Blame view
mm/vmscan.c
81.7 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
/* * 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> |
5a0e3ad6a include cleanup: ... |
16 |
#include <linux/gfp.h> |
1da177e4c Linux-2.6.12-rc2 |
17 18 19 20 21 |
#include <linux/kernel_stat.h> #include <linux/swap.h> #include <linux/pagemap.h> #include <linux/init.h> #include <linux/highmem.h> |
e129b5c23 [PATCH] vm: add p... |
22 |
#include <linux/vmstat.h> |
1da177e4c Linux-2.6.12-rc2 |
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 [PATCH] mm: make ... |
37 |
#include <linux/delay.h> |
3218ae14b [PATCH] pgdat all... |
38 |
#include <linux/kthread.h> |
7dfb71030 [PATCH] Add inclu... |
39 |
#include <linux/freezer.h> |
66e1707bc Memory controller... |
40 |
#include <linux/memcontrol.h> |
873b47717 per-task-delay-ac... |
41 |
#include <linux/delayacct.h> |
af936a160 vmscan: unevictab... |
42 |
#include <linux/sysctl.h> |
1da177e4c Linux-2.6.12-rc2 |
43 44 45 46 47 |
#include <asm/tlbflush.h> #include <asm/div64.h> #include <linux/swapops.h> |
0f8053a50 [PATCH] mm: make ... |
48 |
#include "internal.h" |
33906bc5c vmscan: tracing: ... |
49 50 |
#define CREATE_TRACE_POINTS #include <trace/events/vmscan.h> |
1da177e4c Linux-2.6.12-rc2 |
51 |
struct scan_control { |
1da177e4c Linux-2.6.12-rc2 |
52 53 |
/* Incremented by the number of inactive pages that were scanned */ unsigned long nr_scanned; |
a79311c14 vmscan: bail out ... |
54 55 |
/* Number of pages freed so far during a call to shrink_zones() */ unsigned long nr_reclaimed; |
22fba3354 vmscan: separate ... |
56 57 |
/* How many pages shrink_list() should reclaim */ unsigned long nr_to_reclaim; |
7b51755c3 vmscan: kill hibe... |
58 |
unsigned long hibernation_mode; |
1da177e4c Linux-2.6.12-rc2 |
59 |
/* This context's GFP mask */ |
6daa0e286 [PATCH] gfp_t: mm... |
60 |
gfp_t gfp_mask; |
1da177e4c Linux-2.6.12-rc2 |
61 62 |
int may_writepage; |
a6dc60f89 vmscan: rename sc... |
63 64 |
/* Can mapped pages be reclaimed? */ int may_unmap; |
f1fd1067e [PATCH] Zone recl... |
65 |
|
2e2e42598 vmscan,memcg: rei... |
66 67 |
/* Can pages be swapped as part of reclaim? */ int may_swap; |
d6277db4a [PATCH] swsusp: r... |
68 |
int swappiness; |
408d85441 [PATCH] oom: use ... |
69 |
|
5ad333eb6 Lumpy Reclaim V4 |
70 |
int order; |
66e1707bc Memory controller... |
71 |
|
5f53e7629 vmscan: page_chec... |
72 73 74 75 76 |
/* * Intend to reclaim enough contenious memory rather than to reclaim * enough amount memory. I.e, it's the mode for high order allocation. */ bool lumpy_reclaim_mode; |
66e1707bc Memory controller... |
77 78 |
/* Which cgroup do we reclaim from */ struct mem_cgroup *mem_cgroup; |
327c0e968 vmscan: fix it to... |
79 80 81 82 83 |
/* * Nodemask of nodes allowed by the caller. If NULL, all nodes * are scanned. */ nodemask_t *nodemask; |
1da177e4c Linux-2.6.12-rc2 |
84 |
}; |
1da177e4c Linux-2.6.12-rc2 |
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 [PATCH] initialis... |
119 |
long vm_total_pages; /* The total number of pages which the VM controls */ |
1da177e4c Linux-2.6.12-rc2 |
120 121 122 |
static LIST_HEAD(shrinker_list); static DECLARE_RWSEM(shrinker_rwsem); |
00f0b8259 Memory controller... |
123 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR |
e72e2bd67 memcg: rename sca... |
124 |
#define scanning_global_lru(sc) (!(sc)->mem_cgroup) |
91a45470f per-zone and recl... |
125 |
#else |
e72e2bd67 memcg: rename sca... |
126 |
#define scanning_global_lru(sc) (1) |
91a45470f per-zone and recl... |
127 |
#endif |
6e9015716 mm: introduce zon... |
128 129 130 |
static struct zone_reclaim_stat *get_reclaim_stat(struct zone *zone, struct scan_control *sc) { |
e72e2bd67 memcg: rename sca... |
131 |
if (!scanning_global_lru(sc)) |
3e2f41f1f memcg: add zone_r... |
132 |
return mem_cgroup_get_reclaim_stat(sc->mem_cgroup, zone); |
6e9015716 mm: introduce zon... |
133 134 |
return &zone->reclaim_stat; } |
0b2176763 mm/vmscan: rename... |
135 136 |
static unsigned long zone_nr_lru_pages(struct zone *zone, struct scan_control *sc, enum lru_list lru) |
c9f299d98 mm: add zone nr_p... |
137 |
{ |
e72e2bd67 memcg: rename sca... |
138 |
if (!scanning_global_lru(sc)) |
a3d8e0549 memcg: add mem_cg... |
139 |
return mem_cgroup_zone_nr_pages(sc->mem_cgroup, zone, lru); |
c9f299d98 mm: add zone nr_p... |
140 141 |
return zone_page_state(zone, NR_LRU_BASE + lru); } |
1da177e4c Linux-2.6.12-rc2 |
142 143 144 |
/* * Add a shrinker callback to be called from the vm */ |
8e1f936b7 mm: clean up and ... |
145 |
void register_shrinker(struct shrinker *shrinker) |
1da177e4c Linux-2.6.12-rc2 |
146 |
{ |
8e1f936b7 mm: clean up and ... |
147 148 149 150 |
shrinker->nr = 0; down_write(&shrinker_rwsem); list_add_tail(&shrinker->list, &shrinker_list); up_write(&shrinker_rwsem); |
1da177e4c Linux-2.6.12-rc2 |
151 |
} |
8e1f936b7 mm: clean up and ... |
152 |
EXPORT_SYMBOL(register_shrinker); |
1da177e4c Linux-2.6.12-rc2 |
153 154 155 156 |
/* * Remove one */ |
8e1f936b7 mm: clean up and ... |
157 |
void unregister_shrinker(struct shrinker *shrinker) |
1da177e4c Linux-2.6.12-rc2 |
158 159 160 161 |
{ down_write(&shrinker_rwsem); list_del(&shrinker->list); up_write(&shrinker_rwsem); |
1da177e4c Linux-2.6.12-rc2 |
162 |
} |
8e1f936b7 mm: clean up and ... |
163 |
EXPORT_SYMBOL(unregister_shrinker); |
1da177e4c Linux-2.6.12-rc2 |
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 spelling fixes: mm/ |
174 |
* If the vm encountered mapped pages on the LRU it increase the pressure on |
1da177e4c Linux-2.6.12-rc2 |
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 [PATCH] vmscan: n... |
182 183 |
* * Returns the number of slab objects which we shrunk. |
1da177e4c Linux-2.6.12-rc2 |
184 |
*/ |
69e05944a [PATCH] vmscan: u... |
185 186 |
unsigned long shrink_slab(unsigned long scanned, gfp_t gfp_mask, unsigned long lru_pages) |
1da177e4c Linux-2.6.12-rc2 |
187 188 |
{ struct shrinker *shrinker; |
69e05944a [PATCH] vmscan: u... |
189 |
unsigned long ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
190 191 192 193 194 |
if (scanned == 0) scanned = SWAP_CLUSTER_MAX; if (!down_read_trylock(&shrinker_rwsem)) |
b15e0905f [PATCH] vmscan: n... |
195 |
return 1; /* Assume we'll be able to shrink next time */ |
1da177e4c Linux-2.6.12-rc2 |
196 197 198 199 |
list_for_each_entry(shrinker, &shrinker_list, list) { unsigned long long delta; unsigned long total_scan; |
7f8275d0d mm: add context a... |
200 |
unsigned long max_pass; |
1da177e4c Linux-2.6.12-rc2 |
201 |
|
7f8275d0d mm: add context a... |
202 |
max_pass = (*shrinker->shrink)(shrinker, 0, gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
203 |
delta = (4 * scanned) / shrinker->seeks; |
ea164d73a [PATCH] shrinker-... |
204 |
delta *= max_pass; |
1da177e4c Linux-2.6.12-rc2 |
205 206 |
do_div(delta, lru_pages + 1); shrinker->nr += delta; |
ea164d73a [PATCH] shrinker-... |
207 |
if (shrinker->nr < 0) { |
88c3bd707 vmscan: print shr... |
208 209 210 211 |
printk(KERN_ERR "shrink_slab: %pF negative objects to " "delete nr=%ld ", shrinker->shrink, shrinker->nr); |
ea164d73a [PATCH] shrinker-... |
212 213 214 215 216 217 218 219 220 221 |
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 Linux-2.6.12-rc2 |
222 223 224 225 226 227 228 |
total_scan = shrinker->nr; shrinker->nr = 0; while (total_scan >= SHRINK_BATCH) { long this_scan = SHRINK_BATCH; int shrink_ret; |
b15e0905f [PATCH] vmscan: n... |
229 |
int nr_before; |
1da177e4c Linux-2.6.12-rc2 |
230 |
|
7f8275d0d mm: add context a... |
231 232 233 |
nr_before = (*shrinker->shrink)(shrinker, 0, gfp_mask); shrink_ret = (*shrinker->shrink)(shrinker, this_scan, gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
234 235 |
if (shrink_ret == -1) break; |
b15e0905f [PATCH] vmscan: n... |
236 237 |
if (shrink_ret < nr_before) ret += nr_before - shrink_ret; |
f8891e5e1 [PATCH] Light wei... |
238 |
count_vm_events(SLABS_SCANNED, this_scan); |
1da177e4c Linux-2.6.12-rc2 |
239 240 241 242 243 244 245 246 |
total_scan -= this_scan; cond_resched(); } shrinker->nr += total_scan; } up_read(&shrinker_rwsem); |
b15e0905f [PATCH] vmscan: n... |
247 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
248 |
} |
1da177e4c Linux-2.6.12-rc2 |
249 250 |
static inline int is_page_cache_freeable(struct page *page) { |
ceddc3a52 mm: document is_p... |
251 252 253 254 255 |
/* * 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 mm: return boolea... |
256 |
return page_count(page) - page_has_private(page) == 2; |
1da177e4c Linux-2.6.12-rc2 |
257 258 259 260 |
} static int may_write_to_queue(struct backing_dev_info *bdi) { |
930d91525 [PATCH] Swap Migr... |
261 |
if (current->flags & PF_SWAPWRITE) |
1da177e4c Linux-2.6.12-rc2 |
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 |
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) { |
a6aa62a09 mm/vmscan.c: fix ... |
285 |
lock_page_nosync(page); |
3e9f45bd1 Factor outstandin... |
286 287 |
if (page_mapping(page) == mapping) mapping_set_error(mapping, error); |
1da177e4c Linux-2.6.12-rc2 |
288 289 |
unlock_page(page); } |
c661b078f synchronous lumpy... |
290 291 292 293 294 |
/* Request for sync pageout. */ enum pageout_io { PAGEOUT_IO_ASYNC, PAGEOUT_IO_SYNC, }; |
04e62a29b [PATCH] More page... |
295 296 297 298 299 300 301 302 303 304 305 |
/* 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 Linux-2.6.12-rc2 |
306 |
/* |
1742f19fa [PATCH] vmscan: r... |
307 308 |
* pageout is called by shrink_page_list() for each dirty page. * Calls ->writepage(). |
1da177e4c Linux-2.6.12-rc2 |
309 |
*/ |
c661b078f synchronous lumpy... |
310 311 |
static pageout_t pageout(struct page *page, struct address_space *mapping, enum pageout_io sync_writeback) |
1da177e4c Linux-2.6.12-rc2 |
312 313 314 315 316 317 318 319 |
{ /* * 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 mm/vmscan: change... |
320 |
* If this process is currently in __generic_file_aio_write() against |
1da177e4c Linux-2.6.12-rc2 |
321 322 323 324 325 326 327 |
* 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 Linux-2.6.12-rc2 |
328 329 330 331 332 333 334 335 |
*/ 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 FS-Cache: Recruit... |
336 |
if (page_has_private(page)) { |
1da177e4c Linux-2.6.12-rc2 |
337 338 |
if (try_to_free_buffers(page)) { ClearPageDirty(page); |
d40cee245 mm: remove remain... |
339 340 |
printk("%s: orphaned page ", __func__); |
1da177e4c Linux-2.6.12-rc2 |
341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 |
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 [PATCH] writeback... |
356 357 |
.range_start = 0, .range_end = LLONG_MAX, |
1da177e4c Linux-2.6.12-rc2 |
358 359 360 361 362 363 364 365 |
.nonblocking = 1, .for_reclaim = 1, }; SetPageReclaim(page); res = mapping->a_ops->writepage(page, &wbc); if (res < 0) handle_write_error(mapping, page, res); |
994fc28c7 [PATCH] add AOP_T... |
366 |
if (res == AOP_WRITEPAGE_ACTIVATE) { |
1da177e4c Linux-2.6.12-rc2 |
367 368 369 |
ClearPageReclaim(page); return PAGE_ACTIVATE; } |
c661b078f synchronous lumpy... |
370 371 372 373 374 375 376 377 |
/* * 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 Linux-2.6.12-rc2 |
378 379 380 381 |
if (!PageWriteback(page)) { /* synchronous write or broken a_ops? */ ClearPageReclaim(page); } |
755f0225e vmscan: tracing: ... |
