Blame view
mm/swap.c
23.3 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 5 6 7 |
/* * linux/mm/swap.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds */ /* |
183ff22bb spelling fixes: mm/ |
8 |
* This file contains the default values for the operation of the |
1da177e4c Linux-2.6.12-rc2 |
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 mm: Map most file... |
24 |
#include <linux/export.h> |
1da177e4c Linux-2.6.12-rc2 |
25 |
#include <linux/mm_inline.h> |
1da177e4c Linux-2.6.12-rc2 |
26 27 28 29 |
#include <linux/percpu_counter.h> #include <linux/percpu.h> #include <linux/cpu.h> #include <linux/notifier.h> |
e0bf68dde mm: bdi init hooks |
30 |
#include <linux/backing-dev.h> |
66e1707bc Memory controller... |
31 |
#include <linux/memcontrol.h> |
5a0e3ad6a include cleanup: ... |
32 |
#include <linux/gfp.h> |
1da177e4c Linux-2.6.12-rc2 |
33 |
|
64d6519dd swap: cull unevic... |
34 |
#include "internal.h" |
1da177e4c Linux-2.6.12-rc2 |
35 36 |
/* How many pages do we try to swap or page in/out together? */ int page_cluster; |
f04e9ebbe swap: use an arra... |
37 |
static DEFINE_PER_CPU(struct pagevec[NR_LRU_LISTS], lru_add_pvecs); |
f84f9504b mm: remove initia... |
38 |
static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs); |
315601809 mm: deactivate in... |
39 |
static DEFINE_PER_CPU(struct pagevec, lru_deactivate_pvecs); |
902aaed0d mm: use pagevec t... |
40 |
|
b221385bc [PATCH] mm/: make... |
41 42 43 44 |
/* * This path almost never happens for VM activity - pages are normally * freed via pagevecs. But it gets used by networking. */ |
920c7a5d0 mm: remove fastca... |
45 |
static void __page_cache_release(struct page *page) |
b221385bc [PATCH] mm/: make... |
46 47 |
{ if (PageLRU(page)) { |
b221385bc [PATCH] mm/: make... |
48 |
struct zone *zone = page_zone(page); |
fa9add641 mm/memcg: apply a... |
49 50 |
struct lruvec *lruvec; unsigned long flags; |
b221385bc [PATCH] mm/: make... |
51 52 |
spin_lock_irqsave(&zone->lru_lock, flags); |
fa9add641 mm/memcg: apply a... |
53 |
lruvec = mem_cgroup_page_lruvec(page, zone); |
b221385bc [PATCH] mm/: make... |
54 55 |
VM_BUG_ON(!PageLRU(page)); __ClearPageLRU(page); |
fa9add641 mm/memcg: apply a... |
56 |
del_page_from_lru_list(page, lruvec, page_off_lru(page)); |
b221385bc [PATCH] mm/: make... |
57 58 |
spin_unlock_irqrestore(&zone->lru_lock, flags); } |
918070634 thp: alter compou... |
59 60 61 62 63 |
} static void __put_single_page(struct page *page) { __page_cache_release(page); |
fc91668ea mm: remove free_h... |
64 |
free_hot_cold_page(page, 0); |
b221385bc [PATCH] mm/: make... |
65 |
} |
918070634 thp: alter compou... |
66 |
static void __put_compound_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
67 |
{ |
918070634 thp: alter compou... |
68 |
compound_page_dtor *dtor; |
1da177e4c Linux-2.6.12-rc2 |
69 |
|
918070634 thp: alter compou... |
70 71 72 73 74 75 76 77 78 |
__page_cache_release(page); dtor = get_compound_page_dtor(page); (*dtor)(page); } static void put_compound_page(struct page *page) { if (unlikely(PageTail(page))) { /* __split_huge_page_refcount can run under us */ |
70b50f94f mm: thp: tail pag... |
79 80 81 82 |
struct page *page_head = compound_trans_head(page); if (likely(page != page_head && get_page_unless_zero(page_head))) { |
918070634 thp: alter compou... |
83 |
unsigned long flags; |
5bf5f03c2 mm: fix slab->pag... |
84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 |
/* * THP can not break up slab pages so avoid taking * compound_lock(). Slab performs non-atomic bit ops * on page->flags for better performance. In particular * slab_unlock() in slub used to be a hot path. It is * still hot on arches that do not support * this_cpu_cmpxchg_double(). */ if (PageSlab(page_head)) { if (PageTail(page)) { if (put_page_testzero(page_head)) VM_BUG_ON(1); atomic_dec(&page->_mapcount); goto skip_lock_tail; } else goto skip_lock; } |
