking. */ |
920c7a5d0
|
55 |
static void __page_cache_release(struct page *page) |
b221385bc
|
56 57 |
{ if (PageLRU(page)) { |
b221385bc
|
58 |
struct zone *zone = page_zone(page); |
fa9add641
|
59 60 |
struct lruvec *lruvec; unsigned long flags; |
b221385bc
|
61 |
|
a52633d8e
|
62 |
spin_lock_irqsave(zone_lru_lock(zone), flags); |
599d0c954
|
63 |
lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat); |
309381fea
|
64 |
VM_BUG_ON_PAGE(!PageLRU(page), page); |
b221385bc
|
65 |
__ClearPageLRU(page); |
fa9add641
|
66 |
del_page_from_lru_list(page, lruvec, page_off_lru(page)); |
a52633d8e
|
67 |
spin_unlock_irqrestore(zone_lru_lock(zone), flags); |
b221385bc
|
68 |
} |
629060270
|
69 |
__ClearPageWaiters(page); |
0a31bc97c
|
70 |
mem_cgroup_uncharge(page); |
918070634
|
71 72 73 74 75 |
} static void __put_single_page(struct page *page) { __page_cache_release(page); |
2d4894b5d
|
76 |
free_unref_page(page); |
b221385bc
|
77 |
} |
918070634
|
78 |
static void __put_compound_page(struct page *page) |
1da177e4c
|
79 |
{ |
918070634
|
80 |
compound_page_dtor *dtor; |
1da177e4c
|
81 |
|
822fc6136
|
82 83 84 85 86 87 88 89 |
/* * __page_cache_release() is supposed to be called for thp, not for * hugetlb. This is because hugetlb page does never have PageLRU set * (it's never listed to any LRU lists) and no memcg routines should * be called for hugetlb (it has a separate hugetlb_cgroup.) */ if (!PageHuge(page)) __page_cache_release(page); |
918070634
|
90 91 92 |
dtor = get_compound_page_dtor(page); (*dtor)(page); } |
ddc58f27f
|
93 |
void __put_page(struct page *page) |
8519fb30e
|
94 |
{ |
713897038
|
95 96 97 98 99 100 101 102 103 |
if (is_zone_device_page(page)) { put_dev_pagemap(page->pgmap); /* * The page belongs to the device that created pgmap. Do * not return it to page allocator. */ return; } |
8519fb30e
|
104 |
if (unlikely(PageCompound(page))) |
ddc58f27f
|
105 106 |
__put_compound_page(page); else |
918070634
|
107 |
__put_single_page(page); |
1da177e4c
|
108 |
} |
ddc58f27f
|
109 |
EXPORT_SYMBOL(__put_page); |
70b50f94f
|
110 |
|
1d7ea7324
|
111 |
/** |
7682486b3
|
112 113 |
* put_pages_list() - release a list of pages * @pages: list of pages threaded on page->lru |
1d7ea7324
|
114 115 116 |
* * Release a list of pages which are strung together on page.lru. Currently * used by read_cache_pages() and related error recovery code. |
1d7ea7324
|
117 118 119 120 121 122 123 124 |
*/ void put_pages_list(struct list_head *pages) { while (!list_empty(pages)) { struct page *victim; victim = list_entry(pages->prev, struct page, lru); list_del(&victim->lru); |
09cbfeaf1
|
125 |
put_page(victim); |
1d7ea7324
|
126 127 128 |
} } EXPORT_SYMBOL(put_pages_list); |
18022c5d8
|
129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 |
/* * get_kernel_pages() - pin kernel pages in memory * @kiov: An array of struct kvec structures * @nr_segs: number of segments to pin * @write: pinning for read/write, currently ignored * @pages: array that receives pointers to the pages pinned. * Should be at least nr_segs long. * * Returns number of pages pinned. This may be fewer than the number * requested. If nr_pages is 0 or negative, returns 0. If no pages * were pinned, returns -errno. Each page returned must be released * with a put_page() call when it is finished with. */ int get_kernel_pages(const struct kvec *kiov, int nr_segs, int write, struct page **pages) { int seg; for (seg = 0; seg < nr_segs; seg++) { if (WARN_ON(kiov[seg].iov_len != PAGE_SIZE)) return seg; |
5a178119b
|
150 |
pages[seg] = kmap_to_page(kiov[seg].iov_base); |
09cbfeaf1
|
151 |
get_page(pages[seg]); |
18022c5d8
|
152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 |
} return seg; } EXPORT_SYMBOL_GPL(get_kernel_pages); /* * get_kernel_page() - pin a kernel page in memory * @start: starting kernel address * @write: pinning for read/write, currently ignored * @pages: array that receives pointer to the page pinned. * Must be at least nr_segs long. * * Returns 1 if page is pinned. If the page was not pinned, returns * -errno. The page returned must be released with a put_page() call * when it is finished with. */ int get_kernel_page(unsigned long start, int write, struct page **pages) { const struct kvec kiov = { .iov_base = (void *)start, .iov_len = PAGE_SIZE }; return get_kernel_pages(&kiov, 1, write, pages); } EXPORT_SYMBOL_GPL(get_kernel_page); |
3dd7ae8ec
|
179 |
static void pagevec_lru_move_fn(struct pagevec *pvec, |
fa9add641
|
