vmscan.c 57.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*
 *  linux/mm/vmscan.c
 *
 *  Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
 *
 *  Swap reorganised 29.12.95, Stephen Tweedie.
 *  kswapd added: 7.1.96  sct
 *  Removed kswapd_ctl limits, and swap out as many pages as needed
 *  to bring the system back to freepages.high: 2.4.97, Rik van Riel.
 *  Zone aware kswapd started 02/00, Kanoj Sarcar (kanoj@sgi.com).
 *  Multiqueue VM started 5.8.00, Rik van Riel.
 */

#include <linux/mm.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/kernel_stat.h>
#include <linux/swap.h>
#include <linux/pagemap.h>
#include <linux/init.h>
#include <linux/highmem.h>
22
#include <linux/vmstat.h>
L
Linus Torvalds 已提交
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>
37
#include <linux/delay.h>
38
#include <linux/kthread.h>
39
#include <linux/freezer.h>
40
#include <linux/memcontrol.h>
L
Linus Torvalds 已提交
41 42 43 44 45 46

#include <asm/tlbflush.h>
#include <asm/div64.h>

#include <linux/swapops.h>

47 48
#include "internal.h"

L
Linus Torvalds 已提交
49 50 51 52 53
struct scan_control {
	/* Incremented by the number of inactive pages that were scanned */
	unsigned long nr_scanned;

	/* This context's GFP mask */
A
Al Viro 已提交
54
	gfp_t gfp_mask;
L
Linus Torvalds 已提交
55 56 57

	int may_writepage;

58 59 60
	/* Can pages be swapped as part of reclaim? */
	int may_swap;

L
Linus Torvalds 已提交
61 62 63 64 65
	/* This context's SWAP_CLUSTER_MAX. If freeing memory for
	 * suspend, we effectively ignore SWAP_CLUSTER_MAX.
	 * In this context, it doesn't matter that we scan the
	 * whole list at once. */
	int swap_cluster_max;
66 67

	int swappiness;
68 69

	int all_unreclaimable;
A
Andy Whitcroft 已提交
70 71

	int order;
72 73 74 75 76 77 78 79 80

	/* Which cgroup do we reclaim from */
	struct mem_cgroup *mem_cgroup;

	/* Pluggable isolate pages callback */
	unsigned long (*isolate_pages)(unsigned long nr, struct list_head *dst,
			unsigned long *scanned, int order, int mode,
			struct zone *z, struct mem_cgroup *mem_cont,
			int active);
L
Linus Torvalds 已提交
81 82 83 84 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
};

#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;
117
long vm_total_pages;	/* The total number of pages which the VM controls */
L
Linus Torvalds 已提交
118 119 120 121

static LIST_HEAD(shrinker_list);
static DECLARE_RWSEM(shrinker_rwsem);

122
#ifdef CONFIG_CGROUP_MEM_RES_CTLR
123 124 125 126 127
#define scan_global_lru(sc)	(!(sc)->mem_cgroup)
#else
#define scan_global_lru(sc)	(1)
#endif

L
Linus Torvalds 已提交
128 129 130
/*
 * Add a shrinker callback to be called from the vm
 */
131
void register_shrinker(struct shrinker *shrinker)
L
Linus Torvalds 已提交
132
{
133 134 135 136
	shrinker->nr = 0;
	down_write(&shrinker_rwsem);
	list_add_tail(&shrinker->list, &shrinker_list);
	up_write(&shrinker_rwsem);
L
Linus Torvalds 已提交
137
}
138
EXPORT_SYMBOL(register_shrinker);
L
Linus Torvalds 已提交
139 140 141 142

/*
 * Remove one
 */
143
void unregister_shrinker(struct shrinker *shrinker)
L
Linus Torvalds 已提交
144 145 146 147 148
{
	down_write(&shrinker_rwsem);
	list_del(&shrinker->list);
	up_write(&shrinker_rwsem);
}
149
EXPORT_SYMBOL(unregister_shrinker);
L
Linus Torvalds 已提交
150 151 152 153 154 155 156 157 158 159

