Blame view
mm/swap.c
31 KB
457c89965 treewide: Add SPD... |
1 |
// SPDX-License-Identifier: GPL-2.0-only |
1da177e4c Linux-2.6.12-rc2 |
2 3 4 5 6 7 8 |
/* * linux/mm/swap.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds */ /* |
183ff22bb spelling fixes: mm/ |
9 |
* This file contains the default values for the operation of the |
1da177e4c Linux-2.6.12-rc2 |
10 |
* Linux VM subsystem. Fine-tuning documentation can be found in |
570432470 docs: admin-guide... |
11 |
* Documentation/admin-guide/sysctl/vm.rst. |
1da177e4c Linux-2.6.12-rc2 |
12 13 14 15 16 17 18 19 20 21 22 23 24 |
* 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... |
25 |
#include <linux/export.h> |
1da177e4c Linux-2.6.12-rc2 |
26 |
#include <linux/mm_inline.h> |
1da177e4c Linux-2.6.12-rc2 |
27 |
#include <linux/percpu_counter.h> |
3565fce3a mm, x86: get_user... |
28 |
#include <linux/memremap.h> |
1da177e4c Linux-2.6.12-rc2 |
29 30 31 |
#include <linux/percpu.h> #include <linux/cpu.h> #include <linux/notifier.h> |
e0bf68dde mm: bdi init hooks |
32 |
#include <linux/backing-dev.h> |
66e1707bc Memory controller... |
33 |
#include <linux/memcontrol.h> |
5a0e3ad6a include cleanup: ... |
34 |
#include <linux/gfp.h> |
a27bb332c aio: don't includ... |
35 |
#include <linux/uio.h> |
822fc6136 mm: don't call __... |
36 |
#include <linux/hugetlb.h> |
33c3fc71c mm: introduce idl... |
37 |
#include <linux/page_idle.h> |
1da177e4c Linux-2.6.12-rc2 |
38 |
|
64d6519dd swap: cull unevic... |
39 |
#include "internal.h" |
c6286c983 mm: add tracepoin... |
40 41 |
#define CREATE_TRACE_POINTS #include <trace/events/pagemap.h> |
1da177e4c Linux-2.6.12-rc2 |
42 43 |
/* How many pages do we try to swap or page in/out together? */ int page_cluster; |
13f7f7898 mm: pagevec: defe... |
44 |
static DEFINE_PER_CPU(struct pagevec, lru_add_pvec); |
f84f9504b mm: remove initia... |
45 |
static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs); |
cc5993bd7 mm: rename deacti... |
46 |
static DEFINE_PER_CPU(struct pagevec, lru_deactivate_file_pvecs); |
9c276cc65 mm: introduce MAD... |
47 |
static DEFINE_PER_CPU(struct pagevec, lru_deactivate_pvecs); |
f7ad2a6cb mm: move MADV_FRE... |
48 |
static DEFINE_PER_CPU(struct pagevec, lru_lazyfree_pvecs); |
a4a921aa5 mm/swap.c: put ac... |
49 50 51 |
#ifdef CONFIG_SMP static DEFINE_PER_CPU(struct pagevec, activate_page_pvecs); #endif |
902aaed0d mm: use pagevec t... |
52 |
|
b221385bc [PATCH] mm/: make... |
53 54 55 56 |
/* * This path almost never happens for VM activity - pages are normally * freed via pagevecs. But it gets used by networking. */ |
920c7a5d0 mm: remove fastca... |
57 |
static void __page_cache_release(struct page *page) |
b221385bc [PATCH] mm/: make... |
58 59 |
{ if (PageLRU(page)) { |
f4b7e272b mm: remove zone_l... |
60 |
pg_data_t *pgdat = page_pgdat(page); |
fa9add641 mm/memcg: apply a... |
61 62 |
struct lruvec *lruvec; unsigned long flags; |
b221385bc [PATCH] mm/: make... |
63 |
|
f4b7e272b mm: remove zone_l... |
64 65 |
spin_lock_irqsave(&pgdat->lru_lock, flags); lruvec = mem_cgroup_page_lruvec(page, pgdat); |
309381fea mm: dump page whe... |
66 |
VM_BUG_ON_PAGE(!PageLRU(page), page); |
b221385bc [PATCH] mm/: make... |
67 |
__ClearPageLRU(page); |
fa9add641 mm/memcg: apply a... |
68 |
del_page_from_lru_list(page, lruvec, page_off_lru(page)); |
f4b7e272b mm: remove zone_l... |
69 |
spin_unlock_irqrestore(&pgdat->lru_lock, flags); |
b221385bc [PATCH] mm/: make... |
70 |
} |
629060270 mm: add PageWaite... |
71 |
__ClearPageWaiters(page); |
918070634 thp: alter compou... |
72 73 74 75 76 |
} static void __put_single_page(struct page *page) { __page_cache_release(page); |
7ae88534c mm: move mem_cgro... |
77 |
mem_cgroup_uncharge(page); |
2d4894b5d mm: remove cold p... |
78 |
free_unref_page(page); |
b221385bc [PATCH] mm/: make... |
79 |
} |
918070634 thp: alter compou... |
80 |
static void __put_compound_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
81 |
{ |
918070634 thp: alter compou... |
82 |
compound_page_dtor *dtor; |
1da177e4c Linux-2.6.12-rc2 |
83 |
|
822fc6136 mm: don't call __... |
84 85 86 87 88 89 90 91 |
/* * __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 thp: alter compou... |
92 93 94 |
dtor = get_compound_page_dtor(page); (*dtor)(page); } |
ddc58f27f mm: drop tail pag... |
95 |
void __put_page(struct page *page) |
8519fb30e [PATCH] mm: compo... |
96 |
{ |
713897038 mm, zone_device: ... |
97 98 99 100 101 102 103 104 105 |
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 [PATCH] mm: compo... |
106 |
if (unlikely(PageCompound(page))) |
ddc58f27f mm: drop tail pag... |
107 108 |
__put_compound_page(page); else |
918070634 thp: alter compou... |
109 |
__put_single_page(page); |
1da177e4c Linux-2.6.12-rc2 |
110 |
} |
ddc58f27f mm: drop tail pag... |
111 |
EXPORT_SYMBOL(__put_page); |
70b50f94f mm: thp: tail pag... |
112 |
|
1d7ea7324 [PATCH] fuse: fix... |
113 |
/** |
7682486b3 mm: fix various k... |
114 115 |
* put_pages_list() - release a list of pages * @pages: list of pages threaded on page->lru |
1d7ea7324 [PATCH] fuse: fix... |
116 117 118 |
* * 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... |
119 120 121 122 123 |
*/ void put_pages_list(struct list_head *pages) { while (!list_empty(pages)) { struct page *victim; |
f86196ea8 fs: don't open co... |
124 |
victim = lru_to_page(pages); |
1d7ea7324 [PATCH] fuse: fix... |
125 |