180 181 |
void (*move_fn)(struct page *page, struct lruvec *lruvec, void *arg), void *arg) |
902aaed0d
|
182 183 |
{ int i; |
68eb0731c
|
184 |
struct pglist_data *pgdat = NULL; |
fa9add641
|
185 |
struct lruvec *lruvec; |
3dd7ae8ec
|
186 |
unsigned long flags = 0; |
902aaed0d
|
187 188 189 |
for (i = 0; i < pagevec_count(pvec); i++) { struct page *page = pvec->pages[i]; |
68eb0731c
|
190 |
struct pglist_data *pagepgdat = page_pgdat(page); |
902aaed0d
|
191 |
|
68eb0731c
|
192 193 194 195 196 |
if (pagepgdat != pgdat) { if (pgdat) spin_unlock_irqrestore(&pgdat->lru_lock, flags); pgdat = pagepgdat; spin_lock_irqsave(&pgdat->lru_lock, flags); |
902aaed0d
|
197 |
} |
3dd7ae8ec
|
198 |
|
68eb0731c
|
199 |
lruvec = mem_cgroup_page_lruvec(page, pgdat); |
fa9add641
|
200 |
(*move_fn)(page, lruvec, arg); |
902aaed0d
|
201 |
} |
68eb0731c
|
202 203 |
if (pgdat) spin_unlock_irqrestore(&pgdat->lru_lock, flags); |
c6f92f9fb
|
204 |
release_pages(pvec->pages, pvec->nr); |
83896fb5e
|
205 |
pagevec_reinit(pvec); |
d8505dee1
|
206 |
} |
fa9add641
|
207 208 |
static void pagevec_move_tail_fn(struct page *page, struct lruvec *lruvec, void *arg) |
3dd7ae8ec
|
209 210 |
{ int *pgmoved = arg; |
3dd7ae8ec
|
211 |
|
c55e8d035
|
212 213 214 215 |
if (PageLRU(page) && !PageUnevictable(page)) { del_page_from_lru_list(page, lruvec, page_lru(page)); ClearPageActive(page); add_page_to_lru_list_tail(page, lruvec, page_lru(page)); |
3dd7ae8ec
|
216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 |
(*pgmoved)++; } } /* * pagevec_move_tail() must be called with IRQ disabled. * Otherwise this may cause nasty races. */ static void pagevec_move_tail(struct pagevec *pvec) { int pgmoved = 0; pagevec_lru_move_fn(pvec, pagevec_move_tail_fn, &pgmoved); __count_vm_events(PGROTATED, pgmoved); } |
902aaed0d
|
231 |
/* |
1da177e4c
|
232 233 |
* Writeback is about to end against a page which has been marked for immediate * reclaim. If it still appears to be reclaimable, move it to the tail of the |
902aaed0d
|
234 |
* inactive list. |
1da177e4c
|
235 |
*/ |
3dd7ae8ec
|
236 |
void rotate_reclaimable_page(struct page *page) |
1da177e4c
|
237 |
{ |
c55e8d035
|
238 |
if (!PageLocked(page) && !PageDirty(page) && |
894bc3104
|
239 |
!PageUnevictable(page) && PageLRU(page)) { |
ac6aadb24
|
240 241 |
struct pagevec *pvec; unsigned long flags; |
09cbfeaf1
|
242 |
get_page(page); |
ac6aadb24
|
243 |
local_irq_save(flags); |
7c8e0181e
|
244 |
pvec = this_cpu_ptr(&lru_rotate_pvecs); |
8f182270d
|
245 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
ac6aadb24
|
246 247 248 |
pagevec_move_tail(pvec); local_irq_restore(flags); } |
1da177e4c
|
249 |
} |
fa9add641
|
250 |
static void update_page_reclaim_stat(struct lruvec *lruvec, |
3e2f41f1f
|
251 252 |
int file, int rotated) { |
fa9add641
|
253 |
struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat; |
3e2f41f1f
|
254 255 256 257 |
reclaim_stat->recent_scanned[file]++; if (rotated) reclaim_stat->recent_rotated[file]++; |
3e2f41f1f
|
258 |
} |
fa9add641
|
259 260 |
static void __activate_page(struct page *page, struct lruvec *lruvec, void *arg) |
1da177e4c
|
261 |
{ |
744ed1442
|
262 |
if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { |
7a608572a
|
263 264 |
int file = page_is_file_cache(page); int lru = page_lru_base_type(page); |
744ed1442
|
265 |
|
fa9add641
|
266 |
del_page_from_lru_list(page, lruvec, lru); |
7a608572a
|
267 268 |
SetPageActive(page); lru += LRU_ACTIVE; |
fa9add641
|
269 |
add_page_to_lru_list(page, lruvec, lru); |
24b7e5819
|
270 |
trace_mm_lru_activate(page); |
4f98a2fee
|
271 |
|
fa9add641
|
272 273 |
__count_vm_event(PGACTIVATE); update_page_reclaim_stat(lruvec, file, 1); |
1da177e4c
|
274 |
} |
eb709b0d0
|
275 276 277 |
} #ifdef CONFIG_SMP |
eb709b0d0
|
278 279 280 281 282 283 284 |
static void activate_page_drain(int cpu) { struct pagevec *pvec = &per_cpu(activate_page_pvecs, cpu); if (pagevec_count(pvec)) pagevec_lru_move_fn(pvec, __activate_page, NULL); } |
5fbc46163
|
285 286 287 288 |
static bool need_activate_page_drain(int cpu) { return pagevec_count(&per_cpu(activate_page_pvecs, cpu)) != 0; } |
eb709b0d0
|
289 290 |
void activate_page(struct page *page) { |
800d8c63b
|
291 |
page = compound_head(page); |
eb709b0d0
|
292 293 |
if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { struct pagevec *pvec = &get_cpu_var(activate_page_pvecs); |
09cbfeaf1
|
294 |
get_page(page); |
8f182270d
|
295 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
eb709b0d0