#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.
 *
S
Simon Arlott 已提交
160
 * If the vm encountered mapped pages on the LRU it increase the pressure on
L
Linus Torvalds 已提交
161 162 163 164 165 166 167
 * 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.
168 169
 *
 * Returns the number of slab objects which we shrunk.
L
Linus Torvalds 已提交
170
 */
171 172
unsigned long shrink_slab(unsigned long scanned, gfp_t gfp_mask,
			unsigned long lru_pages)
L
Linus Torvalds 已提交
173 174
{
	struct shrinker *shrinker;
175
	unsigned long ret = 0;
L
Linus Torvalds 已提交
176 177 178 179 180

	if (scanned == 0)
		scanned = SWAP_CLUSTER_MAX;

	if (!down_read_trylock(&shrinker_rwsem))
181
		return 1;	/* Assume we'll be able to shrink next time */
L
Linus Torvalds 已提交
182 183 184 185

	list_for_each_entry(shrinker, &shrinker_list, list) {
		unsigned long long delta;
		unsigned long total_scan;
186
		unsigned long max_pass = (*shrinker->shrink)(0, gfp_mask);
L
Linus Torvalds 已提交
187 188

		delta = (4 * scanned) / shrinker->seeks;
189
		delta *= max_pass;
L
Linus Torvalds 已提交
190 191
		do_div(delta, lru_pages + 1);
		shrinker->nr += delta;
192 193 194 195 196 197 198 199 200 201 202 203 204
		if (shrinker->nr < 0) {
			printk(KERN_ERR "%s: nr=%ld\n",
					__FUNCTION__, shrinker->nr);
			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;
L
Linus Torvalds 已提交
205 206 207 208 209 210 211

		total_scan = shrinker->nr;
		shrinker->nr = 0;

		while (total_scan >= SHRINK_BATCH) {
			long this_scan = SHRINK_BATCH;
			int shrink_ret;
212
			int nr_before;
L
Linus Torvalds 已提交
213

214 215
			nr_before = (*shrinker->shrink)(0, gfp_mask);
			shrink_ret = (*shrinker->shrink)(this_scan, gfp_mask);
L
Linus Torvalds 已提交
216 217
			if (shrink_ret == -1)
				break;
218 219
			if (shrink_ret < nr_before)
				ret += nr_before - shrink_ret;
220
			count_vm_events(SLABS_SCANNED, this_scan);
L
Linus Torvalds 已提交
221 222 223 224 225 226 227 228
			total_scan -= this_scan;

			cond_resched();
		}

		shrinker->nr += total_scan;
	}
	up_read(&shrinker_rwsem);
229
	return ret;
L
Linus Torvalds 已提交
230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259
}

/* Called without lock on whether page is mapped, so answer is unstable */
static inline int page_mapping_inuse(struct page *page)
{
	struct address_space *mapping;

	/* Page is in somebody's page tables. */
	if (page_mapped(page))
		return 1;

	/* Be more reluctant to reclaim swapcache than pagecache */
	if (PageSwapCache(page))
		return 1;

	mapping = page_mapping(page);
	if (!mapping)
		return 0;

	/* File is mmap'd by somebody? */
	return mapping_mapped(mapping);
}

static inline int is_page_cache_freeable(struct page *page)
{
	return page_count(page) - !!PagePrivate(page) == 2;
}

static int may_write_to_queue(struct backing_dev_info *bdi)
{
260
	if (current->flags & PF_SWAPWRITE)
L
Linus Torvalds 已提交
261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284
		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)
{
	lock_page(page);
285 286
	if (page_mapping(page) == mapping)
		mapping_set_error(mapping, error);
L
Linus Torvalds 已提交
287 288 289
	unlock_page(page);
}

290 291 292 293 294 295
/* Request for sync pageout. */
enum pageout_io {
	PAGEOUT_IO_ASYNC,
	PAGEOUT_IO_SYNC,
};

296 297 298 299 300 301 302 303 304 305 306 307
/* 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;