382 383 |
trace_mm_vmscan_writepage(page, trace_reclaim_flags(page, sync_writeback)); |
e129b5c23 [PATCH] vm: add p... |
384 |
inc_zone_page_state(page, NR_VMSCAN_WRITE); |
1da177e4c Linux-2.6.12-rc2 |
385 386 387 388 389 |
return PAGE_SUCCESS; } return PAGE_CLEAN; } |
a649fd927 [PATCH] invalidat... |
390 |
/* |
e286781d5 mm: speculative p... |
391 392 |
* Same as remove_mapping, but if the page is removed from the mapping, it * gets returned with a refcount of 0. |
a649fd927 [PATCH] invalidat... |
393 |
*/ |
e286781d5 mm: speculative p... |
394 |
static int __remove_mapping(struct address_space *mapping, struct page *page) |
49d2e9cc4 [PATCH] Swap Migr... |
395 |
{ |
28e4d965e [PATCH] mm: remov... |
396 397 |
BUG_ON(!PageLocked(page)); BUG_ON(mapping != page_mapping(page)); |
49d2e9cc4 [PATCH] Swap Migr... |
398 |
|
19fd62312 mm: spinlock tree... |
399 |
spin_lock_irq(&mapping->tree_lock); |
49d2e9cc4 [PATCH] Swap Migr... |
400 |
/* |
0fd0e6b05 [PATCH] page inva... |
401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 |
* 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 [PATCH] Swap Migr... |
424 |
*/ |
e286781d5 mm: speculative p... |
425 |
if (!page_freeze_refs(page, 2)) |
49d2e9cc4 [PATCH] Swap Migr... |
426 |
goto cannot_free; |
e286781d5 mm: speculative p... |
427 428 429 |
/* note: atomic_cmpxchg in page_freeze_refs provides the smp_rmb */ if (unlikely(PageDirty(page))) { page_unfreeze_refs(page, 2); |
49d2e9cc4 [PATCH] Swap Migr... |
430 |
goto cannot_free; |
e286781d5 mm: speculative p... |
431 |
} |
49d2e9cc4 [PATCH] Swap Migr... |
432 433 434 435 |
if (PageSwapCache(page)) { swp_entry_t swap = { .val = page_private(page) }; __delete_from_swap_cache(page); |
19fd62312 mm: spinlock tree... |
436 |
spin_unlock_irq(&mapping->tree_lock); |
cb4b86ba4 mm: add swap cach... |
437 |
swapcache_free(swap, page); |
e286781d5 mm: speculative p... |
438 439 |
} else { __remove_from_page_cache(page); |
19fd62312 mm: spinlock tree... |
440 |
spin_unlock_irq(&mapping->tree_lock); |
e767e0561 memcg: fix deadlo... |
441 |
mem_cgroup_uncharge_cache_page(page); |
49d2e9cc4 [PATCH] Swap Migr... |
442 |
} |
49d2e9cc4 [PATCH] Swap Migr... |
443 444 445 |
return 1; cannot_free: |
19fd62312 mm: spinlock tree... |
446 |
spin_unlock_irq(&mapping->tree_lock); |
49d2e9cc4 [PATCH] Swap Migr... |
447 448 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
449 |
/* |
e286781d5 mm: speculative p... |
450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 |
* 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 Unevictable LRU I... |
468 469 470 471 472 473 474 475 476 |
/** * 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 Unevictable LRU I... |
477 478 479 480 |
void putback_lru_page(struct page *page) { int lru; int active = !!TestClearPageActive(page); |
bbfd28eee unevictable lru: ... |
481 |
int was_unevictable = PageUnevictable(page); |
894bc3104 Unevictable LRU I... |
482 483 484 485 486 487 488 489 490 491 492 493 494 |
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 mm: introduce pag... |
495 |
lru = active + page_lru_base_type(page); |
894bc3104 Unevictable LRU I... |
496 497 498 499 500 501 502 503 |
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 vmscan: order evi... |
504 505 506 507 508 509 510 511 512 513 |
/* * 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 Unevictable LRU I... |
514 |
} |
894bc3104 Unevictable LRU I... |
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 |
/* * 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 unevictable lru: ... |
531 532 533 534 |
if (was_unevictable && lru != LRU_UNEVICTABLE) count_vm_event(UNEVICTABLE_PGRESCUED); else if (!was_unevictable && lru == LRU_UNEVICTABLE) count_vm_event(UNEVICTABLE_PGCULLED); |
894bc3104 Unevictable LRU I... |
535 536 |
put_page(page); /* drop ref from isolate */ } |
dfc8d636c vmscan: factor ou... |
537 538 539 |
enum page_references { PAGEREF_RECLAIM, PAGEREF_RECLAIM_CLEAN, |
645747462 vmscan: detect ma... |
540 |
PAGEREF_KEEP, |
dfc8d636c vmscan: factor ou... |
541 542 543 544 545 546 |
PAGEREF_ACTIVATE, }; static enum page_references page_check_references(struct page *page, struct scan_control *sc) { |
645747462 vmscan: detect ma... |
547 |
int referenced_ptes, referenced_page; |
dfc8d636c vmscan: factor ou... |
548 |
unsigned long vm_flags; |
dfc8d636c vmscan: factor ou... |
549 |
|
645747462 vmscan: detect ma... |
550 551 |
referenced_ptes = page_referenced(page, 1, sc->mem_cgroup, &vm_flags); referenced_page = TestClearPageReferenced(page); |
dfc8d636c vmscan: factor ou... |
552 553 |
/* Lumpy reclaim - ignore references */ |
5f53e7629 vmscan: page_chec... |
554 |
if (sc->lumpy_reclaim_mode) |
dfc8d636c vmscan: factor ou... |
555 556 557 558 559 560 561 562 |
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 vmscan: detect ma... |
563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 |
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 vmscan: factor ou... |
587 588 |
/* Reclaim if clean, defer dirty pages to writeback */ |
645747462 vmscan: detect ma... |
589 590 591 592 |
if (referenced_page) return PAGEREF_RECLAIM_CLEAN; return PAGEREF_RECLAIM; |
dfc8d636c vmscan: factor ou... |
593 |
} |
abe4c3b50 vmscan: set up pa... |
594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 |
static noinline_for_stack void free_page_list(struct list_head *free_pages) { struct pagevec freed_pvec; struct page *page, *tmp; pagevec_init(&freed_pvec, 1); list_for_each_entry_safe(page, tmp, free_pages, lru) { list_del(&page->lru); if (!pagevec_add(&freed_pvec, page)) { __pagevec_free(&freed_pvec); pagevec_reinit(&freed_pvec); } } pagevec_free(&freed_pvec); } |
e286781d5 mm: speculative p... |
611 |
/* |
1742f19fa [PATCH] vmscan: r... |
612 |
* shrink_page_list() returns the number of reclaimed pages |
1da177e4c Linux-2.6.12-rc2 |
613 |
*/ |
1742f19fa [PATCH] vmscan: r... |
614 |
static unsigned long shrink_page_list(struct list_head *page_list, |
c661b078f synchronous lumpy... |
615 616 |
struct scan_control *sc, enum pageout_io sync_writeback) |
1da177e4c Linux-2.6.12-rc2 |
617 618 |
{ LIST_HEAD(ret_pages); |
abe4c3b50 vmscan: set up pa... |
619 |
LIST_HEAD(free_pages); |
1da177e4c Linux-2.6.12-rc2 |
620 |
int pgactivate = 0; |
05ff51376 [PATCH] vmscan re... |
621 |
unsigned long nr_reclaimed = 0; |
1da177e4c Linux-2.6.12-rc2 |
622 623 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
624 |
while (!list_empty(page_list)) { |
dfc8d636c vmscan: factor ou... |
625 |
enum page_references references; |
1da177e4c Linux-2.6.12-rc2 |
626 627 628 |
struct address_space *mapping; struct page *page; int may_enter_fs; |
1da177e4c Linux-2.6.12-rc2 |
629 630 631 632 633 |
cond_resched(); page = lru_to_page(page_list); list_del(&page->lru); |
529ae9aaa mm: rename page t... |
634 |
if (!trylock_page(page)) |
1da177e4c Linux-2.6.12-rc2 |
635 |
goto keep; |
725d704ec [PATCH] mm: VM_BU... |
636 |
VM_BUG_ON(PageActive(page)); |
1da177e4c Linux-2.6.12-rc2 |
637 638 |
sc->nr_scanned++; |
80e434260 [PATCH] zone recl... |
639 |
|
b291f0003 mlock: mlocked pa... |
640 641 |
if (unlikely(!page_evictable(page, NULL))) goto cull_mlocked; |
894bc3104 Unevictable LRU I... |
642 |
|
a6dc60f89 vmscan: rename sc... |
643 |
if (!sc->may_unmap && page_mapped(page)) |
80e434260 [PATCH] zone recl... |
644 |
goto keep_locked; |
1da177e4c Linux-2.6.12-rc2 |
645 646 647 |
/* Double the slab pressure for mapped and swapcache pages */ if (page_mapped(page) || PageSwapCache(page)) sc->nr_scanned++; |
c661b078f synchronous lumpy... |
648 649 650 651 652 653 654 655 656 657 658 659 660 661 |
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 revert "kswapd sh... |
662 |
else |
c661b078f synchronous lumpy... |
663 664 |
goto keep_locked; } |
1da177e4c Linux-2.6.12-rc2 |
665 |
|
dfc8d636c vmscan: factor ou... |
666 667 668 |
references = page_check_references(page, sc); switch (references) { case PAGEREF_ACTIVATE: |
1da177e4c Linux-2.6.12-rc2 |
669 |
goto activate_locked; |
645747462 vmscan: detect ma... |
670 671 |
case PAGEREF_KEEP: goto keep_locked; |
dfc8d636c vmscan: factor ou... |
672 673 674 675 |
case PAGEREF_RECLAIM: case PAGEREF_RECLAIM_CLEAN: ; /* try to reclaim the page below */ } |
1da177e4c Linux-2.6.12-rc2 |
676 |
|
1da177e4c Linux-2.6.12-rc2 |
677 678 679 680 |
/* * Anonymous process memory has backing store? * Try to allocate it some swap space here. */ |
b291f0003 mlock: mlocked pa... |
681 |
if (PageAnon(page) && !PageSwapCache(page)) { |
63eb6b93c vmscan: let GFP_N... |
682 683 |
if (!(sc->gfp_mask & __GFP_IO)) goto keep_locked; |
ac47b003d mm: remove gfp_ma... |
684 |
if (!add_to_swap(page)) |
1da177e4c Linux-2.6.12-rc2 |
685 |
goto activate_locked; |
63eb6b93c vmscan: let GFP_N... |
686 |
may_enter_fs = 1; |
b291f0003 mlock: mlocked pa... |
687 |
} |
1da177e4c Linux-2.6.12-rc2 |
688 689 |
mapping = page_mapping(page); |
1da177e4c Linux-2.6.12-rc2 |
690 691 692 693 694 695 |
/* * The page is mapped into the page tables of one or more * processes. Try to unmap it here. */ if (page_mapped(page) && mapping) { |
14fa31b89 HWPOISON: Use bit... |
696 |
switch (try_to_unmap(page, TTU_UNMAP)) { |
1da177e4c Linux-2.6.12-rc2 |
697 698 699 700 |
case SWAP_FAIL: goto activate_locked; case SWAP_AGAIN: goto keep_locked; |
b291f0003 mlock: mlocked pa... |
701 702 |
case SWAP_MLOCK: goto cull_mlocked; |
1da177e4c Linux-2.6.12-rc2 |
703 704 705 706 707 708 |
case SWAP_SUCCESS: ; /* try to free the page below */ } } if (PageDirty(page)) { |
dfc8d636c vmscan: factor ou... |
709 |
if (references == PAGEREF_RECLAIM_CLEAN) |
1da177e4c Linux-2.6.12-rc2 |
710 |
goto keep_locked; |
4dd4b9202 revert "kswapd sh... |
711 |
if (!may_enter_fs) |
1da177e4c Linux-2.6.12-rc2 |
712 |
goto keep_locked; |
52a8363ea [PATCH] mm: impro... |
713 |
if (!sc->may_writepage) |
1da177e4c Linux-2.6.12-rc2 |
714 715 716 |
goto keep_locked; /* Page is dirty, try to write it out here */ |
c661b078f synchronous lumpy... |
717 |
switch (pageout(page, mapping, sync_writeback)) { |
1da177e4c Linux-2.6.12-rc2 |
718 719 720 721 722 |
case PAGE_KEEP: goto keep_locked; case PAGE_ACTIVATE: goto activate_locked; case PAGE_SUCCESS: |
4dd4b9202 revert "kswapd sh... |
723 |
if (PageWriteback(page) || PageDirty(page)) |
1da177e4c Linux-2.6.12-rc2 |
724 725 726 727 728 |
goto keep; /* * A synchronous write - probably a ramdisk. Go * ahead and try to reclaim the page. */ |
529ae9aaa mm: rename page t... |
729 |
if (!trylock_page(page)) |
1da177e4c Linux-2.6.12-rc2 |
730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 |
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 Unevictable LRU I... |
749 |
* will do this, as well as the blockdev mapping. |
1da177e4c Linux-2.6.12-rc2 |
750 751 752 753 754 755 756 757 758 759 |
* 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 FS-Cache: Recruit... |
760 |
if (page_has_private(page)) { |
1da177e4c Linux-2.6.12-rc2 |
761 762 |
if (!try_to_release_page(page, sc->gfp_mask)) goto activate_locked; |
e286781d5 mm: speculative p... |
763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 |
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 Linux-2.6.12-rc2 |
779 |
} |
e286781d5 mm: speculative p... |
780 |
if (!mapping || !__remove_mapping(mapping, page)) |
49d2e9cc4 [PATCH] Swap Migr... |