918070634 thp: alter compou... |
103 |
/* |
70b50f94f mm: thp: tail pag... |
104 105 106 107 |
* page_head wasn't a dangling pointer but it * may not be a head page anymore by the time * we obtain the lock. That is ok as long as it * can't be freed from under us. |
918070634 thp: alter compou... |
108 |
*/ |
918070634 thp: alter compou... |
109 110 111 112 |
flags = compound_lock_irqsave(page_head); if (unlikely(!PageTail(page))) { /* __split_huge_page_refcount run before us */ compound_unlock_irqrestore(page_head, flags); |
5bf5f03c2 mm: fix slab->pag... |
113 |
skip_lock: |
918070634 thp: alter compou... |
114 115 |
if (put_page_testzero(page_head)) __put_single_page(page_head); |
5bf5f03c2 mm: fix slab->pag... |
116 |
out_put_single: |
918070634 thp: alter compou... |
117 118 119 120 121 122 123 |
if (put_page_testzero(page)) __put_single_page(page); return; } VM_BUG_ON(page_head != page->first_page); /* * We can release the refcount taken by |
70b50f94f mm: thp: tail pag... |
124 125 126 |
* get_page_unless_zero() now that * __split_huge_page_refcount() is blocked on * the compound_lock. |
918070634 thp: alter compou... |
127 128 129 130 |
*/ if (put_page_testzero(page_head)) VM_BUG_ON(1); /* __split_huge_page_refcount will wait now */ |
70b50f94f mm: thp: tail pag... |
131 132 |
VM_BUG_ON(page_mapcount(page) <= 0); atomic_dec(&page->_mapcount); |
918070634 thp: alter compou... |
133 |
VM_BUG_ON(atomic_read(&page_head->_count) <= 0); |
70b50f94f mm: thp: tail pag... |
134 |
VM_BUG_ON(atomic_read(&page->_count) != 0); |
918070634 thp: alter compou... |
135 |
compound_unlock_irqrestore(page_head, flags); |
5bf5f03c2 mm: fix slab->pag... |
136 137 |
skip_lock_tail: |
a95a82e96 thp: put_page: re... |
138 139 140 141 142 143 |
if (put_page_testzero(page_head)) { if (PageHead(page_head)) __put_compound_page(page_head); else __put_single_page(page_head); } |
918070634 thp: alter compou... |
144 145 146 147 148 149 150 151 152 153 |
} else { /* page_head is a dangling pointer */ VM_BUG_ON(PageTail(page)); goto out_put_single; } } else if (put_page_testzero(page)) { if (PageHead(page)) __put_compound_page(page); else __put_single_page(page); |
1da177e4c Linux-2.6.12-rc2 |
154 |
} |
8519fb30e [PATCH] mm: compo... |
155 156 157 158 159 160 161 |
} void put_page(struct page *page) { if (unlikely(PageCompound(page))) put_compound_page(page); else if (put_page_testzero(page)) |
918070634 thp: alter compou... |
162 |
__put_single_page(page); |
1da177e4c Linux-2.6.12-rc2 |
163 164 |
} EXPORT_SYMBOL(put_page); |
1da177e4c Linux-2.6.12-rc2 |
165 |
|
70b50f94f mm: thp: tail pag... |
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 |
/* * This function is exported but must not be called by anything other * than get_page(). It implements the slow path of get_page(). */ bool __get_page_tail(struct page *page) { /* * This takes care of get_page() if run on a tail page * returned by one of the get_user_pages/follow_page variants. * get_user_pages/follow_page itself doesn't need the compound * lock because it runs __get_page_tail_foll() under the * proper PT lock that already serializes against * split_huge_page(). */ unsigned long flags; bool got = false; struct page *page_head = compound_trans_head(page); if (likely(page != page_head && get_page_unless_zero(page_head))) { |
5bf5f03c2 mm: fix slab->pag... |
185 186 187 188 189 190 191 192 193 194 195 |
/* Ref to put_compound_page() comment. */ if (PageSlab(page_head)) { if (likely(PageTail(page))) { __get_page_tail_foll(page, false); return true; } else { put_page(page_head); return false; } } |
70b50f94f mm: thp: tail pag... |
196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 |