|
296 297 298 299 300 301 302 303 304 |
pagevec_lru_move_fn(pvec, __activate_page, NULL); put_cpu_var(activate_page_pvecs); } } #else static inline void activate_page_drain(int cpu) { } |
5fbc46163
|
305 306 307 308 |
static bool need_activate_page_drain(int cpu) { return false; } |
eb709b0d0
|
309 310 311 |
void activate_page(struct page *page) { struct zone *zone = page_zone(page); |
800d8c63b
|
312 |
page = compound_head(page); |
a52633d8e
|
313 |
spin_lock_irq(zone_lru_lock(zone)); |
599d0c954
|
314 |
__activate_page(page, mem_cgroup_page_lruvec(page, zone->zone_pgdat), NULL); |
a52633d8e
|
315 |
spin_unlock_irq(zone_lru_lock(zone)); |
1da177e4c
|
316 |
} |
eb709b0d0
|
317 |
#endif |
1da177e4c
|
318 |
|
059285a25
|
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 |
static void __lru_cache_activate_page(struct page *page) { struct pagevec *pvec = &get_cpu_var(lru_add_pvec); int i; /* * Search backwards on the optimistic assumption that the page being * activated has just been added to this pagevec. Note that only * the local pagevec is examined as a !PageLRU page could be in the * process of being released, reclaimed, migrated or on a remote * pagevec that is currently being drained. Furthermore, marking * a remote pagevec's page PageActive potentially hits a race where * a page is marked PageActive just after it is added to the inactive * list causing accounting errors and BUG_ON checks to trigger. */ for (i = pagevec_count(pvec) - 1; i >= 0; i--) { struct page *pagevec_page = pvec->pages[i]; if (pagevec_page == page) { SetPageActive(page); break; } } put_cpu_var(lru_add_pvec); } |
1da177e4c
|
345 346 347 348 349 350 |
/* * Mark a page as having seen activity. * * inactive,unreferenced -> inactive,referenced * inactive,referenced -> active,unreferenced * active,unreferenced -> active,referenced |
eb39d618f
|
351 352 353 |
* * When a newly allocated page is not yet visible, so safe for non-atomic ops, * __SetPageReferenced(page) may be substituted for mark_page_accessed(page). |
1da177e4c
|
354 |
*/ |
920c7a5d0
|
355 |
void mark_page_accessed(struct page *page) |
1da177e4c
|
356 |
{ |
e90309c9f
|
357 |
page = compound_head(page); |
894bc3104
|
358 |
if (!PageActive(page) && !PageUnevictable(page) && |
059285a25
|
359 360 361 362 363 364 365 366 367 368 369 370 |
PageReferenced(page)) { /* * If the page is on the LRU, queue it for activation via * activate_page_pvecs. Otherwise, assume the page is on a * pagevec, mark it active and it'll be moved to the active * LRU on the next drain. */ if (PageLRU(page)) activate_page(page); else __lru_cache_activate_page(page); |
1da177e4c
|
371 |
ClearPageReferenced(page); |
a528910e1
|
372 373 |
if (page_is_file_cache(page)) workingset_activation(page); |
1da177e4c
|
374 375 376 |
} else if (!PageReferenced(page)) { SetPageReferenced(page); } |
33c3fc71c
|
377 378 |
if (page_is_idle(page)) clear_page_idle(page); |
1da177e4c
|
379 |
} |
1da177e4c
|
380 |
EXPORT_SYMBOL(mark_page_accessed); |
2329d3751
|
381 |
static void __lru_cache_add(struct page *page) |
1da177e4c
|
382 |
{ |
13f7f7898
|
383 |
struct pagevec *pvec = &get_cpu_var(lru_add_pvec); |
09cbfeaf1
|
384 |
get_page(page); |
8f182270d
|
385 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
a0b8cab3b
|
386 |
__pagevec_lru_add(pvec); |
13f7f7898
|
387 |
put_cpu_var(lru_add_pvec); |
1da177e4c
|
388 |
} |
2329d3751
|
389 390 391 392 393 394 395 |
/** * lru_cache_add: add a page to the page lists * @page: the page to add */ void lru_cache_add_anon(struct page *page) { |
6fb81a17d
|
396 397 |
if (PageActive(page)) ClearPageActive(page); |
2329d3751
|
398 399 400 401 402 |
__lru_cache_add(page); } void lru_cache_add_file(struct page *page) { |
6fb81a17d
|
403 404 |
if (PageActive(page)) ClearPageActive(page); |
2329d3751
|
405 406 407 |
__lru_cache_add(page); } EXPORT_SYMBOL(lru_cache_add_file); |
1da177e4c
|
408 |
|
f04e9ebbe
|
409 |
/** |
c53954a09
|
410 |
* lru_cache_add - add a page to a page list |
f04e9ebbe
|
411 |
* @page: the page to be added to the LRU. |
2329d3751
|
412 413 414 415 416 |
* * Queue the page for addition to the LRU via pagevec. The decision on whether * to add the page to the [in]active [file|anon] list is deferred until the * pagevec is drained. This gives a chance for the caller of lru_cache_add() * have the page added to the active list using mark_page_accessed(). |
f04e9ebbe
|
417 |
*/ |
c53954a09
|
418 |
void lru_cache_add(struct page *page) |
1da177e4c
|
419 |