781 |
goto keep_locked; |
1da177e4c Linux-2.6.12-rc2 |
782 |
|
a978d6f52 mm: unlockless re... |
783 784 785 786 787 788 789 790 |
/* * 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 mm: speculative p... |
791 |
free_it: |
05ff51376 [PATCH] vmscan re... |
792 |
nr_reclaimed++; |
abe4c3b50 vmscan: set up pa... |
793 794 795 796 797 798 |
/* * Is there need to periodically free_page_list? It would * appear not as the counts should be low */ list_add(&page->lru, &free_pages); |
1da177e4c Linux-2.6.12-rc2 |
799 |
continue; |
b291f0003 mlock: mlocked pa... |
800 |
cull_mlocked: |
63d6c5ad7 mm: remove try_to... |
801 802 |
if (PageSwapCache(page)) try_to_free_swap(page); |
b291f0003 mlock: mlocked pa... |
803 804 805 |
unlock_page(page); putback_lru_page(page); continue; |
1da177e4c Linux-2.6.12-rc2 |
806 |
activate_locked: |
68a22394c vmscan: free swap... |
807 808 |
/* Not a candidate for swapping, so reclaim swap space. */ if (PageSwapCache(page) && vm_swap_full()) |
a2c43eed8 mm: try_to_free_s... |
809 |
try_to_free_swap(page); |
894bc3104 Unevictable LRU I... |
810 |
VM_BUG_ON(PageActive(page)); |
1da177e4c Linux-2.6.12-rc2 |
811 812 813 814 815 816 |
SetPageActive(page); pgactivate++; keep_locked: unlock_page(page); keep: list_add(&page->lru, &ret_pages); |
b291f0003 mlock: mlocked pa... |
817 |
VM_BUG_ON(PageLRU(page) || PageUnevictable(page)); |
1da177e4c Linux-2.6.12-rc2 |
818 |
} |
abe4c3b50 vmscan: set up pa... |
819 820 |
free_page_list(&free_pages); |
1da177e4c Linux-2.6.12-rc2 |
821 |
list_splice(&ret_pages, page_list); |
f8891e5e1 [PATCH] Light wei... |
822 |
count_vm_events(PGACTIVATE, pgactivate); |
05ff51376 [PATCH] vmscan re... |
823 |
return nr_reclaimed; |
1da177e4c Linux-2.6.12-rc2 |
824 |
} |
5ad333eb6 Lumpy Reclaim V4 |
825 826 827 828 829 830 831 832 833 834 |
/* * 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 vmscan: split LRU... |
835 |
int __isolate_lru_page(struct page *page, int mode, int file) |
5ad333eb6 Lumpy Reclaim V4 |
836 837 838 839 840 841 842 843 844 845 846 847 848 849 |
{ 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 mm: return boolea... |
850 |
if (mode != ISOLATE_BOTH && page_is_file_cache(page) != file) |
4f98a2fee vmscan: split LRU... |
851 |
return ret; |
894bc3104 Unevictable LRU I... |
852 853 854 855 856 857 858 |
/* * 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 Lumpy Reclaim V4 |
859 |
ret = -EBUSY; |
08e552c69 memcg: synchroniz... |
860 |
|
5ad333eb6 Lumpy Reclaim V4 |
861 862 863 864 865 866 867 868 869 870 871 872 |
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 [PATCH] Swap Migr... |
873 |
/* |
1da177e4c Linux-2.6.12-rc2 |
874 875 876 877 878 879 880 881 882 883 884 885 886 |
* 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 Lumpy Reclaim V4 |
887 888 |
* @order: The caller's attempted allocation order * @mode: One of the LRU isolation modes |
4f98a2fee vmscan: split LRU... |
889 |
* @file: True [1] if isolating file [!anon] pages |
1da177e4c Linux-2.6.12-rc2 |
890 891 892 |
* * returns how many pages were moved onto *@dst. */ |
69e05944a [PATCH] vmscan: u... |
893 894 |
static unsigned long isolate_lru_pages(unsigned long nr_to_scan, struct list_head *src, struct list_head *dst, |
4f98a2fee vmscan: split LRU... |
895 |
unsigned long *scanned, int order, int mode, int file) |
1da177e4c Linux-2.6.12-rc2 |
896 |
{ |
69e05944a [PATCH] vmscan: u... |
897 |
unsigned long nr_taken = 0; |
a8a94d151 vmscan: tracing: ... |
898 899 900 |
unsigned long nr_lumpy_taken = 0; unsigned long nr_lumpy_dirty = 0; unsigned long nr_lumpy_failed = 0; |
c9b02d970 [PATCH] mm: isola... |
901 |
unsigned long scan; |
1da177e4c Linux-2.6.12-rc2 |
902 |
|
c9b02d970 [PATCH] mm: isola... |
903 |
for (scan = 0; scan < nr_to_scan && !list_empty(src); scan++) { |
5ad333eb6 Lumpy Reclaim V4 |
904 905 906 907 908 |
struct page *page; unsigned long pfn; unsigned long end_pfn; unsigned long page_pfn; int zone_id; |
1da177e4c Linux-2.6.12-rc2 |
909 910 |
page = lru_to_page(src); prefetchw_prev_lru_page(page, src, flags); |
725d704ec [PATCH] mm: VM_BU... |
911 |
VM_BUG_ON(!PageLRU(page)); |
8d438f96d [PATCH] mm: PageL... |
912 |
|
4f98a2fee vmscan: split LRU... |
913 |
switch (__isolate_lru_page(page, mode, file)) { |
5ad333eb6 Lumpy Reclaim V4 |
914 915 |
case 0: list_move(&page->lru, dst); |
2ffebca6a memcg: fix lru ro... |
916 |
mem_cgroup_del_lru(page); |
7c8ee9a86 [PATCH] mm: simpl... |
917 |
nr_taken++; |
5ad333eb6 Lumpy Reclaim V4 |
918 919 920 921 922 |
break; case -EBUSY: /* else it is being freed elsewhere */ list_move(&page->lru, src); |
2ffebca6a memcg: fix lru ro... |
923 |
mem_cgroup_rotate_lru_list(page, page_lru(page)); |
5ad333eb6 Lumpy Reclaim V4 |
924 |
continue; |
46453a6e1 [PATCH] mm: never... |
925 |
|
5ad333eb6 Lumpy Reclaim V4 |
926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 |
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 vmscan: split LRU... |
958 |
|
5ad333eb6 Lumpy Reclaim V4 |
959 960 961 |
/* Check that we have not crossed a zone boundary. */ if (unlikely(page_zone_id(cursor_page) != zone_id)) continue; |
de2e7567c vmscan: don't att... |
962 963 964 965 966 967 968 969 970 |
/* * 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 mm: fix lumpy rec... |
971 |
if (__isolate_lru_page(cursor_page, mode, file) == 0) { |
5ad333eb6 Lumpy Reclaim V4 |
972 |
list_move(&cursor_page->lru, dst); |
cb4cbcf6b mm: fix incorrect... |
973 |
mem_cgroup_del_lru(cursor_page); |
5ad333eb6 Lumpy Reclaim V4 |
974 |
nr_taken++; |
a8a94d151 vmscan: tracing: ... |
975 976 977 |
nr_lumpy_taken++; if (PageDirty(cursor_page)) nr_lumpy_dirty++; |
5ad333eb6 Lumpy Reclaim V4 |
978 |
scan++; |
a8a94d151 vmscan: tracing: ... |
979 980 981 982 |
} else { if (mode == ISOLATE_BOTH && page_count(cursor_page)) nr_lumpy_failed++; |
5ad333eb6 Lumpy Reclaim V4 |
983 984 |
} } |
1da177e4c Linux-2.6.12-rc2 |
985 986 987 |
} *scanned = scan; |
a8a94d151 vmscan: tracing: ... |
988 989 990 991 992 993 |
trace_mm_vmscan_lru_isolate(order, nr_to_scan, scan, nr_taken, nr_lumpy_taken, nr_lumpy_dirty, nr_lumpy_failed, mode); |
1da177e4c Linux-2.6.12-rc2 |
994 995 |
return nr_taken; } |
66e1707bc Memory controller... |
996 997 998 999 |
static unsigned long isolate_pages_global(unsigned long nr, struct list_head *dst, unsigned long *scanned, int order, int mode, struct zone *z, |
4f98a2fee vmscan: split LRU... |
1000 |
int active, int file) |
66e1707bc Memory controller... |
1001 |
{ |
4f98a2fee vmscan: split LRU... |
1002 |
int lru = LRU_BASE; |
66e1707bc Memory controller... |
1003 |
if (active) |
4f98a2fee vmscan: split LRU... |
1004 1005 1006 1007 |
lru += LRU_ACTIVE; if (file) lru += LRU_FILE; return isolate_lru_pages(nr, &z->lru[lru].list, dst, scanned, order, |
b7c46d151 mm: drop unneeded... |
1008 |
mode, file); |
66e1707bc Memory controller... |
1009 |
} |
1da177e4c Linux-2.6.12-rc2 |
1010 |
/* |
5ad333eb6 Lumpy Reclaim V4 |
1011 1012 1013 |
* clear_active_flags() is a helper for shrink_active_list(), clearing * any active bits from the pages in the list. */ |
4f98a2fee vmscan: split LRU... |
1014 1015 |
static unsigned long clear_active_flags(struct list_head *page_list, unsigned int *count) |
5ad333eb6 Lumpy Reclaim V4 |
1016 1017 |
{ int nr_active = 0; |
4f98a2fee vmscan: split LRU... |
1018 |
int lru; |
5ad333eb6 Lumpy Reclaim V4 |
1019 |
struct page *page; |
4f98a2fee vmscan: split LRU... |
1020 |
list_for_each_entry(page, page_list, lru) { |
401a8e1c1 mm: introduce pag... |
1021 |
lru = page_lru_base_type(page); |
5ad333eb6 Lumpy Reclaim V4 |
1022 |
if (PageActive(page)) { |
4f98a2fee vmscan: split LRU... |
1023 |
lru += LRU_ACTIVE; |
5ad333eb6 Lumpy Reclaim V4 |
1024 1025 1026 |
ClearPageActive(page); nr_active++; } |
1489fa14c vmscan: update is... |
1027 1028 |
if (count) count[lru]++; |
4f98a2fee vmscan: split LRU... |
1029 |
} |
5ad333eb6 Lumpy Reclaim V4 |
1030 1031 1032 |
return nr_active; } |
62695a84e vmscan: move isol... |
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 |
/** * 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 Unevictable LRU I... |
1044 1045 1046 |
* 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 vmscan: move isol... |
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 |
* * 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 Unevictable LRU I... |
1067 |
int lru = page_lru(page); |
62695a84e vmscan: move isol... |
1068 1069 |
ret = 0; ClearPageLRU(page); |
4f98a2fee vmscan: split LRU... |
1070 |
|
4f98a2fee vmscan: split LRU... |
1071 |
del_page_from_lru_list(zone, page, lru); |
62695a84e vmscan: move isol... |
1072 1073 1074 1075 1076 |
} spin_unlock_irq(&zone->lru_lock); } return ret; } |
5ad333eb6 Lumpy Reclaim V4 |
1077 |
/* |
35cd78156 vmscan: throttle ... |
1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 |
* 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; } /* |
666356297 vmscan: set up pa... |
1103 1104 1105 |
* TODO: Try merging with migrations version of putback_lru_pages */ static noinline_for_stack void |
1489fa14c vmscan: update is... |
1106 |
putback_lru_pages(struct zone *zone, struct scan_control *sc, |
666356297 vmscan: set up pa... |
1107 1108 1109 1110 1111 |
unsigned long nr_anon, unsigned long nr_file, struct list_head *page_list) { struct page *page; struct pagevec pvec; |
1489fa14c vmscan: update is... |
1112 |
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); |
666356297 vmscan: set up pa... |
1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 |
pagevec_init(&pvec, 1); /* * Put back any unfreeable pages. */ spin_lock(&zone->lru_lock); while (!list_empty(page_list)) { int lru; page = lru_to_page(page_list); VM_BUG_ON(PageLRU(page)); list_del(&page->lru); 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); if (is_active_lru(lru)) { int file = is_file_lru(lru); reclaim_stat->recent_rotated[file]++; } if (!pagevec_add(&pvec, page)) { spin_unlock_irq(&zone->lru_lock); __pagevec_release(&pvec); spin_lock_irq(&zone->lru_lock); } } __mod_zone_page_state(zone, NR_ISOLATED_ANON, -nr_anon); __mod_zone_page_state(zone, NR_ISOLATED_FILE, -nr_file); spin_unlock_irq(&zone->lru_lock); pagevec_release(&pvec); } |
1489fa14c vmscan: update is... |
1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 |
static noinline_for_stack void update_isolated_counts(struct zone *zone, struct scan_control *sc, unsigned long *nr_anon, unsigned long *nr_file, struct list_head *isolated_list) { unsigned long nr_active; unsigned int count[NR_LRU_LISTS] = { 0, }; struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); nr_active = clear_active_flags(isolated_list, count); __count_vm_events(PGDEACTIVATE, nr_active); __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]); *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); reclaim_stat->recent_scanned[0] += *nr_anon; reclaim_stat->recent_scanned[1] += *nr_file; } |
666356297 vmscan: set up pa... |
1180 |
/* |
e31f3698c vmscan: raise the... |
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 |