/* * page_head wasn't a dangling pointer but it * may not be a head page anymore by the time * we obtain the lock. That is ok as long as it * can't be freed from under us. */ flags = compound_lock_irqsave(page_head); /* here __split_huge_page_refcount won't run anymore */ if (likely(PageTail(page))) { __get_page_tail_foll(page, false); got = true; } compound_unlock_irqrestore(page_head, flags); if (unlikely(!got)) put_page(page_head); } return got; } EXPORT_SYMBOL(__get_page_tail); |
1d7ea7324 [PATCH] fuse: fix... |
215 |
/** |
7682486b3 mm: fix various k... |
216 217 |
* put_pages_list() - release a list of pages * @pages: list of pages threaded on page->lru |
1d7ea7324 [PATCH] fuse: fix... |
218 219 220 |
* * Release a list of pages which are strung together on page.lru. Currently * used by read_cache_pages() and related error recovery code. |
1d7ea7324 [PATCH] fuse: fix... |
221 222 223 224 225 226 227 228 229 230 231 232 |
*/ 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); page_cache_release(victim); } } EXPORT_SYMBOL(put_pages_list); |
18022c5d8 mm: add get_kerne... |
233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 |
/* * 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 mm: add support f... |
254 |
pages[seg] = kmap_to_page(kiov[seg].iov_base); |
18022c5d8 mm: add get_kerne... |
255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 |
page_cache_get(pages[seg]); } 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 mm: simplify code... |
283 |
static void pagevec_lru_move_fn(struct pagevec *pvec, |
fa9add641 mm/memcg: apply a... |
284 285 |
void (*move_fn)(struct page *page, struct lruvec *lruvec, void *arg), void *arg) |
902aaed0d mm: use pagevec t... |
286 287 |
{ int i; |
902aaed0d mm: use pagevec t... |
288 |
struct zone *zone = NULL; |
fa9add641 mm/memcg: apply a... |
289 |
struct lruvec *lruvec; |
3dd7ae8ec mm: simplify code... |
290 |
unsigned long flags = 0; |
902aaed0d mm: use pagevec t... |
291 292 293 294 295 296 297 |
for (i = 0; i < pagevec_count(pvec); i++) { struct page *page = pvec->pages[i]; struct zone *pagezone = page_zone(page); if (pagezone != zone) { if (zone) |
3dd7ae8ec mm: simplify code... |
298 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
902aaed0d mm: use pagevec t... |
299 |
zone = pagezone; |
3dd7ae8ec mm: simplify code... |
300 |
spin_lock_irqsave(&zone->lru_lock, flags); |
902aaed0d mm: use pagevec t... |
301 |
} |
3dd7ae8ec mm: simplify code... |
302 |
|
fa9add641 mm/memcg: apply a... |
303 304 |
lruvec = mem_cgroup_page_lruvec(page, zone); (*move_fn)(page, lruvec, arg); |
902aaed0d mm: use pagevec t... |
305 306 |
} if (zone) |
3dd7ae8ec mm: simplify code... |
307 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
83896fb5e Revert "mm: simpl... |
308 309 |
release_pages(pvec->pages, pvec->nr, pvec->cold); pagevec_reinit(pvec); |
d8505dee1 mm: simplify code... |
310 |
} |
fa9add641 mm/memcg: apply a... |
311 312 |
static void pagevec_move_tail_fn(struct page *page, struct lruvec *lruvec, void *arg) |
3dd7ae8ec mm: simplify code... |
313 314 |
{ int *pgmoved = arg; |
3dd7ae8ec mm: simplify code... |
315 316 317 |
if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { enum lru_list lru = page_lru_base_type(page); |
925b7673c mm: make per-memc... |
318 |
list_move_tail(&page->lru, &lruvec->lists[lru]); |
3dd7ae8ec mm: simplify code... |
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 |
(*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 mm: use pagevec t... |
334 |
/* |
1da177e4c Linux-2.6.12-rc2 |
335 336 |
* 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 mm: use pagevec t... |
337 |
* inactive list. |
1da177e4c Linux-2.6.12-rc2 |
338 |
*/ |
3dd7ae8ec mm: simplify code... |
339 |
void rotate_reclaimable_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
340 |
{ |
ac6aadb24 mm: rotate_reclai... |
341 |
if (!PageLocked(page) && !PageDirty(page) && !PageActive(page) && |
894bc3104 Unevictable LRU I... |
342 |
!PageUnevictable(page) && PageLRU(page)) { |
ac6aadb24 mm: rotate_reclai... |
343 344 345 346 347 348 349 350 351 352 |