{ |
309381fea
|
420 421 |
VM_BUG_ON_PAGE(PageActive(page) && PageUnevictable(page), page); VM_BUG_ON_PAGE(PageLRU(page), page); |
c53954a09
|
422 |
__lru_cache_add(page); |
1da177e4c
|
423 |
} |
894bc3104
|
424 425 426 427 428 429 430 431 432 433 434 435 |
/** * add_page_to_unevictable_list - add a page to the unevictable list * @page: the page to be added to the unevictable list * * Add page directly to its zone's unevictable list. To avoid races with * tasks that might be making the page evictable, through eg. munlock, * munmap or exit, while it's not on the lru, we want to add the page * while it's locked or otherwise "invisible" to other tasks. This is * difficult to do when using the pagevec cache, so bypass that. */ void add_page_to_unevictable_list(struct page *page) { |
599d0c954
|
436 |
struct pglist_data *pgdat = page_pgdat(page); |
fa9add641
|
437 |
struct lruvec *lruvec; |
894bc3104
|
438 |
|
599d0c954
|
439 440 |
spin_lock_irq(&pgdat->lru_lock); lruvec = mem_cgroup_page_lruvec(page, pgdat); |
ef2a2cbdd
|
441 |
ClearPageActive(page); |
894bc3104
|
442 443 |
SetPageUnevictable(page); SetPageLRU(page); |
fa9add641
|
444 |
add_page_to_lru_list(page, lruvec, LRU_UNEVICTABLE); |
599d0c954
|
445 |
spin_unlock_irq(&pgdat->lru_lock); |
894bc3104
|
446 |
} |
00501b531
|
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 |
/** * lru_cache_add_active_or_unevictable * @page: the page to be added to LRU * @vma: vma in which page is mapped for determining reclaimability * * Place @page on the active or unevictable LRU list, depending on its * evictability. Note that if the page is not evictable, it goes * directly back onto it's zone's unevictable list, it does NOT use a * per cpu pagevec. */ void lru_cache_add_active_or_unevictable(struct page *page, struct vm_area_struct *vma) { VM_BUG_ON_PAGE(PageLRU(page), page); if (likely((vma->vm_flags & (VM_LOCKED | VM_SPECIAL)) != VM_LOCKED)) { SetPageActive(page); lru_cache_add(page); return; } if (!TestSetPageMlocked(page)) { /* * We use the irq-unsafe __mod_zone_page_stat because this * counter is not modified from interrupt context, and the pte * lock is held(spinlock), which implies preemption disabled. */ __mod_zone_page_state(page_zone(page), NR_MLOCK, hpage_nr_pages(page)); count_vm_event(UNEVICTABLE_PGMLOCKED); } add_page_to_unevictable_list(page); } |
902aaed0d
|
480 |
/* |
315601809
|
481 482 483 484 485 |
* If the page can not be invalidated, it is moved to the * inactive list to speed up its reclaim. It is moved to the * head of the list, rather than the tail, to give the flusher * threads some time to write it out, as this is much more * effective than the single-page writeout from reclaim. |
278df9f45
|
486 487 488 489 490 491 492 493 494 495 496 497 498 499 |
* * If the page isn't page_mapped and dirty/writeback, the page * could reclaim asap using PG_reclaim. * * 1. active, mapped page -> none * 2. active, dirty/writeback page -> inactive, head, PG_reclaim * 3. inactive, mapped page -> none * 4. inactive, dirty/writeback page -> inactive, head, PG_reclaim * 5. inactive, clean -> inactive, tail * 6. Others -> none * * In 4, why it moves inactive's head, the VM expects the page would * be write it out by flusher threads as this is much more effective * than the single-page writeout from reclaim. |
315601809
|
500 |
*/ |
cc5993bd7
|
501 |
static void lru_deactivate_file_fn(struct page *page, struct lruvec *lruvec, |
fa9add641
|
502 |
void *arg) |
315601809
|
503 504 |
{ int lru, file; |
278df9f45
|
505 |
bool active; |
315601809
|
506 |
|
278df9f45
|
507 |
if (!PageLRU(page)) |
315601809
|
508 |
return; |
bad49d9c8
|
509 510 |
if (PageUnevictable(page)) return; |
315601809
|
511 512 513 |
/* Some processes are using the page */ if (page_mapped(page)) return; |
278df9f45
|
514 |
active = PageActive(page); |
315601809
|
515 516 |
file = page_is_file_cache(page); lru = page_lru_base_type(page); |
fa9add641
|
517 518 |
del_page_from_lru_list(page, lruvec, lru + active); |
315601809
|
519 520 |
ClearPageActive(page); ClearPageReferenced(page); |
fa9add641
|
521 |
add_page_to_lru_list(page, lruvec, lru); |
315601809
|
522 |
|
278df9f45
|
523 524 525 526 527 528 529 530 531 532 533 534 |
if (PageWriteback(page) || PageDirty(page)) { /* * PG_reclaim could be raced with end_page_writeback * It can make readahead confusing. But race window * is _really_ small and it's non-critical problem. */ SetPageReclaim(page); } else { /* * The page's writeback ends up during pagevec * We moves tha page into tail of inactive. */ |