* Returns true if the caller should wait to clean dirty/writeback pages. * * If we are direct reclaiming for contiguous pages and we do not reclaim * everything in the list, try again and wait for writeback IO to complete. * This will stall high-order allocations noticeably. Only do that when really * need to free the pages under high memory pressure. */ static inline bool should_reclaim_stall(unsigned long nr_taken, unsigned long nr_freed, int priority, struct scan_control *sc) { int lumpy_stall_priority; /* kswapd should not stall on sync IO */ if (current_is_kswapd()) return false; /* Only stall on lumpy reclaim */ if (!sc->lumpy_reclaim_mode) return false; /* If we have relaimed everything on the isolated list, no stall */ if (nr_freed == nr_taken) return false; /* * For high-order allocations, there are two stall thresholds. * High-cost allocations stall immediately where as lower * order allocations such as stacks require the scanning * priority to be much higher before stalling. */ if (sc->order > PAGE_ALLOC_COSTLY_ORDER) lumpy_stall_priority = DEF_PRIORITY; else lumpy_stall_priority = DEF_PRIORITY / 3; return priority <= lumpy_stall_priority; } /* |
1742f19fa [PATCH] vmscan: r... |
1222 1223 |
* shrink_inactive_list() is a helper for shrink_zone(). It returns the number * of reclaimed pages |
1da177e4c Linux-2.6.12-rc2 |
1224 |
*/ |
666356297 vmscan: set up pa... |
1225 1226 1227 |
static noinline_for_stack unsigned long shrink_inactive_list(unsigned long nr_to_scan, struct zone *zone, struct scan_control *sc, int priority, int file) |
1da177e4c Linux-2.6.12-rc2 |
1228 1229 |
{ LIST_HEAD(page_list); |
e247dbce5 vmscan: simplify ... |
1230 |
unsigned long nr_scanned; |
05ff51376 [PATCH] vmscan re... |
1231 |
unsigned long nr_reclaimed = 0; |
e247dbce5 vmscan: simplify ... |
1232 1233 |
unsigned long nr_taken; unsigned long nr_active; |
e247dbce5 vmscan: simplify ... |
1234 1235 |
unsigned long nr_anon; unsigned long nr_file; |
78dc583d3 vmscan: low order... |
1236 |
|
35cd78156 vmscan: throttle ... |
1237 |
while (unlikely(too_many_isolated(zone, file, sc))) { |
58355c787 congestion_wait()... |
1238 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
35cd78156 vmscan: throttle ... |
1239 1240 1241 1242 1243 |
/* We are about to die and free our memory. Return now. */ if (fatal_signal_pending(current)) return SWAP_CLUSTER_MAX; } |
1da177e4c Linux-2.6.12-rc2 |
1244 |
|
1da177e4c Linux-2.6.12-rc2 |
1245 1246 |
lru_add_drain(); spin_lock_irq(&zone->lru_lock); |
b35ea17b7 mm: shrink_inacti... |
1247 |
|
e247dbce5 vmscan: simplify ... |
1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 |
if (scanning_global_lru(sc)) { nr_taken = isolate_pages_global(nr_to_scan, &page_list, &nr_scanned, sc->order, sc->lumpy_reclaim_mode ? ISOLATE_BOTH : ISOLATE_INACTIVE, zone, 0, file); zone->pages_scanned += nr_scanned; if (current_is_kswapd()) __count_zone_vm_events(PGSCAN_KSWAPD, zone, nr_scanned); else __count_zone_vm_events(PGSCAN_DIRECT, zone, nr_scanned); } else { nr_taken = mem_cgroup_isolate_pages(nr_to_scan, &page_list, &nr_scanned, sc->order, sc->lumpy_reclaim_mode ? ISOLATE_BOTH : ISOLATE_INACTIVE, zone, sc->mem_cgroup, 0, file); /* * mem_cgroup_isolate_pages() keeps track of * scanned pages on its own. */ } |
b35ea17b7 mm: shrink_inacti... |
1273 |
|
666356297 vmscan: set up pa... |
1274 1275 1276 1277 |
if (nr_taken == 0) { spin_unlock_irq(&zone->lru_lock); return 0; } |
5ad333eb6 Lumpy Reclaim V4 |
1278 |
|
1489fa14c vmscan: update is... |
1279 |
update_isolated_counts(zone, sc, &nr_anon, &nr_file, &page_list); |
1da177e4c Linux-2.6.12-rc2 |
1280 |
|
e247dbce5 vmscan: simplify ... |
1281 |
spin_unlock_irq(&zone->lru_lock); |
c661b078f synchronous lumpy... |
1282 |
|
e247dbce5 vmscan: simplify ... |
1283 |
nr_reclaimed = shrink_page_list(&page_list, sc, PAGEOUT_IO_ASYNC); |
c661b078f synchronous lumpy... |
1284 |
|
e31f3698c vmscan: raise the... |
1285 1286 |
/* Check if we should syncronously wait for writeback */ if (should_reclaim_stall(nr_taken, nr_reclaimed, priority, sc)) { |
e247dbce5 vmscan: simplify ... |
1287 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
c661b078f synchronous lumpy... |
1288 |
|
e247dbce5 vmscan: simplify ... |
1289 1290 1291 1292 |
/* * The attempt at page out may have made some * of the pages active, mark them inactive again. */ |
1489fa14c vmscan: update is... |
1293 |
nr_active = clear_active_flags(&page_list, NULL); |
e247dbce5 vmscan: simplify ... |
1294 |
count_vm_events(PGDEACTIVATE, nr_active); |
c661b078f synchronous lumpy... |
1295 |
|
e247dbce5 vmscan: simplify ... |
1296 1297 |
nr_reclaimed += shrink_page_list(&page_list, sc, PAGEOUT_IO_SYNC); } |
b35ea17b7 mm: shrink_inacti... |
1298 |
|
e247dbce5 vmscan: simplify ... |
1299 1300 1301 1302 |
local_irq_disable(); if (current_is_kswapd()) __count_vm_events(KSWAPD_STEAL, nr_reclaimed); __count_zone_vm_events(PGSTEAL, zone, nr_reclaimed); |
a74609faf [PATCH] mm: page_... |
1303 |
|
1489fa14c vmscan: update is... |
1304 |
putback_lru_pages(zone, sc, nr_anon, nr_file, &page_list); |
05ff51376 [PATCH] vmscan re... |
1305 |
return nr_reclaimed; |
1da177e4c Linux-2.6.12-rc2 |
1306 |
} |
3bb1a852a [PATCH] vmscan: F... |
1307 |
/* |
1da177e4c Linux-2.6.12-rc2 |
1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 |
* 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 per-zone and recl... |
1324 |
|
3eb4140f0 vmscan: merge dup... |
1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 |
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 vmscan: merge dup... |
1337 1338 1339 |
VM_BUG_ON(PageLRU(page)); SetPageLRU(page); |
3eb4140f0 vmscan: merge dup... |
1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 |
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 per-zone and recl... |
1356 |
|
1742f19fa [PATCH] vmscan: r... |
1357 |
static void shrink_active_list(unsigned long nr_pages, struct zone *zone, |
4f98a2fee vmscan: split LRU... |
1358 |
struct scan_control *sc, int priority, int file) |
1da177e4c Linux-2.6.12-rc2 |
1359 |
{ |
44c241f16 mm: rename pgmove... |
1360 |
unsigned long nr_taken; |
69e05944a [PATCH] vmscan: u... |
1361 |
unsigned long pgscanned; |
6fe6b7e35 vmscan: report vm... |
1362 |
unsigned long vm_flags; |
1da177e4c Linux-2.6.12-rc2 |
1363 |
LIST_HEAD(l_hold); /* The pages which were snipped off */ |
8cab4754d vmscan: make mapp... |
1364 |
LIST_HEAD(l_active); |
b69408e88 vmscan: Use an in... |
1365 |
LIST_HEAD(l_inactive); |
1da177e4c Linux-2.6.12-rc2 |
1366 |
struct page *page; |
6e9015716 mm: introduce zon... |
1367 |
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); |
44c241f16 mm: rename pgmove... |
1368 |
unsigned long nr_rotated = 0; |
1da177e4c Linux-2.6.12-rc2 |
1369 1370 1371 |
lru_add_drain(); spin_lock_irq(&zone->lru_lock); |
e72e2bd67 memcg: rename sca... |
1372 |
if (scanning_global_lru(sc)) { |
8b25c6d22 vmscan: remove is... |
1373 1374 1375 1376 |
nr_taken = isolate_pages_global(nr_pages, &l_hold, &pgscanned, sc->order, ISOLATE_ACTIVE, zone, 1, file); |
1cfb419b3 per-zone and recl... |
1377 |
zone->pages_scanned += pgscanned; |
8b25c6d22 vmscan: remove is... |
1378 1379 1380 1381 1382 1383 1384 1385 1386 |
} else { nr_taken = mem_cgroup_isolate_pages(nr_pages, &l_hold, &pgscanned, sc->order, ISOLATE_ACTIVE, zone, sc->mem_cgroup, 1, file); /* * mem_cgroup_isolate_pages() keeps track of * scanned pages on its own. */ |
4f98a2fee vmscan: split LRU... |
1387 |
} |
8b25c6d22 vmscan: remove is... |
1388 |
|
b7c46d151 mm: drop unneeded... |
1389 |
reclaim_stat->recent_scanned[file] += nr_taken; |
1cfb419b3 per-zone and recl... |
1390 |
|
3eb4140f0 vmscan: merge dup... |
1391 |
__count_zone_vm_events(PGREFILL, zone, pgscanned); |
4f98a2fee vmscan: split LRU... |
1392 |
if (file) |
44c241f16 mm: rename pgmove... |
1393 |
__mod_zone_page_state(zone, NR_ACTIVE_FILE, -nr_taken); |
4f98a2fee vmscan: split LRU... |
1394 |
else |
44c241f16 mm: rename pgmove... |
1395 |
__mod_zone_page_state(zone, NR_ACTIVE_ANON, -nr_taken); |
a731286de mm: vmstat: add i... |
1396 |
__mod_zone_page_state(zone, NR_ISOLATED_ANON + file, nr_taken); |
1da177e4c Linux-2.6.12-rc2 |
1397 |
spin_unlock_irq(&zone->lru_lock); |
1da177e4c Linux-2.6.12-rc2 |
1398 1399 1400 1401 |
while (!list_empty(&l_hold)) { cond_resched(); page = lru_to_page(&l_hold); list_del(&page->lru); |
7e9cd4842 vmscan: fix pagec... |
1402 |
|
894bc3104 Unevictable LRU I... |
1403 1404 1405 1406 |
if (unlikely(!page_evictable(page, NULL))) { putback_lru_page(page); continue; } |
645747462 vmscan: detect ma... |
1407 |
if (page_referenced(page, 0, sc->mem_cgroup, &vm_flags)) { |
44c241f16 mm: rename pgmove... |
1408 |
nr_rotated++; |
8cab4754d vmscan: make mapp... |
1409 1410 1411 1412 1413 1414 1415 1416 1417 |
/* * 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 vmscan: limit VM_... |
1418 |
if ((vm_flags & VM_EXEC) && page_is_file_cache(page)) { |
8cab4754d vmscan: make mapp... |
1419 1420 1421 1422 |
list_add(&page->lru, &l_active); continue; } } |
7e9cd4842 vmscan: fix pagec... |
1423 |
|
5205e56ee vmscan: move Clea... |
1424 |
ClearPageActive(page); /* we are de-activating */ |
1da177e4c Linux-2.6.12-rc2 |
1425 1426 |
list_add(&page->lru, &l_inactive); } |
b555749aa vmscan: shrink_ac... |
1427 |
/* |
8cab4754d vmscan: make mapp... |
1428 |
* Move pages back to the lru list. |
b555749aa vmscan: shrink_ac... |
1429 |
*/ |
2a1dc5097 vmscan: protect z... |
1430 |
spin_lock_irq(&zone->lru_lock); |
4f98a2fee vmscan: split LRU... |
1431 |
/* |
8cab4754d vmscan: make mapp... |
1432 1433 1434 1435 |
* 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 vmscan: fix pagec... |
1436 |
*/ |
b7c46d151 mm: drop unneeded... |
1437 |
reclaim_stat->recent_rotated[file] += nr_rotated; |
556adecba vmscan: second ch... |
1438 |
|
3eb4140f0 vmscan: merge dup... |
1439 1440 1441 1442 |
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 mm: vmstat: add i... |
1443 |
__mod_zone_page_state(zone, NR_ISOLATED_ANON + file, -nr_taken); |
f8891e5e1 [PATCH] Light wei... |
1444 |
spin_unlock_irq(&zone->lru_lock); |
1da177e4c Linux-2.6.12-rc2 |
1445 |
} |
14797e236 memcg: add inacti... |
1446 |
static int inactive_anon_is_low_global(struct zone *zone) |
f89eb90e3 inactive_anon_is_... |
1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 |
{ 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 memcg: add inacti... |
1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 |
/** * 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 memcg: rename sca... |
1469 |
if (scanning_global_lru(sc)) |
14797e236 memcg: add inacti... |
1470 1471 |
low = inactive_anon_is_low_global(zone); else |
c772be939 memcg: fix calcul... |
1472 |
low = mem_cgroup_inactive_anon_is_low(sc->mem_cgroup); |
14797e236 memcg: add inacti... |
1473 1474 |
return low; } |
56e49d218 vmscan: evict use... |
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 |
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 vmscan: do not ev... |
1510 1511 1512 1513 1514 1515 1516 1517 |
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 vmscan: split LRU... |
1518 |
static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan, |
b69408e88 vmscan: Use an in... |
1519 1520 |
struct zone *zone, struct scan_control *sc, int priority) { |
4f98a2fee vmscan: split LRU... |
1521 |
int file = is_file_lru(lru); |
b39415b27 vmscan: do not ev... |
1522 1523 1524 |
if (is_active_lru(lru)) { if (inactive_list_is_low(zone, sc, file)) shrink_active_list(nr_to_scan, zone, sc, priority, file); |
556adecba vmscan: second ch... |
1525 1526 |