struct pagevec *pvec; unsigned long flags; page_cache_get(page); local_irq_save(flags); pvec = &__get_cpu_var(lru_rotate_pvecs); if (!pagevec_add(pvec, page)) pagevec_move_tail(pvec); local_irq_restore(flags); } |
1da177e4c Linux-2.6.12-rc2 |
353 |
} |
fa9add641 mm/memcg: apply a... |
354 |
static void update_page_reclaim_stat(struct lruvec *lruvec, |
3e2f41f1f memcg: add zone_r... |
355 356 |
int file, int rotated) { |
fa9add641 mm/memcg: apply a... |
357 |
struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat; |
3e2f41f1f memcg: add zone_r... |
358 359 360 361 |
reclaim_stat->recent_scanned[file]++; if (rotated) reclaim_stat->recent_rotated[file]++; |
3e2f41f1f memcg: add zone_r... |
362 |
} |
fa9add641 mm/memcg: apply a... |
363 364 |
static void __activate_page(struct page *page, struct lruvec *lruvec, void *arg) |
1da177e4c Linux-2.6.12-rc2 |
365 |
{ |
744ed1442 mm: batch activat... |
366 |
if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { |
7a608572a Revert "mm: batch... |
367 368 |
int file = page_is_file_cache(page); int lru = page_lru_base_type(page); |
744ed1442 mm: batch activat... |
369 |
|
fa9add641 mm/memcg: apply a... |
370 |
del_page_from_lru_list(page, lruvec, lru); |
7a608572a Revert "mm: batch... |
371 372 |
SetPageActive(page); lru += LRU_ACTIVE; |
fa9add641 mm/memcg: apply a... |
373 |
add_page_to_lru_list(page, lruvec, lru); |
4f98a2fee vmscan: split LRU... |
374 |
|
fa9add641 mm/memcg: apply a... |
375 376 |
__count_vm_event(PGACTIVATE); update_page_reclaim_stat(lruvec, file, 1); |
1da177e4c Linux-2.6.12-rc2 |
377 |
} |
eb709b0d0 mm: batch activat... |
378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 |
} #ifdef CONFIG_SMP static DEFINE_PER_CPU(struct pagevec, activate_page_pvecs); 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); } void activate_page(struct page *page) { if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { struct pagevec *pvec = &get_cpu_var(activate_page_pvecs); page_cache_get(page); if (!pagevec_add(pvec, page)) pagevec_lru_move_fn(pvec, __activate_page, NULL); put_cpu_var(activate_page_pvecs); } } #else static inline void activate_page_drain(int cpu) { } void activate_page(struct page *page) { struct zone *zone = page_zone(page); spin_lock_irq(&zone->lru_lock); |
fa9add641 mm/memcg: apply a... |
413 |
__activate_page(page, mem_cgroup_page_lruvec(page, zone), NULL); |
1da177e4c Linux-2.6.12-rc2 |
414 415 |
spin_unlock_irq(&zone->lru_lock); } |
eb709b0d0 mm: batch activat... |
416 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
417 418 419 420 421 422 423 424 |
/* * Mark a page as having seen activity. * * inactive,unreferenced -> inactive,referenced * inactive,referenced -> active,unreferenced * active,unreferenced -> active,referenced */ |
920c7a5d0 mm: remove fastca... |
425 |
void mark_page_accessed(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
426 |
{ |
894bc3104 Unevictable LRU I... |
427 428 |
if (!PageActive(page) && !PageUnevictable(page) && PageReferenced(page) && PageLRU(page)) { |
1da177e4c Linux-2.6.12-rc2 |
429 430 431 432 433 434 |
activate_page(page); ClearPageReferenced(page); } else if (!PageReferenced(page)) { SetPageReferenced(page); } } |
1da177e4c Linux-2.6.12-rc2 |
435 |
EXPORT_SYMBOL(mark_page_accessed); |
d741c9cde mm: fix nonunifor... |
436 437 438 439 440 441 442 443 |
/* * Order of operations is important: flush the pagevec when it's already * full, not when adding the last page, to make sure that last page is * not added to the LRU directly when passed to this function. Because * mark_page_accessed() (called after this when writing) only activates * pages that are on the LRU, linear writes in subpage chunks would see * every PAGEVEC_SIZE page activated, which is unexpected. */ |
f04e9ebbe swap: use an arra... |
444 |
void __lru_cache_add(struct page *page, enum lru_list lru) |
1da177e4c Linux-2.6.12-rc2 |
445 |
