Blame view
mm/vmscan.c
80.6 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 |
/* * linux/mm/vmscan.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds * * Swap reorganised 29.12.95, Stephen Tweedie. * kswapd added: 7.1.96 sct * Removed kswapd_ctl limits, and swap out as many pages as needed * to bring the system back to freepages.high: 2.4.97, Rik van Riel. * Zone aware kswapd started 02/00, Kanoj Sarcar (kanoj@sgi.com). * Multiqueue VM started 5.8.00, Rik van Riel. */ #include <linux/mm.h> #include <linux/module.h> #include <linux/slab.h> #include <linux/kernel_stat.h> #include <linux/swap.h> #include <linux/pagemap.h> #include <linux/init.h> #include <linux/highmem.h> |
e129b5c23
|
22 |
#include <linux/vmstat.h> |
1da177e4c
|
23 24 25 26 27 28 29 30 31 32 33 34 35 36 |
#include <linux/file.h> #include <linux/writeback.h> #include <linux/blkdev.h> #include <linux/buffer_head.h> /* for try_to_release_page(), buffer_heads_over_limit */ #include <linux/mm_inline.h> #include <linux/pagevec.h> #include <linux/backing-dev.h> #include <linux/rmap.h> #include <linux/topology.h> #include <linux/cpu.h> #include <linux/cpuset.h> #include <linux/notifier.h> #include <linux/rwsem.h> |
248a0301e
|
37 |
#include <linux/delay.h> |
3218ae14b
|
38 |
#include <linux/kthread.h> |
7dfb71030
|
39 |
#include <linux/freezer.h> |
66e1707bc
|
40 |
#include <linux/memcontrol.h> |
873b47717
|
41 |
#include <linux/delayacct.h> |
af936a160
|
42 |
#include <linux/sysctl.h> |
1da177e4c
|
43 44 45 46 47 |
#include <asm/tlbflush.h> #include <asm/div64.h> #include <linux/swapops.h> |
0f8053a50
|
48 |
#include "internal.h" |
1da177e4c
|
49 |
struct scan_control { |
1da177e4c
|
50 51 |
/* Incremented by the number of inactive pages that were scanned */ unsigned long nr_scanned; |
a79311c14
|
52 53 |
/* Number of pages freed so far during a call to shrink_zones() */ unsigned long nr_reclaimed; |
22fba3354
|
54 55 |
/* How many pages shrink_list() should reclaim */ unsigned long nr_to_reclaim; |
7b51755c3
|
56 |
unsigned long hibernation_mode; |
1da177e4c
|
57 |
/* This context's GFP mask */ |
6daa0e286
|
58 |
gfp_t gfp_mask; |
1da177e4c
|
59 60 |
int may_writepage; |
a6dc60f89
|
61 62 |
/* Can mapped pages be reclaimed? */ int may_unmap; |
f1fd1067e
|
63 |
|
2e2e42598
|
64 65 |
/* Can pages be swapped as part of reclaim? */ int may_swap; |
d6277db4a
|
66 |
int swappiness; |
408d85441
|
67 68 |
int all_unreclaimable; |
5ad333eb6
|
69 70 |
int order; |
66e1707bc
|
71 72 73 |
/* Which cgroup do we reclaim from */ struct mem_cgroup *mem_cgroup; |
327c0e968
|
74 75 76 77 78 |
/* * Nodemask of nodes allowed by the caller. If NULL, all nodes * are scanned. */ nodemask_t *nodemask; |
66e1707bc
|
79 80 81 82 |
/* Pluggable isolate pages callback */ unsigned long (*isolate_pages)(unsigned long nr, struct list_head *dst, unsigned long *scanned, int order, int mode, struct zone *z, struct mem_cgroup *mem_cont, |
4f98a2fee
|
83 |
int active, int file); |
1da177e4c
|
84 |
}; |
1da177e4c
|
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 |
#define lru_to_page(_head) (list_entry((_head)->prev, struct page, lru)) #ifdef ARCH_HAS_PREFETCH #define prefetch_prev_lru_page(_page, _base, _field) \ do { \ if ((_page)->lru.prev != _base) { \ struct page *prev; \ \ prev = lru_to_page(&(_page->lru)); \ prefetch(&prev->_field); \ } \ } while (0) #else #define prefetch_prev_lru_page(_page, _base, _field) do { } while (0) #endif #ifdef ARCH_HAS_PREFETCHW #define prefetchw_prev_lru_page(_page, _base, _field) \ do { \ if ((_page)->lru.prev != _base) { \ struct page *prev; \ \ prev = lru_to_page(&(_page->lru)); \ prefetchw(&prev->_field); \ } \ } while (0) #else #define prefetchw_prev_lru_page(_page, _base, _field) do { } while (0) #endif /* * From 0 .. 100. Higher means more swappy. */ int vm_swappiness = 60; |
bd1e22b8e
|
119 |
long vm_total_pages; /* The total number of pages which the VM controls */ |
1da177e4c
|
120 121 122 |
static LIST_HEAD(shrinker_list); static DECLARE_RWSEM(shrinker_rwsem); |
00f0b8259
|
123 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR |
e72e2bd67
|
124 |
#define scanning_global_lru(sc) (!(sc)->mem_cgroup) |
91a45470f
|
125 |
#else |
e72e2bd67
|
126 |
#define scanning_global_lru(sc) (1) |
91a45470f
|
127 |
#endif |
6e9015716
|
128 129 130 |
static struct zone_reclaim_stat *get_reclaim_stat(struct zone *zone, struct scan_control *sc) { |
e72e2bd67
|
131 |
if (!scanning_global_lru(sc)) |
3e2f41f1f
|
132 |
return mem_cgroup_get_reclaim_stat(sc->mem_cgroup, zone); |
6e9015716
|
133 134 |
return &zone->reclaim_stat; } |
0b2176763
|
135 136 |
static unsigned long zone_nr_lru_pages(struct zone *zone, struct scan_control *sc, enum lru_list lru) |
c9f299d98
|
137 |
{ |
e72e2bd67
|
138 |
if (!scanning_global_lru(sc)) |
a3d8e0549
|
139 |
return mem_cgroup_zone_nr_pages(sc->mem_cgroup, zone, lru); |
c9f299d98
|
140 141 |
return zone_page_state(zone, NR_LRU_BASE + lru); } |
1da177e4c
|
142 143 144 |
/* * Add a shrinker callback to be called from the vm */ |
8e1f936b7
|
145 |
void register_shrinker(struct shrinker *shrinker) |
1da177e4c
|
146 |
{ |
8e1f936b7
|
147 148 149 150 |
shrinker->nr = 0; down_write(&shrinker_rwsem); list_add_tail(&shrinker->list, &shrinker_list); up_write(&shrinker_rwsem); |
1da177e4c
|
151 |
} |
8e1f936b7
|
152 |
EXPORT_SYMBOL(register_shrinker); |
1da177e4c
|
153 154 155 156 |
/* * Remove one */ |
8e1f936b7
|
157 |
void unregister_shrinker(struct shrinker *shrinker) |
1da177e4c
|
158 159 160 161 |
{ down_write(&shrinker_rwsem); list_del(&shrinker->list); up_write(&shrinker_rwsem); |
1da177e4c
|
162 |
} |
8e1f936b7
|
163 |
EXPORT_SYMBOL(unregister_shrinker); |
1da177e4c
|
164 165 166 167 168 169 170 171 172 173 |
#define SHRINK_BATCH 128 /* * Call the shrink functions to age shrinkable caches * * Here we assume it costs one seek to replace a lru page and that it also * takes a seek to recreate a cache object. With this in mind we age equal * percentages of the lru and ageable caches. This should balance the seeks * generated by these structures. * |
183ff22bb
|
174 |
* If the vm encountered mapped pages on the LRU it increase the pressure on |
1da177e4c
|
175 176 177 178 179 180 181 |
* slab to avoid swapping. * * We do weird things to avoid (scanned*seeks*entries) overflowing 32 bits. * * `lru_pages' represents the number of on-LRU pages in all the zones which * are eligible for the caller's allocation attempt. It is used for balancing * slab reclaim versus page reclaim. |
b15e0905f
|
182 183 |
* * Returns the number of slab objects which we shrunk. |
1da177e4c
|
184 |
*/ |
69e05944a
|
185 186 |
unsigned long shrink_slab(unsigned long scanned, gfp_t gfp_mask, unsigned long lru_pages) |
1da177e4c
|
187 188 |
{ struct shrinker *shrinker; |
69e05944a
|
189 |
unsigned long ret = 0; |
1da177e4c
|
190 191 192 193 194 |
if (scanned == 0) scanned = SWAP_CLUSTER_MAX; if (!down_read_trylock(&shrinker_rwsem)) |
b15e0905f
|
195 |
return 1; /* Assume we'll be able to shrink next time */ |
1da177e4c
|
196 197 198 199 |
list_for_each_entry(shrinker, &shrinker_list, list) { unsigned long long delta; unsigned long total_scan; |
8e1f936b7
|
200 |
unsigned long max_pass = (*shrinker->shrink)(0, gfp_mask); |
1da177e4c
|
201 202 |
delta = (4 * scanned) / shrinker->seeks; |
ea164d73a
|
203 |
delta *= max_pass; |
1da177e4c
|
204 205 |
do_div(delta, lru_pages + 1); shrinker->nr += delta; |
ea164d73a
|
206 |
if (shrinker->nr < 0) { |
88c3bd707
|
207 208 209 210 |
printk(KERN_ERR "shrink_slab: %pF negative objects to " "delete nr=%ld ", shrinker->shrink, shrinker->nr); |
ea164d73a
|
211 212 213 214 215 216 217 218 219 220 |
shrinker->nr = max_pass; } /* * Avoid risking looping forever due to too large nr value: * never try to free more than twice the estimate number of * freeable entries. */ if (shrinker->nr > max_pass * 2) shrinker->nr = max_pass * 2; |
1da177e4c
|
221 222 223 224 225 226 227 |
total_scan = shrinker->nr; shrinker->nr = 0; while (total_scan >= SHRINK_BATCH) { long this_scan = SHRINK_BATCH; int shrink_ret; |
b15e0905f
|
228 |
int nr_before; |
1da177e4c
|
229 |
|
8e1f936b7
|
230 231 |
nr_before = (*shrinker->shrink)(0, gfp_mask); shrink_ret = (*shrinker->shrink)(this_scan, gfp_mask); |
1da177e4c
|
232 233 |
if (shrink_ret == -1) break; |
b15e0905f
|
234 235 |
if (shrink_ret < nr_before) ret += nr_before - shrink_ret; |
f8891e5e1
|
236 |
count_vm_events(SLABS_SCANNED, this_scan); |
1da177e4c
|
237 238 239 240 241 242 243 244 |
total_scan -= this_scan; cond_resched(); } shrinker->nr += total_scan; } up_read(&shrinker_rwsem); |
b15e0905f
|
245 |
return ret; |
1da177e4c
|
246 |
} |
1da177e4c
|
247 248 |
static inline int is_page_cache_freeable(struct page *page) { |
ceddc3a52
|
249 250 251 252 253 |
/* * A freeable page cache page is referenced only by the caller * that isolated the page, the page cache radix tree and * optional buffer heads at page->private. */ |
edcf4748c
|
254 |
return page_count(page) - page_has_private(page) == 2; |
1da177e4c
|
255 256 257 258 |
} static int may_write_to_queue(struct backing_dev_info *bdi) { |
930d91525
|
259 |
if (current->flags & PF_SWAPWRITE) |
1da177e4c
|
260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 |
return 1; if (!bdi_write_congested(bdi)) return 1; if (bdi == current->backing_dev_info) return 1; return 0; } /* * We detected a synchronous write error writing a page out. Probably * -ENOSPC. We need to propagate that into the address_space for a subsequent * fsync(), msync() or close(). * * The tricky part is that after writepage we cannot touch the mapping: nothing * prevents it from being freed up. But we have a ref on the page and once * that page is locked, the mapping is pinned. * * We're allowed to run sleeping lock_page() here because we know the caller has * __GFP_FS. */ static void handle_write_error(struct address_space *mapping, struct page *page, int error) { lock_page(page); |
3e9f45bd1
|
284 285 |
if (page_mapping(page) == mapping) mapping_set_error(mapping, error); |
1da177e4c
|
286 287 |
unlock_page(page); } |
c661b078f
|
288 289 290 291 292 |
/* Request for sync pageout. */ enum pageout_io { PAGEOUT_IO_ASYNC, PAGEOUT_IO_SYNC, }; |
04e62a29b
|
293 294 295 296 297 298 299 300 301 302 303 |
/* possible outcome of pageout() */ typedef enum { /* failed to write page out, page is locked */ PAGE_KEEP, /* move page to the active list, page is locked */ PAGE_ACTIVATE, /* page has been sent to the disk successfully, page is unlocked */ PAGE_SUCCESS, /* page is clean and locked */ PAGE_CLEAN, } pageout_t; |
1da177e4c
|
304 |
/* |
1742f19fa
|
305 306 |
* pageout is called by shrink_page_list() for each dirty page. * Calls ->writepage(). |
1da177e4c
|
307 |
*/ |
c661b078f
|
308 309 |
static pageout_t pageout(struct page *page, struct address_space *mapping, enum pageout_io sync_writeback) |
1da177e4c
|
310 311 312 313 314 315 316 317 |
{ /* * If the page is dirty, only perform writeback if that write * will be non-blocking. To prevent this allocation from being * stalled by pagecache activity. But note that there may be * stalls if we need to run get_block(). We could test * PagePrivate for that. * |
6aceb53be
|
318 |
* If this process is currently in __generic_file_aio_write() against |
1da177e4c
|
319 320 321 322 323 324 325 |
* this page's queue, we can perform writeback even if that * will block. * * If the page is swapcache, write it back even if that would * block, for some throttling. This happens by accident, because * swap_backing_dev_info is bust: it doesn't reflect the * congestion state of the swapdevs. Easy to fix, if needed. |
1da177e4c
|
326 327 328 329 330 331 332 333 |
*/ if (!is_page_cache_freeable(page)) return PAGE_KEEP; if (!mapping) { /* * Some data journaling orphaned pages can have * page->mapping == NULL while being dirty with clean buffers. */ |
266cf658e
|
334 |
if (page_has_private(page)) { |
1da177e4c
|
335 336 |
if (try_to_free_buffers(page)) { ClearPageDirty(page); |
d40cee245
|
337 338 |
printk("%s: orphaned page ", __func__); |
1da177e4c
|
339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 |
return PAGE_CLEAN; } } return PAGE_KEEP; } if (mapping->a_ops->writepage == NULL) return PAGE_ACTIVATE; if (!may_write_to_queue(mapping->backing_dev_info)) return PAGE_KEEP; if (clear_page_dirty_for_io(page)) { int res; struct writeback_control wbc = { .sync_mode = WB_SYNC_NONE, .nr_to_write = SWAP_CLUSTER_MAX, |
111ebb6e6
|
354 355 |
.range_start = 0, .range_end = LLONG_MAX, |
1da177e4c
|
356 357 358 359 360 361 362 363 |
.nonblocking = 1, .for_reclaim = 1, }; SetPageReclaim(page); res = mapping->a_ops->writepage(page, &wbc); if (res < 0) handle_write_error(mapping, page, res); |
994fc28c7
|
364 |
if (res == AOP_WRITEPAGE_ACTIVATE) { |
1da177e4c
|
365 366 367 |
ClearPageReclaim(page); return PAGE_ACTIVATE; } |
c661b078f
|
368 369 370 371 372 373 374 375 |
/* * Wait on writeback if requested to. This happens when * direct reclaiming a large contiguous area and the * first attempt to free a range of pages fails. */ if (PageWriteback(page) && sync_writeback == PAGEOUT_IO_SYNC) wait_on_page_writeback(page); |