925b7673c
|
535 |
list_move_tail(&page->lru, &lruvec->lists[lru]); |
278df9f45
|
536 537 538 539 540 |
__count_vm_event(PGROTATED); } if (active) __count_vm_event(PGDEACTIVATE); |
fa9add641
|
541 |
update_page_reclaim_stat(lruvec, file, 0); |
315601809
|
542 |
} |
10853a039
|
543 |
|
f7ad2a6cb
|
544 |
static void lru_lazyfree_fn(struct page *page, struct lruvec *lruvec, |
10853a039
|
545 546 |
void *arg) { |
f7ad2a6cb
|
547 |
if (PageLRU(page) && PageAnon(page) && PageSwapBacked(page) && |
24c92eb7d
|
548 |
!PageSwapCache(page) && !PageUnevictable(page)) { |
f7ad2a6cb
|
549 |
bool active = PageActive(page); |
10853a039
|
550 |
|
f7ad2a6cb
|
551 552 |
del_page_from_lru_list(page, lruvec, LRU_INACTIVE_ANON + active); |
10853a039
|
553 554 |
ClearPageActive(page); ClearPageReferenced(page); |
f7ad2a6cb
|
555 556 557 558 559 560 561 |
/* * lazyfree pages are clean anonymous pages. They have * SwapBacked flag cleared to distinguish normal anonymous * pages */ ClearPageSwapBacked(page); add_page_to_lru_list(page, lruvec, LRU_INACTIVE_FILE); |
10853a039
|
562 |
|
f7ad2a6cb
|
563 |
__count_vm_events(PGLAZYFREE, hpage_nr_pages(page)); |
2262185c5
|
564 |
count_memcg_page_event(page, PGLAZYFREE); |
f7ad2a6cb
|
565 |
update_page_reclaim_stat(lruvec, 1, 0); |
10853a039
|
566 567 |
} } |
315601809
|
568 |
/* |
902aaed0d
|
569 570 571 572 |
* Drain pages out of the cpu's pagevecs. * Either "cpu" is the current CPU, and preemption has already been * disabled; or "cpu" is being hot-unplugged, and is already dead. */ |
f0cb3c76a
|
573 |
void lru_add_drain_cpu(int cpu) |
1da177e4c
|
574 |
{ |
13f7f7898
|
575 |
struct pagevec *pvec = &per_cpu(lru_add_pvec, cpu); |
1da177e4c
|
576 |
|
13f7f7898
|
577 |
if (pagevec_count(pvec)) |
a0b8cab3b
|
578 |
__pagevec_lru_add(pvec); |
902aaed0d
|
579 580 581 582 583 584 585 586 587 588 |
pvec = &per_cpu(lru_rotate_pvecs, cpu); if (pagevec_count(pvec)) { unsigned long flags; /* No harm done if a racing interrupt already did this */ local_irq_save(flags); pagevec_move_tail(pvec); local_irq_restore(flags); } |
315601809
|
589 |
|
cc5993bd7
|
590 |
pvec = &per_cpu(lru_deactivate_file_pvecs, cpu); |
315601809
|
591 |
if (pagevec_count(pvec)) |
cc5993bd7
|
592 |
pagevec_lru_move_fn(pvec, lru_deactivate_file_fn, NULL); |
eb709b0d0
|
593 |
|
f7ad2a6cb
|
594 |
pvec = &per_cpu(lru_lazyfree_pvecs, cpu); |
10853a039
|
595 |
if (pagevec_count(pvec)) |
f7ad2a6cb
|
596 |
pagevec_lru_move_fn(pvec, lru_lazyfree_fn, NULL); |
10853a039
|
597 |
|
eb709b0d0
|
598 |
activate_page_drain(cpu); |
315601809
|
599 600 601 |
} /** |
cc5993bd7
|
602 |
* deactivate_file_page - forcefully deactivate a file page |
315601809
|
603 604 605 606 607 608 |
* @page: page to deactivate * * This function hints the VM that @page is a good reclaim candidate, * for example if its invalidation fails due to the page being dirty * or under writeback. */ |
cc5993bd7
|
609 |
void deactivate_file_page(struct page *page) |
315601809
|
610 |
{ |
821ed6bbe
|
611 |
/* |
cc5993bd7
|
612 613 |
* In a workload with many unevictable page such as mprotect, * unevictable page deactivation for accelerating reclaim is pointless. |
821ed6bbe
|
614 615 616 |
*/ if (PageUnevictable(page)) return; |
315601809
|
617 |
if (likely(get_page_unless_zero(page))) { |
cc5993bd7
|
618 |
struct pagevec *pvec = &get_cpu_var(lru_deactivate_file_pvecs); |
315601809
|
619 |
|
8f182270d
|
620 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
cc5993bd7
|
621 622 |
pagevec_lru_move_fn(pvec, lru_deactivate_file_fn, NULL); put_cpu_var(lru_deactivate_file_pvecs); |
315601809
|
623 |
} |
80bfed904
|
624 |
} |
10853a039
|
625 |
/** |
f7ad2a6cb
|
626 |
* mark_page_lazyfree - make an anon page lazyfree |
10853a039
|
627 628 |
* @page: page to deactivate * |
f7ad2a6cb
|
629 630 |
* mark_page_lazyfree() moves @page to the inactive file list. * This is done to accelerate the reclaim of @page. |
10853a039
|
631 |
*/ |
f7ad2a6cb
|
632 |
void mark_page_lazyfree(struct page *page) |
10853a039
|
633 |
{ |
f7ad2a6cb
|
634 |
if (PageLRU(page) && PageAnon(page) && PageSwapBacked(page) && |
24c92eb7d
|
635 |
!PageSwapCache(page) && !PageUnevictable(page)) { |
f7ad2a6cb
|
636 |
struct pagevec *pvec = &get_cpu_var(lru_lazyfree_pvecs); |
10853a039
|
637 |
|
09cbfeaf1
|
638 |
get_page(page); |
8f182270d
|
639 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
f7ad2a6cb
|
640 641 |