L
Linus Torvalds 已提交
308
/*
A
Andrew Morton 已提交
309 310
 * pageout is called by shrink_page_list() for each dirty page.
 * Calls ->writepage().
L
Linus Torvalds 已提交
311
 */
312 313
static pageout_t pageout(struct page *page, struct address_space *mapping,
						enum pageout_io sync_writeback)
L
Linus Torvalds 已提交
314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338
{
	/*
	 * 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.
	 *
	 * If this process is currently in generic_file_write() against
	 * 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.
	 * See swapfile.c:page_queue_congested().
	 */
	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.
		 */
339
		if (PagePrivate(page)) {
L
Linus Torvalds 已提交
340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357
			if (try_to_free_buffers(page)) {
				ClearPageDirty(page);
				printk("%s: orphaned page\n", __FUNCTION__);
				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,
358 359
			.range_start = 0,
			.range_end = LLONG_MAX,
L
Linus Torvalds 已提交
360 361 362 363 364 365 366 367
			.nonblocking = 1,
			.for_reclaim = 1,
		};

		SetPageReclaim(page);
		res = mapping->a_ops->writepage(page, &wbc);
		if (res < 0)
			handle_write_error(mapping, page, res);
368
		if (res == AOP_WRITEPAGE_ACTIVATE) {
L
Linus Torvalds 已提交
369 370 371
			ClearPageReclaim(page);
			return PAGE_ACTIVATE;
		}
372 373 374 375 376 377 378 379 380

		/*
		 * 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);

L
Linus Torvalds 已提交
381 382 383 384
		if (!PageWriteback(page)) {
			/* synchronous write or broken a_ops? */
			ClearPageReclaim(page);
		}
385
		inc_zone_page_state(page, NR_VMSCAN_WRITE);
L
Linus Torvalds 已提交
386 387 388 389 390 391
		return PAGE_SUCCESS;
	}

	return PAGE_CLEAN;
}

392 393 394 395 396 397
/*
 * 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.
 */
C
Christoph Lameter 已提交
398
int remove_mapping(struct address_space *mapping, struct page *page)
399
{
400 401
	BUG_ON(!PageLocked(page));
	BUG_ON(mapping != page_mapping(page));
402 403 404

	write_lock_irq(&mapping->tree_lock);
	/*
N
Nick Piggin 已提交
405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427
	 * 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.
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
	 */
	if (unlikely(page_count(page) != 2))
		goto cannot_free;
	smp_rmb();
	if (unlikely(PageDirty(page)))
		goto cannot_free;

	if (PageSwapCache(page)) {
		swp_entry_t swap = { .val = page_private(page) };
		__delete_from_swap_cache(page);
		write_unlock_irq(&mapping->tree_lock);
		swap_free(swap);
		__put_page(page);	/* The pagecache ref */
		return 1;
	}

	__remove_from_page_cache(page);
	write_unlock_irq(&mapping->tree_lock);
	__put_page(page);
	return 1;

cannot_free:
	write_unlock_irq(&mapping->tree_lock);
	return 0;
}

L
Linus Torvalds 已提交
454
/*
A
Andrew Morton 已提交
455
 * shrink_page_list() returns the number of reclaimed pages
L
Linus Torvalds 已提交
456
 */
A
Andrew Morton 已提交
457
static unsigned long shrink_page_list(struct list_head *page_list,
458 459
					struct scan_control *sc,
					enum pageout_io sync_writeback)
L
Linus Torvalds 已提交
460 461 462 463
{
	LIST_HEAD(ret_pages);
	struct pagevec freed_pvec;
	int pgactivate = 0;
464
	unsigned long nr_reclaimed = 0;
L
Linus Torvalds 已提交
465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482

	cond_resched();

	pagevec_init(&freed_pvec, 1);
	while (!list_empty(page_list)) {
		struct address_space *mapping;
		struct page *page;
		int may_enter_fs;
		int referenced;

		cond_resched();

		page = lru_to_page(page_list);
		list_del(&page->lru);

		if (TestSetPageLocked(page))
			goto keep;

N
Nick Piggin 已提交
483
		VM_BUG_ON(PageActive(page));
L
Linus Torvalds 已提交
484 485

		sc->nr_scanned++;
486 487 488 489

		if (!sc->may_swap && page_mapped(page))
			goto keep_locked;