{ |
f04e9ebbe swap: use an arra... |
446 |
struct pagevec *pvec = &get_cpu_var(lru_add_pvecs)[lru]; |
1da177e4c Linux-2.6.12-rc2 |
447 448 |
page_cache_get(page); |
d741c9cde mm: fix nonunifor... |
449 |
if (!pagevec_space(pvec)) |
5095ae837 mm: fewer undersc... |
450 |
__pagevec_lru_add(pvec, lru); |
d741c9cde mm: fix nonunifor... |
451 |
pagevec_add(pvec, page); |
1da177e4c Linux-2.6.12-rc2 |
452 453 |
put_cpu_var(lru_add_pvecs); } |
47846b065 mm: export lru_ca... |
454 |
EXPORT_SYMBOL(__lru_cache_add); |
1da177e4c Linux-2.6.12-rc2 |
455 |
|
f04e9ebbe swap: use an arra... |
456 457 458 459 460 461 |
/** * lru_cache_add_lru - add a page to a page list * @page: the page to be added to the LRU. * @lru: the LRU list to which the page is added. */ void lru_cache_add_lru(struct page *page, enum lru_list lru) |
1da177e4c Linux-2.6.12-rc2 |
462 |
{ |
f04e9ebbe swap: use an arra... |
463 |
if (PageActive(page)) { |
894bc3104 Unevictable LRU I... |
464 |
VM_BUG_ON(PageUnevictable(page)); |
f04e9ebbe swap: use an arra... |
465 |
ClearPageActive(page); |
894bc3104 Unevictable LRU I... |
466 467 468 |
} else if (PageUnevictable(page)) { VM_BUG_ON(PageActive(page)); ClearPageUnevictable(page); |
f04e9ebbe swap: use an arra... |
469 |
} |
1da177e4c Linux-2.6.12-rc2 |
470 |
|
894bc3104 Unevictable LRU I... |
471 |
VM_BUG_ON(PageLRU(page) || PageActive(page) || PageUnevictable(page)); |
f04e9ebbe swap: use an arra... |
472 |
__lru_cache_add(page, lru); |
1da177e4c Linux-2.6.12-rc2 |
473 |
} |
894bc3104 Unevictable LRU I... |
474 475 476 477 478 479 480 481 482 483 484 485 486 |
/** * 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) { struct zone *zone = page_zone(page); |
fa9add641 mm/memcg: apply a... |
487 |
struct lruvec *lruvec; |
894bc3104 Unevictable LRU I... |
488 489 |
spin_lock_irq(&zone->lru_lock); |
fa9add641 mm/memcg: apply a... |
490 |
lruvec = mem_cgroup_page_lruvec(page, zone); |
894bc3104 Unevictable LRU I... |
491 492 |
SetPageUnevictable(page); SetPageLRU(page); |
fa9add641 mm/memcg: apply a... |
493 |
add_page_to_lru_list(page, lruvec, LRU_UNEVICTABLE); |
894bc3104 Unevictable LRU I... |
494 495 |
spin_unlock_irq(&zone->lru_lock); } |
902aaed0d mm: use pagevec t... |
496 |
/* |
315601809 mm: deactivate in... |
497 498 499 500 501 |
* 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 mm: reclaim inval... |
502 503 504 505 506 507 508 509 510 511 512 513 514 515 |
* * 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 mm: deactivate in... |
516 |
*/ |
fa9add641 mm/memcg: apply a... |
517 518 |
static void lru_deactivate_fn(struct page *page, struct lruvec *lruvec, void *arg) |
315601809 mm: deactivate in... |
519 520 |
{ int lru, file; |
278df9f45 mm: reclaim inval... |
521 |
bool active; |
315601809 mm: deactivate in... |
522 |
|
278df9f45 mm: reclaim inval... |
523 |
if (!PageLRU(page)) |
315601809 mm: deactivate in... |
524 |
return; |
bad49d9c8 mm: check PageUne... |
525 526 |
if (PageUnevictable(page)) return; |
315601809 mm: deactivate in... |
527 528 529 |
/* Some processes are using the page */ if (page_mapped(page)) return; |
278df9f45 mm: reclaim inval... |
530 |
active = PageActive(page); |
315601809 mm: deactivate in... |
531 532 |
file = page_is_file_cache(page); lru = page_lru_base_type(page); |
fa9add641 mm/memcg: apply a... |
533 534 |
del_page_from_lru_list(page, lruvec, lru + active); |
315601809 mm: deactivate in... |
535 536 |
ClearPageActive(page); ClearPageReferenced(page); |
fa9add641 mm/memcg: apply a... |
537 |
add_page_to_lru_list(page, lruvec, lru); |
315601809 mm: deactivate in... |
538 |
|
278df9f45 mm: reclaim inval... |
539 540 541 542 543 544 545 546 547 548 549 550 |
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 mm: make per-memc... |
551 |
list_move_tail(&page->lru, &lruvec->lists[lru]); |
278df9f45 mm: reclaim inval... |
552 553 554 555 556 |