list_del(&victim->lru); |
09cbfeaf1 mm, fs: get rid o... |
126 |
put_page(victim); |
1d7ea7324 [PATCH] fuse: fix... |
127 128 129 |
} } EXPORT_SYMBOL(put_pages_list); |
18022c5d8 mm: add get_kerne... |
130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 |
/* * 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... |
151 |
pages[seg] = kmap_to_page(kiov[seg].iov_base); |
09cbfeaf1 mm, fs: get rid o... |
152 |
get_page(pages[seg]); |
18022c5d8 mm: add get_kerne... |
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 |
} 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... |
180 |
static void pagevec_lru_move_fn(struct pagevec *pvec, |
fa9add641 mm/memcg: apply a... |
181 182 |
void (*move_fn)(struct page *page, struct lruvec *lruvec, void *arg), void *arg) |
902aaed0d mm: use pagevec t... |
183 184 |
{ int i; |
68eb0731c mm, pagevec: rele... |
185 |
struct pglist_data *pgdat = NULL; |
fa9add641 mm/memcg: apply a... |
186 |
struct lruvec *lruvec; |
3dd7ae8ec mm: simplify code... |
187 |
unsigned long flags = 0; |
902aaed0d mm: use pagevec t... |
188 189 190 |
for (i = 0; i < pagevec_count(pvec); i++) { struct page *page = pvec->pages[i]; |
68eb0731c mm, pagevec: rele... |
191 |
struct pglist_data *pagepgdat = page_pgdat(page); |
902aaed0d mm: use pagevec t... |
192 |
|
68eb0731c mm, pagevec: rele... |
193 194 195 196 197 |
if (pagepgdat != pgdat) { if (pgdat) spin_unlock_irqrestore(&pgdat->lru_lock, flags); pgdat = pagepgdat; spin_lock_irqsave(&pgdat->lru_lock, flags); |
902aaed0d mm: use pagevec t... |
198 |
} |
3dd7ae8ec mm: simplify code... |
199 |
|
68eb0731c mm, pagevec: rele... |
200 |
lruvec = mem_cgroup_page_lruvec(page, pgdat); |
fa9add641 mm/memcg: apply a... |
201 |
(*move_fn)(page, lruvec, arg); |
902aaed0d mm: use pagevec t... |
202 |
} |
68eb0731c mm, pagevec: rele... |
203 204 |
if (pgdat) spin_unlock_irqrestore(&pgdat->lru_lock, flags); |
c6f92f9fb mm: remove cold p... |
205 |
release_pages(pvec->pages, pvec->nr); |
83896fb5e Revert "mm: simpl... |
206 |
pagevec_reinit(pvec); |
d8505dee1 mm: simplify code... |
207 |
} |
fa9add641 mm/memcg: apply a... |
208 209 |
static void pagevec_move_tail_fn(struct page *page, struct lruvec *lruvec, void *arg) |
3dd7ae8ec mm: simplify code... |
210 211 |
{ int *pgmoved = arg; |
3dd7ae8ec mm: simplify code... |
212 |
|
c55e8d035 mm: vmscan: move ... |
213 214 215 216 |
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 mm: simplify code... |
217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 |
(*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... |
232 |
/* |
1da177e4c Linux-2.6.12-rc2 |
233 234 |
* 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... |
235 |
* inactive list. |
1da177e4c Linux-2.6.12-rc2 |
236 |
*/ |
3dd7ae8ec mm: simplify code... |
237 |
void rotate_reclaimable_page(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
238 |
{ |
c55e8d035 mm: vmscan: move ... |
239 |
if (!PageLocked(page) && !PageDirty(page) && |
894bc3104 Unevictable LRU I... |
240 |
!PageUnevictable(page) && PageLRU(page)) { |
ac6aadb24 mm: rotate_reclai... |
241 242 |
struct pagevec *pvec; unsigned long flags; |
09cbfeaf1 mm, fs: get rid o... |
243 |
get_page(page); |
ac6aadb24 mm: rotate_reclai... |
244 |
local_irq_save(flags); |
7c8e0181e mm: replace __get... |
245 |
pvec = this_cpu_ptr(&lru_rotate_pvecs); |
8f182270d mm/swap.c: flush ... |
246 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
ac6aadb24 mm: rotate_reclai... |
247 248 249 |
pagevec_move_tail(pvec); local_irq_restore(flags); } |
1da177e4c Linux-2.6.12-rc2 |
250 |
} |
fa9add641 mm/memcg: apply a... |
251 |
static void update_page_reclaim_stat(struct lruvec *lruvec, |
3e2f41f1f memcg: add zone_r... |
252 253 |
int file, int rotated) { |
fa9add641 mm/memcg: apply a... |
254 |
struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat; |
3e2f41f1f memcg: add zone_r... |
255 256 257 258 |
reclaim_stat->recent_scanned[file]++; if (rotated) reclaim_stat->recent_rotated[file]++; |
3e2f41f1f memcg: add zone_r... |
259 |
} |
fa9add641 mm/memcg: apply a... |
260 261 |
static void __activate_page(struct page *page, struct lruvec *lruvec, void *arg) |
1da177e4c Linux-2.6.12-rc2 |
262 |
{ |
744ed1442 mm: batch activat... |
263 |
if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { |
7a608572a Revert "mm: batch... |
264 265 |
int file = page_is_file_cache(page); int lru = page_lru_base_type(page); |
744ed1442 mm: batch activat... |
266 |
|
fa9add641 mm/memcg: apply a... |
267 |
del_page_from_lru_list(page, lruvec, lru); |
7a608572a Revert "mm: batch... |
268 269 |
SetPageActive(page); lru += LRU_ACTIVE; |
fa9add641 mm/memcg: apply a... |
270 |
add_page_to_lru_list(page, lruvec, lru); |
24b7e5819 mm: pagemap: avoi... |
271 |
trace_mm_lru_activate(page); |
4f98a2fee vmscan: split LRU... |
272 |
|
fa9add641 mm/memcg: apply a... |
273 274 |
__count_vm_event(PGACTIVATE); update_page_reclaim_stat(lruvec, file, 1); |
1da177e4c Linux-2.6.12-rc2 |
275 |
} |
eb709b0d0 mm: batch activat... |
276 277 278 |
} #ifdef CONFIG_SMP |
eb709b0d0 mm: batch activat... |
279 280 281 282 283 284 285 |
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 mm: make lru_add_... |
286 287 288 289 |
static bool need_activate_page_drain(int cpu) { return pagevec_count(&per_cpu(activate_page_pvecs, cpu)) != 0; } |
eb709b0d0 mm: batch activat... |
290 291 |
void activate_page(struct page *page) { |