pagevec_lru_move_fn(pvec, lru_lazyfree_fn, NULL); put_cpu_var(lru_lazyfree_pvecs); |
10853a039
|
642 643 |
} } |
80bfed904
|
644 645 |
void lru_add_drain(void) { |
f0cb3c76a
|
646 |
lru_add_drain_cpu(get_cpu()); |
80bfed904
|
647 |
put_cpu(); |
1da177e4c
|
648 |
} |
c4028958b
|
649 |
static void lru_add_drain_per_cpu(struct work_struct *dummy) |
053837fce
|
650 651 652 |
{ lru_add_drain(); } |
5fbc46163
|
653 |
static DEFINE_PER_CPU(struct work_struct, lru_add_drain_work); |
a47fed5b5
|
654 |
void lru_add_drain_all_cpuslocked(void) |
053837fce
|
655 |
{ |
5fbc46163
|
656 657 658 |
static DEFINE_MUTEX(lock); static struct cpumask has_work; int cpu; |
ce612879d
|
659 660 661 662 663 664 |
/* * Make sure nobody triggers this path before mm_percpu_wq is fully * initialized. */ if (WARN_ON(!mm_percpu_wq)) return; |
5fbc46163
|
665 |
mutex_lock(&lock); |
5fbc46163
|
666 667 668 669 670 671 672 |
cpumask_clear(&has_work); for_each_online_cpu(cpu) { struct work_struct *work = &per_cpu(lru_add_drain_work, cpu); if (pagevec_count(&per_cpu(lru_add_pvec, cpu)) || pagevec_count(&per_cpu(lru_rotate_pvecs, cpu)) || |
cc5993bd7
|
673 |
pagevec_count(&per_cpu(lru_deactivate_file_pvecs, cpu)) || |
f7ad2a6cb
|
674 |
pagevec_count(&per_cpu(lru_lazyfree_pvecs, cpu)) || |
5fbc46163
|
675 676 |
need_activate_page_drain(cpu)) { INIT_WORK(work, lru_add_drain_per_cpu); |
ce612879d
|
677 |
queue_work_on(cpu, mm_percpu_wq, work); |
5fbc46163
|
678 679 680 681 682 683 |
cpumask_set_cpu(cpu, &has_work); } } for_each_cpu(cpu, &has_work) flush_work(&per_cpu(lru_add_drain_work, cpu)); |
5fbc46163
|
684 |
mutex_unlock(&lock); |
053837fce
|
685 |
} |
a47fed5b5
|
686 687 688 689 690 691 |
void lru_add_drain_all(void) { get_online_cpus(); lru_add_drain_all_cpuslocked(); put_online_cpus(); } |
aabfb5729
|
692 |
/** |
ea1754a08
|
693 |
* release_pages - batched put_page() |
aabfb5729
|
694 695 696 |
* @pages: array of pages to release * @nr: number of pages * @cold: whether the pages are cache cold |
1da177e4c
|
697 |
* |
aabfb5729
|
698 699 |
* Decrement the reference count on all the pages in @pages. If it * fell to zero, remove the page from the LRU and free it. |
1da177e4c
|
700 |
*/ |
c6f92f9fb
|
701 |
void release_pages(struct page **pages, int nr) |
1da177e4c
|
702 703 |
{ int i; |
cc59850ef
|
704 |
LIST_HEAD(pages_to_free); |
599d0c954
|
705 |
struct pglist_data *locked_pgdat = NULL; |
fa9add641
|
706 |
struct lruvec *lruvec; |
902aaed0d
|
707 |
unsigned long uninitialized_var(flags); |
aabfb5729
|
708 |
unsigned int uninitialized_var(lock_batch); |
1da177e4c
|
709 |
|
1da177e4c
|
710 711 |
for (i = 0; i < nr; i++) { struct page *page = pages[i]; |
1da177e4c
|
712 |
|
aabfb5729
|
713 714 715 |
/* * Make sure the IRQ-safe lock-holding time does not get * excessive with a continuous string of pages from the |
599d0c954
|
716 |
* same pgdat. The lock is held only if pgdat != NULL. |
aabfb5729
|
717 |
*/ |
599d0c954
|
718 719 720 |
if (locked_pgdat && ++lock_batch == SWAP_CLUSTER_MAX) { spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); locked_pgdat = NULL; |
aabfb5729
|
721 |
} |
6fcb52a56
|
722 |
if (is_huge_zero_page(page)) |
aa88b68c3
|
723 |
continue; |
aa88b68c3
|
724 |
|
df6ad6983
|
725 726 727 728 729 730 731 732 733 734 |
/* Device public page can not be huge page */ if (is_device_public_page(page)) { if (locked_pgdat) { spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); locked_pgdat = NULL; } put_zone_device_private_or_public_page(page); continue; } |
ddc58f27f
|
735 |
page = compound_head(page); |
b5810039a
|
736 |
if (!put_page_testzero(page)) |
1da177e4c
|
737 |
continue; |
ddc58f27f
|
738 |
if (PageCompound(page)) { |
599d0c954
|
739 740 741 |
if (locked_pgdat) { spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); locked_pgdat = NULL; |
ddc58f27f
|
742 743 744 745 |
} __put_compound_page(page); continue; } |
46453a6e1
|
746 |
if (PageLRU(page)) { |
599d0c954
|
747 |
struct pglist_data *pgdat = page_pgdat(page); |
894bc3104
|
748 |
|
599d0c954
|
749 750 751 |
if (pgdat != locked_pgdat) { if (locked_pgdat) spin_unlock_irqrestore(&locked_pgdat->lru_lock, |
902aaed0d
|
752 |
flags); |
aabfb5729
|
753 |
lock_batch = 0; |
599d0c954
|
754 755 |
locked_pgdat = pgdat; spin_lock_irqsave(&locked_pgdat->lru_lock, flags); |
46453a6e1
|
756 |
} |
fa9add641
|
757 |
|
599d0c954
|
758 |
lruvec = mem_cgroup_page_lruvec(page, locked_pgdat); |
309381fea
|
759 |
VM_BUG_ON_PAGE(!PageLRU(page), page); |
674539115
|
760 |