__count_vm_event(PGROTATED); } if (active) __count_vm_event(PGDEACTIVATE); |
fa9add641 mm/memcg: apply a... |
557 |
update_page_reclaim_stat(lruvec, file, 0); |
315601809 mm: deactivate in... |
558 |
} |
315601809 mm: deactivate in... |
559 |
/* |
902aaed0d mm: use pagevec t... |
560 561 562 563 |
* 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 mm: drain percpu ... |
564 |
void lru_add_drain_cpu(int cpu) |
1da177e4c Linux-2.6.12-rc2 |
565 |
{ |
f04e9ebbe swap: use an arra... |
566 |
struct pagevec *pvecs = per_cpu(lru_add_pvecs, cpu); |
902aaed0d mm: use pagevec t... |
567 |
struct pagevec *pvec; |
f04e9ebbe swap: use an arra... |
568 |
int lru; |
1da177e4c Linux-2.6.12-rc2 |
569 |
|
f04e9ebbe swap: use an arra... |
570 571 572 |
for_each_lru(lru) { pvec = &pvecs[lru - LRU_BASE]; if (pagevec_count(pvec)) |
5095ae837 mm: fewer undersc... |
573 |
__pagevec_lru_add(pvec, lru); |
f04e9ebbe swap: use an arra... |
574 |
} |
902aaed0d mm: use pagevec t... |
575 576 577 578 579 580 581 582 583 584 |
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 mm: deactivate in... |
585 586 587 |
pvec = &per_cpu(lru_deactivate_pvecs, cpu); if (pagevec_count(pvec)) |
3dd7ae8ec mm: simplify code... |
588 |
pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL); |
eb709b0d0 mm: batch activat... |
589 590 |
activate_page_drain(cpu); |
315601809 mm: deactivate in... |
591 592 593 594 595 596 597 598 599 600 601 602 |
} /** * deactivate_page - forcefully deactivate a page * @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. */ void deactivate_page(struct page *page) { |
821ed6bbe mm: filter unevic... |
603 604 605 606 607 608 |
/* * In a workload with many unevictable page such as mprotect, unevictable * page deactivation for accelerating reclaim is pointless. */ if (PageUnevictable(page)) return; |
315601809 mm: deactivate in... |
609 610 611 612 |
if (likely(get_page_unless_zero(page))) { struct pagevec *pvec = &get_cpu_var(lru_deactivate_pvecs); if (!pagevec_add(pvec, page)) |
3dd7ae8ec mm: simplify code... |
613 |
pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL); |
315601809 mm: deactivate in... |
614 615 |
put_cpu_var(lru_deactivate_pvecs); } |
80bfed904 [PATCH] consolida... |
616 617 618 619 |
} void lru_add_drain(void) { |
f0cb3c76a mm: drain percpu ... |
620 |
lru_add_drain_cpu(get_cpu()); |
80bfed904 [PATCH] consolida... |
621 |
put_cpu(); |
1da177e4c Linux-2.6.12-rc2 |
622 |
} |
c4028958b WorkStruct: make ... |
623 |
static void lru_add_drain_per_cpu(struct work_struct *dummy) |
053837fce [PATCH] mm: migra... |
624 625 626 627 628 629 630 631 632 |
{ lru_add_drain(); } /* * Returns 0 for success */ int lru_add_drain_all(void) { |
c4028958b WorkStruct: make ... |
633 |
return schedule_on_each_cpu(lru_add_drain_per_cpu); |
053837fce [PATCH] mm: migra... |
634 |
} |
1da177e4c Linux-2.6.12-rc2 |
635 |
/* |
1da177e4c Linux-2.6.12-rc2 |
636 637 638 639 640 641 642 |
* Batched page_cache_release(). Decrement the reference count on all the * passed pages. If it fell to zero then remove the page from the LRU and * free it. * * Avoid taking zone->lru_lock if possible, but if it is taken, retain it * for the remainder of the operation. * |
ab33dc09a swap: update func... |
643 644 645 646 |
* The locking in this function is against shrink_inactive_list(): we recheck * the page count inside the lock to see whether shrink_inactive_list() * grabbed the page via the LRU. If it did, give up: shrink_inactive_list() * will free it. |
1da177e4c Linux-2.6.12-rc2 |
647 648 649 650 |
*/ void release_pages(struct page **pages, int nr, int cold) { int i; |
cc59850ef mm: add free_hot_... |
651 |
LIST_HEAD(pages_to_free); |
1da177e4c Linux-2.6.12-rc2 |
652 |
struct zone *zone = NULL; |
fa9add641 mm/memcg: apply a... |
653 |
struct lruvec *lruvec; |
902aaed0d mm: use pagevec t... |
654 |
unsigned long uninitialized_var(flags); |
1da177e4c Linux-2.6.12-rc2 |