1da177e4c
|
376 377 378 379 |
if (!PageWriteback(page)) { /* synchronous write or broken a_ops? */ ClearPageReclaim(page); } |
e129b5c23
|
380 |
inc_zone_page_state(page, NR_VMSCAN_WRITE); |
1da177e4c
|
381 382 383 384 385 |
return PAGE_SUCCESS; } return PAGE_CLEAN; } |
a649fd927
|
386 |
/* |
e286781d5
|
387 388 |
* Same as remove_mapping, but if the page is removed from the mapping, it * gets returned with a refcount of 0. |
a649fd927
|
389 |
*/ |
e286781d5
|
390 |
static int __remove_mapping(struct address_space *mapping, struct page *page) |
49d2e9cc4
|
391 |
{ |
28e4d965e
|
392 393 |
BUG_ON(!PageLocked(page)); BUG_ON(mapping != page_mapping(page)); |
49d2e9cc4
|
394 |
|
19fd62312
|
395 |
spin_lock_irq(&mapping->tree_lock); |
49d2e9cc4
|
396 |
/* |
0fd0e6b05
|
397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 |
* The non racy check for a busy page. * * Must be careful with the order of the tests. When someone has * a ref to the page, it may be possible that they dirty it then * drop the reference. So if PageDirty is tested before page_count * here, then the following race may occur: * * get_user_pages(&page); * [user mapping goes away] * write_to(page); * !PageDirty(page) [good] * SetPageDirty(page); * put_page(page); * !page_count(page) [good, discard it] * * [oops, our write_to data is lost] * * Reversing the order of the tests ensures such a situation cannot * escape unnoticed. The smp_rmb is needed to ensure the page->flags * load is not satisfied before that of page->_count. * * Note that if SetPageDirty is always performed via set_page_dirty, * and thus under tree_lock, then this ordering is not required. |
49d2e9cc4
|
420 |
*/ |
e286781d5
|
421 |
if (!page_freeze_refs(page, 2)) |
49d2e9cc4
|
422 |
goto cannot_free; |
e286781d5
|
423 424 425 |
/* note: atomic_cmpxchg in page_freeze_refs provides the smp_rmb */ if (unlikely(PageDirty(page))) { page_unfreeze_refs(page, 2); |
49d2e9cc4
|
426 |
goto cannot_free; |
e286781d5
|
427 |
} |
49d2e9cc4
|
428 429 430 431 |
if (PageSwapCache(page)) { swp_entry_t swap = { .val = page_private(page) }; __delete_from_swap_cache(page); |
19fd62312
|
432 |
spin_unlock_irq(&mapping->tree_lock); |
cb4b86ba4
|
433 |
swapcache_free(swap, page); |
e286781d5
|
434 435 |
} else { __remove_from_page_cache(page); |
19fd62312
|
436 |
spin_unlock_irq(&mapping->tree_lock); |
e767e0561
|
437 |
mem_cgroup_uncharge_cache_page(page); |
49d2e9cc4
|
438 |
} |
49d2e9cc4
|
439 440 441 |
return 1; cannot_free: |
19fd62312
|
442 |
spin_unlock_irq(&mapping->tree_lock); |
49d2e9cc4
|
443 444 |
return 0; } |
1da177e4c
|
445 |
/* |
e286781d5
|
446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 |
* Attempt to detach a locked page from its ->mapping. If it is dirty or if * someone else has a ref on the page, abort and return 0. If it was * successfully detached, return 1. Assumes the caller has a single ref on * this page. */ int remove_mapping(struct address_space *mapping, struct page *page) { if (__remove_mapping(mapping, page)) { /* * Unfreezing the refcount with 1 rather than 2 effectively * drops the pagecache ref for us without requiring another * atomic operation. */ page_unfreeze_refs(page, 1); return 1; } return 0; } |
894bc3104
|
464 465 466 467 468 469 470 471 472 |
/** * putback_lru_page - put previously isolated page onto appropriate LRU list * @page: page to be put back to appropriate lru list * * Add previously isolated @page to appropriate LRU list. * Page may still be unevictable for other reasons. * * lru_lock must not be held, interrupts must be enabled. */ |
894bc3104
|
473 474 475 476 |
void putback_lru_page(struct page *page) { int lru; int active = !!TestClearPageActive(page); |
bbfd28eee
|
477 |
int was_unevictable = PageUnevictable(page); |
894bc3104
|
478 479 480 481 482 483 484 485 486 487 488 489 490 |
VM_BUG_ON(PageLRU(page)); redo: ClearPageUnevictable(page); if (page_evictable(page, NULL)) { /* * For evictable pages, we can use the cache. * In event of a race, worst case is we end up with an * unevictable page on [in]active list. * We know how to handle that. */ |
401a8e1c1
|
491 |
lru = active + page_lru_base_type(page); |
894bc3104
|
492 493 494 495 496 497 498 499 |
lru_cache_add_lru(page, lru); } else { /* * Put unevictable pages directly on zone's unevictable * list. */ lru = LRU_UNEVICTABLE; add_page_to_unevictable_list(page); |
6a7b95481
|
500 501 502 503 504 505 506 507 508 509 |
/* * When racing with an mlock clearing (page is * unlocked), make sure that if the other thread does * not observe our setting of PG_lru and fails * isolation, we see PG_mlocked cleared below and move * the page back to the evictable list. * * The other side is TestClearPageMlocked(). */ smp_mb(); |
894bc3104
|
510 |
} |
894bc3104
|
511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 |
/* * page's status can change while we move it among lru. If an evictable * page is on unevictable list, it never be freed. To avoid that, * check after we added it to the list, again. */ if (lru == LRU_UNEVICTABLE && page_evictable(page, NULL)) { if (!isolate_lru_page(page)) { put_page(page); goto redo; } /* This means someone else dropped this page from LRU * So, it will be freed or putback to LRU again. There is * nothing to do here. */ } |
bbfd28eee
|
527 528 529 530 |
if (was_unevictable && lru != LRU_UNEVICTABLE) count_vm_event(UNEVICTABLE_PGRESCUED); else if (!was_unevictable && lru == LRU_UNEVICTABLE) count_vm_event(UNEVICTABLE_PGCULLED); |
894bc3104
|
531 532 |
put_page(page); /* drop ref from isolate */ } |
dfc8d636c
|
533 534 535 |
enum page_references { PAGEREF_RECLAIM, PAGEREF_RECLAIM_CLEAN, |
645747462
|
536 |
PAGEREF_KEEP, |
dfc8d636c
|
537 538 539 540 541 542 |
PAGEREF_ACTIVATE, }; static enum page_references page_check_references(struct page *page, struct scan_control *sc) { |
645747462
|
543 |
int referenced_ptes, referenced_page; |
dfc8d636c
|
544 |
unsigned long vm_flags; |
dfc8d636c
|
545 |
|
645747462
|
546 547 |
referenced_ptes = page_referenced(page, 1, sc->mem_cgroup, &vm_flags); referenced_page = TestClearPageReferenced(page); |
dfc8d636c
|
548 549 550 551 552 553 554 555 556 557 558 |
/* Lumpy reclaim - ignore references */ if (sc->order > PAGE_ALLOC_COSTLY_ORDER) return PAGEREF_RECLAIM; /* * Mlock lost the isolation race with us. Let try_to_unmap() * move the page to the unevictable list. */ if (vm_flags & VM_LOCKED) return PAGEREF_RECLAIM; |
645747462
|
559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 |
if (referenced_ptes) { if (PageAnon(page)) return PAGEREF_ACTIVATE; /* * All mapped pages start out with page table * references from the instantiating fault, so we need * to look twice if a mapped file page is used more * than once. * * Mark it and spare it for another trip around the * inactive list. Another page table reference will * lead to its activation. * * Note: the mark is set for activated pages as well * so that recently deactivated but used pages are * quickly recovered. */ SetPageReferenced(page); if (referenced_page) return PAGEREF_ACTIVATE; return PAGEREF_KEEP; } |
dfc8d636c
|
583 584 |
/* Reclaim if clean, defer dirty pages to writeback */ |
645747462
|
585 586 587 588 |
if (referenced_page) return PAGEREF_RECLAIM_CLEAN; return PAGEREF_RECLAIM; |
dfc8d636c
|
589 |
} |
e286781d5
|
590 |
/* |
1742f19fa
|
591 |
* shrink_page_list() returns the number of reclaimed pages |
1da177e4c
|
592 |
*/ |
1742f19fa
|
593 |
static unsigned long shrink_page_list(struct list_head *page_list, |
c661b078f
|
594 595 |
struct scan_control *sc, enum pageout_io sync_writeback) |
1da177e4c
|
596 597 598 599 |
{ LIST_HEAD(ret_pages); struct pagevec freed_pvec; int pgactivate = 0; |
05ff51376
|
600 |
unsigned long nr_reclaimed = 0; |
1da177e4c
|
601 602 603 604 605 |
cond_resched(); pagevec_init(&freed_pvec, 1); while (!list_empty(page_list)) { |
dfc8d636c
|
606 |
enum page_references references; |
1da177e4c
|
607 608 609 |
struct address_space *mapping; struct page *page; int may_enter_fs; |
1da177e4c
|
610 611 612 613 614 |
cond_resched(); page = lru_to_page(page_list); list_del(&page->lru); |
529ae9aaa
|
615 |
if (!trylock_page(page)) |
1da177e4c
|
616 |
goto keep; |
725d704ec
|
617 |
VM_BUG_ON(PageActive(page)); |
1da177e4c
|
618 619 |
sc->nr_scanned++; |
80e434260
|
620 |
|
b291f0003
|
621 622 |
if (unlikely(!page_evictable(page, NULL))) goto cull_mlocked; |
894bc3104
|
623 |
|
a6dc60f89
|
624 |
if (!sc->may_unmap && page_mapped(page)) |
80e434260
|
625 |
goto keep_locked; |
1da177e4c
|
626 627 628 |
/* Double the slab pressure for mapped and swapcache pages */ if (page_mapped(page) || PageSwapCache(page)) sc->nr_scanned++; |
c661b078f
|
629 630 631 632 633 634 635 636 637 638 639 640 641 642 |
may_enter_fs = (sc->gfp_mask & __GFP_FS) || (PageSwapCache(page) && (sc->gfp_mask & __GFP_IO)); if (PageWriteback(page)) { /* * Synchronous reclaim is performed in two passes, * first an asynchronous pass over the list to * start parallel writeback, and a second synchronous * pass to wait for the IO to complete. Wait here * for any page for which writeback has already * started. */ if (sync_writeback == PAGEOUT_IO_SYNC && may_enter_fs) wait_on_page_writeback(page); |
4dd4b9202
|
643 |
else |
c661b078f
|
644 645 |
goto keep_locked; } |
1da177e4c
|
646 |
|
dfc8d636c
|
647 648 649 |
references = page_check_references(page, sc); switch (references) { case PAGEREF_ACTIVATE: |
1da177e4c
|
650 |
goto activate_locked; |
645747462
|
651 652 |
case PAGEREF_KEEP: goto keep_locked; |
dfc8d636c
|
653 654 655 656 |
case PAGEREF_RECLAIM: case PAGEREF_RECLAIM_CLEAN: ; /* try to reclaim the page below */ } |
1da177e4c
|
657 |
|
1da177e4c
|
658 659 660 661 |
/* * Anonymous process memory has backing store? * Try to allocate it some swap space here. */ |
b291f0003
|
662 |
if (PageAnon(page) && !PageSwapCache(page)) { |
63eb6b93c
|
663 664 |
if (!(sc->gfp_mask & __GFP_IO)) goto keep_locked; |
ac47b003d
|
665 |
if (!add_to_swap(page)) |
1da177e4c
|
666 |
goto activate_locked; |
63eb6b93c
|
667 |
may_enter_fs = 1; |
b291f0003
|
668 |
} |
1da177e4c
|
669 670 |
mapping = page_mapping(page); |
1da177e4c
|
671 672 673 674 675 676 |
/* * The page is mapped into the page tables of one or more * processes. Try to unmap it here. */ if (page_mapped(page) && mapping) { |
14fa31b89
|
677 |
switch (try_to_unmap(page, TTU_UNMAP)) { |
1da177e4c
|
678 679 680 681 |
case SWAP_FAIL: goto activate_locked; case SWAP_AGAIN: goto keep_locked; |
b291f0003
|
682 683 |
case SWAP_MLOCK: goto cull_mlocked; |
1da177e4c
|
684 685 686 687 688 689 |
case SWAP_SUCCESS: ; /* try to free the page below */ } } if (PageDirty(page)) { |
dfc8d636c
|
690 |
if (references == PAGEREF_RECLAIM_CLEAN) |
1da177e4c
|
691 |
goto keep_locked; |
4dd4b9202
|
692 |
if (!may_enter_fs) |
1da177e4c
|
693 |
goto keep_locked; |
52a8363ea
|
694 |
if (!sc->may_writepage) |
1da177e4c
|
695 696 697 |
goto keep_locked; /* Page is dirty, try to write it out here */ |
c661b078f
|
698 |
switch (pageout(page, mapping, sync_writeback)) { |
1da177e4c
|
699 700 701 702 703 |
case PAGE_KEEP: goto keep_locked; case PAGE_ACTIVATE: goto activate_locked; case PAGE_SUCCESS: |
4dd4b9202
|
704 |
if (PageWriteback(page) || PageDirty(page)) |
1da177e4c
|
705 706 707 708 709 |
goto keep; /* * A synchronous write - probably a ramdisk. Go * ahead and try to reclaim the page. */ |
529ae9aaa
|
710 |
if (!trylock_page(page)) |
1da177e4c
|
711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 |
goto keep; if (PageDirty(page) || PageWriteback(page)) goto keep_locked; mapping = page_mapping(page); case PAGE_CLEAN: ; /* try to free the page below */ } } /* * If the page has buffers, try to free the buffer mappings * associated with this page. If we succeed we try to free * the page as well. * * We do this even if the page is PageDirty(). * try_to_release_page() does not perform I/O, but it is * possible for a page to have PageDirty set, but it is actually * clean (all its buffers are clean). This happens if the * buffers were written out directly, with submit_bh(). ext3 |
894bc3104
|
730 |
* will do this, as well as the blockdev mapping. |
1da177e4c
|
731 732 733 734 735 736 737 738 739 740 |
* try_to_release_page() will discover that cleanness and will * drop the buffers and mark the page clean - it can be freed. * * Rarely, pages can have buffers and no ->mapping. These are * the pages which were not successfully invalidated in * truncate_complete_page(). We try to drop those buffers here * and if that worked, and the page is no longer mapped into * process address space (page_count == 1) it can be freed. * Otherwise, leave the page on the LRU so it is swappable. */ |
266cf658e
|
741 |
if (page_has_private(page)) { |
1da177e4c
|
742 743 |
if (!try_to_release_page(page, sc->gfp_mask)) goto activate_locked; |
e286781d5
|
744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 |