return 0; } |
33c120ed2 more aggressively... |
1527 |
return shrink_inactive_list(nr_to_scan, zone, sc, priority, file); |
4f98a2fee vmscan: split LRU... |
1528 1529 1530 |
} /* |
76a33fc38 vmscan: prevent g... |
1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 |
* 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, unsigned long *nr_saved_scan) { unsigned long nr; *nr_saved_scan += nr_to_scan; nr = *nr_saved_scan; if (nr >= SWAP_CLUSTER_MAX) *nr_saved_scan = 0; else nr = 0; return nr; } /* |
4f98a2fee vmscan: split LRU... |
1551 1552 1553 1554 1555 |
* 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. * |
76a33fc38 vmscan: prevent g... |
1556 |
* nr[0] = anon pages to scan; nr[1] = file pages to scan |
4f98a2fee vmscan: split LRU... |
1557 |
*/ |
76a33fc38 vmscan: prevent g... |
1558 1559 |
static void get_scan_count(struct zone *zone, struct scan_control *sc, unsigned long *nr, int priority) |
4f98a2fee vmscan: split LRU... |
1560 1561 1562 1563 |
{ unsigned long anon, file, free; unsigned long anon_prio, file_prio; unsigned long ap, fp; |
6e9015716 mm: introduce zon... |
1564 |
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); |
76a33fc38 vmscan: prevent g... |
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 |
u64 fraction[2], denominator; enum lru_list l; int noswap = 0; /* If we have no swap space, do not bother scanning anon pages. */ if (!sc->may_swap || (nr_swap_pages <= 0)) { noswap = 1; fraction[0] = 0; fraction[1] = 1; denominator = 1; goto out; } |
4f98a2fee vmscan: split LRU... |
1577 |
|
0b2176763 mm/vmscan: rename... |
1578 1579 1580 1581 |
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 mm: optimize get_... |
1582 |
|
e72e2bd67 memcg: rename sca... |
1583 |
if (scanning_global_lru(sc)) { |
eeee9a8cd mm: make get_scan... |
1584 1585 1586 |
free = zone_page_state(zone, NR_FREE_PAGES); /* If we have very few page cache pages, force-scan anon pages. */ |
418589663 page allocator: u... |
1587 |
if (unlikely(file + free <= high_wmark_pages(zone))) { |
76a33fc38 vmscan: prevent g... |
1588 1589 1590 1591 |
fraction[0] = 1; fraction[1] = 0; denominator = 1; goto out; |
eeee9a8cd mm: make get_scan... |
1592 |
} |
4f98a2fee vmscan: split LRU... |
1593 1594 1595 |
} /* |
58c37f6e0 vmscan: protect r... |
1596 1597 1598 1599 1600 1601 1602 |
* 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; /* |
4f98a2fee vmscan: split LRU... |
1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 |
* 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] */ |
58c37f6e0 vmscan: protect r... |
1613 |
spin_lock_irq(&zone->lru_lock); |
6e9015716 mm: introduce zon... |
1614 |
if (unlikely(reclaim_stat->recent_scanned[0] > anon / 4)) { |
6e9015716 mm: introduce zon... |
1615 1616 |
reclaim_stat->recent_scanned[0] /= 2; reclaim_stat->recent_rotated[0] /= 2; |
4f98a2fee vmscan: split LRU... |
1617 |
} |
6e9015716 mm: introduce zon... |
1618 |
if (unlikely(reclaim_stat->recent_scanned[1] > file / 4)) { |
6e9015716 mm: introduce zon... |
1619 1620 |
reclaim_stat->recent_scanned[1] /= 2; reclaim_stat->recent_rotated[1] /= 2; |
4f98a2fee vmscan: split LRU... |
1621 1622 1623 |
} /* |
00d8089c5 vmscan: fix get_s... |
1624 1625 1626 |
* 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 vmscan: split LRU... |
1627 |
*/ |
6e9015716 mm: introduce zon... |
1628 1629 |
ap = (anon_prio + 1) * (reclaim_stat->recent_scanned[0] + 1); ap /= reclaim_stat->recent_rotated[0] + 1; |
4f98a2fee vmscan: split LRU... |
1630 |
|
6e9015716 mm: introduce zon... |
1631 1632 |
fp = (file_prio + 1) * (reclaim_stat->recent_scanned[1] + 1); fp /= reclaim_stat->recent_rotated[1] + 1; |
58c37f6e0 vmscan: protect r... |
1633 |
spin_unlock_irq(&zone->lru_lock); |
4f98a2fee vmscan: split LRU... |
1634 |
|
76a33fc38 vmscan: prevent g... |
1635 1636 1637 1638 1639 1640 1641 |
fraction[0] = ap; fraction[1] = fp; denominator = ap + fp + 1; out: for_each_evictable_lru(l) { int file = is_file_lru(l); unsigned long scan; |
6e08a369e vmscan: cleanup t... |
1642 |
|
76a33fc38 vmscan: prevent g... |
1643 1644 1645 1646 1647 1648 1649 1650 |
scan = zone_nr_lru_pages(zone, sc, l); if (priority || noswap) { scan >>= priority; scan = div64_u64(scan * fraction[file], denominator); } nr[l] = nr_scan_try_batch(scan, &reclaim_stat->nr_saved_scan[l]); } |
6e08a369e vmscan: cleanup t... |
1651 |
} |
4f98a2fee vmscan: split LRU... |
1652 |
|
5f53e7629 vmscan: page_chec... |
1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 |
static void set_lumpy_reclaim_mode(int priority, struct scan_control *sc) { /* * 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. */ if (sc->order > PAGE_ALLOC_COSTLY_ORDER) sc->lumpy_reclaim_mode = 1; else if (sc->order && priority < DEF_PRIORITY - 2) sc->lumpy_reclaim_mode = 1; else sc->lumpy_reclaim_mode = 0; } |
1da177e4c Linux-2.6.12-rc2 |
1667 1668 1669 |
/* * This is a basic per-zone page freer. Used by both kswapd and direct reclaim. */ |
a79311c14 vmscan: bail out ... |
1670 |
static void shrink_zone(int priority, struct zone *zone, |
05ff51376 [PATCH] vmscan re... |
1671 |
struct scan_control *sc) |
1da177e4c Linux-2.6.12-rc2 |
1672 |
{ |
b69408e88 vmscan: Use an in... |
1673 |
unsigned long nr[NR_LRU_LISTS]; |
8695949a1 [PATCH] Thin out ... |
1674 |
unsigned long nr_to_scan; |
b69408e88 vmscan: Use an in... |
1675 |
enum lru_list l; |
01dbe5c9b vmscan: improve r... |
1676 |
unsigned long nr_reclaimed = sc->nr_reclaimed; |
22fba3354 vmscan: separate ... |
1677 |
unsigned long nr_to_reclaim = sc->nr_to_reclaim; |
e0f79b8f1 vmscan: don't acc... |
1678 |
|
76a33fc38 vmscan: prevent g... |
1679 |
get_scan_count(zone, sc, nr, priority); |
1da177e4c Linux-2.6.12-rc2 |
1680 |
|
5f53e7629 vmscan: page_chec... |
1681 |
set_lumpy_reclaim_mode(priority, sc); |
556adecba vmscan: second ch... |
1682 1683 |
while (nr[LRU_INACTIVE_ANON] || nr[LRU_ACTIVE_FILE] || nr[LRU_INACTIVE_FILE]) { |
894bc3104 Unevictable LRU I... |
1684 |
for_each_evictable_lru(l) { |
b69408e88 vmscan: Use an in... |
1685 |
if (nr[l]) { |
ece74b2e7 vmscan: kill sc.s... |
1686 1687 |
nr_to_scan = min_t(unsigned long, nr[l], SWAP_CLUSTER_MAX); |
b69408e88 vmscan: Use an in... |
1688 |
nr[l] -= nr_to_scan; |
1da177e4c Linux-2.6.12-rc2 |
1689 |
|
01dbe5c9b vmscan: improve r... |
1690 1691 |
nr_reclaimed += shrink_list(l, nr_to_scan, zone, sc, priority); |
b69408e88 vmscan: Use an in... |
1692 |
} |
1da177e4c Linux-2.6.12-rc2 |
1693 |
} |
a79311c14 vmscan: bail out ... |
1694 1695 1696 1697 1698 1699 1700 1701 |
/* * 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 vmscan: make cons... |
1702 |
if (nr_reclaimed >= nr_to_reclaim && priority < DEF_PRIORITY) |
a79311c14 vmscan: bail out ... |
1703 |
break; |
1da177e4c Linux-2.6.12-rc2 |
1704 |
} |
01dbe5c9b vmscan: improve r... |
1705 |
sc->nr_reclaimed = nr_reclaimed; |
556adecba vmscan: second ch... |
1706 1707 1708 1709 |
/* * Even if we did not try to evict anon pages at all, we want to * rebalance the anon lru active/inactive ratio. */ |
69c854817 vmscan: prevent s... |
1710 |
if (inactive_anon_is_low(zone, sc) && nr_swap_pages > 0) |
556adecba vmscan: second ch... |
1711 |
shrink_active_list(SWAP_CLUSTER_MAX, zone, sc, priority, 0); |
232ea4d69 [PATCH] throttle_... |
1712 |
throttle_vm_writeout(sc->gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
1713 1714 1715 1716 1717 1718 1719 |
} /* * 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 page allocator: u... |
1720 1721 |
* We reclaim from a zone even if that zone is over high_wmark_pages(zone). * Because: |
1da177e4c Linux-2.6.12-rc2 |
1722 1723 |
* a) The caller may be trying to free *extra* pages to satisfy a higher-order * allocation or |
418589663 page allocator: u... |
1724 1725 1726 |
* 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 Linux-2.6.12-rc2 |
1727 |
* |
1da177e4c Linux-2.6.12-rc2 |
1728 1729 1730 |
* If a zone is deemed to be full of pinned pages then just give it a light * scan then give up on it. */ |
bb21c7ce1 vmscan: fix do_tr... |
1731 |
static bool shrink_zones(int priority, struct zonelist *zonelist, |
05ff51376 [PATCH] vmscan re... |
1732 |
struct scan_control *sc) |
1da177e4c Linux-2.6.12-rc2 |
1733 |
{ |
dd1a239f6 mm: have zonelist... |
1734 |
struct zoneref *z; |
54a6eb5c4 mm: use two zonel... |
1735 |
struct zone *zone; |
bb21c7ce1 vmscan: fix do_tr... |
1736 |
bool all_unreclaimable = true; |
1cfb419b3 per-zone and recl... |
1737 |
|
d4debc66d vmscan: remove un... |
1738 1739 |
for_each_zone_zonelist_nodemask(zone, z, zonelist, gfp_zone(sc->gfp_mask), sc->nodemask) { |
f3fe65122 [PATCH] mm: add p... |
1740 |
if (!populated_zone(zone)) |
1da177e4c Linux-2.6.12-rc2 |
1741 |
continue; |
1cfb419b3 per-zone and recl... |
1742 1743 1744 1745 |
/* * Take care memory controller reclaiming has small influence * to global LRU. */ |
e72e2bd67 memcg: rename sca... |
1746 |
if (scanning_global_lru(sc)) { |
1cfb419b3 per-zone and recl... |
1747 1748 |
if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) continue; |
93e4a89a8 mm: restore zone-... |
1749 |
if (zone->all_unreclaimable && priority != DEF_PRIORITY) |
1cfb419b3 per-zone and recl... |
1750 |
continue; /* Let kswapd poll it */ |
1cfb419b3 per-zone and recl... |
1751 |
} |
408d85441 [PATCH] oom: use ... |
1752 |
|
a79311c14 vmscan: bail out ... |
1753 |
shrink_zone(priority, zone, sc); |
bb21c7ce1 vmscan: fix do_tr... |
1754 |
all_unreclaimable = false; |
1da177e4c Linux-2.6.12-rc2 |
1755 |
} |
bb21c7ce1 vmscan: fix do_tr... |
1756 |
return all_unreclaimable; |
1da177e4c Linux-2.6.12-rc2 |
1757 |
} |
4f98a2fee vmscan: split LRU... |
1758 |
|
1da177e4c Linux-2.6.12-rc2 |
1759 1760 1761 1762 1763 1764 1765 1766 |
/* * 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 writeback: get ri... |
1767 1768 1769 1770 |
* 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 page allocator: s... |
1771 1772 1773 |
* * returns: 0, if no pages reclaimed * else, the number of pages reclaimed |
1da177e4c Linux-2.6.12-rc2 |
1774 |
*/ |
dac1d27bc mm: use zonelists... |
1775 |
static unsigned long do_try_to_free_pages(struct zonelist *zonelist, |
dd1a239f6 mm: have zonelist... |
1776 |
struct scan_control *sc) |
1da177e4c Linux-2.6.12-rc2 |
1777 1778 |
{ int priority; |
bb21c7ce1 vmscan: fix do_tr... |
1779 |
bool all_unreclaimable; |
69e05944a [PATCH] vmscan: u... |
1780 |
unsigned long total_scanned = 0; |
1da177e4c Linux-2.6.12-rc2 |
1781 |
struct reclaim_state *reclaim_state = current->reclaim_state; |
dd1a239f6 mm: have zonelist... |
1782 |
struct zoneref *z; |
54a6eb5c4 mm: use two zonel... |
1783 |
struct zone *zone; |
22fba3354 vmscan: separate ... |
1784 |
unsigned long writeback_threshold; |
1da177e4c Linux-2.6.12-rc2 |
1785 |
|
c0ff7453b cpuset,mm: fix no... |
1786 |
get_mems_allowed(); |
873b47717 per-task-delay-ac... |
1787 |
delayacct_freepages_start(); |
e72e2bd67 memcg: rename sca... |
1788 |
if (scanning_global_lru(sc)) |
1cfb419b3 per-zone and recl... |
1789 |
count_vm_event(ALLOCSTALL); |
1da177e4c Linux-2.6.12-rc2 |
1790 1791 |
for (priority = DEF_PRIORITY; priority >= 0; priority--) { |
66e1707bc Memory controller... |
1792 |
sc->nr_scanned = 0; |
f7b7fd8f3 [PATCH] temporari... |
1793 1794 |
if (!priority) disable_swap_token(); |
bb21c7ce1 vmscan: fix do_tr... |
1795 |
all_unreclaimable = shrink_zones(priority, zonelist, sc); |