L
Linus Torvalds 已提交
490 491 492 493
		/* Double the slab pressure for mapped and swapcache pages */
		if (page_mapped(page) || PageSwapCache(page))
			sc->nr_scanned++;

494 495 496 497 498 499 500 501 502 503 504 505 506 507
		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);
508
			else
509 510
				goto keep_locked;
		}
L
Linus Torvalds 已提交
511

512
		referenced = page_referenced(page, 1, sc->mem_cgroup);
L
Linus Torvalds 已提交
513
		/* In active use or really unfreeable?  Activate it. */
A
Andy Whitcroft 已提交
514 515
		if (sc->order <= PAGE_ALLOC_COSTLY_ORDER &&
					referenced && page_mapping_inuse(page))
L
Linus Torvalds 已提交
516 517 518 519 520 521 522
			goto activate_locked;

#ifdef CONFIG_SWAP
		/*
		 * Anonymous process memory has backing store?
		 * Try to allocate it some swap space here.
		 */
523
		if (PageAnon(page) && !PageSwapCache(page))
524
			if (!add_to_swap(page, GFP_ATOMIC))
L
Linus Torvalds 已提交
525 526 527 528 529 530 531 532 533 534
				goto activate_locked;
#endif /* CONFIG_SWAP */

		mapping = page_mapping(page);

		/*
		 * The page is mapped into the page tables of one or more
		 * processes. Try to unmap it here.
		 */
		if (page_mapped(page) && mapping) {
535
			switch (try_to_unmap(page, 0)) {
L
Linus Torvalds 已提交
536 537 538 539 540 541 542 543 544 545
			case SWAP_FAIL:
				goto activate_locked;
			case SWAP_AGAIN:
				goto keep_locked;
			case SWAP_SUCCESS:
				; /* try to free the page below */
			}
		}

		if (PageDirty(page)) {
A
Andy Whitcroft 已提交
546
			if (sc->order <= PAGE_ALLOC_COSTLY_ORDER && referenced)
L
Linus Torvalds 已提交
547
				goto keep_locked;
548
			if (!may_enter_fs)
L
Linus Torvalds 已提交
549
				goto keep_locked;
550
			if (!sc->may_writepage)
L
Linus Torvalds 已提交
551 552 553
				goto keep_locked;

			/* Page is dirty, try to write it out here */
554
			switch (pageout(page, mapping, sync_writeback)) {
L
Linus Torvalds 已提交
555 556 557 558 559
			case PAGE_KEEP:
				goto keep_locked;
			case PAGE_ACTIVATE:
				goto activate_locked;
			case PAGE_SUCCESS:
560
				if (PageWriteback(page) || PageDirty(page))
L
Linus Torvalds 已提交
561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603
					goto keep;
				/*
				 * A synchronous write - probably a ramdisk.  Go
				 * ahead and try to reclaim the page.
				 */
				if (TestSetPageLocked(page))
					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
		 * will do this, as well as the blockdev mapping. 
		 * 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.
		 */
		if (PagePrivate(page)) {
			if (!try_to_release_page(page, sc->gfp_mask))
				goto activate_locked;
			if (!mapping && page_count(page) == 1)
				goto free_it;
		}

604
		if (!mapping || !remove_mapping(mapping, page))
605
			goto keep_locked;
L
Linus Torvalds 已提交
606 607 608

free_it:
		unlock_page(page);
609
		nr_reclaimed++;
L
Linus Torvalds 已提交
610 611 612 613 614 615 616 617 618 619 620
		if (!pagevec_add(&freed_pvec, page))
			__pagevec_release_nonlru(&freed_pvec);
		continue;

activate_locked:
		SetPageActive(page);
		pgactivate++;
keep_locked:
		unlock_page(page);
keep:
		list_add(&page->lru, &ret_pages);
N
Nick Piggin 已提交
621
		VM_BUG_ON(PageLRU(page));
L
Linus Torvalds 已提交
622 623 624 625
	}
	list_splice(&ret_pages, page_list);
	if (pagevec_count(&freed_pvec))
		__pagevec_release_nonlru(&freed_pvec);
626
	count_vm_events(PGACTIVATE, pgactivate);
627
	return nr_reclaimed;
L
Linus Torvalds 已提交
628 629
}