__ClearPageLRU(page); |
fa9add641
|
761 |
del_page_from_lru_list(page, lruvec, page_off_lru(page)); |
46453a6e1
|
762 |
} |
c53954a09
|
763 |
/* Clear Active bit in case of parallel mark_page_accessed */ |
e3741b506
|
764 |
__ClearPageActive(page); |
629060270
|
765 |
__ClearPageWaiters(page); |
c53954a09
|
766 |
|
cc59850ef
|
767 |
list_add(&page->lru, &pages_to_free); |
1da177e4c
|
768 |
} |
599d0c954
|
769 770 |
if (locked_pgdat) spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); |
1da177e4c
|
771 |
|
747db954c
|
772 |
mem_cgroup_uncharge_list(&pages_to_free); |
2d4894b5d
|
773 |
free_unref_page_list(&pages_to_free); |
1da177e4c
|
774 |
} |
0be8557bc
|
775 |
EXPORT_SYMBOL(release_pages); |
1da177e4c
|
776 777 778 779 780 781 782 783 784 785 786 787 788 |
/* * The pages which we're about to release may be in the deferred lru-addition * queues. That would prevent them from really being freed right now. That's * OK from a correctness point of view but is inefficient - those pages may be * cache-warm and we want to give them back to the page allocator ASAP. * * So __pagevec_release() will drain those queues here. __pagevec_lru_add() * and __pagevec_lru_add_active() call release_pages() directly to avoid * mutual recursion. */ void __pagevec_release(struct pagevec *pvec) { |
7f0b5fb95
|
789 |
if (!pvec->percpu_pvec_drained) { |
d9ed0d08b
|
790 |
lru_add_drain(); |
7f0b5fb95
|
791 |
pvec->percpu_pvec_drained = true; |
d9ed0d08b
|
792 |
} |
c6f92f9fb
|
793 |
release_pages(pvec->pages, pagevec_count(pvec)); |
1da177e4c
|
794 795 |
pagevec_reinit(pvec); } |
7f2857018
|
796 |
EXPORT_SYMBOL(__pagevec_release); |
12d271078
|
797 |
#ifdef CONFIG_TRANSPARENT_HUGEPAGE |
71e3aac07
|
798 |
/* used by __split_huge_page_refcount() */ |
fa9add641
|
799 |
void lru_add_page_tail(struct page *page, struct page *page_tail, |
5bc7b8aca
|
800 |
struct lruvec *lruvec, struct list_head *list) |
71e3aac07
|
801 |
{ |
71e3aac07
|
802 |
const int file = 0; |
71e3aac07
|
803 |
|
309381fea
|
804 805 806 |
VM_BUG_ON_PAGE(!PageHead(page), page); VM_BUG_ON_PAGE(PageCompound(page_tail), page); VM_BUG_ON_PAGE(PageLRU(page_tail), page); |
fa9add641
|
807 |
VM_BUG_ON(NR_CPUS != 1 && |
599d0c954
|
808 |
!spin_is_locked(&lruvec_pgdat(lruvec)->lru_lock)); |
71e3aac07
|
809 |
|
5bc7b8aca
|
810 811 |
if (!list) SetPageLRU(page_tail); |
71e3aac07
|
812 |
|
12d271078
|
813 814 |
if (likely(PageLRU(page))) list_add_tail(&page_tail->lru, &page->lru); |
5bc7b8aca
|
815 816 817 818 819 |
else if (list) { /* page reclaim is reclaiming a huge page */ get_page(page_tail); list_add_tail(&page_tail->lru, list); } else { |
12d271078
|
820 821 822 823 824 825 826 827 |
struct list_head *list_head; /* * Head page has not yet been counted, as an hpage, * so we must account for each subpage individually. * * Use the standard add function to put page_tail on the list, * but then correct its position so they all end up in order. */ |
e180cf806
|
828 |
add_page_to_lru_list(page_tail, lruvec, page_lru(page_tail)); |
12d271078
|
829 830 |
list_head = page_tail->lru.prev; list_move_tail(&page_tail->lru, list_head); |
71e3aac07
|
831 |
} |
7512102cf
|
832 833 |
if (!PageUnevictable(page)) |
e180cf806
|
834 |
update_page_reclaim_stat(lruvec, file, PageActive(page_tail)); |
71e3aac07
|
835 |
} |
12d271078
|
836 |
#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ |
71e3aac07
|
837 |
|
fa9add641
|
838 839 |
static void __pagevec_lru_add_fn(struct page *page, struct lruvec *lruvec, void *arg) |
3dd7ae8ec
|
840 |
{ |
13f7f7898
|
841 842 843 |
int file = page_is_file_cache(page); int active = PageActive(page); enum lru_list lru = page_lru(page); |
3dd7ae8ec
|
844 |
|
309381fea
|
845 |
VM_BUG_ON_PAGE(PageLRU(page), page); |
3dd7ae8ec
|
846 847 |
SetPageLRU(page); |
fa9add641
|
848 849 |
add_page_to_lru_list(page, lruvec, lru); update_page_reclaim_stat(lruvec, file, active); |
24b7e5819
|
850 |
trace_mm_lru_insertion(page, lru); |
3dd7ae8ec
|
851 |
} |
1da177e4c
|
852 |
/* |
1da177e4c
|
853 854 855 |
* Add the passed pages to the LRU, then drop the caller's refcount * on them. Reinitialises the caller's pagevec. */ |
a0b8cab3b
|
856 |
void __pagevec_lru_add(struct pagevec *pvec) |
1da177e4c
|
857 |
{ |
a0b8cab3b
|
858 |
pagevec_lru_move_fn(pvec, __pagevec_lru_add_fn, NULL); |
1da177e4c
|
859 |
} |
5095ae837
|
860 |
EXPORT_SYMBOL(__pagevec_lru_add); |
1da177e4c
|
861 |
|
1da177e4c
|
862 |
/** |
0cd6144aa
|
863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 |