66e1707bc Memory controller... |
1796 1797 1798 1799 |
/* * Don't shrink slabs when reclaiming memory from * over limit cgroups */ |
e72e2bd67 memcg: rename sca... |
1800 |
if (scanning_global_lru(sc)) { |
c6a8a8c58 vmscan: recalcula... |
1801 |
unsigned long lru_pages = 0; |
d4debc66d vmscan: remove un... |
1802 1803 |
for_each_zone_zonelist(zone, z, zonelist, gfp_zone(sc->gfp_mask)) { |
c6a8a8c58 vmscan: recalcula... |
1804 1805 1806 1807 1808 |
if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) continue; lru_pages += zone_reclaimable_pages(zone); } |
dd1a239f6 mm: have zonelist... |
1809 |
shrink_slab(sc->nr_scanned, sc->gfp_mask, lru_pages); |
91a45470f per-zone and recl... |
1810 |
if (reclaim_state) { |
a79311c14 vmscan: bail out ... |
1811 |
sc->nr_reclaimed += reclaim_state->reclaimed_slab; |
91a45470f per-zone and recl... |
1812 1813 |
reclaim_state->reclaimed_slab = 0; } |
1da177e4c Linux-2.6.12-rc2 |
1814 |
} |
66e1707bc Memory controller... |
1815 |
total_scanned += sc->nr_scanned; |
bb21c7ce1 vmscan: fix do_tr... |
1816 |
if (sc->nr_reclaimed >= sc->nr_to_reclaim) |
1da177e4c Linux-2.6.12-rc2 |
1817 |
goto out; |
1da177e4c Linux-2.6.12-rc2 |
1818 1819 1820 1821 1822 1823 1824 1825 |
/* * 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 vmscan: separate ... |
1826 1827 |
writeback_threshold = sc->nr_to_reclaim + sc->nr_to_reclaim / 2; if (total_scanned > writeback_threshold) { |
03ba3782e writeback: switch... |
1828 |
wakeup_flusher_threads(laptop_mode ? 0 : total_scanned); |
66e1707bc Memory controller... |
1829 |
sc->may_writepage = 1; |
1da177e4c Linux-2.6.12-rc2 |
1830 1831 1832 |
} /* Take a nap, wait for some writeback to complete */ |
7b51755c3 vmscan: kill hibe... |
1833 1834 |
if (!sc->hibernation_mode && sc->nr_scanned && priority < DEF_PRIORITY - 2) |
8aa7e847d Fix congestion_wa... |
1835 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
1da177e4c Linux-2.6.12-rc2 |
1836 |
} |
bb21c7ce1 vmscan: fix do_tr... |
1837 |
|
1da177e4c Linux-2.6.12-rc2 |
1838 |
out: |
3bb1a852a [PATCH] vmscan: F... |
1839 1840 1841 1842 1843 1844 1845 1846 1847 |
/* * 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 Linux-2.6.12-rc2 |
1848 |
|
873b47717 per-task-delay-ac... |
1849 |
delayacct_freepages_end(); |
c0ff7453b cpuset,mm: fix no... |
1850 |
put_mems_allowed(); |
873b47717 per-task-delay-ac... |
1851 |
|
bb21c7ce1 vmscan: fix do_tr... |
1852 1853 1854 1855 1856 1857 1858 1859 |
if (sc->nr_reclaimed) return sc->nr_reclaimed; /* top priority shrink_zones still had more to do? don't OOM, then */ if (scanning_global_lru(sc) && !all_unreclaimable) return 1; return 0; |
1da177e4c Linux-2.6.12-rc2 |
1860 |
} |
dac1d27bc mm: use zonelists... |
1861 |
unsigned long try_to_free_pages(struct zonelist *zonelist, int order, |
327c0e968 vmscan: fix it to... |
1862 |
gfp_t gfp_mask, nodemask_t *nodemask) |
66e1707bc Memory controller... |
1863 |
{ |
33906bc5c vmscan: tracing: ... |
1864 |
unsigned long nr_reclaimed; |
66e1707bc Memory controller... |
1865 1866 1867 |
struct scan_control sc = { .gfp_mask = gfp_mask, .may_writepage = !laptop_mode, |
22fba3354 vmscan: separate ... |
1868 |
.nr_to_reclaim = SWAP_CLUSTER_MAX, |
a6dc60f89 vmscan: rename sc... |
1869 |
.may_unmap = 1, |
2e2e42598 vmscan,memcg: rei... |
1870 |
.may_swap = 1, |
66e1707bc Memory controller... |
1871 1872 1873 |
.swappiness = vm_swappiness, .order = order, .mem_cgroup = NULL, |
327c0e968 vmscan: fix it to... |
1874 |
.nodemask = nodemask, |
66e1707bc Memory controller... |
1875 |
}; |
33906bc5c vmscan: tracing: ... |
1876 1877 1878 1879 1880 1881 1882 1883 1884 |
trace_mm_vmscan_direct_reclaim_begin(order, sc.may_writepage, gfp_mask); nr_reclaimed = do_try_to_free_pages(zonelist, &sc); trace_mm_vmscan_direct_reclaim_end(nr_reclaimed); return nr_reclaimed; |
66e1707bc Memory controller... |
1885 |
} |
00f0b8259 Memory controller... |
1886 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR |
66e1707bc Memory controller... |
1887 |
|
4e4169535 memory controller... |
1888 1889 1890 |
unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem, gfp_t gfp_mask, bool noswap, unsigned int swappiness, |
14fec7968 memcg: mem_cgroup... |
1891 |
struct zone *zone) |
4e4169535 memory controller... |
1892 1893 |
{ struct scan_control sc = { |
b8f5c5664 memcg: sc.nr_to_r... |
1894 |
.nr_to_reclaim = SWAP_CLUSTER_MAX, |
4e4169535 memory controller... |
1895 1896 1897 |
.may_writepage = !laptop_mode, .may_unmap = 1, .may_swap = !noswap, |
4e4169535 memory controller... |
1898 1899 1900 |
.swappiness = swappiness, .order = 0, .mem_cgroup = mem, |
4e4169535 memory controller... |
1901 |
}; |
4e4169535 memory controller... |
1902 1903 |
sc.gfp_mask = (gfp_mask & GFP_RECLAIM_MASK) | (GFP_HIGHUSER_MOVABLE & ~GFP_RECLAIM_MASK); |
bdce6d9eb memcg, vmscan: ad... |
1904 1905 1906 1907 |
trace_mm_vmscan_memcg_softlimit_reclaim_begin(0, sc.may_writepage, sc.gfp_mask); |
4e4169535 memory controller... |
1908 1909 1910 1911 1912 1913 1914 1915 |
/* * 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); |
bdce6d9eb memcg, vmscan: ad... |
1916 1917 |
trace_mm_vmscan_memcg_softlimit_reclaim_end(sc.nr_reclaimed); |
4e4169535 memory controller... |
1918 1919 |
return sc.nr_reclaimed; } |
e1a1cd590 Memory controller... |
1920 |
unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont, |
a7885eb8a memcg: swappiness |
1921 1922 1923 |
gfp_t gfp_mask, bool noswap, unsigned int swappiness) |
66e1707bc Memory controller... |
1924 |
{ |
4e4169535 memory controller... |
1925 |
struct zonelist *zonelist; |
bdce6d9eb memcg, vmscan: ad... |
1926 |
unsigned long nr_reclaimed; |
66e1707bc Memory controller... |
1927 |
struct scan_control sc = { |
66e1707bc Memory controller... |
1928 |
.may_writepage = !laptop_mode, |
a6dc60f89 vmscan: rename sc... |
1929 |
.may_unmap = 1, |
2e2e42598 vmscan,memcg: rei... |
1930 |
.may_swap = !noswap, |
22fba3354 vmscan: separate ... |
1931 |
.nr_to_reclaim = SWAP_CLUSTER_MAX, |
a7885eb8a memcg: swappiness |
1932 |
.swappiness = swappiness, |
66e1707bc Memory controller... |
1933 1934 |
.order = 0, .mem_cgroup = mem_cont, |
327c0e968 vmscan: fix it to... |
1935 |
.nodemask = NULL, /* we don't care the placement */ |
66e1707bc Memory controller... |
1936 |
}; |
66e1707bc Memory controller... |
1937 |
|
dd1a239f6 mm: have zonelist... |
1938 1939 1940 |
sc.gfp_mask = (gfp_mask & GFP_RECLAIM_MASK) | (GFP_HIGHUSER_MOVABLE & ~GFP_RECLAIM_MASK); zonelist = NODE_DATA(numa_node_id())->node_zonelists; |
bdce6d9eb memcg, vmscan: ad... |
1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 |
trace_mm_vmscan_memcg_reclaim_begin(0, sc.may_writepage, sc.gfp_mask); nr_reclaimed = do_try_to_free_pages(zonelist, &sc); trace_mm_vmscan_memcg_reclaim_end(nr_reclaimed); return nr_reclaimed; |
66e1707bc Memory controller... |
1951 1952 |
} #endif |
f50de2d38 vmscan: have kswa... |
1953 |
/* is kswapd sleeping prematurely? */ |
bb3ab5968 vmscan: stop kswa... |
1954 |
static int sleeping_prematurely(pg_data_t *pgdat, int order, long remaining) |
f50de2d38 vmscan: have kswa... |
1955 |
{ |
bb3ab5968 vmscan: stop kswa... |
1956 |
int i; |
f50de2d38 vmscan: have kswa... |
1957 1958 1959 1960 1961 1962 |
/* 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 vmscan: stop kswa... |
1963 1964 1965 1966 1967 |
for (i = 0; i < pgdat->nr_zones; i++) { struct zone *zone = pgdat->node_zones + i; if (!populated_zone(zone)) continue; |
93e4a89a8 mm: restore zone-... |
1968 |
if (zone->all_unreclaimable) |
de3fab393 vmscan: kswapd: d... |
1969 |
continue; |
f50de2d38 vmscan: have kswa... |
1970 1971 1972 |
if (!zone_watermark_ok(zone, order, high_wmark_pages(zone), 0, 0)) return 1; |
bb3ab5968 vmscan: stop kswa... |
1973 |
} |
f50de2d38 vmscan: have kswa... |
1974 1975 1976 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
1977 1978 |
/* * For kswapd, balance_pgdat() will work across all this node's zones until |
418589663 page allocator: u... |
1979 |
* they are all at high_wmark_pages(zone). |
1da177e4c Linux-2.6.12-rc2 |
1980 |
* |
1da177e4c Linux-2.6.12-rc2 |
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 |
* 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 page allocator: u... |
1992 1993 1994 1995 1996 |
* 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 Linux-2.6.12-rc2 |
1997 |
*/ |
d6277db4a [PATCH] swsusp: r... |
1998 |
static unsigned long balance_pgdat(pg_data_t *pgdat, int order) |
1da177e4c Linux-2.6.12-rc2 |
1999 |
{ |
1da177e4c Linux-2.6.12-rc2 |
2000 2001 2002 |
int all_zones_ok; int priority; int i; |
69e05944a [PATCH] vmscan: u... |
2003 |
unsigned long total_scanned; |
1da177e4c Linux-2.6.12-rc2 |
2004 |
struct reclaim_state *reclaim_state = current->reclaim_state; |
179e96395 [PATCH] vmscan: s... |
2005 2006 |
struct scan_control sc = { .gfp_mask = GFP_KERNEL, |
a6dc60f89 vmscan: rename sc... |
2007 |
.may_unmap = 1, |
2e2e42598 vmscan,memcg: rei... |
2008 |
.may_swap = 1, |
22fba3354 vmscan: separate ... |
2009 2010 2011 2012 2013 |
/* * 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 [PATCH] swsusp: r... |
2014 |
.swappiness = vm_swappiness, |
5ad333eb6 Lumpy Reclaim V4 |
2015 |
.order = order, |
66e1707bc Memory controller... |
2016 |
.mem_cgroup = NULL, |
179e96395 [PATCH] vmscan: s... |
2017 |
}; |
1da177e4c Linux-2.6.12-rc2 |
2018 2019 |
loop_again: total_scanned = 0; |
a79311c14 vmscan: bail out ... |
2020 |
sc.nr_reclaimed = 0; |
c0bbbc73d [PATCH] typo in v... |
2021 |
sc.may_writepage = !laptop_mode; |
f8891e5e1 [PATCH] Light wei... |
2022 |
count_vm_event(PAGEOUTRUN); |
1da177e4c Linux-2.6.12-rc2 |
2023 |
|
1da177e4c Linux-2.6.12-rc2 |
2024 2025 2026 |
for (priority = DEF_PRIORITY; priority >= 0; priority--) { int end_zone = 0; /* Inclusive. 0 = ZONE_DMA */ unsigned long lru_pages = 0; |
bb3ab5968 vmscan: stop kswa... |
2027 |
int has_under_min_watermark_zone = 0; |
1da177e4c Linux-2.6.12-rc2 |
2028 |
|
f7b7fd8f3 [PATCH] temporari... |
2029 2030 2031 |
/* The swap token gets in the way of swapout... */ if (!priority) disable_swap_token(); |
1da177e4c Linux-2.6.12-rc2 |
2032 |
all_zones_ok = 1; |
d6277db4a [PATCH] swsusp: r... |
2033 2034 2035 2036 2037 2038 |
/* * 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 Linux-2.6.12-rc2 |
2039 |
|
d6277db4a [PATCH] swsusp: r... |
2040 2041 |
if (!populated_zone(zone)) continue; |
1da177e4c Linux-2.6.12-rc2 |
2042 |
|
93e4a89a8 mm: restore zone-... |
2043 |
if (zone->all_unreclaimable && priority != DEF_PRIORITY) |
d6277db4a [PATCH] swsusp: r... |
2044 |
continue; |
1da177e4c Linux-2.6.12-rc2 |
2045 |
|
556adecba vmscan: second ch... |
2046 2047 2048 2049 |
/* * Do some background aging of the anon list, to give * pages a chance to be referenced before reclaiming. */ |
14797e236 memcg: add inacti... |
2050 |
if (inactive_anon_is_low(zone, &sc)) |
556adecba vmscan: second ch... |
2051 2052 |
shrink_active_list(SWAP_CLUSTER_MAX, zone, &sc, priority, 0); |
418589663 page allocator: u... |
2053 2054 |
if (!zone_watermark_ok(zone, order, high_wmark_pages(zone), 0, 0)) { |
d6277db4a [PATCH] swsusp: r... |
2055 |
end_zone = i; |
e1dbeda60 [PATCH] balance_p... |
2056 |
break; |
1da177e4c Linux-2.6.12-rc2 |
2057 |
} |
1da177e4c Linux-2.6.12-rc2 |
2058 |
} |
e1dbeda60 [PATCH] balance_p... |
2059 2060 |