800d8c63b shmem: add huge p... |
292 |
page = compound_head(page); |
eb709b0d0 mm: batch activat... |
293 294 |
if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { struct pagevec *pvec = &get_cpu_var(activate_page_pvecs); |
09cbfeaf1 mm, fs: get rid o... |
295 |
get_page(page); |
8f182270d mm/swap.c: flush ... |
296 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
eb709b0d0 mm: batch activat... |
297 298 299 300 301 302 303 304 305 306 307 308 |
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) { |
f4b7e272b mm: remove zone_l... |
309 |
pg_data_t *pgdat = page_pgdat(page); |
eb709b0d0 mm: batch activat... |
310 |
|
800d8c63b shmem: add huge p... |
311 |
page = compound_head(page); |
f4b7e272b mm: remove zone_l... |
312 313 314 |
spin_lock_irq(&pgdat->lru_lock); __activate_page(page, mem_cgroup_page_lruvec(page, pgdat), NULL); spin_unlock_irq(&pgdat->lru_lock); |
1da177e4c Linux-2.6.12-rc2 |
315 |
} |
eb709b0d0 mm: batch activat... |
316 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
317 |
|
059285a25 mm: activate !Pag... |
318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 |
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 Linux-2.6.12-rc2 |
344 345 346 347 348 349 |
/* * Mark a page as having seen activity. * * inactive,unreferenced -> inactive,referenced * inactive,referenced -> active,unreferenced * active,unreferenced -> active,referenced |
eb39d618f mm: replace init_... |
350 351 352 |
* * 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 Linux-2.6.12-rc2 |
353 |
*/ |
920c7a5d0 mm: remove fastca... |
354 |
void mark_page_accessed(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
355 |
{ |
e90309c9f thp: allow mlocke... |
356 |
page = compound_head(page); |
059285a25 mm: activate !Pag... |
357 |
|
a1100a740 mm/swap.c: trivia... |
358 359 360 361 362 363 364 365 366 |
if (!PageReferenced(page)) { SetPageReferenced(page); } else if (PageUnevictable(page)) { /* * Unevictable pages are on the "LRU_UNEVICTABLE" list. But, * this list is never rotated or maintained, so marking an * evictable page accessed has no effect. */ } else if (!PageActive(page)) { |
059285a25 mm: activate !Pag... |
367 368 369 370 371 372 373 374 375 376 |
/* * 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 Linux-2.6.12-rc2 |
377 |
ClearPageReferenced(page); |
a528910e1 mm: thrash detect... |
378 379 |
if (page_is_file_cache(page)) workingset_activation(page); |
1da177e4c Linux-2.6.12-rc2 |
380 |
} |
33c3fc71c mm: introduce idl... |
381 382 |
if (page_is_idle(page)) clear_page_idle(page); |
1da177e4c Linux-2.6.12-rc2 |
383 |
} |
1da177e4c Linux-2.6.12-rc2 |
384 |
EXPORT_SYMBOL(mark_page_accessed); |
2329d3751 mm/swap.c: clean ... |
385 |
static void __lru_cache_add(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
386 |
{ |
13f7f7898 mm: pagevec: defe... |
387 |
struct pagevec *pvec = &get_cpu_var(lru_add_pvec); |
09cbfeaf1 mm, fs: get rid o... |
388 |
get_page(page); |
8f182270d mm/swap.c: flush ... |
389 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
a0b8cab3b mm: remove lru pa... |
390 |
__pagevec_lru_add(pvec); |
13f7f7898 mm: pagevec: defe... |
391 |
put_cpu_var(lru_add_pvec); |
1da177e4c Linux-2.6.12-rc2 |
392 |
} |
2329d3751 mm/swap.c: clean ... |
393 394 |
/** |
e02a9f048 mm/swap.c: make f... |
395 |
* lru_cache_add_anon - add a page to the page lists |
2329d3751 mm/swap.c: clean ... |
396 397 398 399 |
* @page: the page to add */ void lru_cache_add_anon(struct page *page) { |
6fb81a17d mm: do not use un... |
400 401 |
if (PageActive(page)) ClearPageActive(page); |
2329d3751 mm/swap.c: clean ... |
402 403 404 405 406 |
__lru_cache_add(page); } void lru_cache_add_file(struct page *page) { |
6fb81a17d mm: do not use un... |
407 408 |
if (PageActive(page)) ClearPageActive(page); |
2329d3751 mm/swap.c: clean ... |
409 410 411 |
__lru_cache_add(page); } EXPORT_SYMBOL(lru_cache_add_file); |
1da177e4c Linux-2.6.12-rc2 |
412 |
|
f04e9ebbe swap: use an arra... |
413 |
/** |
c53954a09 mm: remove lru pa... |
414 |
* lru_cache_add - add a page to a page list |
f04e9ebbe swap: use an arra... |
415 |
* @page: the page to be added to the LRU. |
2329d3751 mm/swap.c: clean ... |
416 417 418 419 420 |
* * 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 swap: use an arra... |
421 |
*/ |
c53954a09 mm: remove lru pa... |
422 |
void lru_cache_add(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
423 |
{ |
309381fea mm: dump page whe... |
424 425 |
VM_BUG_ON_PAGE(PageActive(page) && PageUnevictable(page), page); VM_BUG_ON_PAGE(PageLRU(page), page); |
c53954a09 mm: remove lru pa... |
426 |
__lru_cache_add(page); |
1da177e4c Linux-2.6.12-rc2 |
427 |
} |
894bc3104 Unevictable LRU I... |
428 |
/** |
00501b531 mm: memcontrol: r... |
429 430 431 432 433 434 435 436 437 438 439 440 441 |
* 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); |
9c4e6b1a7 mm, mlock, vmscan... |
442 |
if (likely((vma->vm_flags & (VM_LOCKED | VM_SPECIAL)) != VM_LOCKED)) |
00501b531 mm: memcontrol: r... |
443 |
SetPageActive(page); |
9c4e6b1a7 mm, mlock, vmscan... |
444 |
else if (!TestSetPageMlocked(page)) { |
00501b531 mm: memcontrol: r... |
445 446 447 448 449 450 451 452 453 |
/* * 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); } |
9c4e6b1a7 mm, mlock, vmscan... |
454 |
lru_cache_add(page); |
00501b531 mm: memcontrol: r... |
455 |
} |
902aaed0d mm: use pagevec t... |