if (!mapping && page_count(page) == 1) { unlock_page(page); if (put_page_testzero(page)) goto free_it; else { /* * rare race with speculative reference. * the speculative reference will free * this page shortly, so we may * increment nr_reclaimed here (and * leave it off the LRU). */ nr_reclaimed++; continue; } } |
1da177e4c
|
760 |
} |
e286781d5
|
761 |
if (!mapping || !__remove_mapping(mapping, page)) |
49d2e9cc4
|
762 |
goto keep_locked; |
1da177e4c
|
763 |
|
a978d6f52
|
764 765 766 767 768 769 770 771 |
/* * At this point, we have no other references and there is * no way to pick any more up (removed from LRU, removed * from pagecache). Can use non-atomic bitops now (and * we obviously don't have to worry about waking up a process * waiting on the page lock, because there are no references. */ __clear_page_locked(page); |
e286781d5
|
772 |
free_it: |
05ff51376
|
773 |
nr_reclaimed++; |
e286781d5
|
774 775 776 777 |
if (!pagevec_add(&freed_pvec, page)) { __pagevec_free(&freed_pvec); pagevec_reinit(&freed_pvec); } |
1da177e4c
|
778 |
continue; |
b291f0003
|
779 |
cull_mlocked: |
63d6c5ad7
|
780 781 |
if (PageSwapCache(page)) try_to_free_swap(page); |
b291f0003
|
782 783 784 |
unlock_page(page); putback_lru_page(page); continue; |
1da177e4c
|
785 |
activate_locked: |
68a22394c
|
786 787 |
/* Not a candidate for swapping, so reclaim swap space. */ if (PageSwapCache(page) && vm_swap_full()) |
a2c43eed8
|
788 |
try_to_free_swap(page); |
894bc3104
|
789 |
VM_BUG_ON(PageActive(page)); |
1da177e4c
|
790 791 792 793 794 795 |
SetPageActive(page); pgactivate++; keep_locked: unlock_page(page); keep: list_add(&page->lru, &ret_pages); |
b291f0003
|
796 |
VM_BUG_ON(PageLRU(page) || PageUnevictable(page)); |
1da177e4c
|
797 798 799 |
} list_splice(&ret_pages, page_list); if (pagevec_count(&freed_pvec)) |
e286781d5
|
800 |
__pagevec_free(&freed_pvec); |
f8891e5e1
|
801 |
count_vm_events(PGACTIVATE, pgactivate); |
05ff51376
|
802 |
return nr_reclaimed; |
1da177e4c
|
803 |
} |
5ad333eb6
|
804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 |
/* LRU Isolation modes. */ #define ISOLATE_INACTIVE 0 /* Isolate inactive pages. */ #define ISOLATE_ACTIVE 1 /* Isolate active pages. */ #define ISOLATE_BOTH 2 /* Isolate both active and inactive pages. */ /* * Attempt to remove the specified page from its LRU. Only take this page * if it is of the appropriate PageActive status. Pages which are being * freed elsewhere are also ignored. * * page: page to consider * mode: one of the LRU isolation modes defined above * * returns 0 on success, -ve errno on failure. */ |
4f98a2fee
|
819 |
int __isolate_lru_page(struct page *page, int mode, int file) |
5ad333eb6
|
820 821 822 823 824 825 826 827 828 829 830 831 832 833 |
{ int ret = -EINVAL; /* Only take pages on the LRU. */ if (!PageLRU(page)) return ret; /* * When checking the active state, we need to be sure we are * dealing with comparible boolean values. Take the logical not * of each. */ if (mode != ISOLATE_BOTH && (!PageActive(page) != !mode)) return ret; |
6c0b13519
|
834 |
if (mode != ISOLATE_BOTH && page_is_file_cache(page) != file) |
4f98a2fee
|
835 |
return ret; |
894bc3104
|
836 837 838 839 840 841 842 |
/* * When this function is being called for lumpy reclaim, we * initially look into all LRU pages, active, inactive and * unevictable; only give shrink_page_list evictable pages. */ if (PageUnevictable(page)) return ret; |
5ad333eb6
|
843 |
ret = -EBUSY; |
08e552c69
|
844 |
|
5ad333eb6
|
845 846 847 848 849 850 851 852 853 854 855 856 |
if (likely(get_page_unless_zero(page))) { /* * Be careful not to clear PageLRU until after we're * sure the page is not being freed elsewhere -- the * page release code relies on it. */ ClearPageLRU(page); ret = 0; } return ret; } |
49d2e9cc4
|
857 |
/* |
1da177e4c
|
858 859 860 861 862 863 864 865 866 867 868 869 870 |
* zone->lru_lock is heavily contended. Some of the functions that * shrink the lists perform better by taking out a batch of pages * and working on them outside the LRU lock. * * For pagecache intensive workloads, this function is the hottest * spot in the kernel (apart from copy_*_user functions). * * Appropriate locks must be held before calling this function. * * @nr_to_scan: The number of pages to look through on the list. * @src: The LRU list to pull pages off. * @dst: The temp list to put pages on to. * @scanned: The number of pages that were scanned. |
5ad333eb6
|
871 872 |
* @order: The caller's attempted allocation order * @mode: One of the LRU isolation modes |
4f98a2fee
|
873 |
* @file: True [1] if isolating file [!anon] pages |
1da177e4c
|
874 875 876 |
* * returns how many pages were moved onto *@dst. */ |
69e05944a
|
877 878 |
static unsigned long isolate_lru_pages(unsigned long nr_to_scan, struct list_head *src, struct list_head *dst, |
4f98a2fee
|
879 |
unsigned long *scanned, int order, int mode, int file) |
1da177e4c
|
880 |
{ |
69e05944a
|
881 |
unsigned long nr_taken = 0; |
c9b02d970
|
882 |
unsigned long scan; |
1da177e4c
|
883 |
|
c9b02d970
|
884 |
for (scan = 0; scan < nr_to_scan && !list_empty(src); scan++) { |
5ad333eb6
|
885 886 887 888 889 |
struct page *page; unsigned long pfn; unsigned long end_pfn; unsigned long page_pfn; int zone_id; |
1da177e4c
|
890 891 |
page = lru_to_page(src); prefetchw_prev_lru_page(page, src, flags); |
725d704ec
|
892 |
VM_BUG_ON(!PageLRU(page)); |
8d438f96d
|
893 |
|
4f98a2fee
|
894 |
switch (__isolate_lru_page(page, mode, file)) { |
5ad333eb6
|
895 896 |
case 0: list_move(&page->lru, dst); |
2ffebca6a
|
897 |
mem_cgroup_del_lru(page); |
7c8ee9a86
|
898 |
nr_taken++; |
5ad333eb6
|
899 900 901 902 903 |
break; case -EBUSY: /* else it is being freed elsewhere */ list_move(&page->lru, src); |
2ffebca6a
|
904 |
mem_cgroup_rotate_lru_list(page, page_lru(page)); |
5ad333eb6
|
905 |
continue; |
46453a6e1
|
906 |
|
5ad333eb6
|
907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 |
default: BUG(); } if (!order) continue; /* * Attempt to take all pages in the order aligned region * surrounding the tag page. Only take those pages of * the same active state as that tag page. We may safely * round the target page pfn down to the requested order * as the mem_map is guarenteed valid out to MAX_ORDER, * where that page is in a different zone we will detect * it from its zone id and abort this block scan. */ zone_id = page_zone_id(page); page_pfn = page_to_pfn(page); pfn = page_pfn & ~((1 << order) - 1); end_pfn = pfn + (1 << order); for (; pfn < end_pfn; pfn++) { struct page *cursor_page; /* The target page is in the block, ignore it. */ if (unlikely(pfn == page_pfn)) continue; /* Avoid holes within the zone. */ if (unlikely(!pfn_valid_within(pfn))) break; cursor_page = pfn_to_page(pfn); |
4f98a2fee
|
939 |
|
5ad333eb6
|
940 941 942 |
/* Check that we have not crossed a zone boundary. */ if (unlikely(page_zone_id(cursor_page) != zone_id)) continue; |
de2e7567c
|
943 944 945 946 947 948 949 950 951 |
/* * If we don't have enough swap space, reclaiming of * anon page which don't already have a swap slot is * pointless. */ if (nr_swap_pages <= 0 && PageAnon(cursor_page) && !PageSwapCache(cursor_page)) continue; |
ee993b135
|
952 |
if (__isolate_lru_page(cursor_page, mode, file) == 0) { |
5ad333eb6
|
953 |
list_move(&cursor_page->lru, dst); |
cb4cbcf6b
|
954 |
mem_cgroup_del_lru(cursor_page); |
5ad333eb6
|
955 956 |
nr_taken++; scan++; |
5ad333eb6
|
957 958 |
} } |
1da177e4c
|
959 960 961 962 963 |
} *scanned = scan; return nr_taken; } |
66e1707bc
|
964 965 966 967 968 |
static unsigned long isolate_pages_global(unsigned long nr, struct list_head *dst, unsigned long *scanned, int order, int mode, struct zone *z, struct mem_cgroup *mem_cont, |
4f98a2fee
|
969 |
int active, int file) |
66e1707bc
|
970 |
{ |
4f98a2fee
|
971 |
int lru = LRU_BASE; |
66e1707bc
|
972 |
if (active) |
4f98a2fee
|
973 974 975 976 |
lru += LRU_ACTIVE; if (file) lru += LRU_FILE; return isolate_lru_pages(nr, &z->lru[lru].list, dst, scanned, order, |
b7c46d151
|
977 |
mode, file); |
66e1707bc
|
978 |
} |
1da177e4c
|
979 |
/* |
5ad333eb6
|
980 981 982 |
* clear_active_flags() is a helper for shrink_active_list(), clearing * any active bits from the pages in the list. */ |
4f98a2fee
|
983 984 |
static unsigned long clear_active_flags(struct list_head *page_list, unsigned int *count) |
5ad333eb6
|
985 986 |
{ int nr_active = 0; |
4f98a2fee
|
987 |
int lru; |
5ad333eb6
|
988 |
struct page *page; |
4f98a2fee
|
989 |
list_for_each_entry(page, page_list, lru) { |
401a8e1c1
|
990 |
lru = page_lru_base_type(page); |
5ad333eb6
|
991 |
if (PageActive(page)) { |
4f98a2fee
|
992 |
lru += LRU_ACTIVE; |
5ad333eb6
|
993 994 995 |
ClearPageActive(page); nr_active++; } |
4f98a2fee
|
996 997 |
count[lru]++; } |
5ad333eb6
|
998 999 1000 |
return nr_active; } |
62695a84e
|
1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 |
/** * isolate_lru_page - tries to isolate a page from its LRU list * @page: page to isolate from its LRU list * * Isolates a @page from an LRU list, clears PageLRU and adjusts the * vmstat statistic corresponding to whatever LRU list the page was on. * * Returns 0 if the page was removed from an LRU list. * Returns -EBUSY if the page was not on an LRU list. * * The returned page will have PageLRU() cleared. If it was found on |
894bc3104
|
1012 1013 1014 |
* the active list, it will have PageActive set. If it was found on * the unevictable list, it will have the PageUnevictable bit set. That flag * may need to be cleared by the caller before letting the page go. |
62695a84e
|
1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 |
* * The vmstat statistic corresponding to the list on which the page was * found will be decremented. * * Restrictions: * (1) Must be called with an elevated refcount on the page. This is a * fundamentnal difference from isolate_lru_pages (which is called * without a stable reference). * (2) the lru_lock must not be held. * (3) interrupts must be enabled. */ int isolate_lru_page(struct page *page) { int ret = -EBUSY; if (PageLRU(page)) { struct zone *zone = page_zone(page); spin_lock_irq(&zone->lru_lock); if (PageLRU(page) && get_page_unless_zero(page)) { |
894bc3104
|
1035 |
int lru = page_lru(page); |
62695a84e
|
1036 1037 |
ret = 0; ClearPageLRU(page); |
4f98a2fee
|
1038 |
|
4f98a2fee
|
1039 |
del_page_from_lru_list(zone, page, lru); |
62695a84e
|
1040 1041 1042 1043 1044 |
} spin_unlock_irq(&zone->lru_lock); } return ret; } |
5ad333eb6
|
1045 |
/* |
35cd78156
|
1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 |
* Are there way too many processes in the direct reclaim path already? */ static int too_many_isolated(struct zone *zone, int file, struct scan_control *sc) { unsigned long inactive, isolated; if (current_is_kswapd()) return 0; if (!scanning_global_lru(sc)) return 0; if (file) { inactive = zone_page_state(zone, NR_INACTIVE_FILE); isolated = zone_page_state(zone, NR_ISOLATED_FILE); } else { inactive = zone_page_state(zone, NR_INACTIVE_ANON); isolated = zone_page_state(zone, NR_ISOLATED_ANON); } return isolated > inactive; } /* |
1742f19fa
|
1071 1072 |
* shrink_inactive_list() is a helper for shrink_zone(). It returns the number * of reclaimed pages |
1da177e4c
|
1073 |
*/ |
1742f19fa
|
1074 |
static unsigned long shrink_inactive_list(unsigned long max_scan, |
33c120ed2
|
1075 1076 |
struct zone *zone, struct scan_control *sc, int priority, int file) |
1da177e4c
|
1077 1078 1079 |
{ LIST_HEAD(page_list); struct pagevec pvec; |
69e05944a
|
1080 |
unsigned long nr_scanned = 0; |
05ff51376
|
1081 |
unsigned long nr_reclaimed = 0; |
6e9015716
|
1082 |
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); |
78dc583d3
|
1083 |
int lumpy_reclaim = 0; |
35cd78156
|
1084 |
while (unlikely(too_many_isolated(zone, file, sc))) { |
58355c787
|
1085 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
35cd78156
|
1086 1087 1088 1089 1090 |
/* We are about to die and free our memory. Return now. */ if (fatal_signal_pending(current)) return SWAP_CLUSTER_MAX; } |
78dc583d3
|
1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 |
/* * If we need a large contiguous chunk of memory, or have * trouble getting a small set of contiguous pages, we * will reclaim both active and inactive pages. * * We use the same threshold as pageout congestion_wait below. */ if (sc->order > PAGE_ALLOC_COSTLY_ORDER) lumpy_reclaim = 1; else if (sc->order && priority < DEF_PRIORITY - 2) lumpy_reclaim = 1; |
1da177e4c
|
1102 1103 1104 1105 1106 |
pagevec_init(&pvec, 1); lru_add_drain(); spin_lock_irq(&zone->lru_lock); |
69e05944a
|
1107 |
do { |
1da177e4c
|
1108 |
struct page *page; |
69e05944a
|
1109 1110 1111 |
unsigned long nr_taken; unsigned long nr_scan; unsigned long nr_freed; |
5ad333eb6
|
1112 |
unsigned long nr_active; |
4f98a2fee
|
1113 |
unsigned int count[NR_LRU_LISTS] = { 0, }; |
78dc583d3
|
1114 |
int mode = lumpy_reclaim ? ISOLATE_BOTH : ISOLATE_INACTIVE; |
a731286de
|
1115 1116 |
unsigned long nr_anon; unsigned long nr_file; |
1da177e4c
|
1117 |
|
ece74b2e7
|
1118 |
nr_taken = sc->isolate_pages(SWAP_CLUSTER_MAX, |