655 |
|
1da177e4c Linux-2.6.12-rc2 |
656 657 |
for (i = 0; i < nr; i++) { struct page *page = pages[i]; |
1da177e4c Linux-2.6.12-rc2 |
658 |
|
8519fb30e [PATCH] mm: compo... |
659 660 |
if (unlikely(PageCompound(page))) { if (zone) { |
902aaed0d mm: use pagevec t... |
661 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
8519fb30e [PATCH] mm: compo... |
662 663 664 665 666 |
zone = NULL; } put_compound_page(page); continue; } |
b5810039a [PATCH] core remo... |
667 |
if (!put_page_testzero(page)) |
1da177e4c Linux-2.6.12-rc2 |
668 |
continue; |
46453a6e1 [PATCH] mm: never... |
669 670 |
if (PageLRU(page)) { struct zone *pagezone = page_zone(page); |
894bc3104 Unevictable LRU I... |
671 |
|
46453a6e1 [PATCH] mm: never... |
672 673 |
if (pagezone != zone) { if (zone) |
902aaed0d mm: use pagevec t... |
674 675 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
46453a6e1 [PATCH] mm: never... |
676 |
zone = pagezone; |
902aaed0d mm: use pagevec t... |
677 |
spin_lock_irqsave(&zone->lru_lock, flags); |
46453a6e1 [PATCH] mm: never... |
678 |
} |
fa9add641 mm/memcg: apply a... |
679 680 |
lruvec = mem_cgroup_page_lruvec(page, zone); |
725d704ec [PATCH] mm: VM_BU... |
681 |
VM_BUG_ON(!PageLRU(page)); |
674539115 [PATCH] mm: less ... |
682 |
__ClearPageLRU(page); |
fa9add641 mm/memcg: apply a... |
683 |
del_page_from_lru_list(page, lruvec, page_off_lru(page)); |
46453a6e1 [PATCH] mm: never... |
684 |
} |
cc59850ef mm: add free_hot_... |
685 |
list_add(&page->lru, &pages_to_free); |
1da177e4c Linux-2.6.12-rc2 |
686 687 |
} if (zone) |
902aaed0d mm: use pagevec t... |
688 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
1da177e4c Linux-2.6.12-rc2 |
689 |
|
cc59850ef mm: add free_hot_... |
690 |
free_hot_cold_page_list(&pages_to_free, cold); |
1da177e4c Linux-2.6.12-rc2 |
691 |
} |
0be8557bc fuse: use release... |
692 |
EXPORT_SYMBOL(release_pages); |
1da177e4c Linux-2.6.12-rc2 |
693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 |
/* * 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) { lru_add_drain(); release_pages(pvec->pages, pagevec_count(pvec), pvec->cold); pagevec_reinit(pvec); } |
7f2857018 Export __pagevec_... |
710 |
EXPORT_SYMBOL(__pagevec_release); |
12d271078 memcg: fix split_... |
711 |
#ifdef CONFIG_TRANSPARENT_HUGEPAGE |
71e3aac07 thp: transparent ... |
712 |
/* used by __split_huge_page_refcount() */ |
fa9add641 mm/memcg: apply a... |
713 714 |
void lru_add_page_tail(struct page *page, struct page *page_tail, struct lruvec *lruvec) |
71e3aac07 thp: transparent ... |
715 |
{ |
7512102cf memcg: fix GPF wh... |
716 |
int uninitialized_var(active); |
71e3aac07 thp: transparent ... |
717 718 |
enum lru_list lru; const int file = 0; |
71e3aac07 thp: transparent ... |
719 720 721 722 |
VM_BUG_ON(!PageHead(page)); VM_BUG_ON(PageCompound(page_tail)); VM_BUG_ON(PageLRU(page_tail)); |
fa9add641 mm/memcg: apply a... |
723 724 |
VM_BUG_ON(NR_CPUS != 1 && !spin_is_locked(&lruvec_zone(lruvec)->lru_lock)); |
71e3aac07 thp: transparent ... |
725 726 |
SetPageLRU(page_tail); |
39b5f29ac mm: remove vma ar... |
727 |
if (page_evictable(page_tail)) { |
71e3aac07 thp: transparent ... |
728 729 730 731 732 733 734 735 |
if (PageActive(page)) { SetPageActive(page_tail); active = 1; lru = LRU_ACTIVE_ANON; } else { active = 0; lru = LRU_INACTIVE_ANON; } |
71e3aac07 thp: transparent ... |
736 737 |
} else { SetPageUnevictable(page_tail); |
12d271078 memcg: fix split_... |
738 739 740 741 742 743 744 745 746 747 748 749 750 751 |
lru = LRU_UNEVICTABLE; } if (likely(PageLRU(page))) list_add_tail(&page_tail->lru, &page->lru); else { 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. */ |
fa9add641 mm/memcg: apply a... |
752 |
add_page_to_lru_list(page_tail, lruvec, lru); |
12d271078 memcg: fix split_... |
753 754 |
list_head = page_tail->lru.prev; list_move_tail(&page_tail->lru, list_head); |
71e3aac07 thp: transparent ... |