456 |
/* |
315601809 mm: deactivate in... |
457 458 459 460 461 |
* 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... |
462 463 464 465 466 467 468 469 470 471 472 473 474 475 |
* * 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... |
476 |
*/ |
cc5993bd7 mm: rename deacti... |
477 |
static void lru_deactivate_file_fn(struct page *page, struct lruvec *lruvec, |
fa9add641 mm/memcg: apply a... |
478 |
void *arg) |
315601809 mm: deactivate in... |
479 480 |
{ int lru, file; |
278df9f45 mm: reclaim inval... |
481 |
bool active; |
315601809 mm: deactivate in... |
482 |
|
278df9f45 mm: reclaim inval... |
483 |
if (!PageLRU(page)) |
315601809 mm: deactivate in... |
484 |
return; |
bad49d9c8 mm: check PageUne... |
485 486 |
if (PageUnevictable(page)) return; |
315601809 mm: deactivate in... |
487 488 489 |
/* Some processes are using the page */ if (page_mapped(page)) return; |
278df9f45 mm: reclaim inval... |
490 |
active = PageActive(page); |
315601809 mm: deactivate in... |
491 492 |
file = page_is_file_cache(page); lru = page_lru_base_type(page); |
fa9add641 mm/memcg: apply a... |
493 494 |
del_page_from_lru_list(page, lruvec, lru + active); |
315601809 mm: deactivate in... |
495 496 |
ClearPageActive(page); ClearPageReferenced(page); |
315601809 mm: deactivate in... |
497 |
|
278df9f45 mm: reclaim inval... |
498 499 500 501 502 503 |
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. */ |
e7a1aaf28 mm: replace list_... |
504 |
add_page_to_lru_list(page, lruvec, lru); |
278df9f45 mm: reclaim inval... |
505 506 507 508 509 510 |
SetPageReclaim(page); } else { /* * The page's writeback ends up during pagevec * We moves tha page into tail of inactive. */ |
e7a1aaf28 mm: replace list_... |
511 |
add_page_to_lru_list_tail(page, lruvec, lru); |
278df9f45 mm: reclaim inval... |
512 513 514 515 516 |
__count_vm_event(PGROTATED); } if (active) __count_vm_event(PGDEACTIVATE); |
fa9add641 mm/memcg: apply a... |
517 |
update_page_reclaim_stat(lruvec, file, 0); |
315601809 mm: deactivate in... |
518 |
} |
9c276cc65 mm: introduce MAD... |
519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 |
static void lru_deactivate_fn(struct page *page, struct lruvec *lruvec, void *arg) { if (PageLRU(page) && PageActive(page) && !PageUnevictable(page)) { int file = page_is_file_cache(page); int lru = page_lru_base_type(page); del_page_from_lru_list(page, lruvec, lru + LRU_ACTIVE); ClearPageActive(page); ClearPageReferenced(page); add_page_to_lru_list(page, lruvec, lru); __count_vm_events(PGDEACTIVATE, hpage_nr_pages(page)); update_page_reclaim_stat(lruvec, file, 0); } } |
10853a039 mm: move lazily f... |
535 |
|
f7ad2a6cb mm: move MADV_FRE... |
536 |
static void lru_lazyfree_fn(struct page *page, struct lruvec *lruvec, |
10853a039 mm: move lazily f... |
537 538 |
void *arg) { |
f7ad2a6cb mm: move MADV_FRE... |
539 |
if (PageLRU(page) && PageAnon(page) && PageSwapBacked(page) && |
24c92eb7d mm: avoid marking... |
540 |
!PageSwapCache(page) && !PageUnevictable(page)) { |
f7ad2a6cb mm: move MADV_FRE... |
541 |
bool active = PageActive(page); |
10853a039 mm: move lazily f... |
542 |
|
f7ad2a6cb mm: move MADV_FRE... |
543 544 |
del_page_from_lru_list(page, lruvec, LRU_INACTIVE_ANON + active); |
10853a039 mm: move lazily f... |
545 546 |
ClearPageActive(page); ClearPageReferenced(page); |
f7ad2a6cb mm: move MADV_FRE... |
547 548 549 550 551 552 553 |
/* * 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 mm: move lazily f... |
554 |
|
f7ad2a6cb mm: move MADV_FRE... |
555 |
__count_vm_events(PGLAZYFREE, hpage_nr_pages(page)); |
2262185c5 mm: per-cgroup me... |
556 |
count_memcg_page_event(page, PGLAZYFREE); |
f7ad2a6cb mm: move MADV_FRE... |
557 |
update_page_reclaim_stat(lruvec, 1, 0); |
10853a039 mm: move lazily f... |
558 559 |
} } |
315601809 mm: deactivate in... |
560 |
/* |
902aaed0d mm: use pagevec t... |
561 562 563 564 |
* 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 ... |
565 |
void lru_add_drain_cpu(int cpu) |
1da177e4c Linux-2.6.12-rc2 |
566 |
{ |
13f7f7898 mm: pagevec: defe... |
567 |
struct pagevec *pvec = &per_cpu(lru_add_pvec, cpu); |
1da177e4c Linux-2.6.12-rc2 |
568 |
|
13f7f7898 mm: pagevec: defe... |
569 |
if (pagevec_count(pvec)) |
a0b8cab3b mm: remove lru pa... |
570 |
__pagevec_lru_add(pvec); |
902aaed0d mm: use pagevec t... |
571 572 573 574 575 576 577 578 579 580 |
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... |
581 |
|
cc5993bd7 mm: rename deacti... |
582 |
pvec = &per_cpu(lru_deactivate_file_pvecs, cpu); |
315601809 mm: deactivate in... |
583 |
if (pagevec_count(pvec)) |
cc5993bd7 mm: rename deacti... |
584 |
pagevec_lru_move_fn(pvec, lru_deactivate_file_fn, NULL); |
eb709b0d0 mm: batch activat... |
585 |
|
9c276cc65 mm: introduce MAD... |
586 587 588 |
pvec = &per_cpu(lru_deactivate_pvecs, cpu); if (pagevec_count(pvec)) pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL); |
f7ad2a6cb mm: move MADV_FRE... |
589 |
pvec = &per_cpu(lru_lazyfree_pvecs, cpu); |
10853a039 mm: move lazily f... |
590 |
if (pagevec_count(pvec)) |
f7ad2a6cb mm: move MADV_FRE... |
591 |
pagevec_lru_move_fn(pvec, lru_lazyfree_fn, NULL); |
10853a039 mm: move lazily f... |
592 |
|
eb709b0d0 mm: batch activat... |
593 |
activate_page_drain(cpu); |
315601809 mm: deactivate in... |
594 595 596 |
} /** |
cc5993bd7 mm: rename deacti... |
597 |