A
Andy Whitcroft 已提交
630 631 632 633 634 635 636 637 638 639 640 641 642 643 644
/* LRU Isolation modes. */
#define ISOLATE_INACTIVE 0	/* Isolate inactive pages. */
#define ISOLATE_ACTIVE 1	/* Isolate active pages. */
#define ISOLATE_BOTH 2		/* Isolate both active and inactive pages. */

/*
 * Attempt to remove the specified page from its LRU.  Only take this page
 * if it is of the appropriate PageActive status.  Pages which are being
 * freed elsewhere are also ignored.
 *
 * page:	page to consider
 * mode:	one of the LRU isolation modes defined above
 *
 * returns 0 on success, -ve errno on failure.
 */
645
int __isolate_lru_page(struct page *page, int mode)
A
Andy Whitcroft 已提交
646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674
{
	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;

	ret = -EBUSY;
	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;
}

L
Linus Torvalds 已提交
675 676 677 678 679 680 681 682 683 684 685 686 687 688
/*
 * 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.
A
Andy Whitcroft 已提交
689 690
 * @order:	The caller's attempted allocation order
 * @mode:	One of the LRU isolation modes
L
Linus Torvalds 已提交
691 692 693
 *
 * returns how many pages were moved onto *@dst.
 */
694 695
static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
		struct list_head *src, struct list_head *dst,
A
Andy Whitcroft 已提交
696
		unsigned long *scanned, int order, int mode)
L
Linus Torvalds 已提交
697
{
698
	unsigned long nr_taken = 0;
699
	unsigned long scan;
L
Linus Torvalds 已提交
700

701
	for (scan = 0; scan < nr_to_scan && !list_empty(src); scan++) {
A
Andy Whitcroft 已提交
702 703 704 705 706 707
		struct page *page;
		unsigned long pfn;
		unsigned long end_pfn;
		unsigned long page_pfn;
		int zone_id;

L
Linus Torvalds 已提交
708 709 710
		page = lru_to_page(src);
		prefetchw_prev_lru_page(page, src, flags);

N
Nick Piggin 已提交
711
		VM_BUG_ON(!PageLRU(page));
N
Nick Piggin 已提交
712

A
Andy Whitcroft 已提交
713 714 715
		switch (__isolate_lru_page(page, mode)) {
		case 0:
			list_move(&page->lru, dst);
716
			nr_taken++;
A
Andy Whitcroft 已提交
717 718 719 720 721 722
			break;

		case -EBUSY:
			/* else it is being freed elsewhere */
			list_move(&page->lru, src);
			continue;
723

A
Andy Whitcroft 已提交
724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772
		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);
			/* Check that we have not crossed a zone boundary. */
			if (unlikely(page_zone_id(cursor_page) != zone_id))
				continue;
			switch (__isolate_lru_page(cursor_page, mode)) {
			case 0:
				list_move(&cursor_page->lru, dst);
				nr_taken++;
				scan++;
				break;

			case -EBUSY:
				/* else it is being freed elsewhere */
				list_move(&cursor_page->lru, src);
			default:
				break;
			}
		}
L
Linus Torvalds 已提交
773 774 775 776 777 778
	}

	*scanned = scan;
	return nr_taken;
}

779 780 781 782 783 784 785 786 787 788 789 790 791 792 793
static unsigned long isolate_pages_global(unsigned long nr,
					struct list_head *dst,
					unsigned long *scanned, int order,
					int mode, struct zone *z,
					struct mem_cgroup *mem_cont,
					int active)
{
	if (active)
		return isolate_lru_pages(nr, &z->active_list, dst,
						scanned, order, mode);
	else
		return isolate_lru_pages(nr, &z->inactive_list, dst,
						scanned, order, mode);
}