755 |
} |
7512102cf memcg: fix GPF wh... |
756 757 |
if (!PageUnevictable(page)) |
fa9add641 mm/memcg: apply a... |
758 |
update_page_reclaim_stat(lruvec, file, active); |
71e3aac07 thp: transparent ... |
759 |
} |
12d271078 memcg: fix split_... |
760 |
#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ |
71e3aac07 thp: transparent ... |
761 |
|
fa9add641 mm/memcg: apply a... |
762 763 |
static void __pagevec_lru_add_fn(struct page *page, struct lruvec *lruvec, void *arg) |
3dd7ae8ec mm: simplify code... |
764 765 |
{ enum lru_list lru = (enum lru_list)arg; |
3dd7ae8ec mm: simplify code... |
766 767 768 769 770 771 772 773 774 775 |
int file = is_file_lru(lru); int active = is_active_lru(lru); VM_BUG_ON(PageActive(page)); VM_BUG_ON(PageUnevictable(page)); VM_BUG_ON(PageLRU(page)); SetPageLRU(page); if (active) SetPageActive(page); |
fa9add641 mm/memcg: apply a... |
776 777 |
add_page_to_lru_list(page, lruvec, lru); update_page_reclaim_stat(lruvec, file, active); |
3dd7ae8ec mm: simplify code... |
778 |
} |
1da177e4c Linux-2.6.12-rc2 |
779 |
/* |
1da177e4c Linux-2.6.12-rc2 |
780 781 782 |
* Add the passed pages to the LRU, then drop the caller's refcount * on them. Reinitialises the caller's pagevec. */ |
5095ae837 mm: fewer undersc... |
783 |
void __pagevec_lru_add(struct pagevec *pvec, enum lru_list lru) |
1da177e4c Linux-2.6.12-rc2 |
784 |
{ |
894bc3104 Unevictable LRU I... |
785 |
VM_BUG_ON(is_unevictable_lru(lru)); |
1da177e4c Linux-2.6.12-rc2 |
786 |
|
5095ae837 mm: fewer undersc... |
787 |
pagevec_lru_move_fn(pvec, __pagevec_lru_add_fn, (void *)lru); |
1da177e4c Linux-2.6.12-rc2 |
788 |
} |
5095ae837 mm: fewer undersc... |
789 |
EXPORT_SYMBOL(__pagevec_lru_add); |
1da177e4c Linux-2.6.12-rc2 |
790 |
|
1da177e4c Linux-2.6.12-rc2 |
791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 |
/** * pagevec_lookup - gang pagecache lookup * @pvec: Where the resulting pages are placed * @mapping: The address_space to search * @start: The starting page index * @nr_pages: The maximum number of pages * * pagevec_lookup() will search for and return a group of up to @nr_pages pages * in the mapping. The pages are placed in @pvec. pagevec_lookup() takes a * reference against the pages in @pvec. * * The search returns a group of mapping-contiguous pages with ascending * indexes. There may be holes in the indices due to not-present pages. * * pagevec_lookup() returns the number of pages which were found. */ unsigned pagevec_lookup(struct pagevec *pvec, struct address_space *mapping, pgoff_t start, unsigned nr_pages) { pvec->nr = find_get_pages(mapping, start, nr_pages, pvec->pages); return pagevec_count(pvec); } |
78539fdfa [XFS] Export page... |
813 |
EXPORT_SYMBOL(pagevec_lookup); |
1da177e4c Linux-2.6.12-rc2 |
814 815 816 817 818 819 820 |
unsigned pagevec_lookup_tag(struct pagevec *pvec, struct address_space *mapping, pgoff_t *index, int tag, unsigned nr_pages) { pvec->nr = find_get_pages_tag(mapping, index, tag, nr_pages, pvec->pages); return pagevec_count(pvec); } |
7f2857018 Export __pagevec_... |
821 |
EXPORT_SYMBOL(pagevec_lookup_tag); |
1da177e4c Linux-2.6.12-rc2 |
822 |
|
1da177e4c Linux-2.6.12-rc2 |
823 824 825 826 827 |
/* * Perform any setup for the swap system */ void __init swap_setup(void) { |
4481374ce mm: replace vario... |
828 |
unsigned long megs = totalram_pages >> (20 - PAGE_SHIFT); |
e0bf68dde mm: bdi init hooks |
829 |
#ifdef CONFIG_SWAP |
33806f06d swap: make each s... |
830 831 832 833 834 835 836 |
int i; bdi_init(swapper_spaces[0].backing_dev_info); for (i = 0; i < MAX_SWAPFILES; i++) { spin_lock_init(&swapper_spaces[i].tree_lock); INIT_LIST_HEAD(&swapper_spaces[i].i_mmap_nonlinear); } |
e0bf68dde mm: bdi init hooks |
837 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
838 839 840 841 842 843 844 845 846 |
/* 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 Linux-2.6.12-rc2 |
847 |
} |