* deactivate_file_page - forcefully deactivate a file page |
315601809 mm: deactivate in... |
598 599 600 601 602 603 |
* @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 mm: rename deacti... |
604 |
void deactivate_file_page(struct page *page) |
315601809 mm: deactivate in... |
605 |
{ |
821ed6bbe mm: filter unevic... |
606 |
/* |
cc5993bd7 mm: rename deacti... |
607 608 |
* In a workload with many unevictable page such as mprotect, * unevictable page deactivation for accelerating reclaim is pointless. |
821ed6bbe mm: filter unevic... |
609 610 611 |
*/ if (PageUnevictable(page)) return; |
315601809 mm: deactivate in... |
612 |
if (likely(get_page_unless_zero(page))) { |
cc5993bd7 mm: rename deacti... |
613 |
struct pagevec *pvec = &get_cpu_var(lru_deactivate_file_pvecs); |
315601809 mm: deactivate in... |
614 |
|
8f182270d mm/swap.c: flush ... |
615 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
cc5993bd7 mm: rename deacti... |
616 617 |
pagevec_lru_move_fn(pvec, lru_deactivate_file_fn, NULL); put_cpu_var(lru_deactivate_file_pvecs); |
315601809 mm: deactivate in... |
618 |
} |
80bfed904 [PATCH] consolida... |
619 |
} |
9c276cc65 mm: introduce MAD... |
620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 |
/* * deactivate_page - deactivate a page * @page: page to deactivate * * deactivate_page() moves @page to the inactive list if @page was on the active * list and was not an unevictable page. This is done to accelerate the reclaim * of @page. */ void deactivate_page(struct page *page) { if (PageLRU(page) && PageActive(page) && !PageUnevictable(page)) { struct pagevec *pvec = &get_cpu_var(lru_deactivate_pvecs); get_page(page); if (!pagevec_add(pvec, page) || PageCompound(page)) pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL); put_cpu_var(lru_deactivate_pvecs); } } |
10853a039 mm: move lazily f... |
639 |
/** |
f7ad2a6cb mm: move MADV_FRE... |
640 |
* mark_page_lazyfree - make an anon page lazyfree |
10853a039 mm: move lazily f... |
641 642 |
* @page: page to deactivate * |
f7ad2a6cb mm: move MADV_FRE... |
643 644 |
* mark_page_lazyfree() moves @page to the inactive file list. * This is done to accelerate the reclaim of @page. |
10853a039 mm: move lazily f... |
645 |
*/ |
f7ad2a6cb mm: move MADV_FRE... |
646 |
void mark_page_lazyfree(struct page *page) |
10853a039 mm: move lazily f... |
647 |
{ |
f7ad2a6cb mm: move MADV_FRE... |
648 |
if (PageLRU(page) && PageAnon(page) && PageSwapBacked(page) && |
24c92eb7d mm: avoid marking... |
649 |
!PageSwapCache(page) && !PageUnevictable(page)) { |
f7ad2a6cb mm: move MADV_FRE... |
650 |
struct pagevec *pvec = &get_cpu_var(lru_lazyfree_pvecs); |
10853a039 mm: move lazily f... |
651 |
|
09cbfeaf1 mm, fs: get rid o... |
652 |
get_page(page); |
8f182270d mm/swap.c: flush ... |
653 |
if (!pagevec_add(pvec, page) || PageCompound(page)) |
f7ad2a6cb mm: move MADV_FRE... |
654 655 |
pagevec_lru_move_fn(pvec, lru_lazyfree_fn, NULL); put_cpu_var(lru_lazyfree_pvecs); |
10853a039 mm: move lazily f... |
656 657 |
} } |
80bfed904 [PATCH] consolida... |
658 659 |
void lru_add_drain(void) { |
f0cb3c76a mm: drain percpu ... |
660 |
lru_add_drain_cpu(get_cpu()); |
80bfed904 [PATCH] consolida... |
661 |
put_cpu(); |
1da177e4c Linux-2.6.12-rc2 |
662 |
} |
6ea183d60 mm: handle lru_ad... |
663 664 665 |
#ifdef CONFIG_SMP static DEFINE_PER_CPU(struct work_struct, lru_add_drain_work); |
c4028958b WorkStruct: make ... |
666 |
static void lru_add_drain_per_cpu(struct work_struct *dummy) |
053837fce [PATCH] mm: migra... |
667 668 669 |
{ lru_add_drain(); } |
9852a7212 mm: drop hotplug ... |
670 671 672 673 674 675 676 677 |
/* * Doesn't need any cpu hotplug locking because we do rely on per-cpu * kworkers being shut down before our page_alloc_cpu_dead callback is * executed on the offlined cpu. * Calling this function with cpu hotplug locks held can actually lead * to obscure indirect dependencies via WQ context. */ void lru_add_drain_all(void) |
053837fce [PATCH] mm: migra... |
678 |
{ |
eef1a429f mm/swap.c: piggyb... |
679 |
static seqcount_t seqcount = SEQCNT_ZERO(seqcount); |
5fbc46163 mm: make lru_add_... |
680 681 |
static DEFINE_MUTEX(lock); static struct cpumask has_work; |
eef1a429f mm/swap.c: piggyb... |
682 |
int cpu, seq; |
5fbc46163 mm: make lru_add_... |
683 |
|
ce612879d mm: move pcp and ... |
684 685 686 687 688 689 |
/* * Make sure nobody triggers this path before mm_percpu_wq is fully * initialized. */ if (WARN_ON(!mm_percpu_wq)) return; |
eef1a429f mm/swap.c: piggyb... |
690 |
seq = raw_read_seqcount_latch(&seqcount); |
5fbc46163 mm: make lru_add_... |
691 |
mutex_lock(&lock); |
eef1a429f mm/swap.c: piggyb... |
692 693 694 695 696 697 698 699 700 |
/* * Piggyback on drain started and finished while we waited for lock: * all pages pended at the time of our enter were drained from vectors. */ if (__read_seqcount_retry(&seqcount, seq)) goto done; raw_write_seqcount_latch(&seqcount); |
5fbc46163 mm: make lru_add_... |
701 702 703 704 705 706 707 |
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 mm: rename deacti... |
708 |
pagevec_count(&per_cpu(lru_deactivate_file_pvecs, cpu)) || |
9c276cc65 mm: introduce MAD... |
709 |
pagevec_count(&per_cpu(lru_deactivate_pvecs, cpu)) || |
f7ad2a6cb mm: move MADV_FRE... |
710 |
pagevec_count(&per_cpu(lru_lazyfree_pvecs, cpu)) || |
5fbc46163 mm: make lru_add_... |
711 712 |
need_activate_page_drain(cpu)) { INIT_WORK(work, lru_add_drain_per_cpu); |