4f98a2fee
|
1119 1120 |
&page_list, &nr_scan, sc->order, mode, zone, sc->mem_cgroup, 0, file); |
b35ea17b7
|
1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 |
if (scanning_global_lru(sc)) { zone->pages_scanned += nr_scan; if (current_is_kswapd()) __count_zone_vm_events(PGSCAN_KSWAPD, zone, nr_scan); else __count_zone_vm_events(PGSCAN_DIRECT, zone, nr_scan); } if (nr_taken == 0) goto done; |
4f98a2fee
|
1134 |
nr_active = clear_active_flags(&page_list, count); |
e9187bdcb
|
1135 |
__count_vm_events(PGDEACTIVATE, nr_active); |
5ad333eb6
|
1136 |
|
4f98a2fee
|
1137 1138 1139 1140 1141 1142 1143 1144 |
__mod_zone_page_state(zone, NR_ACTIVE_FILE, -count[LRU_ACTIVE_FILE]); __mod_zone_page_state(zone, NR_INACTIVE_FILE, -count[LRU_INACTIVE_FILE]); __mod_zone_page_state(zone, NR_ACTIVE_ANON, -count[LRU_ACTIVE_ANON]); __mod_zone_page_state(zone, NR_INACTIVE_ANON, -count[LRU_INACTIVE_ANON]); |
a731286de
|
1145 1146 1147 1148 |
nr_anon = count[LRU_ACTIVE_ANON] + count[LRU_INACTIVE_ANON]; nr_file = count[LRU_ACTIVE_FILE] + count[LRU_INACTIVE_FILE]; __mod_zone_page_state(zone, NR_ISOLATED_ANON, nr_anon); __mod_zone_page_state(zone, NR_ISOLATED_FILE, nr_file); |
3e2f41f1f
|
1149 |
|
62c0c2f19
|
1150 1151 |
reclaim_stat->recent_scanned[0] += nr_anon; reclaim_stat->recent_scanned[1] += nr_file; |
3e2f41f1f
|
1152 |
|
1da177e4c
|
1153 |
spin_unlock_irq(&zone->lru_lock); |
69e05944a
|
1154 |
nr_scanned += nr_scan; |
c661b078f
|
1155 1156 1157 1158 1159 1160 1161 1162 1163 |
nr_freed = shrink_page_list(&page_list, sc, PAGEOUT_IO_ASYNC); /* * If we are direct reclaiming for contiguous pages and we do * not reclaim everything in the list, try again and wait * for IO to complete. This will stall high-order allocations * but that should be acceptable to the caller */ if (nr_freed < nr_taken && !current_is_kswapd() && |
78dc583d3
|
1164 |
lumpy_reclaim) { |
8aa7e847d
|
1165 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
c661b078f
|
1166 1167 1168 1169 1170 |
/* * The attempt at page out may have made some * of the pages active, mark them inactive again. */ |
4f98a2fee
|
1171 |
nr_active = clear_active_flags(&page_list, count); |
c661b078f
|
1172 1173 1174 1175 1176 |
count_vm_events(PGDEACTIVATE, nr_active); nr_freed += shrink_page_list(&page_list, sc, PAGEOUT_IO_SYNC); } |
05ff51376
|
1177 |
nr_reclaimed += nr_freed; |
b35ea17b7
|
1178 |
|
a74609faf
|
1179 |
local_irq_disable(); |
b35ea17b7
|
1180 |
if (current_is_kswapd()) |
f8891e5e1
|
1181 |
__count_vm_events(KSWAPD_STEAL, nr_freed); |
918d3f90e
|
1182 |
__count_zone_vm_events(PGSTEAL, zone, nr_freed); |
a74609faf
|
1183 1184 |
spin_lock(&zone->lru_lock); |
1da177e4c
|
1185 1186 1187 1188 |
/* * Put back any unfreeable pages. */ while (!list_empty(&page_list)) { |
894bc3104
|
1189 |
int lru; |
1da177e4c
|
1190 |
page = lru_to_page(&page_list); |
725d704ec
|
1191 |
VM_BUG_ON(PageLRU(page)); |
1da177e4c
|
1192 |
list_del(&page->lru); |
894bc3104
|
1193 1194 1195 1196 1197 1198 1199 1200 1201 |
if (unlikely(!page_evictable(page, NULL))) { spin_unlock_irq(&zone->lru_lock); putback_lru_page(page); spin_lock_irq(&zone->lru_lock); continue; } SetPageLRU(page); lru = page_lru(page); add_page_to_lru_list(zone, page, lru); |
74a1c48fb
|
1202 |
if (is_active_lru(lru)) { |
b7c46d151
|
1203 |
int file = is_file_lru(lru); |
6e9015716
|
1204 |
reclaim_stat->recent_rotated[file]++; |
4f98a2fee
|
1205 |
} |
1da177e4c
|
1206 1207 1208 1209 1210 1211 |
if (!pagevec_add(&pvec, page)) { spin_unlock_irq(&zone->lru_lock); __pagevec_release(&pvec); spin_lock_irq(&zone->lru_lock); } } |
a731286de
|
1212 1213 |
__mod_zone_page_state(zone, NR_ISOLATED_ANON, -nr_anon); __mod_zone_page_state(zone, NR_ISOLATED_FILE, -nr_file); |
69e05944a
|
1214 |
} while (nr_scanned < max_scan); |
b35ea17b7
|
1215 |
|
1da177e4c
|
1216 |
done: |
b35ea17b7
|
1217 |
spin_unlock_irq(&zone->lru_lock); |
1da177e4c
|
1218 |
pagevec_release(&pvec); |
05ff51376
|
1219 |
return nr_reclaimed; |
1da177e4c
|
1220 |
} |
3bb1a852a
|
1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 |
/* * We are about to scan this zone at a certain priority level. If that priority * level is smaller (ie: more urgent) than the previous priority, then note * that priority level within the zone. This is done so that when the next * process comes in to scan this zone, it will immediately start out at this * priority level rather than having to build up its own scanning priority. * Here, this priority affects only the reclaim-mapped threshold. */ static inline void note_zone_scanning_priority(struct zone *zone, int priority) { if (priority < zone->prev_priority) zone->prev_priority = priority; } |
1cfb419b3
|
1234 |
/* |
1da177e4c
|
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 |
* This moves pages from the active list to the inactive list. * * We move them the other way if the page is referenced by one or more * processes, from rmap. * * If the pages are mostly unmapped, the processing is fast and it is * appropriate to hold zone->lru_lock across the whole operation. But if * the pages are mapped, the processing is slow (page_referenced()) so we * should drop zone->lru_lock around each page. It's impossible to balance * this, so instead we remove the pages from the LRU while processing them. * It is safe to rely on PG_active against the non-LRU pages in here because * nobody will play with that bit on a non-LRU page. * * The downside is that we have to touch page->_count against each page. * But we had to alter page->flags anyway. */ |
1cfb419b3
|
1251 |
|
3eb4140f0
|
1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 |
static void move_active_pages_to_lru(struct zone *zone, struct list_head *list, enum lru_list lru) { unsigned long pgmoved = 0; struct pagevec pvec; struct page *page; pagevec_init(&pvec, 1); while (!list_empty(list)) { page = lru_to_page(list); |
3eb4140f0
|
1264 1265 1266 |
VM_BUG_ON(PageLRU(page)); SetPageLRU(page); |
3eb4140f0
|
1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 |
list_move(&page->lru, &zone->lru[lru].list); mem_cgroup_add_lru_list(page, lru); pgmoved++; if (!pagevec_add(&pvec, page) || list_empty(list)) { spin_unlock_irq(&zone->lru_lock); if (buffer_heads_over_limit) pagevec_strip(&pvec); __pagevec_release(&pvec); spin_lock_irq(&zone->lru_lock); } } __mod_zone_page_state(zone, NR_LRU_BASE + lru, pgmoved); if (!is_active_lru(lru)) __count_vm_events(PGDEACTIVATE, pgmoved); } |
1cfb419b3
|
1283 |
|
1742f19fa
|
1284 |
static void shrink_active_list(unsigned long nr_pages, struct zone *zone, |
4f98a2fee
|
1285 |
struct scan_control *sc, int priority, int file) |
1da177e4c
|
1286 |
{ |
44c241f16
|
1287 |
unsigned long nr_taken; |
69e05944a
|
1288 |
unsigned long pgscanned; |
6fe6b7e35
|
1289 |
unsigned long vm_flags; |
1da177e4c
|
1290 |
LIST_HEAD(l_hold); /* The pages which were snipped off */ |
8cab4754d
|
1291 |
LIST_HEAD(l_active); |
b69408e88
|
1292 |
LIST_HEAD(l_inactive); |
1da177e4c
|
1293 |
struct page *page; |
6e9015716
|
1294 |
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); |
44c241f16
|
1295 |
unsigned long nr_rotated = 0; |
1da177e4c
|
1296 1297 1298 |
lru_add_drain(); spin_lock_irq(&zone->lru_lock); |
44c241f16
|
1299 |
nr_taken = sc->isolate_pages(nr_pages, &l_hold, &pgscanned, sc->order, |
66e1707bc
|
1300 |
ISOLATE_ACTIVE, zone, |
4f98a2fee
|
1301 |
sc->mem_cgroup, 1, file); |
1cfb419b3
|
1302 1303 1304 1305 |
/* * zone->pages_scanned is used for detect zone's oom * mem_cgroup remembers nr_scan by itself. */ |
e72e2bd67
|
1306 |
if (scanning_global_lru(sc)) { |
1cfb419b3
|
1307 |
zone->pages_scanned += pgscanned; |
4f98a2fee
|
1308 |
} |
b7c46d151
|
1309 |
reclaim_stat->recent_scanned[file] += nr_taken; |
1cfb419b3
|
1310 |
|
3eb4140f0
|
1311 |
__count_zone_vm_events(PGREFILL, zone, pgscanned); |
4f98a2fee
|
1312 |
if (file) |
44c241f16
|
1313 |
__mod_zone_page_state(zone, NR_ACTIVE_FILE, -nr_taken); |
4f98a2fee
|
1314 |
else |
44c241f16
|
1315 |
__mod_zone_page_state(zone, NR_ACTIVE_ANON, -nr_taken); |
a731286de
|
1316 |
__mod_zone_page_state(zone, NR_ISOLATED_ANON + file, nr_taken); |
1da177e4c
|
1317 |
spin_unlock_irq(&zone->lru_lock); |
1da177e4c
|
1318 1319 1320 1321 |
while (!list_empty(&l_hold)) { cond_resched(); page = lru_to_page(&l_hold); list_del(&page->lru); |
7e9cd4842
|
1322 |
|
894bc3104
|
1323 1324 1325 1326 |
if (unlikely(!page_evictable(page, NULL))) { putback_lru_page(page); continue; } |
645747462
|
1327 |
if (page_referenced(page, 0, sc->mem_cgroup, &vm_flags)) { |
44c241f16
|
1328 |
nr_rotated++; |
8cab4754d
|
1329 1330 1331 1332 1333 1334 1335 1336 1337 |
/* * Identify referenced, file-backed active pages and * give them one more trip around the active list. So * that executable code get better chances to stay in * memory under moderate memory pressure. Anon pages * are not likely to be evicted by use-once streaming * IO, plus JVM can create lots of anon VM_EXEC pages, * so we ignore them here. */ |
41e20983f
|
1338 |
if ((vm_flags & VM_EXEC) && page_is_file_cache(page)) { |
8cab4754d
|
1339 1340 1341 1342 |
list_add(&page->lru, &l_active); continue; } } |
7e9cd4842
|
1343 |
|
5205e56ee
|
1344 |
ClearPageActive(page); /* we are de-activating */ |
1da177e4c
|
1345 1346 |
list_add(&page->lru, &l_inactive); } |
b555749aa
|
1347 |
/* |
8cab4754d
|
1348 |
* Move pages back to the lru list. |
b555749aa
|
1349 |
*/ |
2a1dc5097
|
1350 |
spin_lock_irq(&zone->lru_lock); |
4f98a2fee
|
1351 |
/* |
8cab4754d
|
1352 1353 1354 1355 |
* Count referenced pages from currently used mappings as rotated, * even though only some of them are actually re-activated. This * helps balance scan pressure between file and anonymous pages in * get_scan_ratio. |
7e9cd4842
|
1356 |
*/ |
b7c46d151
|
1357 |
reclaim_stat->recent_rotated[file] += nr_rotated; |
556adecba
|
1358 |
|
3eb4140f0
|
1359 1360 1361 1362 |
move_active_pages_to_lru(zone, &l_active, LRU_ACTIVE + file * LRU_FILE); move_active_pages_to_lru(zone, &l_inactive, LRU_BASE + file * LRU_FILE); |
a731286de
|
1363 |
__mod_zone_page_state(zone, NR_ISOLATED_ANON + file, -nr_taken); |
f8891e5e1
|
1364 |
spin_unlock_irq(&zone->lru_lock); |
1da177e4c
|
1365 |
} |
14797e236
|
1366 |
static int inactive_anon_is_low_global(struct zone *zone) |
f89eb90e3
|
1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 |
{ unsigned long active, inactive; active = zone_page_state(zone, NR_ACTIVE_ANON); inactive = zone_page_state(zone, NR_INACTIVE_ANON); if (inactive * zone->inactive_ratio < active) return 1; return 0; } |
14797e236
|
1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 |
/** * inactive_anon_is_low - check if anonymous pages need to be deactivated * @zone: zone to check * @sc: scan control of this context * * Returns true if the zone does not have enough inactive anon pages, * meaning some active anon pages need to be deactivated. */ static int inactive_anon_is_low(struct zone *zone, struct scan_control *sc) { int low; |
e72e2bd67
|
1389 |
if (scanning_global_lru(sc)) |
14797e236
|
1390 1391 |
low = inactive_anon_is_low_global(zone); else |
c772be939
|
1392 |
low = mem_cgroup_inactive_anon_is_low(sc->mem_cgroup); |
14797e236
|
1393 1394 |
return low; } |
56e49d218
|
1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 |
static int inactive_file_is_low_global(struct zone *zone) { unsigned long active, inactive; active = zone_page_state(zone, NR_ACTIVE_FILE); inactive = zone_page_state(zone, NR_INACTIVE_FILE); return (active > inactive); } /** * inactive_file_is_low - check if file pages need to be deactivated * @zone: zone to check * @sc: scan control of this context * * When the system is doing streaming IO, memory pressure here * ensures that active file pages get deactivated, until more * than half of the file pages are on the inactive list. * * Once we get to that situation, protect the system's working * set from being evicted by disabling active file page aging. * * This uses a different ratio than the anonymous pages, because * the page cache uses a use-once replacement algorithm. */ static int inactive_file_is_low(struct zone *zone, struct scan_control *sc) { int low; if (scanning_global_lru(sc)) low = inactive_file_is_low_global(zone); else low = mem_cgroup_inactive_file_is_low(sc->mem_cgroup); return low; } |
b39415b27
|
1430 1431 1432 1433 1434 1435 1436 1437 |
static int inactive_list_is_low(struct zone *zone, struct scan_control *sc, int file) { if (file) return inactive_file_is_low(zone, sc); else return inactive_anon_is_low(zone, sc); } |
4f98a2fee
|
1438 |
static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan, |
b69408e88
|
1439 1440 |
struct zone *zone, struct scan_control *sc, int priority) { |
4f98a2fee
|
1441 |
int file = is_file_lru(lru); |
b39415b27
|
1442 1443 1444 |
if (is_active_lru(lru)) { if (inactive_list_is_low(zone, sc, file)) shrink_active_list(nr_to_scan, zone, sc, priority, file); |
556adecba
|
1445 1446 |
return 0; } |
33c120ed2
|
1447 |
return shrink_inactive_list(nr_to_scan, zone, sc, priority, file); |
4f98a2fee
|