if (i < 0) goto out; |
1da177e4c Linux-2.6.12-rc2 |
2061 2062 |
for (i = 0; i <= end_zone; i++) { struct zone *zone = pgdat->node_zones + i; |
adea02a1b mm: count only re... |
2063 |
lru_pages += zone_reclaimable_pages(zone); |
1da177e4c Linux-2.6.12-rc2 |
2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 |
} /* * 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 [PATCH] vmscan: n... |
2077 |
int nr_slab; |
1da177e4c Linux-2.6.12-rc2 |
2078 |
|
f3fe65122 [PATCH] mm: add p... |
2079 |
if (!populated_zone(zone)) |
1da177e4c Linux-2.6.12-rc2 |
2080 |
continue; |
93e4a89a8 mm: restore zone-... |
2081 |
if (zone->all_unreclaimable && priority != DEF_PRIORITY) |
1da177e4c Linux-2.6.12-rc2 |
2082 |
continue; |
1da177e4c Linux-2.6.12-rc2 |
2083 |
sc.nr_scanned = 0; |
4e4169535 memory controller... |
2084 |
|
4e4169535 memory controller... |
2085 2086 2087 2088 |
/* * Call soft limit reclaim before calling shrink_zone. * For now we ignore the return value */ |
00918b6ab memcg: remove nid... |
2089 |
mem_cgroup_soft_limit_reclaim(zone, order, sc.gfp_mask); |
32a4330d4 mm: prevent kswap... |
2090 2091 2092 2093 |
/* * We put equal pressure on every zone, unless one * zone has way too many pages free already. */ |
418589663 page allocator: u... |
2094 2095 |
if (!zone_watermark_ok(zone, order, 8*high_wmark_pages(zone), end_zone, 0)) |
a79311c14 vmscan: bail out ... |
2096 |
shrink_zone(priority, zone, &sc); |
1da177e4c Linux-2.6.12-rc2 |
2097 |
reclaim_state->reclaimed_slab = 0; |
b15e0905f [PATCH] vmscan: n... |
2098 2099 |
nr_slab = shrink_slab(sc.nr_scanned, GFP_KERNEL, lru_pages); |
a79311c14 vmscan: bail out ... |
2100 |
sc.nr_reclaimed += reclaim_state->reclaimed_slab; |
1da177e4c Linux-2.6.12-rc2 |
2101 |
total_scanned += sc.nr_scanned; |
93e4a89a8 mm: restore zone-... |
2102 |
if (zone->all_unreclaimable) |
1da177e4c Linux-2.6.12-rc2 |
2103 |
continue; |
93e4a89a8 mm: restore zone-... |
2104 2105 2106 |
if (nr_slab == 0 && zone->pages_scanned >= (zone_reclaimable_pages(zone) * 6)) zone->all_unreclaimable = 1; |
1da177e4c Linux-2.6.12-rc2 |
2107 2108 2109 2110 2111 2112 |
/* * 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 vmscan: bail out ... |
2113 |
total_scanned > sc.nr_reclaimed + sc.nr_reclaimed / 2) |
1da177e4c Linux-2.6.12-rc2 |
2114 |
sc.may_writepage = 1; |
bb3ab5968 vmscan: stop kswa... |
2115 |
|
45973d74f vmscan: check hig... |
2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 |
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 vmscan: stop kswa... |
2128 |
|
1da177e4c Linux-2.6.12-rc2 |
2129 |
} |
1da177e4c Linux-2.6.12-rc2 |
2130 2131 2132 2133 2134 2135 |
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 vmscan: stop kswa... |
2136 2137 2138 2139 2140 2141 |
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 Linux-2.6.12-rc2 |
2142 2143 2144 2145 2146 2147 2148 |
/* * 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 vmscan: bail out ... |
2149 |
if (sc.nr_reclaimed >= SWAP_CLUSTER_MAX) |
1da177e4c Linux-2.6.12-rc2 |
2150 2151 2152 |
break; } out: |
1da177e4c Linux-2.6.12-rc2 |
2153 2154 |
if (!all_zones_ok) { cond_resched(); |
8357376d3 [PATCH] swsusp: I... |
2155 2156 |
try_to_freeze(); |
73ce02e96 mm: stop kswapd's... |
2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 |
/* * 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 Linux-2.6.12-rc2 |
2173 2174 |
goto loop_again; } |
a79311c14 vmscan: bail out ... |
2175 |
return sc.nr_reclaimed; |
1da177e4c Linux-2.6.12-rc2 |
2176 2177 2178 2179 |
} /* * The background pageout daemon, started as a kernel thread |
4f98a2fee vmscan: split LRU... |
2180 |
* from the init process. |
1da177e4c Linux-2.6.12-rc2 |
2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 |
* * 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 cpumask: replace ... |
2200 |
const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id); |
1da177e4c Linux-2.6.12-rc2 |
2201 |
|
cf40bd16f lockdep: annotate... |
2202 |
lockdep_set_current_reclaim_state(GFP_KERNEL); |
174596a0b cpumask: convert mm/ |
2203 |
if (!cpumask_empty(cpumask)) |
c5f59f083 nodemask: use new... |
2204 |
set_cpus_allowed_ptr(tsk, cpumask); |
1da177e4c Linux-2.6.12-rc2 |
2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 |
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 [PATCH] Swap Migr... |
2219 |
tsk->flags |= PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD; |
831441862 Freezer: make ker... |
2220 |
set_freezable(); |
1da177e4c Linux-2.6.12-rc2 |
2221 2222 2223 2224 |
order = 0; for ( ; ; ) { unsigned long new_order; |
8fe23e057 mm: clear node in... |
2225 |
int ret; |
3e1d1d28d [PATCH] Cleanup p... |
2226 |
|
1da177e4c Linux-2.6.12-rc2 |
2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 |
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 vmscan: have kswa... |
2237 2238 2239 2240 |
if (!freezing(current) && !kthread_should_stop()) { long remaining = 0; /* Try to sleep for a short interval */ |
bb3ab5968 vmscan: stop kswa... |
2241 |
if (!sleeping_prematurely(pgdat, order, remaining)) { |
f50de2d38 vmscan: have kswa... |
2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 |
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 */ |
33906bc5c vmscan: tracing: ... |
2252 2253 |
if (!sleeping_prematurely(pgdat, order, remaining)) { trace_mm_vmscan_kswapd_sleep(pgdat->node_id); |
f50de2d38 vmscan: have kswa... |
2254 |
schedule(); |
33906bc5c vmscan: tracing: ... |
2255 |
} else { |
f50de2d38 vmscan: have kswa... |
2256 |
if (remaining) |
bb3ab5968 vmscan: stop kswa... |
2257 |
count_vm_event(KSWAPD_LOW_WMARK_HIT_QUICKLY); |
f50de2d38 vmscan: have kswa... |
2258 |
else |
bb3ab5968 vmscan: stop kswa... |
2259 |
count_vm_event(KSWAPD_HIGH_WMARK_HIT_QUICKLY); |
f50de2d38 vmscan: have kswa... |
2260 2261 |
} } |
b1296cc48 freezer: fix racy... |
2262 |
|
1da177e4c Linux-2.6.12-rc2 |
2263 2264 2265 |
order = pgdat->kswapd_max_order; } finish_wait(&pgdat->kswapd_wait, &wait); |
8fe23e057 mm: clear node in... |
2266 2267 2268 2269 2270 2271 2272 2273 |
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 */ |
33906bc5c vmscan: tracing: ... |
2274 2275 |
if (!ret) { trace_mm_vmscan_kswapd_wake(pgdat->node_id, order); |
b1296cc48 freezer: fix racy... |
2276 |
balance_pgdat(pgdat, order); |
33906bc5c vmscan: tracing: ... |
2277 |
} |
1da177e4c Linux-2.6.12-rc2 |
2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 |
} 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 [PATCH] mm: add p... |
2288 |
if (!populated_zone(zone)) |
1da177e4c Linux-2.6.12-rc2 |
2289 2290 2291 |
return; pgdat = zone->zone_pgdat; |
418589663 page allocator: u... |
2292 |
if (zone_watermark_ok(zone, order, low_wmark_pages(zone), 0, 0)) |
1da177e4c Linux-2.6.12-rc2 |
2293 2294 2295 |
return; if (pgdat->kswapd_max_order < order) pgdat->kswapd_max_order = order; |
33906bc5c vmscan: tracing: ... |
2296 |
trace_mm_vmscan_wakeup_kswapd(pgdat->node_id, zone_idx(zone), order); |
02a0e53d8 [PATCH] cpuset: r... |
2297 |
if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) |
1da177e4c Linux-2.6.12-rc2 |
2298 |
return; |
8d0986e28 [PATCH] vm: kswap... |
2299 |
if (!waitqueue_active(&pgdat->kswapd_wait)) |
1da177e4c Linux-2.6.12-rc2 |
2300 |
return; |
8d0986e28 [PATCH] vm: kswap... |
2301 |
wake_up_interruptible(&pgdat->kswapd_wait); |
1da177e4c Linux-2.6.12-rc2 |
2302 |
} |
adea02a1b mm: count only re... |
2303 2304 2305 2306 2307 2308 2309 2310 |
/* * 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 vmscan: split LRU... |
2311 |
{ |
adea02a1b mm: count only re... |
2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 |
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 vmscan: split LRU... |
2336 |
} |
c6f37f121 PM/Suspend: Do no... |
2337 |
#ifdef CONFIG_HIBERNATION |
1da177e4c Linux-2.6.12-rc2 |
2338 |
/* |
7b51755c3 vmscan: kill hibe... |
2339 |
* Try to free `nr_to_reclaim' of memory, system-wide, and return the number of |
d6277db4a [PATCH] swsusp: r... |
2340 2341 2342 2343 2344 |
* 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 Linux-2.6.12-rc2 |
2345 |
*/ |
7b51755c3 vmscan: kill hibe... |
2346 |
unsigned long shrink_all_memory(unsigned long nr_to_reclaim) |
1da177e4c Linux-2.6.12-rc2 |
2347 |
{ |
d6277db4a [PATCH] swsusp: r... |
2348 |
struct reclaim_state reclaim_state; |
d6277db4a [PATCH] swsusp: r... |
2349 |
struct scan_control sc = { |
7b51755c3 vmscan: kill hibe... |
2350 2351 2352 |
.gfp_mask = GFP_HIGHUSER_MOVABLE, .may_swap = 1, .may_unmap = 1, |
d6277db4a [PATCH] swsusp: r... |
2353 |
.may_writepage = 1, |
7b51755c3 vmscan: kill hibe... |
2354 2355 2356 2357 |
.nr_to_reclaim = nr_to_reclaim, .hibernation_mode = 1, .swappiness = vm_swappiness, .order = 0, |
1da177e4c Linux-2.6.12-rc2 |
2358 |
}; |
7b51755c3 vmscan: kill hibe... |
2359 2360 2361 |
struct zonelist * zonelist = node_zonelist(numa_node_id(), sc.gfp_mask); struct task_struct *p = current; unsigned long nr_reclaimed; |
1da177e4c Linux-2.6.12-rc2 |
2362 |
|
7b51755c3 vmscan: kill hibe... |
2363 2364 2365 2366 |
p->flags |= PF_MEMALLOC; lockdep_set_current_reclaim_state(sc.gfp_mask); reclaim_state.reclaimed_slab = 0; p->reclaim_state = &reclaim_state; |
d6277db4a [PATCH] swsusp: r... |
2367 |
|
7b51755c3 vmscan: kill hibe... |
2368 |
nr_reclaimed = do_try_to_free_pages(zonelist, &sc); |
d979677c4 mm: shrink_all_me... |
2369 |
|
7b51755c3 vmscan: kill hibe... |
2370 2371 2372 |
p->reclaim_state = NULL; lockdep_clear_current_reclaim_state(); p->flags &= ~PF_MEMALLOC; |
d6277db4a [PATCH] swsusp: r... |
2373 |
|
7b51755c3 vmscan: kill hibe... |
2374 |
return nr_reclaimed; |
1da177e4c Linux-2.6.12-rc2 |
2375 |
} |
c6f37f121 PM/Suspend: Do no... |
2376 |
#endif /* CONFIG_HIBERNATION */ |
1da177e4c Linux-2.6.12-rc2 |
2377 |
|
1da177e4c Linux-2.6.12-rc2 |
2378 2379 2380 2381 |
/* 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 [PATCH] cpu hotpl... |
2382 |
static int __devinit cpu_callback(struct notifier_block *nfb, |
69e05944a [PATCH] vmscan: u... |
2383 |
unsigned long action, void *hcpu) |
1da177e4c Linux-2.6.12-rc2 |
2384 |
{ |
58c0a4a78 Fix panic of cpu ... |
2385 |
int nid; |
1da177e4c Linux-2.6.12-rc2 |
2386 |
|
8bb784428 Add suspend-relat... |
2387 |
if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN) { |
58c0a4a78 Fix panic of cpu ... |
2388 |
for_each_node_state(nid, N_HIGH_MEMORY) { |
c5f59f083 nodemask: use new... |
2389 |
pg_data_t *pgdat = NODE_DATA(nid); |
a70f73028 cpumask: replace ... |
2390 2391 2392 |
const struct cpumask *mask; mask = cpumask_of_node(pgdat->node_id); |
c5f59f083 nodemask: use new... |
2393 |
|
3e5979453 cpumask: remove a... |
2394 |
if (cpumask_any_and(cpu_online_mask, mask) < nr_cpu_ids) |
1da177e4c Linux-2.6.12-rc2 |
2395 |
/* One of our CPUs online: restore mask */ |
c5f59f083 nodemask: use new... |
2396 |
set_cpus_allowed_ptr(pgdat->kswapd, mask); |
1da177e4c Linux-2.6.12-rc2 |
2397 2398 2399 2400 |
} } return NOTIFY_OK; } |
1da177e4c Linux-2.6.12-rc2 |
2401 |
|
3218ae14b [PATCH] pgdat all... |
2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 |
/* * 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 mm: clear node in... |
2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 |
/* * 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 Linux-2.6.12-rc2 |
2434 2435 |