ce612879d mm: move pcp and ... |
713 |
queue_work_on(cpu, mm_percpu_wq, work); |
5fbc46163 mm: make lru_add_... |
714 715 716 717 718 719 |
cpumask_set_cpu(cpu, &has_work); } } for_each_cpu(cpu, &has_work) flush_work(&per_cpu(lru_add_drain_work, cpu)); |
eef1a429f mm/swap.c: piggyb... |
720 |
done: |
5fbc46163 mm: make lru_add_... |
721 |
mutex_unlock(&lock); |
053837fce [PATCH] mm: migra... |
722 |
} |
6ea183d60 mm: handle lru_ad... |
723 724 725 726 727 728 |
#else void lru_add_drain_all(void) { lru_add_drain(); } #endif |
053837fce [PATCH] mm: migra... |
729 |
|
aabfb5729 mm: memcontrol: d... |
730 |
/** |
ea1754a08 mm, fs: remove re... |
731 |
* release_pages - batched put_page() |
aabfb5729 mm: memcontrol: d... |
732 733 |
* @pages: array of pages to release * @nr: number of pages |
1da177e4c Linux-2.6.12-rc2 |
734 |
* |
aabfb5729 mm: memcontrol: d... |
735 736 |
* 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 Linux-2.6.12-rc2 |
737 |
*/ |
c6f92f9fb mm: remove cold p... |
738 |
void release_pages(struct page **pages, int nr) |
1da177e4c Linux-2.6.12-rc2 |
739 740 |
{ int i; |
cc59850ef mm: add free_hot_... |
741 |
LIST_HEAD(pages_to_free); |
599d0c954 mm, vmscan: move ... |
742 |
struct pglist_data *locked_pgdat = NULL; |
fa9add641 mm/memcg: apply a... |
743 |
struct lruvec *lruvec; |
902aaed0d mm: use pagevec t... |
744 |
unsigned long uninitialized_var(flags); |
aabfb5729 mm: memcontrol: d... |
745 |
unsigned int uninitialized_var(lock_batch); |
1da177e4c Linux-2.6.12-rc2 |
746 |
|
1da177e4c Linux-2.6.12-rc2 |
747 748 |
for (i = 0; i < nr; i++) { struct page *page = pages[i]; |
1da177e4c Linux-2.6.12-rc2 |
749 |
|
aabfb5729 mm: memcontrol: d... |
750 751 752 |
/* * Make sure the IRQ-safe lock-holding time does not get * excessive with a continuous string of pages from the |
599d0c954 mm, vmscan: move ... |
753 |
* same pgdat. The lock is held only if pgdat != NULL. |
aabfb5729 mm: memcontrol: d... |
754 |
*/ |
599d0c954 mm, vmscan: move ... |
755 756 757 |
if (locked_pgdat && ++lock_batch == SWAP_CLUSTER_MAX) { spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); locked_pgdat = NULL; |
aabfb5729 mm: memcontrol: d... |
758 |
} |
6fcb52a56 thp: reduce usage... |
759 |
if (is_huge_zero_page(page)) |
aa88b68c3 thp: keep huge ze... |
760 |
continue; |
aa88b68c3 thp: keep huge ze... |
761 |
|
c5d6c45e9 mm/swap: fix rele... |
762 |
if (is_zone_device_page(page)) { |
df6ad6983 mm/device-public-... |
763 764 765 766 767 |
if (locked_pgdat) { spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); locked_pgdat = NULL; } |
c5d6c45e9 mm/swap: fix rele... |
768 769 770 771 772 773 |
/* * ZONE_DEVICE pages that return 'false' from * put_devmap_managed_page() do not require special * processing, and instead, expect a call to * put_page_testzero(). */ |
07d802699 mm: devmap: refac... |
774 775 |
if (page_is_devmap_managed(page)) { put_devmap_managed_page(page); |
c5d6c45e9 mm/swap: fix rele... |
776 |
continue; |
07d802699 mm: devmap: refac... |
777 |
} |
df6ad6983 mm/device-public-... |
778 |
} |
ddc58f27f mm: drop tail pag... |
779 |
page = compound_head(page); |
b5810039a [PATCH] core remo... |
780 |
if (!put_page_testzero(page)) |
1da177e4c Linux-2.6.12-rc2 |
781 |
continue; |
ddc58f27f mm: drop tail pag... |
782 |
if (PageCompound(page)) { |
599d0c954 mm, vmscan: move ... |
783 784 785 |
if (locked_pgdat) { spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); locked_pgdat = NULL; |
ddc58f27f mm: drop tail pag... |
786 787 788 789 |
} __put_compound_page(page); continue; } |
46453a6e1 [PATCH] mm: never... |
790 |
if (PageLRU(page)) { |
599d0c954 mm, vmscan: move ... |
791 |
struct pglist_data *pgdat = page_pgdat(page); |
894bc3104 Unevictable LRU I... |
792 |
|
599d0c954 mm, vmscan: move ... |
793 794 795 |
if (pgdat != locked_pgdat) { if (locked_pgdat) spin_unlock_irqrestore(&locked_pgdat->lru_lock, |
902aaed0d mm: use pagevec t... |
796 |
flags); |
aabfb5729 mm: memcontrol: d... |
797 |
lock_batch = 0; |
599d0c954 mm, vmscan: move ... |
798 799 |
locked_pgdat = pgdat; spin_lock_irqsave(&locked_pgdat->lru_lock, flags); |
46453a6e1 [PATCH] mm: never... |
800 |
} |
fa9add641 mm/memcg: apply a... |
801 |
|
599d0c954 mm, vmscan: move ... |
802 |
lruvec = mem_cgroup_page_lruvec(page, locked_pgdat); |
309381fea mm: dump page whe... |
803 |
VM_BUG_ON_PAGE(!PageLRU(page), page); |
674539115 [PATCH] mm: less ... |
804 |
__ClearPageLRU(page); |
fa9add641 mm/memcg: apply a... |
805 |
del_page_from_lru_list(page, lruvec, page_off_lru(page)); |
46453a6e1 [PATCH] mm: never... |
806 |
} |
c53954a09 mm: remove lru pa... |
807 |
/* Clear Active bit in case of parallel mark_page_accessed */ |
e3741b506 mm: do not use at... |
808 |
__ClearPageActive(page); |
629060270 mm: add PageWaite... |
809 |
__ClearPageWaiters(page); |
c53954a09 mm: remove lru pa... |
810 |
|
cc59850ef mm: add free_hot_... |
811 |
list_add(&page->lru, &pages_to_free); |
1da177e4c Linux-2.6.12-rc2 |
812 |
} |
599d0c954 mm, vmscan: move ... |
813 814 |
if (locked_pgdat) spin_unlock_irqrestore(&locked_pgdat->lru_lock, flags); |
1da177e4c Linux-2.6.12-rc2 |
815 |
|
747db954c mm: memcontrol: u... |
816 |
mem_cgroup_uncharge_list(&pages_to_free); |
2d4894b5d mm: remove cold p... |
817 |
free_unref_page_list(&pages_to_free); |
1da177e4c Linux-2.6.12-rc2 |
818 |
} |
0be8557bc fuse: use release... |