1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 |
} /* * Determine how aggressively the anon and file LRU lists should be * scanned. The relative value of each set of LRU lists is determined * by looking at the fraction of the pages scanned we did rotate back * onto the active list instead of evict. * * percent[0] specifies how much pressure to put on ram/swap backed * memory, while percent[1] determines pressure on the file LRUs. */ static void get_scan_ratio(struct zone *zone, struct scan_control *sc, unsigned long *percent) { unsigned long anon, file, free; unsigned long anon_prio, file_prio; unsigned long ap, fp; |
6e9015716
|
1465 |
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); |
4f98a2fee
|
1466 |
|
84b18490d
|
1467 1468 1469 1470 1471 1472 |
/* If we have no swap space, do not bother scanning anon pages. */ if (!sc->may_swap || (nr_swap_pages <= 0)) { percent[0] = 0; percent[1] = 100; return; } |
0b2176763
|
1473 1474 1475 1476 |
anon = zone_nr_lru_pages(zone, sc, LRU_ACTIVE_ANON) + zone_nr_lru_pages(zone, sc, LRU_INACTIVE_ANON); file = zone_nr_lru_pages(zone, sc, LRU_ACTIVE_FILE) + zone_nr_lru_pages(zone, sc, LRU_INACTIVE_FILE); |
b962716b4
|
1477 |
|
e72e2bd67
|
1478 |
if (scanning_global_lru(sc)) { |
eeee9a8cd
|
1479 1480 1481 |
free = zone_page_state(zone, NR_FREE_PAGES); /* If we have very few page cache pages, force-scan anon pages. */ |
418589663
|
1482 |
if (unlikely(file + free <= high_wmark_pages(zone))) { |
eeee9a8cd
|
1483 1484 1485 1486 |
percent[0] = 100; percent[1] = 0; return; } |
4f98a2fee
|
1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 |
} /* * OK, so we have swap space and a fair amount of page cache * pages. We use the recently rotated / recently scanned * ratios to determine how valuable each cache is. * * Because workloads change over time (and to avoid overflow) * we keep these statistics as a floating average, which ends * up weighing recent references more than old ones. * * anon in [0], file in [1] */ |
6e9015716
|
1500 |
if (unlikely(reclaim_stat->recent_scanned[0] > anon / 4)) { |
4f98a2fee
|
1501 |
spin_lock_irq(&zone->lru_lock); |
6e9015716
|
1502 1503 |
reclaim_stat->recent_scanned[0] /= 2; reclaim_stat->recent_rotated[0] /= 2; |
4f98a2fee
|
1504 1505 |
spin_unlock_irq(&zone->lru_lock); } |
6e9015716
|
1506 |
if (unlikely(reclaim_stat->recent_scanned[1] > file / 4)) { |
4f98a2fee
|
1507 |
spin_lock_irq(&zone->lru_lock); |
6e9015716
|
1508 1509 |
reclaim_stat->recent_scanned[1] /= 2; reclaim_stat->recent_rotated[1] /= 2; |
4f98a2fee
|
1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 |
spin_unlock_irq(&zone->lru_lock); } /* * With swappiness at 100, anonymous and file have the same priority. * This scanning priority is essentially the inverse of IO cost. */ anon_prio = sc->swappiness; file_prio = 200 - sc->swappiness; /* |
00d8089c5
|
1521 1522 1523 |
* The amount of pressure on anon vs file pages is inversely * proportional to the fraction of recently scanned pages on * each list that were recently referenced and in active use. |
4f98a2fee
|
1524 |
*/ |
6e9015716
|
1525 1526 |
ap = (anon_prio + 1) * (reclaim_stat->recent_scanned[0] + 1); ap /= reclaim_stat->recent_rotated[0] + 1; |
4f98a2fee
|
1527 |
|
6e9015716
|
1528 1529 |
fp = (file_prio + 1) * (reclaim_stat->recent_scanned[1] + 1); fp /= reclaim_stat->recent_rotated[1] + 1; |
4f98a2fee
|
1530 1531 1532 1533 |
/* Normalize to percentages */ percent[0] = 100 * ap / (ap + fp + 1); percent[1] = 100 - percent[0]; |
b69408e88
|
1534 |
} |
6e08a369e
|
1535 1536 1537 1538 1539 |
/* * Smallish @nr_to_scan's are deposited in @nr_saved_scan, * until we collected @swap_cluster_max pages to scan. */ static unsigned long nr_scan_try_batch(unsigned long nr_to_scan, |
ece74b2e7
|
1540 |
unsigned long *nr_saved_scan) |
6e08a369e
|
1541 1542 1543 1544 1545 |
{ unsigned long nr; *nr_saved_scan += nr_to_scan; nr = *nr_saved_scan; |
ece74b2e7
|
1546 |
if (nr >= SWAP_CLUSTER_MAX) |
6e08a369e
|
1547 1548 1549 1550 1551 1552 |
*nr_saved_scan = 0; else nr = 0; return nr; } |
4f98a2fee
|
1553 |
|
1da177e4c
|
1554 1555 1556 |
/* * This is a basic per-zone page freer. Used by both kswapd and direct reclaim. */ |
a79311c14
|
1557 |
static void shrink_zone(int priority, struct zone *zone, |
05ff51376
|
1558 |
struct scan_control *sc) |
1da177e4c
|
1559 |
{ |
b69408e88
|
1560 |
unsigned long nr[NR_LRU_LISTS]; |
8695949a1
|
1561 |
unsigned long nr_to_scan; |
4f98a2fee
|
1562 |
unsigned long percent[2]; /* anon @ 0; file @ 1 */ |
b69408e88
|
1563 |
enum lru_list l; |
01dbe5c9b
|
1564 |
unsigned long nr_reclaimed = sc->nr_reclaimed; |
22fba3354
|
1565 |
unsigned long nr_to_reclaim = sc->nr_to_reclaim; |
f86296317
|
1566 |
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); |
1da177e4c
|
1567 |
|
84b18490d
|
1568 |
get_scan_ratio(zone, sc, percent); |
4f98a2fee
|
1569 |
|
894bc3104
|
1570 |
for_each_evictable_lru(l) { |
9439c1c95
|
1571 |
int file = is_file_lru(l); |
8713e0129
|
1572 |
unsigned long scan; |
e0f79b8f1
|
1573 |
|
84b18490d
|
1574 1575 1576 1577 |
if (percent[file] == 0) { nr[l] = 0; continue; } |
0b2176763
|
1578 |
scan = zone_nr_lru_pages(zone, sc, l); |
84b18490d
|
1579 |
if (priority) { |
9439c1c95
|
1580 1581 1582 |
scan >>= priority; scan = (scan * percent[file]) / 100; } |
f86296317
|
1583 |
nr[l] = nr_scan_try_batch(scan, |
ece74b2e7
|
1584 |
&reclaim_stat->nr_saved_scan[l]); |
1cfb419b3
|
1585 |
} |
1da177e4c
|
1586 |
|
556adecba
|
1587 1588 |
while (nr[LRU_INACTIVE_ANON] || nr[LRU_ACTIVE_FILE] || nr[LRU_INACTIVE_FILE]) { |
894bc3104
|
1589 |
for_each_evictable_lru(l) { |
b69408e88
|
1590 |
if (nr[l]) { |
ece74b2e7
|
1591 1592 |
nr_to_scan = min_t(unsigned long, nr[l], SWAP_CLUSTER_MAX); |
b69408e88
|
1593 |
nr[l] -= nr_to_scan; |
1da177e4c
|
1594 |
|
01dbe5c9b
|
1595 1596 |
nr_reclaimed += shrink_list(l, nr_to_scan, zone, sc, priority); |
b69408e88
|
1597 |
} |
1da177e4c
|
1598 |
} |
a79311c14
|
1599 1600 1601 1602 1603 1604 1605 1606 |
/* * On large memory systems, scan >> priority can become * really large. This is fine for the starting priority; * we want to put equal scanning pressure on each zone. * However, if the VM has a harder time of freeing pages, * with multiple processes reclaiming pages, the total * freeing target can get unreasonably large. */ |
338fde909
|
1607 |
if (nr_reclaimed >= nr_to_reclaim && priority < DEF_PRIORITY) |
a79311c14
|
1608 |
break; |
1da177e4c
|
1609 |
} |
01dbe5c9b
|
1610 |
sc->nr_reclaimed = nr_reclaimed; |
556adecba
|
1611 1612 1613 1614 |
/* * Even if we did not try to evict anon pages at all, we want to * rebalance the anon lru active/inactive ratio. */ |
69c854817
|
1615 |
if (inactive_anon_is_low(zone, sc) && nr_swap_pages > 0) |
556adecba
|
1616 |
shrink_active_list(SWAP_CLUSTER_MAX, zone, sc, priority, 0); |
232ea4d69
|
1617 |
throttle_vm_writeout(sc->gfp_mask); |
1da177e4c
|
1618 1619 1620 1621 1622 1623 1624 |
} /* * This is the direct reclaim path, for page-allocating processes. We only * try to reclaim pages from zones which will satisfy the caller's allocation * request. * |
418589663
|
1625 1626 |
* We reclaim from a zone even if that zone is over high_wmark_pages(zone). * Because: |
1da177e4c
|
1627 1628 |
* a) The caller may be trying to free *extra* pages to satisfy a higher-order * allocation or |
418589663
|
1629 1630 1631 |
* b) The target zone may be at high_wmark_pages(zone) but the lower zones * must go *over* high_wmark_pages(zone) to satisfy the `incremental min' * zone defense algorithm. |
1da177e4c
|
1632 |
* |
1da177e4c
|
1633 1634 1635 |
* If a zone is deemed to be full of pinned pages then just give it a light * scan then give up on it. */ |
a79311c14
|
1636 |
static void shrink_zones(int priority, struct zonelist *zonelist, |
05ff51376
|
1637 |
struct scan_control *sc) |
1da177e4c
|
1638 |
{ |
54a6eb5c4
|
1639 |
enum zone_type high_zoneidx = gfp_zone(sc->gfp_mask); |
dd1a239f6
|
1640 |
struct zoneref *z; |
54a6eb5c4
|
1641 |
struct zone *zone; |
1cfb419b3
|
1642 |
|
408d85441
|
1643 |
sc->all_unreclaimable = 1; |
327c0e968
|
1644 1645 |
for_each_zone_zonelist_nodemask(zone, z, zonelist, high_zoneidx, sc->nodemask) { |
f3fe65122
|
1646 |
if (!populated_zone(zone)) |
1da177e4c
|
1647 |
continue; |
1cfb419b3
|
1648 1649 1650 1651 |
/* * Take care memory controller reclaiming has small influence * to global LRU. */ |
e72e2bd67
|
1652 |
if (scanning_global_lru(sc)) { |
1cfb419b3
|
1653 1654 1655 |
if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) continue; note_zone_scanning_priority(zone, priority); |
1da177e4c
|
1656 |
|
93e4a89a8
|
1657 |
if (zone->all_unreclaimable && priority != DEF_PRIORITY) |
1cfb419b3
|
1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 |
continue; /* Let kswapd poll it */ sc->all_unreclaimable = 0; } else { /* * Ignore cpuset limitation here. We just want to reduce * # of used pages by us regardless of memory shortage. */ sc->all_unreclaimable = 0; mem_cgroup_note_reclaim_priority(sc->mem_cgroup, priority); } |
408d85441
|
1669 |
|
a79311c14
|
1670 |
shrink_zone(priority, zone, sc); |
1da177e4c
|
1671 1672 |
} } |
4f98a2fee
|
1673 |
|
1da177e4c
|
1674 1675 1676 1677 1678 1679 1680 1681 |
/* * This is the main entry point to direct page reclaim. * * If a full scan of the inactive list fails to free enough memory then we * are "out of memory" and something needs to be killed. * * If the caller is !__GFP_FS then the probability of a failure is reasonably * high - the zone may be full of dirty or under-writeback pages, which this |
5b0830cb9
|
1682 1683 1684 1685 |
* caller can't do much about. We kick the writeback threads and take explicit * naps in the hope that some of these pages can be written. But if the * allocating task holds filesystem locks which prevent writeout this might not * work, and the allocation attempt will fail. |
a41f24ea9
|
1686 1687 1688 |
* * returns: 0, if no pages reclaimed * else, the number of pages reclaimed |
1da177e4c
|
1689 |
*/ |
dac1d27bc
|
1690 |
static unsigned long do_try_to_free_pages(struct zonelist *zonelist, |
dd1a239f6
|
1691 |
struct scan_control *sc) |
1da177e4c
|
1692 1693 |
{ int priority; |
c700be3d1
|
1694 |
unsigned long ret = 0; |
69e05944a
|
1695 |
unsigned long total_scanned = 0; |
1da177e4c
|
1696 |
struct reclaim_state *reclaim_state = current->reclaim_state; |
1da177e4c
|
1697 |
unsigned long lru_pages = 0; |
dd1a239f6
|
1698 |
struct zoneref *z; |
54a6eb5c4
|
1699 |
struct zone *zone; |
dd1a239f6
|
1700 |
enum zone_type high_zoneidx = gfp_zone(sc->gfp_mask); |
22fba3354
|
1701 |
unsigned long writeback_threshold; |
1da177e4c
|
1702 |
|
873b47717
|
1703 |
delayacct_freepages_start(); |
e72e2bd67
|
1704 |
if (scanning_global_lru(sc)) |
1cfb419b3
|
1705 1706 1707 1708 |
count_vm_event(ALLOCSTALL); /* * mem_cgroup will not do shrink_slab. */ |
e72e2bd67
|
1709 |
if (scanning_global_lru(sc)) { |
54a6eb5c4
|
1710 |
for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) { |
1da177e4c
|
1711 |
|
1cfb419b3
|
1712 1713 |
if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) continue; |
1da177e4c
|
1714 |
|
adea02a1b
|
1715 |
lru_pages += zone_reclaimable_pages(zone); |
1cfb419b3
|
1716 |
} |
1da177e4c
|
1717 1718 1719 |
} for (priority = DEF_PRIORITY; priority >= 0; priority--) { |
66e1707bc
|
1720 |
sc->nr_scanned = 0; |
f7b7fd8f3
|
1721 1722 |
if (!priority) disable_swap_token(); |
a79311c14
|
1723 |
shrink_zones(priority, zonelist, sc); |
66e1707bc
|
1724 1725 1726 1727 |
/* * Don't shrink slabs when reclaiming memory from * over limit cgroups */ |
e72e2bd67
|
1728 |
if (scanning_global_lru(sc)) { |
dd1a239f6
|
1729 |
shrink_slab(sc->nr_scanned, sc->gfp_mask, lru_pages); |
91a45470f
|
1730 |
if (reclaim_state) { |
a79311c14
|
1731 |
sc->nr_reclaimed += reclaim_state->reclaimed_slab; |
91a45470f
|
1732 1733 |
reclaim_state->reclaimed_slab = 0; } |
1da177e4c
|
1734 |
} |
66e1707bc
|
1735 |
total_scanned += sc->nr_scanned; |
22fba3354
|
1736 |
if (sc->nr_reclaimed >= sc->nr_to_reclaim) { |
a79311c14
|
1737 |
ret = sc->nr_reclaimed; |
1da177e4c
|
1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 |
goto out; } /* * Try to write back as many pages as we just scanned. This * tends to cause slow streaming writers to write data to the * disk smoothly, at the dirtying rate, which is nice. But * that's undesirable in laptop mode, where we *want* lumpy * writeout. So in laptop mode, write out the whole world. */ |
22fba3354
|
1748 1749 |
writeback_threshold = sc->nr_to_reclaim + sc->nr_to_reclaim / 2; if (total_scanned > writeback_threshold) { |
03ba3782e
|
1750 |
wakeup_flusher_threads(laptop_mode ? 0 : total_scanned); |
66e1707bc
|
1751 |
sc->may_writepage = 1; |
1da177e4c
|
1752 1753 1754 |
} /* Take a nap, wait for some writeback to complete */ |
7b51755c3
|
1755 1756 |
if (!sc->hibernation_mode && sc->nr_scanned && priority < DEF_PRIORITY - 2) |
8aa7e847d
|
1757 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
1da177e4c
|
1758 |
} |
87547ee95
|
1759 |
/* top priority shrink_zones still had more to do? don't OOM, then */ |