static int __init kswapd_init(void) { |
3218ae14b [PATCH] pgdat all... |
2436 |
int nid; |
69e05944a [PATCH] vmscan: u... |
2437 |
|
1da177e4c Linux-2.6.12-rc2 |
2438 |
swap_setup(); |
9422ffba4 Memoryless nodes:... |
2439 |
for_each_node_state(nid, N_HIGH_MEMORY) |
3218ae14b [PATCH] pgdat all... |
2440 |
kswapd_run(nid); |
1da177e4c Linux-2.6.12-rc2 |
2441 2442 2443 2444 2445 |
hotcpu_notifier(cpu_callback, 0); return 0; } module_init(kswapd_init) |
9eeff2395 [PATCH] Zone recl... |
2446 2447 2448 2449 2450 2451 2452 |
#ifdef CONFIG_NUMA /* * Zone reclaim mode * * If non-zero call zone_reclaim when the number of free pages falls below * the watermarks. |
9eeff2395 [PATCH] Zone recl... |
2453 2454 |
*/ int zone_reclaim_mode __read_mostly; |
1b2ffb789 [PATCH] Zone recl... |
2455 |
#define RECLAIM_OFF 0 |
7d03431cf swapfile/vmscan: ... |
2456 |
#define RECLAIM_ZONE (1<<0) /* Run shrink_inactive_list on the zone */ |
1b2ffb789 [PATCH] Zone recl... |
2457 2458 |
#define RECLAIM_WRITE (1<<1) /* Writeout pages during reclaim */ #define RECLAIM_SWAP (1<<2) /* Swap pages out during reclaim */ |
9eeff2395 [PATCH] Zone recl... |
2459 |
/* |
a92f71263 [PATCH] zone_recl... |
2460 2461 2462 2463 2464 |
* 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 [PATCH] Zone recl... |
2465 |
/* |
9614634fe [PATCH] ZVC/zone_... |
2466 2467 2468 2469 2470 2471 |
* Percentage of pages in a zone that must be unmapped for zone_reclaim to * occur. */ int sysctl_min_unmapped_ratio = 1; /* |
0ff38490c [PATCH] zone_recl... |
2472 2473 2474 2475 |
* 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 vmscan: properly ... |
2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 |
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 [PATCH] zone_recl... |
2517 |
/* |
9eeff2395 [PATCH] Zone recl... |
2518 2519 |
* Try to free up some pages from this zone through reclaim. */ |
179e96395 [PATCH] vmscan: s... |
2520 |
static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order) |
9eeff2395 [PATCH] Zone recl... |
2521 |
{ |
7fb2d46d3 [PATCH] zone_recl... |
2522 |
/* Minimum pages needed in order to stay on node */ |
69e05944a [PATCH] vmscan: u... |
2523 |
const unsigned long nr_pages = 1 << order; |
9eeff2395 [PATCH] Zone recl... |
2524 2525 |
struct task_struct *p = current; struct reclaim_state reclaim_state; |
8695949a1 [PATCH] Thin out ... |
2526 |
int priority; |
179e96395 [PATCH] vmscan: s... |
2527 2528 |
struct scan_control sc = { .may_writepage = !!(zone_reclaim_mode & RECLAIM_WRITE), |
a6dc60f89 vmscan: rename sc... |
2529 |
.may_unmap = !!(zone_reclaim_mode & RECLAIM_SWAP), |
2e2e42598 vmscan,memcg: rei... |
2530 |
.may_swap = 1, |
22fba3354 vmscan: separate ... |
2531 2532 |
.nr_to_reclaim = max_t(unsigned long, nr_pages, SWAP_CLUSTER_MAX), |
179e96395 [PATCH] vmscan: s... |
2533 |
.gfp_mask = gfp_mask, |
d6277db4a [PATCH] swsusp: r... |
2534 |
.swappiness = vm_swappiness, |
bd2f6199c vmscan: respect h... |
2535 |
.order = order, |
179e96395 [PATCH] vmscan: s... |
2536 |
}; |
157480489 vmscan: avoid sub... |
2537 |
unsigned long nr_slab_pages0, nr_slab_pages1; |
9eeff2395 [PATCH] Zone recl... |
2538 |
|
9eeff2395 [PATCH] Zone recl... |
2539 |
cond_resched(); |
d4f7796e9 [PATCH] vmscan: f... |
2540 2541 2542 2543 2544 2545 |
/* * 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 mm, lockdep: anno... |
2546 |
lockdep_set_current_reclaim_state(gfp_mask); |
9eeff2395 [PATCH] Zone recl... |
2547 2548 |
reclaim_state.reclaimed_slab = 0; p->reclaim_state = &reclaim_state; |
c84db23c6 [PATCH] zone_recl... |
2549 |
|
90afa5de6 vmscan: properly ... |
2550 |
if (zone_pagecache_reclaimable(zone) > zone->min_unmapped_pages) { |
0ff38490c [PATCH] zone_recl... |
2551 2552 2553 2554 2555 2556 |
/* * Free memory by calling shrink zone with increasing * priorities until we have enough memory freed. */ priority = ZONE_RECLAIM_PRIORITY; do { |
a79311c14 vmscan: bail out ... |
2557 |
shrink_zone(priority, zone, &sc); |
0ff38490c [PATCH] zone_recl... |
2558 |
priority--; |
a79311c14 vmscan: bail out ... |
2559 |
} while (priority >= 0 && sc.nr_reclaimed < nr_pages); |
0ff38490c [PATCH] zone_recl... |
2560 |
} |
c84db23c6 [PATCH] zone_recl... |
2561 |
|
157480489 vmscan: avoid sub... |
2562 2563 |
nr_slab_pages0 = zone_page_state(zone, NR_SLAB_RECLAIMABLE); if (nr_slab_pages0 > zone->min_slab_pages) { |
2a16e3f4b [PATCH] Reclaim s... |
2564 |
/* |
7fb2d46d3 [PATCH] zone_recl... |
2565 |
* shrink_slab() does not currently allow us to determine how |
0ff38490c [PATCH] zone_recl... |
2566 2567 2568 2569 |
* 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 [PATCH] Reclaim s... |
2570 |
* |
0ff38490c [PATCH] zone_recl... |
2571 2572 |
* Note that shrink_slab will free memory on all zones and may * take a long time. |
2a16e3f4b [PATCH] Reclaim s... |
2573 |
*/ |
4dc4b3d97 vmscan: shrink_sl... |
2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 |
for (;;) { unsigned long lru_pages = zone_reclaimable_pages(zone); /* No reclaimable slab or very low memory pressure */ if (!shrink_slab(sc.nr_scanned, gfp_mask, lru_pages)) break; /* Freed enough memory */ nr_slab_pages1 = zone_page_state(zone, NR_SLAB_RECLAIMABLE); if (nr_slab_pages1 + nr_pages <= nr_slab_pages0) break; } |
83e33a471 [PATCH] zone recl... |
2587 2588 2589 2590 2591 |
/* * Update nr_reclaimed by the number of slab pages we * reclaimed from this zone. */ |
157480489 vmscan: avoid sub... |
2592 2593 2594 |
nr_slab_pages1 = zone_page_state(zone, NR_SLAB_RECLAIMABLE); if (nr_slab_pages1 < nr_slab_pages0) sc.nr_reclaimed += nr_slab_pages0 - nr_slab_pages1; |
2a16e3f4b [PATCH] Reclaim s... |
2595 |
} |
9eeff2395 [PATCH] Zone recl... |
2596 |
p->reclaim_state = NULL; |
d4f7796e9 [PATCH] vmscan: f... |
2597 |
current->flags &= ~(PF_MEMALLOC | PF_SWAPWRITE); |
76ca542d8 mm, lockdep: anno... |
2598 |
lockdep_clear_current_reclaim_state(); |
a79311c14 vmscan: bail out ... |
2599 |
return sc.nr_reclaimed >= nr_pages; |
9eeff2395 [PATCH] Zone recl... |
2600 |
} |
179e96395 [PATCH] vmscan: s... |
2601 2602 2603 |
int zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order) { |
179e96395 [PATCH] vmscan: s... |
2604 |
int node_id; |
d773ed6b8 mm: test and set ... |
2605 |
int ret; |
179e96395 [PATCH] vmscan: s... |
2606 2607 |
/* |
0ff38490c [PATCH] zone_recl... |
2608 2609 |
* Zone reclaim reclaims unmapped file backed pages and * slab pages if we are over the defined limits. |
34aa1330f [PATCH] zoned vm ... |
2610 |
* |
9614634fe [PATCH] ZVC/zone_... |
2611 2612 2613 2614 2615 |
* 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 [PATCH] vmscan: s... |
2616 |
*/ |
90afa5de6 vmscan: properly ... |
2617 2618 |
if (zone_pagecache_reclaimable(zone) <= zone->min_unmapped_pages && zone_page_state(zone, NR_SLAB_RECLAIMABLE) <= zone->min_slab_pages) |
fa5e084e4 vmscan: do not un... |
2619 |
return ZONE_RECLAIM_FULL; |
179e96395 [PATCH] vmscan: s... |
2620 |
|
93e4a89a8 mm: restore zone-... |
2621 |
if (zone->all_unreclaimable) |
fa5e084e4 vmscan: do not un... |
2622 |
return ZONE_RECLAIM_FULL; |
d773ed6b8 mm: test and set ... |
2623 |
|
179e96395 [PATCH] vmscan: s... |
2624 |
/* |
d773ed6b8 mm: test and set ... |
2625 |
* Do not scan if the allocation should not be delayed. |
179e96395 [PATCH] vmscan: s... |
2626 |
*/ |
d773ed6b8 mm: test and set ... |
2627 |
if (!(gfp_mask & __GFP_WAIT) || (current->flags & PF_MEMALLOC)) |
fa5e084e4 vmscan: do not un... |
2628 |
return ZONE_RECLAIM_NOSCAN; |
179e96395 [PATCH] vmscan: s... |
2629 2630 2631 2632 2633 2634 2635 |
/* * 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 [PATCH] NUMA: Add... |
2636 |
node_id = zone_to_nid(zone); |
37c0708db Memoryless nodes:... |
2637 |
if (node_state(node_id, N_CPU) && node_id != numa_node_id()) |
fa5e084e4 vmscan: do not un... |
2638 |
return ZONE_RECLAIM_NOSCAN; |
d773ed6b8 mm: test and set ... |
2639 2640 |
if (zone_test_and_set_flag(zone, ZONE_RECLAIM_LOCKED)) |
fa5e084e4 vmscan: do not un... |
2641 |
return ZONE_RECLAIM_NOSCAN; |
d773ed6b8 mm: test and set ... |
2642 2643 |
ret = __zone_reclaim(zone, gfp_mask, order); zone_clear_flag(zone, ZONE_RECLAIM_LOCKED); |
24cf72518 vmscan: count the... |
2644 2645 |
if (!ret) count_vm_event(PGSCAN_ZONE_RECLAIM_FAILED); |
d773ed6b8 mm: test and set ... |
2646 |
return ret; |
179e96395 [PATCH] vmscan: s... |
2647 |
} |
9eeff2395 [PATCH] Zone recl... |
2648 |
#endif |
894bc3104 Unevictable LRU I... |
2649 |
|
894bc3104 Unevictable LRU I... |
2650 2651 2652 2653 2654 2655 |
/* * 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 mlock: mlocked pa... |
2656 2657 |
* lists vs unevictable list. The vma argument is !NULL when called from the * fault path to determine how to instantate a new page. |
894bc3104 Unevictable LRU I... |
2658 2659 |
* * Reasons page might not be evictable: |
ba9ddf493 Ramfs and Ram Dis... |
2660 |
* (1) page's mapping marked unevictable |
b291f0003 mlock: mlocked pa... |
2661 |
* (2) page is part of an mlocked VMA |
ba9ddf493 Ramfs and Ram Dis... |
2662 |
* |
894bc3104 Unevictable LRU I... |
2663 2664 2665 |
*/ int page_evictable(struct page *page, struct vm_area_struct *vma) { |
ba9ddf493 Ramfs and Ram Dis... |
2666 2667 |
if (mapping_unevictable(page_mapping(page))) return 0; |
b291f0003 mlock: mlocked pa... |
2668 2669 |
if (PageMlocked(page) || (vma && is_mlocked_vma(vma, page))) return 0; |
894bc3104 Unevictable LRU I... |
2670 2671 2672 |
return 1; } |
89e004ea5 SHM_LOCKED pages ... |
2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 |
/** * 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 mm: introduce pag... |
2692 |
enum lru_list l = page_lru_base_type(page); |
af936a160 vmscan: unevictab... |
2693 |
|
89e004ea5 SHM_LOCKED pages ... |
2694 2695 |
__dec_zone_state(zone, NR_UNEVICTABLE); list_move(&page->lru, &zone->lru[l].list); |
08e552c69 memcg: synchroniz... |
2696 |
mem_cgroup_move_lists(page, LRU_UNEVICTABLE, l); |
89e004ea5 SHM_LOCKED pages ... |
2697 2698 2699 2700 2701 2702 2703 2704 |
__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 memcg: synchroniz... |
2705 |
mem_cgroup_rotate_lru_list(page, LRU_UNEVICTABLE); |
89e004ea5 SHM_LOCKED pages ... |
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 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 |
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 vmscan: unevictab... |
2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 |
/** * 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 mm: make scan_zon... |
2777 |
static void scan_zone_unevictable_pages(struct zone *zone) |
af936a160 vmscan: unevictab... |
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 2818 |
{ 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 mm: make scan_all... |
2819 |
static void scan_all_zones_unevictable_pages(void) |
af936a160 vmscan: unevictab... |
2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 |
{ 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 sysctl: remove "s... |
2835 |
void __user *buffer, |
af936a160 vmscan: unevictab... |
2836 2837 |
size_t *length, loff_t *ppos) { |
8d65af789 sysctl: remove "s... |
2838 |
proc_doulongvec_minmax(table, write, buffer, length, ppos); |
af936a160 vmscan: unevictab... |
2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 |
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); } |