819 |
EXPORT_SYMBOL(release_pages); |
1da177e4c Linux-2.6.12-rc2 |
820 821 822 823 824 825 826 827 828 829 830 831 832 |
/* * 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 mm, pagevec: rena... |
833 |
if (!pvec->percpu_pvec_drained) { |
d9ed0d08b mm: only drain pe... |
834 |
lru_add_drain(); |
7f0b5fb95 mm, pagevec: rena... |
835 |
pvec->percpu_pvec_drained = true; |
d9ed0d08b mm: only drain pe... |
836 |
} |
c6f92f9fb mm: remove cold p... |
837 |
release_pages(pvec->pages, pagevec_count(pvec)); |
1da177e4c Linux-2.6.12-rc2 |
838 839 |
pagevec_reinit(pvec); } |
7f2857018 Export __pagevec_... |
840 |
EXPORT_SYMBOL(__pagevec_release); |
12d271078 memcg: fix split_... |
841 |
#ifdef CONFIG_TRANSPARENT_HUGEPAGE |
71e3aac07 thp: transparent ... |
842 |
/* used by __split_huge_page_refcount() */ |
fa9add641 mm/memcg: apply a... |
843 |
void lru_add_page_tail(struct page *page, struct page *page_tail, |
5bc7b8aca mm: thp: add spli... |
844 |
struct lruvec *lruvec, struct list_head *list) |
71e3aac07 thp: transparent ... |
845 |
{ |
71e3aac07 thp: transparent ... |
846 |
const int file = 0; |
71e3aac07 thp: transparent ... |
847 |
|
309381fea mm: dump page whe... |
848 849 850 |
VM_BUG_ON_PAGE(!PageHead(page), page); VM_BUG_ON_PAGE(PageCompound(page_tail), page); VM_BUG_ON_PAGE(PageLRU(page_tail), page); |
35f3aa39f mm: Replace spin_... |
851 |
lockdep_assert_held(&lruvec_pgdat(lruvec)->lru_lock); |
71e3aac07 thp: transparent ... |
852 |
|
5bc7b8aca mm: thp: add spli... |
853 854 |
if (!list) SetPageLRU(page_tail); |
71e3aac07 thp: transparent ... |
855 |
|
12d271078 memcg: fix split_... |
856 857 |
if (likely(PageLRU(page))) list_add_tail(&page_tail->lru, &page->lru); |
5bc7b8aca mm: thp: add spli... |
858 859 860 861 862 |
else if (list) { /* page reclaim is reclaiming a huge page */ get_page(page_tail); list_add_tail(&page_tail->lru, list); } else { |
12d271078 memcg: fix split_... |
863 864 865 866 |
/* * Head page has not yet been counted, as an hpage, * so we must account for each subpage individually. * |
e7a1aaf28 mm: replace list_... |
867 868 |
* Put page_tail on the list at the correct position * so they all end up in order. |
12d271078 memcg: fix split_... |
869 |
*/ |
e7a1aaf28 mm: replace list_... |
870 871 |
add_page_to_lru_list_tail(page_tail, lruvec, page_lru(page_tail)); |
71e3aac07 thp: transparent ... |
872 |
} |
7512102cf memcg: fix GPF wh... |
873 874 |
if (!PageUnevictable(page)) |
e180cf806 thp, mm: avoid Pa... |
875 |
update_page_reclaim_stat(lruvec, file, PageActive(page_tail)); |
71e3aac07 thp: transparent ... |
876 |
} |
12d271078 memcg: fix split_... |
877 |
#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ |
71e3aac07 thp: transparent ... |
878 |
|
fa9add641 mm/memcg: apply a... |
879 880 |
static void __pagevec_lru_add_fn(struct page *page, struct lruvec *lruvec, void *arg) |
3dd7ae8ec mm: simplify code... |
881 |
{ |
9c4e6b1a7 mm, mlock, vmscan... |
882 883 |
enum lru_list lru; int was_unevictable = TestClearPageUnevictable(page); |
3dd7ae8ec mm: simplify code... |
884 |
|
309381fea mm: dump page whe... |
885 |
VM_BUG_ON_PAGE(PageLRU(page), page); |
3dd7ae8ec mm: simplify code... |
886 887 |
SetPageLRU(page); |
9c4e6b1a7 mm, mlock, vmscan... |
888 889 |
/* * Page becomes evictable in two ways: |
dae966dc8 mm/swap.c: __page... |
890 |
* 1) Within LRU lock [munlock_vma_page() and __munlock_pagevec()]. |
9c4e6b1a7 mm, mlock, vmscan... |
891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 |
* 2) Before acquiring LRU lock to put the page to correct LRU and then * a) do PageLRU check with lock [check_move_unevictable_pages] * b) do PageLRU check before lock [clear_page_mlock] * * (1) & (2a) are ok as LRU lock will serialize them. For (2b), we need * following strict ordering: * * #0: __pagevec_lru_add_fn #1: clear_page_mlock * * SetPageLRU() TestClearPageMlocked() * smp_mb() // explicit ordering // above provides strict * // ordering * PageMlocked() PageLRU() * * * if '#1' does not observe setting of PG_lru by '#0' and fails * isolation, the explicit barrier will make sure that page_evictable * check will put the page in correct LRU. Without smp_mb(), SetPageLRU * can be reordered after PageMlocked check and can make '#1' to fail * the isolation of the page whose Mlocked bit is cleared (#0 is also * looking at the same page) and the evictable page will be stranded * in an unevictable LRU. */ smp_mb(); if (page_evictable(page)) { lru = page_lru(page); update_page_reclaim_stat(lruvec, page_is_file_cache(page), PageActive(page)); if (was_unevictable) count_vm_event(UNEVICTABLE_PGRESCUED); } else { lru = LRU_UNEVICTABLE; ClearPageActive(page); SetPageUnevictable(page); if (!was_unevictable) count_vm_event(UNEVICTABLE_PGCULLED); } |
fa9add641 mm/memcg: apply a... |
929 |
add_page_to_lru_list(page, lruvec, lru); |
24b7e5819 mm: pagemap: avoi... |
930 |
trace_mm_lru_insertion(page, lru); |
3dd7ae8ec mm: simplify code... |
931 |
} |
1da177e4c Linux-2.6.12-rc2 |
932 |
/* |
1da177e4c Linux-2.6.12-rc2 |
933 934 935 |
* Add the passed pages to the LRU, then drop the caller's refcount * on them. Reinitialises the caller's pagevec. */ |
a0b8cab3b mm: remove lru pa... |
936 |
void __pagevec_lru_add(struct pagevec *pvec) |
1da177e4c Linux-2.6.12-rc2 |
937 |
{ |
a0b8cab3b mm: remove lru pa... |
938 |
pagevec_lru_move_fn(pvec, __pagevec_lru_add_fn, NULL); |