e72e2bd67
|
1760 |
if (!sc->all_unreclaimable && scanning_global_lru(sc)) |
a79311c14
|
1761 |
ret = sc->nr_reclaimed; |
1da177e4c
|
1762 |
out: |
3bb1a852a
|
1763 1764 1765 1766 1767 1768 1769 1770 1771 |
/* * Now that we've scanned all the zones at this priority level, note * that level within the zone so that the next thread which performs * scanning of this zone will immediately start out at this priority * level. This affects only the decision whether or not to bring * mapped pages onto the inactive list. */ if (priority < 0) priority = 0; |
1da177e4c
|
1772 |
|
e72e2bd67
|
1773 |
if (scanning_global_lru(sc)) { |
54a6eb5c4
|
1774 |
for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) { |
1cfb419b3
|
1775 1776 1777 1778 1779 1780 1781 1782 |
if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) continue; zone->prev_priority = priority; } } else mem_cgroup_record_reclaim_priority(sc->mem_cgroup, priority); |
1da177e4c
|
1783 |
|
873b47717
|
1784 |
delayacct_freepages_end(); |
1da177e4c
|
1785 1786 |
return ret; } |
dac1d27bc
|
1787 |
unsigned long try_to_free_pages(struct zonelist *zonelist, int order, |
327c0e968
|
1788 |
gfp_t gfp_mask, nodemask_t *nodemask) |
66e1707bc
|
1789 1790 1791 1792 |
{ struct scan_control sc = { .gfp_mask = gfp_mask, .may_writepage = !laptop_mode, |
22fba3354
|
1793 |
.nr_to_reclaim = SWAP_CLUSTER_MAX, |
a6dc60f89
|
1794 |
.may_unmap = 1, |
2e2e42598
|
1795 |
.may_swap = 1, |
66e1707bc
|
1796 1797 1798 1799 |
.swappiness = vm_swappiness, .order = order, .mem_cgroup = NULL, .isolate_pages = isolate_pages_global, |
327c0e968
|
1800 |
.nodemask = nodemask, |
66e1707bc
|
1801 |
}; |
dd1a239f6
|
1802 |
return do_try_to_free_pages(zonelist, &sc); |
66e1707bc
|
1803 |
} |
00f0b8259
|
1804 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR |
66e1707bc
|
1805 |
|
4e4169535
|
1806 1807 1808 1809 1810 1811 1812 1813 1814 |
unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem, gfp_t gfp_mask, bool noswap, unsigned int swappiness, struct zone *zone, int nid) { struct scan_control sc = { .may_writepage = !laptop_mode, .may_unmap = 1, .may_swap = !noswap, |
4e4169535
|
1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 |
.swappiness = swappiness, .order = 0, .mem_cgroup = mem, .isolate_pages = mem_cgroup_isolate_pages, }; nodemask_t nm = nodemask_of_node(nid); sc.gfp_mask = (gfp_mask & GFP_RECLAIM_MASK) | (GFP_HIGHUSER_MOVABLE & ~GFP_RECLAIM_MASK); sc.nodemask = &nm; sc.nr_reclaimed = 0; sc.nr_scanned = 0; /* * NOTE: Although we can get the priority field, using it * here is not a good idea, since it limits the pages we can scan. * if we don't reclaim here, the shrink_zone from balance_pgdat * will pick up pages from other mem cgroup's as well. We hack * the priority and make it zero. */ shrink_zone(0, zone, &sc); return sc.nr_reclaimed; } |
e1a1cd590
|
1837 |
unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont, |
a7885eb8a
|
1838 1839 1840 |
gfp_t gfp_mask, bool noswap, unsigned int swappiness) |
66e1707bc
|
1841 |
{ |
4e4169535
|
1842 |
struct zonelist *zonelist; |
66e1707bc
|
1843 |
struct scan_control sc = { |
66e1707bc
|
1844 |
.may_writepage = !laptop_mode, |
a6dc60f89
|
1845 |
.may_unmap = 1, |
2e2e42598
|
1846 |
.may_swap = !noswap, |
22fba3354
|
1847 |
.nr_to_reclaim = SWAP_CLUSTER_MAX, |
a7885eb8a
|
1848 |
.swappiness = swappiness, |
66e1707bc
|
1849 1850 1851 |
.order = 0, .mem_cgroup = mem_cont, .isolate_pages = mem_cgroup_isolate_pages, |
327c0e968
|
1852 |
.nodemask = NULL, /* we don't care the placement */ |
66e1707bc
|
1853 |
}; |
66e1707bc
|
1854 |
|
dd1a239f6
|
1855 1856 1857 1858 |
sc.gfp_mask = (gfp_mask & GFP_RECLAIM_MASK) | (GFP_HIGHUSER_MOVABLE & ~GFP_RECLAIM_MASK); zonelist = NODE_DATA(numa_node_id())->node_zonelists; return do_try_to_free_pages(zonelist, &sc); |
66e1707bc
|
1859 1860 |
} #endif |
f50de2d38
|
1861 |
/* is kswapd sleeping prematurely? */ |
bb3ab5968
|
1862 |
static int sleeping_prematurely(pg_data_t *pgdat, int order, long remaining) |
f50de2d38
|
1863 |
{ |
bb3ab5968
|
1864 |
int i; |
f50de2d38
|
1865 1866 1867 1868 1869 1870 |
/* If a direct reclaimer woke kswapd within HZ/10, it's premature */ if (remaining) return 1; /* If after HZ/10, a zone is below the high mark, it's premature */ |
bb3ab5968
|
1871 1872 1873 1874 1875 |
for (i = 0; i < pgdat->nr_zones; i++) { struct zone *zone = pgdat->node_zones + i; if (!populated_zone(zone)) continue; |
93e4a89a8
|
1876 |
if (zone->all_unreclaimable) |
de3fab393
|
1877 |
continue; |
f50de2d38
|
1878 1879 1880 |
if (!zone_watermark_ok(zone, order, high_wmark_pages(zone), 0, 0)) return 1; |
bb3ab5968
|
1881 |
} |
f50de2d38
|
1882 1883 1884 |
return 0; } |
1da177e4c
|
1885 1886 |
/* * For kswapd, balance_pgdat() will work across all this node's zones until |
418589663
|
1887 |
* they are all at high_wmark_pages(zone). |
1da177e4c
|
1888 |
* |
1da177e4c
|
1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 |
* Returns the number of pages which were actually freed. * * There is special handling here for zones which are full of pinned pages. * This can happen if the pages are all mlocked, or if they are all used by * device drivers (say, ZONE_DMA). Or if they are all in use by hugetlb. * What we do is to detect the case where all pages in the zone have been * scanned twice and there has been zero successful reclaim. Mark the zone as * dead and from now on, only perform a short scan. Basically we're polling * the zone for when the problem goes away. * * kswapd scans the zones in the highmem->normal->dma direction. It skips |
418589663
|
1900 1901 1902 1903 1904 |
* zones which have free_pages > high_wmark_pages(zone), but once a zone is * found to have free_pages <= high_wmark_pages(zone), we scan that zone and the * lower zones regardless of the number of free pages in the lower zones. This * interoperates with the page allocator fallback scheme to ensure that aging * of pages is balanced across the zones. |
1da177e4c
|
1905 |
*/ |
d6277db4a
|
1906 |
static unsigned long balance_pgdat(pg_data_t *pgdat, int order) |
1da177e4c
|
1907 |
{ |
1da177e4c
|
1908 1909 1910 |
int all_zones_ok; int priority; int i; |
69e05944a
|
1911 |
unsigned long total_scanned; |
1da177e4c
|
1912 |
struct reclaim_state *reclaim_state = current->reclaim_state; |
179e96395
|
1913 1914 |
struct scan_control sc = { .gfp_mask = GFP_KERNEL, |
a6dc60f89
|
1915 |
.may_unmap = 1, |
2e2e42598
|
1916 |
.may_swap = 1, |
22fba3354
|
1917 1918 1919 1920 1921 |
/* * kswapd doesn't want to be bailed out while reclaim. because * we want to put equal scanning pressure on each zone. */ .nr_to_reclaim = ULONG_MAX, |
d6277db4a
|
1922 |
.swappiness = vm_swappiness, |
5ad333eb6
|
1923 |
.order = order, |
66e1707bc
|
1924 1925 |
.mem_cgroup = NULL, .isolate_pages = isolate_pages_global, |
179e96395
|
1926 |
}; |
3bb1a852a
|
1927 1928 |
/* * temp_priority is used to remember the scanning priority at which |
418589663
|
1929 1930 |
* this zone was successfully refilled to * free_pages == high_wmark_pages(zone). |
3bb1a852a
|
1931 1932 |
*/ int temp_priority[MAX_NR_ZONES]; |
1da177e4c
|
1933 1934 1935 |
loop_again: total_scanned = 0; |
a79311c14
|
1936 |
sc.nr_reclaimed = 0; |
c0bbbc73d
|
1937 |
sc.may_writepage = !laptop_mode; |
f8891e5e1
|
1938 |
count_vm_event(PAGEOUTRUN); |
1da177e4c
|
1939 |
|
3bb1a852a
|
1940 1941 |
for (i = 0; i < pgdat->nr_zones; i++) temp_priority[i] = DEF_PRIORITY; |
1da177e4c
|
1942 1943 1944 1945 |
for (priority = DEF_PRIORITY; priority >= 0; priority--) { int end_zone = 0; /* Inclusive. 0 = ZONE_DMA */ unsigned long lru_pages = 0; |
bb3ab5968
|
1946 |
int has_under_min_watermark_zone = 0; |
1da177e4c
|
1947 |
|
f7b7fd8f3
|
1948 1949 1950 |
/* The swap token gets in the way of swapout... */ if (!priority) disable_swap_token(); |
1da177e4c
|
1951 |
all_zones_ok = 1; |
d6277db4a
|
1952 1953 1954 1955 1956 1957 |
/* * Scan in the highmem->dma direction for the highest * zone which needs scanning */ for (i = pgdat->nr_zones - 1; i >= 0; i--) { struct zone *zone = pgdat->node_zones + i; |
1da177e4c
|
1958 |
|
d6277db4a
|
1959 1960 |
if (!populated_zone(zone)) continue; |
1da177e4c
|
1961 |
|
93e4a89a8
|
1962 |
if (zone->all_unreclaimable && priority != DEF_PRIORITY) |
d6277db4a
|
1963 |
continue; |
1da177e4c
|
1964 |
|
556adecba
|
1965 1966 1967 1968 |
/* * Do some background aging of the anon list, to give * pages a chance to be referenced before reclaiming. */ |
14797e236
|
1969 |
if (inactive_anon_is_low(zone, &sc)) |
556adecba
|
1970 1971 |
shrink_active_list(SWAP_CLUSTER_MAX, zone, &sc, priority, 0); |
418589663
|
1972 1973 |
if (!zone_watermark_ok(zone, order, high_wmark_pages(zone), 0, 0)) { |
d6277db4a
|
1974 |
end_zone = i; |
e1dbeda60
|
1975 |
break; |
1da177e4c
|
1976 |
} |
1da177e4c
|
1977 |
} |
e1dbeda60
|
1978 1979 |
if (i < 0) goto out; |
1da177e4c
|
1980 1981 |
for (i = 0; i <= end_zone; i++) { struct zone *zone = pgdat->node_zones + i; |
adea02a1b
|
1982 |
lru_pages += zone_reclaimable_pages(zone); |
1da177e4c
|
1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 |
} /* * Now scan the zone in the dma->highmem direction, stopping * at the last zone which needs scanning. * * We do this because the page allocator works in the opposite * direction. This prevents the page allocator from allocating * pages behind kswapd's direction of progress, which would * cause too much scanning of the lower zones. */ for (i = 0; i <= end_zone; i++) { struct zone *zone = pgdat->node_zones + i; |
b15e0905f
|
1996 |
int nr_slab; |
4e4169535
|
1997 |
int nid, zid; |
1da177e4c
|
1998 |
|
f3fe65122
|
1999 |
if (!populated_zone(zone)) |
1da177e4c
|
2000 |
continue; |
93e4a89a8
|
2001 |
if (zone->all_unreclaimable && priority != DEF_PRIORITY) |
1da177e4c
|
2002 |
continue; |
3bb1a852a
|
2003 |
temp_priority[i] = priority; |
1da177e4c
|
2004 |
sc.nr_scanned = 0; |
3bb1a852a
|
2005 |
note_zone_scanning_priority(zone, priority); |
4e4169535
|
2006 2007 2008 2009 2010 2011 2012 2013 2014 |
nid = pgdat->node_id; zid = zone_idx(zone); /* * Call soft limit reclaim before calling shrink_zone. * For now we ignore the return value */ mem_cgroup_soft_limit_reclaim(zone, order, sc.gfp_mask, nid, zid); |
32a4330d4
|
2015 2016 2017 2018 |
/* * We put equal pressure on every zone, unless one * zone has way too many pages free already. */ |
418589663
|
2019 2020 |
if (!zone_watermark_ok(zone, order, 8*high_wmark_pages(zone), end_zone, 0)) |
a79311c14
|
2021 |
shrink_zone(priority, zone, &sc); |
1da177e4c
|
2022 |
reclaim_state->reclaimed_slab = 0; |
b15e0905f
|
2023 2024 |
nr_slab = shrink_slab(sc.nr_scanned, GFP_KERNEL, lru_pages); |
a79311c14
|
2025 |
sc.nr_reclaimed += reclaim_state->reclaimed_slab; |
1da177e4c
|
2026 |
total_scanned += sc.nr_scanned; |
93e4a89a8
|
2027 |
if (zone->all_unreclaimable) |
1da177e4c
|
2028 |
continue; |
93e4a89a8
|
2029 2030 2031 |
if (nr_slab == 0 && zone->pages_scanned >= (zone_reclaimable_pages(zone) * 6)) zone->all_unreclaimable = 1; |
1da177e4c
|
2032 2033 2034 2035 2036 2037 |
/* * If we've done a decent amount of scanning and * the reclaim ratio is low, start doing writepage * even in laptop mode */ if (total_scanned > SWAP_CLUSTER_MAX * 2 && |
a79311c14
|
2038 |
total_scanned > sc.nr_reclaimed + sc.nr_reclaimed / 2) |
1da177e4c
|
2039 |
sc.may_writepage = 1; |
bb3ab5968
|
2040 |
|
45973d74f
|
2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 |
if (!zone_watermark_ok(zone, order, high_wmark_pages(zone), end_zone, 0)) { all_zones_ok = 0; /* * We are still under min water mark. This * means that we have a GFP_ATOMIC allocation * failure risk. Hurry up! */ if (!zone_watermark_ok(zone, order, min_wmark_pages(zone), end_zone, 0)) has_under_min_watermark_zone = 1; } |
bb3ab5968
|
2053 |
|
1da177e4c
|
2054 |
} |
1da177e4c
|
2055 2056 2057 2058 2059 2060 |
if (all_zones_ok) break; /* kswapd: all done */ /* * OK, kswapd is getting into trouble. Take a nap, then take * another pass across the zones. */ |
bb3ab5968
|
2061 2062 2063 2064 2065 2066 |
if (total_scanned && (priority < DEF_PRIORITY - 2)) { if (has_under_min_watermark_zone) count_vm_event(KSWAPD_SKIP_CONGESTION_WAIT); else congestion_wait(BLK_RW_ASYNC, HZ/10); } |
1da177e4c
|
2067 2068 2069 2070 2071 2072 2073 |
/* * We do this so kswapd doesn't build up large priorities for * example when it is freeing in parallel with allocators. It * matches the direct reclaim path behaviour in terms of impact * on zone->*_priority. */ |
a79311c14
|
2074 |
if (sc.nr_reclaimed >= SWAP_CLUSTER_MAX) |
1da177e4c
|
2075 2076 2077 |
break; } out: |
3bb1a852a
|
2078 2079 2080 2081 2082 |
/* * Note within each zone the priority level at which this zone was * brought into a happy state. So that the next thread which scans this * zone will start out at that priority level. */ |
1da177e4c
|
2083 2084 |
for (i = 0; i < pgdat->nr_zones; i++) { struct zone *zone = pgdat->node_zones + i; |
3bb1a852a
|
2085 |
zone->prev_priority = temp_priority[i]; |
1da177e4c
|
2086 2087 2088 |
} if (!all_zones_ok) { cond_resched(); |