A
Andy Whitcroft 已提交
794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811
/*
 * clear_active_flags() is a helper for shrink_active_list(), clearing
 * any active bits from the pages in the list.
 */
static unsigned long clear_active_flags(struct list_head *page_list)
{
	int nr_active = 0;
	struct page *page;

	list_for_each_entry(page, page_list, lru)
		if (PageActive(page)) {
			ClearPageActive(page);
			nr_active++;
		}

	return nr_active;
}

L
Linus Torvalds 已提交
812
/*
A
Andrew Morton 已提交
813 814
 * shrink_inactive_list() is a helper for shrink_zone().  It returns the number
 * of reclaimed pages
L
Linus Torvalds 已提交
815
 */
A
Andrew Morton 已提交
816 817
static unsigned long shrink_inactive_list(unsigned long max_scan,
				struct zone *zone, struct scan_control *sc)
L
Linus Torvalds 已提交
818 819 820
{
	LIST_HEAD(page_list);
	struct pagevec pvec;
821
	unsigned long nr_scanned = 0;
822
	unsigned long nr_reclaimed = 0;
L
Linus Torvalds 已提交
823 824 825 826 827

	pagevec_init(&pvec, 1);

	lru_add_drain();
	spin_lock_irq(&zone->lru_lock);
828
	do {
L
Linus Torvalds 已提交
829
		struct page *page;
830 831 832
		unsigned long nr_taken;
		unsigned long nr_scan;
		unsigned long nr_freed;
A
Andy Whitcroft 已提交
833
		unsigned long nr_active;
L
Linus Torvalds 已提交
834

835
		nr_taken = sc->isolate_pages(sc->swap_cluster_max,
A
Andy Whitcroft 已提交
836 837
			     &page_list, &nr_scan, sc->order,
			     (sc->order > PAGE_ALLOC_COSTLY_ORDER)?
838 839
					     ISOLATE_BOTH : ISOLATE_INACTIVE,
				zone, sc->mem_cgroup, 0);
A
Andy Whitcroft 已提交
840
		nr_active = clear_active_flags(&page_list);
841
		__count_vm_events(PGDEACTIVATE, nr_active);
A
Andy Whitcroft 已提交
842 843 844 845

		__mod_zone_page_state(zone, NR_ACTIVE, -nr_active);
		__mod_zone_page_state(zone, NR_INACTIVE,
						-(nr_taken - nr_active));
846 847
		if (scan_global_lru(sc))
			zone->pages_scanned += nr_scan;
L
Linus Torvalds 已提交
848 849
		spin_unlock_irq(&zone->lru_lock);

850
		nr_scanned += nr_scan;
851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873
		nr_freed = shrink_page_list(&page_list, sc, PAGEOUT_IO_ASYNC);

		/*
		 * If we are direct reclaiming for contiguous pages and we do
		 * not reclaim everything in the list, try again and wait
		 * for IO to complete. This will stall high-order allocations
		 * but that should be acceptable to the caller
		 */
		if (nr_freed < nr_taken && !current_is_kswapd() &&
					sc->order > PAGE_ALLOC_COSTLY_ORDER) {
			congestion_wait(WRITE, HZ/10);

			/*
			 * The attempt at page out may have made some
			 * of the pages active, mark them inactive again.
			 */
			nr_active = clear_active_flags(&page_list);
			count_vm_events(PGDEACTIVATE, nr_active);

			nr_freed += shrink_page_list(&page_list, sc,
							PAGEOUT_IO_SYNC);
		}

874
		nr_reclaimed += nr_freed;
N
Nick Piggin 已提交
875 876
		local_irq_disable();
		if (current_is_kswapd()) {
877 878
			__count_zone_vm_events(PGSCAN_KSWAPD, zone, nr_scan);
			__count_vm_events(KSWAPD_STEAL, nr_freed);
879
		} else if (scan_global_lru(sc))
880
			__count_zone_vm_events(PGSCAN_DIRECT, zone, nr_scan);
881

S
Shantanu Goel 已提交
882
		__count_zone_vm_events(PGSTEAL, zone, nr_freed);
N
Nick Piggin 已提交
883