1da177e4c Linux-2.6.12-rc2 |
939 |
} |
5095ae837 mm: fewer undersc... |
940 |
EXPORT_SYMBOL(__pagevec_lru_add); |
1da177e4c Linux-2.6.12-rc2 |
941 |
|
1da177e4c Linux-2.6.12-rc2 |
942 |
/** |
0cd6144aa mm + fs: prepare ... |
943 944 945 946 |
* pagevec_lookup_entries - gang pagecache lookup * @pvec: Where the resulting entries are placed * @mapping: The address_space to search * @start: The starting entry index |
cb6f0f348 mm/swap.c: make f... |
947 |
* @nr_entries: The maximum number of pages |
0cd6144aa mm + fs: prepare ... |
948 949 950 |
* @indices: The cache indices corresponding to the entries in @pvec * * pagevec_lookup_entries() will search for and return a group of up |
f144c390f mm: docs: fix par... |
951 |
* to @nr_pages pages and shadow entries in the mapping. All |
0cd6144aa mm + fs: prepare ... |
952 953 954 955 956 957 958 959 960 961 962 963 |
* 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, |
e02a9f048 mm/swap.c: make f... |
964 |
pgoff_t start, unsigned nr_entries, |
0cd6144aa mm + fs: prepare ... |
965 966 |
pgoff_t *indices) { |
e02a9f048 mm/swap.c: make f... |
967 |
pvec->nr = find_get_entries(mapping, start, nr_entries, |
0cd6144aa mm + fs: prepare ... |
968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 |
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]; |
3159f943a xarray: Replace e... |
987 |
if (!xa_is_value(page)) |
0cd6144aa mm + fs: prepare ... |
988 989 990 991 992 993 |
pvec->pages[j++] = page; } pvec->nr = j; } /** |
b947cee4b mm: implement fin... |
994 |
* pagevec_lookup_range - gang pagecache lookup |
1da177e4c Linux-2.6.12-rc2 |
995 996 997 |
* @pvec: Where the resulting pages are placed * @mapping: The address_space to search * @start: The starting page index |
b947cee4b mm: implement fin... |
998 |
* @end: The final page index |
1da177e4c Linux-2.6.12-rc2 |
999 |
* |
e02a9f048 mm/swap.c: make f... |
1000 |
* pagevec_lookup_range() will search for & return a group of up to PAGEVEC_SIZE |
b947cee4b mm: implement fin... |
1001 1002 |
* pages in the mapping starting from index @start and upto index @end * (inclusive). The pages are placed in @pvec. pagevec_lookup() takes a |
1da177e4c Linux-2.6.12-rc2 |
1003 1004 1005 |
* reference against the pages in @pvec. * * The search returns a group of mapping-contiguous pages with ascending |
d72dc8a25 mm: make pagevec_... |
1006 1007 |
* 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 Linux-2.6.12-rc2 |
1008 |
* |
b947cee4b mm: implement fin... |
1009 |
* pagevec_lookup_range() returns the number of pages which were found. If this |
e02a9f048 mm/swap.c: make f... |
1010 |
* number is smaller than PAGEVEC_SIZE, the end of specified range has been |
b947cee4b mm: implement fin... |
1011 |
* reached. |
1da177e4c Linux-2.6.12-rc2 |
1012 |
*/ |
b947cee4b mm: implement fin... |
1013 |
unsigned pagevec_lookup_range(struct pagevec *pvec, |
397162ffa mm: remove nr_pag... |
1014 |
struct address_space *mapping, pgoff_t *start, pgoff_t end) |
1da177e4c Linux-2.6.12-rc2 |
1015 |
{ |
397162ffa mm: remove nr_pag... |
1016 |
pvec->nr = find_get_pages_range(mapping, start, end, PAGEVEC_SIZE, |
b947cee4b mm: implement fin... |
1017 |
pvec->pages); |
1da177e4c Linux-2.6.12-rc2 |
1018 1019 |
return pagevec_count(pvec); } |
b947cee4b mm: implement fin... |
1020 |
EXPORT_SYMBOL(pagevec_lookup_range); |
78539fdfa [XFS] Export page... |
1021 |
|
72b045aec mm: implement fin... |
1022 1023 |
unsigned pagevec_lookup_range_tag(struct pagevec *pvec, struct address_space *mapping, pgoff_t *index, pgoff_t end, |
10bbd2358 pagevec: Use xa_m... |
1024 |
xa_mark_t tag) |
1da177e4c Linux-2.6.12-rc2 |
1025 |
{ |
72b045aec mm: implement fin... |
1026 |
pvec->nr = find_get_pages_range_tag(mapping, index, end, tag, |
67fd707f4 mm: remove nr_pag... |
1027 |
PAGEVEC_SIZE, pvec->pages); |
1da177e4c Linux-2.6.12-rc2 |
1028 1029 |
return pagevec_count(pvec); } |
72b045aec mm: implement fin... |
1030 |
EXPORT_SYMBOL(pagevec_lookup_range_tag); |
1da177e4c Linux-2.6.12-rc2 |
1031 |
|
93d3b7140 mm: add variant o... |
1032 1033 |
unsigned pagevec_lookup_range_nr_tag(struct pagevec *pvec, struct address_space *mapping, pgoff_t *index, pgoff_t end, |
10bbd2358 pagevec: Use xa_m... |
1034 |
xa_mark_t tag, unsigned max_pages) |
93d3b7140 mm: add variant o... |
1035 1036 1037 1038 1039 1040 |
{ 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 Linux-2.6.12-rc2 |
1041 1042 1043 1044 1045 |
/* * Perform any setup for the swap system */ void __init swap_setup(void) { |
ca79b0c21 mm: convert total... |
1046 |
unsigned long megs = totalram_pages() >> (20 - PAGE_SHIFT); |
e0bf68dde mm: bdi init hooks |
1047 |
|
1da177e4c Linux-2.6.12-rc2 |
1048 1049 1050 1051 1052 1053 1054 1055 1056 |
/* 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 |
1057 |
} |
07d802699 mm: devmap: refac... |
1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 |
#ifdef CONFIG_DEV_PAGEMAP_OPS void put_devmap_managed_page(struct page *page) { int count; if (WARN_ON_ONCE(!page_is_devmap_managed(page))) return; count = page_ref_dec_return(page); /* * devmap page refcounts are 1-based, rather than 0-based: if * refcount is 1, then the page is free and the refcount is * stable because nobody holds a reference on the page. */ if (count == 1) free_devmap_managed_page(page); else if (!count) __put_page(page); } EXPORT_SYMBOL(put_devmap_managed_page); #endif |