8357376d3
|
2089 2090 |
try_to_freeze(); |
73ce02e96
|
2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 |
/* * Fragmentation may mean that the system cannot be * rebalanced for high-order allocations in all zones. * At this point, if nr_reclaimed < SWAP_CLUSTER_MAX, * it means the zones have been fully scanned and are still * not balanced. For high-order allocations, there is * little point trying all over again as kswapd may * infinite loop. * * Instead, recheck all watermarks at order-0 as they * are the most important. If watermarks are ok, kswapd will go * back to sleep. High-order users can still perform direct * reclaim if they wish. */ if (sc.nr_reclaimed < SWAP_CLUSTER_MAX) order = sc.order = 0; |
1da177e4c
|
2107 2108 |
goto loop_again; } |
a79311c14
|
2109 |
return sc.nr_reclaimed; |
1da177e4c
|
2110 2111 2112 2113 |
} /* * The background pageout daemon, started as a kernel thread |
4f98a2fee
|
2114 |
* from the init process. |
1da177e4c
|
2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 |
* * This basically trickles out pages so that we have _some_ * free memory available even if there is no other activity * that frees anything up. This is needed for things like routing * etc, where we otherwise might have all activity going on in * asynchronous contexts that cannot page things out. * * If there are applications that are active memory-allocators * (most normal use), this basically shouldn't matter. */ static int kswapd(void *p) { unsigned long order; pg_data_t *pgdat = (pg_data_t*)p; struct task_struct *tsk = current; DEFINE_WAIT(wait); struct reclaim_state reclaim_state = { .reclaimed_slab = 0, }; |
a70f73028
|
2134 |
const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id); |
1da177e4c
|
2135 |
|
cf40bd16f
|
2136 |
lockdep_set_current_reclaim_state(GFP_KERNEL); |
174596a0b
|
2137 |
if (!cpumask_empty(cpumask)) |
c5f59f083
|
2138 |
set_cpus_allowed_ptr(tsk, cpumask); |
1da177e4c
|
2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 |
current->reclaim_state = &reclaim_state; /* * Tell the memory management that we're a "memory allocator", * and that if we need more memory we should get access to it * regardless (see "__alloc_pages()"). "kswapd" should * never get caught in the normal page freeing logic. * * (Kswapd normally doesn't need memory anyway, but sometimes * you need a small amount of memory in order to be able to * page out something else, and this flag essentially protects * us from recursively trying to free more memory as we're * trying to free the first piece of memory in the first place). */ |
930d91525
|
2153 |
tsk->flags |= PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD; |
831441862
|
2154 |
set_freezable(); |
1da177e4c
|
2155 2156 2157 2158 |
order = 0; for ( ; ; ) { unsigned long new_order; |
8fe23e057
|
2159 |
int ret; |
3e1d1d28d
|
2160 |
|
1da177e4c
|
2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 |
prepare_to_wait(&pgdat->kswapd_wait, &wait, TASK_INTERRUPTIBLE); new_order = pgdat->kswapd_max_order; pgdat->kswapd_max_order = 0; if (order < new_order) { /* * Don't sleep if someone wants a larger 'order' * allocation */ order = new_order; } else { |
f50de2d38
|
2171 2172 2173 2174 |
if (!freezing(current) && !kthread_should_stop()) { long remaining = 0; /* Try to sleep for a short interval */ |
bb3ab5968
|
2175 |
if (!sleeping_prematurely(pgdat, order, remaining)) { |
f50de2d38
|
2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 |
remaining = schedule_timeout(HZ/10); finish_wait(&pgdat->kswapd_wait, &wait); prepare_to_wait(&pgdat->kswapd_wait, &wait, TASK_INTERRUPTIBLE); } /* * After a short sleep, check if it was a * premature sleep. If not, then go fully * to sleep until explicitly woken up */ |
bb3ab5968
|
2186 |
if (!sleeping_prematurely(pgdat, order, remaining)) |
f50de2d38
|
2187 2188 2189 |
schedule(); else { if (remaining) |
bb3ab5968
|
2190 |
count_vm_event(KSWAPD_LOW_WMARK_HIT_QUICKLY); |
f50de2d38
|
2191 |
else |
bb3ab5968
|
2192 |
count_vm_event(KSWAPD_HIGH_WMARK_HIT_QUICKLY); |
f50de2d38
|
2193 2194 |
} } |
b1296cc48
|
2195 |
|
1da177e4c
|
2196 2197 2198 |
order = pgdat->kswapd_max_order; } finish_wait(&pgdat->kswapd_wait, &wait); |
8fe23e057
|
2199 2200 2201 2202 2203 2204 2205 2206 2207 |
ret = try_to_freeze(); if (kthread_should_stop()) break; /* * We can speed up thawing tasks if we don't call balance_pgdat * after returning from the refrigerator */ if (!ret) |
b1296cc48
|
2208 |
balance_pgdat(pgdat, order); |
1da177e4c
|
2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 |
} return 0; } /* * A zone is low on free memory, so wake its kswapd task to service it. */ void wakeup_kswapd(struct zone *zone, int order) { pg_data_t *pgdat; |
f3fe65122
|
2219 |
if (!populated_zone(zone)) |
1da177e4c
|
2220 2221 2222 |
return; pgdat = zone->zone_pgdat; |
418589663
|
2223 |
if (zone_watermark_ok(zone, order, low_wmark_pages(zone), 0, 0)) |
1da177e4c
|
2224 2225 2226 |
return; if (pgdat->kswapd_max_order < order) pgdat->kswapd_max_order = order; |
02a0e53d8
|
2227 |
if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) |
1da177e4c
|
2228 |
return; |
8d0986e28
|
2229 |
if (!waitqueue_active(&pgdat->kswapd_wait)) |
1da177e4c
|
2230 |
return; |
8d0986e28
|
2231 |
wake_up_interruptible(&pgdat->kswapd_wait); |
1da177e4c
|
2232 |
} |
adea02a1b
|
2233 2234 2235 2236 2237 2238 2239 2240 |
/* * The reclaimable count would be mostly accurate. * The less reclaimable pages may be * - mlocked pages, which will be moved to unevictable list when encountered * - mapped pages, which may require several travels to be reclaimed * - dirty pages, which is not "instantly" reclaimable */ unsigned long global_reclaimable_pages(void) |
4f98a2fee
|
2241 |
{ |
adea02a1b
|
2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 |
int nr; nr = global_page_state(NR_ACTIVE_FILE) + global_page_state(NR_INACTIVE_FILE); if (nr_swap_pages > 0) nr += global_page_state(NR_ACTIVE_ANON) + global_page_state(NR_INACTIVE_ANON); return nr; } unsigned long zone_reclaimable_pages(struct zone *zone) { int nr; nr = zone_page_state(zone, NR_ACTIVE_FILE) + zone_page_state(zone, NR_INACTIVE_FILE); if (nr_swap_pages > 0) nr += zone_page_state(zone, NR_ACTIVE_ANON) + zone_page_state(zone, NR_INACTIVE_ANON); return nr; |
4f98a2fee
|
2266 |
} |
c6f37f121
|
2267 |
#ifdef CONFIG_HIBERNATION |
1da177e4c
|
2268 |
/* |
7b51755c3
|
2269 |
* Try to free `nr_to_reclaim' of memory, system-wide, and return the number of |
d6277db4a
|
2270 2271 2272 2273 2274 |
* freed pages. * * Rather than trying to age LRUs the aim is to preserve the overall * LRU order by reclaiming preferentially * inactive > active > active referenced > active mapped |
1da177e4c
|
2275 |
*/ |
7b51755c3
|
2276 |
unsigned long shrink_all_memory(unsigned long nr_to_reclaim) |
1da177e4c
|
2277 |
{ |
d6277db4a
|
2278 |
struct reclaim_state reclaim_state; |
d6277db4a
|
2279 |
struct scan_control sc = { |
7b51755c3
|
2280 2281 2282 |
.gfp_mask = GFP_HIGHUSER_MOVABLE, .may_swap = 1, .may_unmap = 1, |
d6277db4a
|
2283 |
.may_writepage = 1, |
7b51755c3
|
2284 2285 2286 2287 |
.nr_to_reclaim = nr_to_reclaim, .hibernation_mode = 1, .swappiness = vm_swappiness, .order = 0, |
66e1707bc
|
2288 |
.isolate_pages = isolate_pages_global, |
1da177e4c
|
2289 |
}; |
7b51755c3
|
2290 2291 2292 |
struct zonelist * zonelist = node_zonelist(numa_node_id(), sc.gfp_mask); struct task_struct *p = current; unsigned long nr_reclaimed; |
1da177e4c
|
2293 |
|
7b51755c3
|
2294 2295 2296 2297 |
p->flags |= PF_MEMALLOC; lockdep_set_current_reclaim_state(sc.gfp_mask); reclaim_state.reclaimed_slab = 0; p->reclaim_state = &reclaim_state; |
d6277db4a
|
2298 |
|
7b51755c3
|
2299 |
nr_reclaimed = do_try_to_free_pages(zonelist, &sc); |
d979677c4
|
2300 |
|
7b51755c3
|
2301 2302 2303 |
p->reclaim_state = NULL; lockdep_clear_current_reclaim_state(); p->flags &= ~PF_MEMALLOC; |
d6277db4a
|
2304 |
|
7b51755c3
|
2305 |
return nr_reclaimed; |
1da177e4c
|
2306 |
} |
c6f37f121
|
2307 |
#endif /* CONFIG_HIBERNATION */ |
1da177e4c
|
2308 |
|
1da177e4c
|
2309 2310 2311 2312 |
/* It's optimal to keep kswapds on the same CPUs as their memory, but not required for correctness. So if the last cpu in a node goes away, we get changed to run anywhere: as the first one comes back, restore their cpu bindings. */ |
9c7b216d2
|
2313 |
static int __devinit cpu_callback(struct notifier_block *nfb, |
69e05944a
|
2314 |
unsigned long action, void *hcpu) |
1da177e4c
|
2315 |
{ |
58c0a4a78
|
2316 |
int nid; |
1da177e4c
|
2317 |
|
8bb784428
|
2318 |
if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN) { |
58c0a4a78
|
2319 |
for_each_node_state(nid, N_HIGH_MEMORY) { |
c5f59f083
|
2320 |
pg_data_t *pgdat = NODE_DATA(nid); |
a70f73028
|
2321 2322 2323 |
const struct cpumask *mask; mask = cpumask_of_node(pgdat->node_id); |
c5f59f083
|
2324 |
|
3e5979453
|
2325 |
if (cpumask_any_and(cpu_online_mask, mask) < nr_cpu_ids) |
1da177e4c
|
2326 |
/* One of our CPUs online: restore mask */ |
c5f59f083
|
2327 |
set_cpus_allowed_ptr(pgdat->kswapd, mask); |
1da177e4c
|
2328 2329 2330 2331 |
} } return NOTIFY_OK; } |
1da177e4c
|
2332 |
|
3218ae14b
|
2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 |
/* * This kswapd start function will be called by init and node-hot-add. * On node-hot-add, kswapd will moved to proper cpus if cpus are hot-added. */ int kswapd_run(int nid) { pg_data_t *pgdat = NODE_DATA(nid); int ret = 0; if (pgdat->kswapd) return 0; pgdat->kswapd = kthread_run(kswapd, pgdat, "kswapd%d", nid); if (IS_ERR(pgdat->kswapd)) { /* failure at boot is fatal */ BUG_ON(system_state == SYSTEM_BOOTING); printk("Failed to start kswapd on node %d ",nid); ret = -1; } return ret; } |
8fe23e057
|
2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 |
/* * Called by memory hotplug when all memory in a node is offlined. */ void kswapd_stop(int nid) { struct task_struct *kswapd = NODE_DATA(nid)->kswapd; if (kswapd) kthread_stop(kswapd); } |
1da177e4c
|
2365 2366 |
static int __init kswapd_init(void) { |
3218ae14b
|
2367 |
int nid; |
69e05944a
|
2368 |
|
1da177e4c
|
2369 |
swap_setup(); |
9422ffba4
|
2370 |
for_each_node_state(nid, N_HIGH_MEMORY) |
3218ae14b
|
2371 |
kswapd_run(nid); |
1da177e4c
|
2372 2373 2374 2375 2376 |
hotcpu_notifier(cpu_callback, 0); return 0; } module_init(kswapd_init) |
9eeff2395
|
2377 2378 2379 2380 2381 2382 2383 |
#ifdef CONFIG_NUMA /* * Zone reclaim mode * * If non-zero call zone_reclaim when the number of free pages falls below * the watermarks. |
9eeff2395
|
2384 2385 |
*/ int zone_reclaim_mode __read_mostly; |
1b2ffb789
|
2386 |
#define RECLAIM_OFF 0 |
7d03431cf
|
2387 |
#define RECLAIM_ZONE (1<<0) /* Run shrink_inactive_list on the zone */ |
1b2ffb789
|
2388 2389 |
#define RECLAIM_WRITE (1<<1) /* Writeout pages during reclaim */ #define RECLAIM_SWAP (1<<2) /* Swap pages out during reclaim */ |
9eeff2395
|
2390 |
/* |
a92f71263
|
2391 2392 2393 2394 2395 |
* Priority for ZONE_RECLAIM. This determines the fraction of pages * of a node considered for each zone_reclaim. 4 scans 1/16th of * a zone. */ #define ZONE_RECLAIM_PRIORITY 4 |
9eeff2395
|
2396 |
/* |
9614634fe
|
2397 2398 2399 2400 2401 2402 |
* Percentage of pages in a zone that must be unmapped for zone_reclaim to * occur. */ int sysctl_min_unmapped_ratio = 1; /* |
0ff38490c
|
2403 2404 2405 2406 |
* If the number of slab pages in a zone grows beyond this percentage then * slab reclaim needs to occur. */ int sysctl_min_slab_ratio = 5; |
90afa5de6
|
2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 |
static inline unsigned long zone_unmapped_file_pages(struct zone *zone) { unsigned long file_mapped = zone_page_state(zone, NR_FILE_MAPPED); unsigned long file_lru = zone_page_state(zone, NR_INACTIVE_FILE) + zone_page_state(zone, NR_ACTIVE_FILE); /* * It's possible for there to be more file mapped pages than * accounted for by the pages on the file LRU lists because * tmpfs pages accounted for as ANON can also be FILE_MAPPED */ return (file_lru > file_mapped) ? (file_lru - file_mapped) : 0; } /* Work out how many page cache pages we can reclaim in this reclaim_mode */ static long zone_pagecache_reclaimable(struct zone *zone) { long nr_pagecache_reclaimable; long delta = 0; /* * If RECLAIM_SWAP is set, then all file pages are considered * potentially reclaimable. Otherwise, we have to worry about * pages like swapcache and zone_unmapped_file_pages() provides * a better estimate */ if (zone_reclaim_mode & RECLAIM_SWAP) nr_pagecache_reclaimable = zone_page_state(zone, NR_FILE_PAGES); else nr_pagecache_reclaimable = zone_unmapped_file_pages(zone); /* If we can't clean pages, remove dirty pages from consideration */ if (!(zone_reclaim_mode & RECLAIM_WRITE)) delta += zone_page_state(zone, NR_FILE_DIRTY); /* Watch for any possible underflows due to delta */ if (unlikely(delta > nr_pagecache_reclaimable)) delta = nr_pagecache_reclaimable; return nr_pagecache_reclaimable - delta; } |
0ff38490c
|
2448 |
/* |
9eeff2395
|
2449 2450 |
* Try to free up some pages from this zone through reclaim. */ |