* pagevec_lookup_entries - gang pagecache lookup * @pvec: Where the resulting entries are placed * @mapping: The address_space to search * @start: The starting entry index * @nr_entries: The maximum number of entries * @indices: The cache indices corresponding to the entries in @pvec * * pagevec_lookup_entries() will search for and return a group of up * to @nr_entries pages and shadow entries in the mapping. All * entries are placed in @pvec. pagevec_lookup_entries() takes a * reference against actual pages in @pvec. * * The search returns a group of mapping-contiguous entries with * ascending indexes. There may be holes in the indices due to * not-present entries. * * pagevec_lookup_entries() returns the number of entries which were * found. */ unsigned pagevec_lookup_entries(struct pagevec *pvec, struct address_space *mapping, pgoff_t start, unsigned nr_pages, pgoff_t *indices) { pvec->nr = find_get_entries(mapping, start, nr_pages, pvec->pages, indices); return pagevec_count(pvec); } /** * pagevec_remove_exceptionals - pagevec exceptionals pruning * @pvec: The pagevec to prune * * pagevec_lookup_entries() fills both pages and exceptional radix * tree entries into the pagevec. This function prunes all * exceptionals from @pvec without leaving holes, so that it can be * passed on to page-only pagevec operations. */ void pagevec_remove_exceptionals(struct pagevec *pvec) { int i, j; for (i = 0, j = 0; i < pagevec_count(pvec); i++) { struct page *page = pvec->pages[i]; if (!radix_tree_exceptional_entry(page)) pvec->pages[j++] = page; } pvec->nr = j; } /** |
b947cee4b
|
914 |
* pagevec_lookup_range - gang pagecache lookup |
1da177e4c
|
915 916 917 |
* @pvec: Where the resulting pages are placed * @mapping: The address_space to search * @start: The starting page index |
b947cee4b
|
918 |
* @end: The final page index |
1da177e4c
|
919 920 |
* @nr_pages: The maximum number of pages * |
b947cee4b
|
921 922 923 |
* pagevec_lookup_range() will search for and return a group of up to @nr_pages * pages in the mapping starting from index @start and upto index @end * (inclusive). The pages are placed in @pvec. pagevec_lookup() takes a |
1da177e4c
|
924 925 926 |
* reference against the pages in @pvec. * * The search returns a group of mapping-contiguous pages with ascending |
d72dc8a25
|
927 928 |
* indexes. There may be holes in the indices due to not-present pages. We * also update @start to index the next page for the traversal. |
1da177e4c
|
929 |
* |
b947cee4b
|
930 931 932 |
* pagevec_lookup_range() returns the number of pages which were found. If this * number is smaller than @nr_pages, the end of specified range has been * reached. |
1da177e4c
|
933 |
*/ |
b947cee4b
|
934 |
unsigned pagevec_lookup_range(struct pagevec *pvec, |
397162ffa
|
935 |
struct address_space *mapping, pgoff_t *start, pgoff_t end) |
1da177e4c
|
936 |
{ |
397162ffa
|
937 |
pvec->nr = find_get_pages_range(mapping, start, end, PAGEVEC_SIZE, |
b947cee4b
|
938 |
pvec->pages); |
1da177e4c
|
939 940 |
return pagevec_count(pvec); } |
b947cee4b
|
941 |
EXPORT_SYMBOL(pagevec_lookup_range); |
78539fdfa
|
942 |
|
72b045aec
|
943 944 |
unsigned pagevec_lookup_range_tag(struct pagevec *pvec, struct address_space *mapping, pgoff_t *index, pgoff_t end, |
67fd707f4
|
945 |
int tag) |
1da177e4c
|
946 |
{ |
72b045aec
|
947 |
pvec->nr = find_get_pages_range_tag(mapping, index, end, tag, |
67fd707f4
|
948 |
PAGEVEC_SIZE, pvec->pages); |
1da177e4c
|
949 950 |
return pagevec_count(pvec); } |
72b045aec
|
951 |
EXPORT_SYMBOL(pagevec_lookup_range_tag); |
1da177e4c
|
952 |
|
93d3b7140
|
953 954 955 956 957 958 959 960 961 |
unsigned pagevec_lookup_range_nr_tag(struct pagevec *pvec, struct address_space *mapping, pgoff_t *index, pgoff_t end, int tag, unsigned max_pages) { pvec->nr = find_get_pages_range_tag(mapping, index, end, tag, min_t(unsigned int, max_pages, PAGEVEC_SIZE), pvec->pages); return pagevec_count(pvec); } EXPORT_SYMBOL(pagevec_lookup_range_nr_tag); |
1da177e4c
|
962 963 964 965 966 |
/* * Perform any setup for the swap system */ void __init swap_setup(void) { |
4481374ce
|
967 |
unsigned long megs = totalram_pages >> (20 - PAGE_SHIFT); |
e0bf68dde
|
968 |
|
1da177e4c
|
969 970 971 972 973 974 975 976 977 |
/* Use a smaller cluster for small-memory machines */ if (megs < 16) page_cluster = 2; else page_cluster = 3; /* * Right now other parts of the system means that we * _really_ don't want to cluster much more */ |
1da177e4c
|
978 |
} |