179e96395
|
2451 |
static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order) |
9eeff2395
|
2452 |
{ |
7fb2d46d3
|
2453 |
/* Minimum pages needed in order to stay on node */ |
69e05944a
|
2454 |
const unsigned long nr_pages = 1 << order; |
9eeff2395
|
2455 2456 |
struct task_struct *p = current; struct reclaim_state reclaim_state; |
8695949a1
|
2457 |
int priority; |
179e96395
|
2458 2459 |
struct scan_control sc = { .may_writepage = !!(zone_reclaim_mode & RECLAIM_WRITE), |
a6dc60f89
|
2460 |
.may_unmap = !!(zone_reclaim_mode & RECLAIM_SWAP), |
2e2e42598
|
2461 |
.may_swap = 1, |
22fba3354
|
2462 2463 |
.nr_to_reclaim = max_t(unsigned long, nr_pages, SWAP_CLUSTER_MAX), |
179e96395
|
2464 |
.gfp_mask = gfp_mask, |
d6277db4a
|
2465 |
.swappiness = vm_swappiness, |
bd2f6199c
|
2466 |
.order = order, |
66e1707bc
|
2467 |
.isolate_pages = isolate_pages_global, |
179e96395
|
2468 |
}; |
83e33a471
|
2469 |
unsigned long slab_reclaimable; |
9eeff2395
|
2470 2471 |
disable_swap_token(); |
9eeff2395
|
2472 |
cond_resched(); |
d4f7796e9
|
2473 2474 2475 2476 2477 2478 |
/* * We need to be able to allocate from the reserves for RECLAIM_SWAP * and we also need to be able to write out pages for RECLAIM_WRITE * and RECLAIM_SWAP. */ p->flags |= PF_MEMALLOC | PF_SWAPWRITE; |
76ca542d8
|
2479 |
lockdep_set_current_reclaim_state(gfp_mask); |
9eeff2395
|
2480 2481 |
reclaim_state.reclaimed_slab = 0; p->reclaim_state = &reclaim_state; |
c84db23c6
|
2482 |
|
90afa5de6
|
2483 |
if (zone_pagecache_reclaimable(zone) > zone->min_unmapped_pages) { |
0ff38490c
|
2484 2485 2486 2487 2488 2489 |
/* * Free memory by calling shrink zone with increasing * priorities until we have enough memory freed. */ priority = ZONE_RECLAIM_PRIORITY; do { |
3bb1a852a
|
2490 |
note_zone_scanning_priority(zone, priority); |
a79311c14
|
2491 |
shrink_zone(priority, zone, &sc); |
0ff38490c
|
2492 |
priority--; |
a79311c14
|
2493 |
} while (priority >= 0 && sc.nr_reclaimed < nr_pages); |
0ff38490c
|
2494 |
} |
c84db23c6
|
2495 |
|
83e33a471
|
2496 2497 |
slab_reclaimable = zone_page_state(zone, NR_SLAB_RECLAIMABLE); if (slab_reclaimable > zone->min_slab_pages) { |
2a16e3f4b
|
2498 |
/* |
7fb2d46d3
|
2499 |
* shrink_slab() does not currently allow us to determine how |
0ff38490c
|
2500 2501 2502 2503 |
* many pages were freed in this zone. So we take the current * number of slab pages and shake the slab until it is reduced * by the same nr_pages that we used for reclaiming unmapped * pages. |
2a16e3f4b
|
2504 |
* |
0ff38490c
|
2505 2506 |
* Note that shrink_slab will free memory on all zones and may * take a long time. |
2a16e3f4b
|
2507 |
*/ |
0ff38490c
|
2508 |
while (shrink_slab(sc.nr_scanned, gfp_mask, order) && |
83e33a471
|
2509 2510 |
zone_page_state(zone, NR_SLAB_RECLAIMABLE) > slab_reclaimable - nr_pages) |
0ff38490c
|
2511 |
; |
83e33a471
|
2512 2513 2514 2515 2516 |
/* * Update nr_reclaimed by the number of slab pages we * reclaimed from this zone. */ |
a79311c14
|
2517 |
sc.nr_reclaimed += slab_reclaimable - |
83e33a471
|
2518 |
zone_page_state(zone, NR_SLAB_RECLAIMABLE); |
2a16e3f4b
|
2519 |
} |
9eeff2395
|
2520 |
p->reclaim_state = NULL; |
d4f7796e9
|
2521 |
current->flags &= ~(PF_MEMALLOC | PF_SWAPWRITE); |
76ca542d8
|
2522 |
lockdep_clear_current_reclaim_state(); |
a79311c14
|
2523 |
return sc.nr_reclaimed >= nr_pages; |
9eeff2395
|
2524 |
} |
179e96395
|
2525 2526 2527 |
int zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order) { |
179e96395
|
2528 |
int node_id; |
d773ed6b8
|
2529 |
int ret; |
179e96395
|
2530 2531 |
/* |
0ff38490c
|
2532 2533 |
* Zone reclaim reclaims unmapped file backed pages and * slab pages if we are over the defined limits. |
34aa1330f
|
2534 |
* |
9614634fe
|
2535 2536 2537 2538 2539 |
* A small portion of unmapped file backed pages is needed for * file I/O otherwise pages read by file I/O will be immediately * thrown out if the zone is overallocated. So we do not reclaim * if less than a specified percentage of the zone is used by * unmapped file backed pages. |
179e96395
|
2540 |
*/ |
90afa5de6
|
2541 2542 |
if (zone_pagecache_reclaimable(zone) <= zone->min_unmapped_pages && zone_page_state(zone, NR_SLAB_RECLAIMABLE) <= zone->min_slab_pages) |
fa5e084e4
|
2543 |
return ZONE_RECLAIM_FULL; |
179e96395
|
2544 |
|
93e4a89a8
|
2545 |
if (zone->all_unreclaimable) |
fa5e084e4
|
2546 |
return ZONE_RECLAIM_FULL; |
d773ed6b8
|
2547 |
|
179e96395
|
2548 |
/* |
d773ed6b8
|
2549 |
* Do not scan if the allocation should not be delayed. |
179e96395
|
2550 |
*/ |
d773ed6b8
|
2551 |
if (!(gfp_mask & __GFP_WAIT) || (current->flags & PF_MEMALLOC)) |
fa5e084e4
|
2552 |
return ZONE_RECLAIM_NOSCAN; |
179e96395
|
2553 2554 2555 2556 2557 2558 2559 |
/* * Only run zone reclaim on the local zone or on zones that do not * have associated processors. This will favor the local processor * over remote processors and spread off node memory allocations * as wide as possible. */ |
89fa30242
|
2560 |
node_id = zone_to_nid(zone); |
37c0708db
|
2561 |
if (node_state(node_id, N_CPU) && node_id != numa_node_id()) |
fa5e084e4
|
2562 |
return ZONE_RECLAIM_NOSCAN; |
d773ed6b8
|
2563 2564 |
if (zone_test_and_set_flag(zone, ZONE_RECLAIM_LOCKED)) |
fa5e084e4
|
2565 |
return ZONE_RECLAIM_NOSCAN; |
d773ed6b8
|
2566 2567 |
ret = __zone_reclaim(zone, gfp_mask, order); zone_clear_flag(zone, ZONE_RECLAIM_LOCKED); |
24cf72518
|
2568 2569 |
if (!ret) count_vm_event(PGSCAN_ZONE_RECLAIM_FAILED); |
d773ed6b8
|
2570 |
return ret; |
179e96395
|
2571 |
} |
9eeff2395
|
2572 |
#endif |
894bc3104
|
2573 |
|
894bc3104
|
2574 2575 2576 2577 2578 2579 |
/* * page_evictable - test whether a page is evictable * @page: the page to test * @vma: the VMA in which the page is or will be mapped, may be NULL * * Test whether page is evictable--i.e., should be placed on active/inactive |
b291f0003
|
2580 2581 |
* lists vs unevictable list. The vma argument is !NULL when called from the * fault path to determine how to instantate a new page. |
894bc3104
|
2582 2583 |
* * Reasons page might not be evictable: |
ba9ddf493
|
2584 |
* (1) page's mapping marked unevictable |
b291f0003
|
2585 |
* (2) page is part of an mlocked VMA |
ba9ddf493
|
2586 |
* |
894bc3104
|
2587 2588 2589 |
*/ int page_evictable(struct page *page, struct vm_area_struct *vma) { |
ba9ddf493
|
2590 2591 |
if (mapping_unevictable(page_mapping(page))) return 0; |
b291f0003
|
2592 2593 |
if (PageMlocked(page) || (vma && is_mlocked_vma(vma, page))) return 0; |
894bc3104
|
2594 2595 2596 |
return 1; } |
89e004ea5
|
2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 |
/** * check_move_unevictable_page - check page for evictability and move to appropriate zone lru list * @page: page to check evictability and move to appropriate lru list * @zone: zone page is in * * Checks a page for evictability and moves the page to the appropriate * zone lru list. * * Restrictions: zone->lru_lock must be held, page must be on LRU and must * have PageUnevictable set. */ static void check_move_unevictable_page(struct page *page, struct zone *zone) { VM_BUG_ON(PageActive(page)); retry: ClearPageUnevictable(page); if (page_evictable(page, NULL)) { |
401a8e1c1
|
2616 |
enum lru_list l = page_lru_base_type(page); |
af936a160
|
2617 |
|
89e004ea5
|
2618 2619 |
__dec_zone_state(zone, NR_UNEVICTABLE); list_move(&page->lru, &zone->lru[l].list); |
08e552c69
|
2620 |
mem_cgroup_move_lists(page, LRU_UNEVICTABLE, l); |
89e004ea5
|
2621 2622 2623 2624 2625 2626 2627 2628 |
__inc_zone_state(zone, NR_INACTIVE_ANON + l); __count_vm_event(UNEVICTABLE_PGRESCUED); } else { /* * rotate unevictable list */ SetPageUnevictable(page); list_move(&page->lru, &zone->lru[LRU_UNEVICTABLE].list); |
08e552c69
|
2629 |
mem_cgroup_rotate_lru_list(page, LRU_UNEVICTABLE); |
89e004ea5
|
2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 |
if (page_evictable(page, NULL)) goto retry; } } /** * scan_mapping_unevictable_pages - scan an address space for evictable pages * @mapping: struct address_space to scan for evictable pages * * Scan all pages in mapping. Check unevictable pages for * evictability and move them to the appropriate zone lru list. */ void scan_mapping_unevictable_pages(struct address_space *mapping) { pgoff_t next = 0; pgoff_t end = (i_size_read(mapping->host) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; struct zone *zone; struct pagevec pvec; if (mapping->nrpages == 0) return; pagevec_init(&pvec, 0); while (next < end && pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE)) { int i; int pg_scanned = 0; zone = NULL; for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; pgoff_t page_index = page->index; struct zone *pagezone = page_zone(page); pg_scanned++; if (page_index > next) next = page_index; next++; if (pagezone != zone) { if (zone) spin_unlock_irq(&zone->lru_lock); zone = pagezone; spin_lock_irq(&zone->lru_lock); } if (PageLRU(page) && PageUnevictable(page)) check_move_unevictable_page(page, zone); } if (zone) spin_unlock_irq(&zone->lru_lock); pagevec_release(&pvec); count_vm_events(UNEVICTABLE_PGSCANNED, pg_scanned); } } |
af936a160
|
2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 |
/** * scan_zone_unevictable_pages - check unevictable list for evictable pages * @zone - zone of which to scan the unevictable list * * Scan @zone's unevictable LRU lists to check for pages that have become * evictable. Move those that have to @zone's inactive list where they * become candidates for reclaim, unless shrink_inactive_zone() decides * to reactivate them. Pages that are still unevictable are rotated * back onto @zone's unevictable list. */ #define SCAN_UNEVICTABLE_BATCH_SIZE 16UL /* arbitrary lock hold batch size */ |
14b90b22e
|
2701 |
static void scan_zone_unevictable_pages(struct zone *zone) |
af936a160
|
2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 |
{ struct list_head *l_unevictable = &zone->lru[LRU_UNEVICTABLE].list; unsigned long scan; unsigned long nr_to_scan = zone_page_state(zone, NR_UNEVICTABLE); while (nr_to_scan > 0) { unsigned long batch_size = min(nr_to_scan, SCAN_UNEVICTABLE_BATCH_SIZE); spin_lock_irq(&zone->lru_lock); for (scan = 0; scan < batch_size; scan++) { struct page *page = lru_to_page(l_unevictable); if (!trylock_page(page)) continue; prefetchw_prev_lru_page(page, l_unevictable, flags); if (likely(PageLRU(page) && PageUnevictable(page))) check_move_unevictable_page(page, zone); unlock_page(page); } spin_unlock_irq(&zone->lru_lock); nr_to_scan -= batch_size; } } /** * scan_all_zones_unevictable_pages - scan all unevictable lists for evictable pages * * A really big hammer: scan all zones' unevictable LRU lists to check for * pages that have become evictable. Move those back to the zones' * inactive list where they become candidates for reclaim. * This occurs when, e.g., we have unswappable pages on the unevictable lists, * and we add swap to the system. As such, it runs in the context of a task * that has possibly/probably made some previously unevictable pages * evictable. */ |
ff30153bf
|
2743 |
static void scan_all_zones_unevictable_pages(void) |
af936a160
|
2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 |
{ struct zone *zone; for_each_zone(zone) { scan_zone_unevictable_pages(zone); } } /* * scan_unevictable_pages [vm] sysctl handler. On demand re-scan of * all nodes' unevictable lists for evictable pages */ unsigned long scan_unevictable_pages; int scan_unevictable_handler(struct ctl_table *table, int write, |
8d65af789
|
2759 |
void __user *buffer, |
af936a160
|
2760 2761 |
size_t *length, loff_t *ppos) { |
8d65af789
|
2762 |
proc_doulongvec_minmax(table, write, buffer, length, ppos); |
af936a160
|
2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 |
if (write && *(unsigned long *)table->data) scan_all_zones_unevictable_pages(); scan_unevictable_pages = 0; return 0; } /* * per node 'scan_unevictable_pages' attribute. On demand re-scan of * a specified node's per zone unevictable lists for evictable pages. */ static ssize_t read_scan_unevictable_node(struct sys_device *dev, struct sysdev_attribute *attr, char *buf) { return sprintf(buf, "0 "); /* always zero; should fit... */ } static ssize_t write_scan_unevictable_node(struct sys_device *dev, struct sysdev_attribute *attr, const char *buf, size_t count) { struct zone *node_zones = NODE_DATA(dev->id)->node_zones; struct zone *zone; unsigned long res; unsigned long req = strict_strtoul(buf, 10, &res); if (!req) return 1; /* zero is no-op */ for (zone = node_zones; zone - node_zones < MAX_NR_ZONES; ++zone) { if (!populated_zone(zone)) continue; scan_zone_unevictable_pages(zone); } return 1; } static SYSDEV_ATTR(scan_unevictable_pages, S_IRUGO | S_IWUSR, read_scan_unevictable_node, write_scan_unevictable_node); int scan_unevictable_register_node(struct node *node) { return sysdev_create_file(&node->sysdev, &attr_scan_unevictable_pages); } void scan_unevictable_unregister_node(struct node *node) { sysdev_remove_file(&node->sysdev, &attr_scan_unevictable_pages); } |