884 885 886
		if (nr_taken == 0)
			goto done;

N
Nick Piggin 已提交
887
		spin_lock(&zone->lru_lock);
L
Linus Torvalds 已提交
888 889 890 891 892
		/*
		 * Put back any unfreeable pages.
		 */
		while (!list_empty(&page_list)) {
			page = lru_to_page(&page_list);
N
Nick Piggin 已提交
893
			VM_BUG_ON(PageLRU(page));
N
Nick Piggin 已提交
894
			SetPageLRU(page);
L
Linus Torvalds 已提交
895 896 897 898 899 900 901 902 903 904 905
			list_del(&page->lru);
			if (PageActive(page))
				add_page_to_active_list(zone, page);
			else
				add_page_to_inactive_list(zone, page);
			if (!pagevec_add(&pvec, page)) {
				spin_unlock_irq(&zone->lru_lock);
				__pagevec_release(&pvec);
				spin_lock_irq(&zone->lru_lock);
			}
		}
906
  	} while (nr_scanned < max_scan);
907
	spin_unlock(&zone->lru_lock);
L
Linus Torvalds 已提交
908
done:
909
	local_irq_enable();
L
Linus Torvalds 已提交
910
	pagevec_release(&pvec);
911
	return nr_reclaimed;
L
Linus Torvalds 已提交
912 913
}

914 915 916 917 918 919 920 921 922 923 924 925 926 927
/*
 * We are about to scan this zone at a certain priority level.  If that priority
 * level is smaller (ie: more urgent) than the previous priority, then note
 * that priority level within the zone.  This is done so that when the next
 * process comes in to scan this zone, it will immediately start out at this
 * priority level rather than having to build up its own scanning priority.
 * Here, this priority affects only the reclaim-mapped threshold.
 */
static inline void note_zone_scanning_priority(struct zone *zone, int priority)
{
	if (priority < zone->prev_priority)
		zone->prev_priority = priority;
}

N
Nick Piggin 已提交
928 929
static inline int zone_is_near_oom(struct zone *zone)
{
930 931
	return zone->pages_scanned >= (zone_page_state(zone, NR_ACTIVE)
				+ zone_page_state(zone, NR_INACTIVE))*3;
N
Nick Piggin 已提交
932 933
}

934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
/*
 * Determine we should try to reclaim mapped pages.
 * This is called only when sc->mem_cgroup is NULL.
 */
static int calc_reclaim_mapped(struct scan_control *sc, struct zone *zone,
				int priority)
{
	long mapped_ratio;
	long distress;
	long swap_tendency;
	long imbalance;
	int reclaim_mapped = 0;
	int prev_priority;

	if (scan_global_lru(sc) && zone_is_near_oom(zone))
		return 1;
	/*
	 * `distress' is a measure of how much trouble we're having
	 * reclaiming pages.  0 -> no problems.  100 -> great trouble.
	 */
	if (scan_global_lru(sc))
		prev_priority = zone->prev_priority;
	else
		prev_priority = mem_cgroup_get_reclaim_priority(sc->mem_cgroup);

	distress = 100 >> min(prev_priority, priority);

	/*
	 * The point of this algorithm is to decide when to start
	 * reclaiming mapped memory instead of just pagecache.  Work out
	 * how much memory
	 * is mapped.
	 */
	if (scan_global_lru(sc))
		mapped_ratio = ((global_page_state(NR_FILE_MAPPED) +
				global_page_state(NR_ANON_PAGES)) * 100) /
					vm_total_pages;
	else
		mapped_ratio = mem_cgroup_calc_mapped_ratio(sc->mem_cgroup);

	/*
	 * Now decide how much we really want to unmap some pages.  The
	 * mapped ratio is downgraded - just because there's a lot of
	 * mapped memory doesn't necessarily mean that page reclaim
	 * isn't succeeding.
	 *
	 * The distress ratio is important - we don't want to start
	 * going oom.
	 *
	 * A 100% value of vm_swappiness overrides this algorithm
	 * altogether.
	 */
	swap_tendency = mapped_ratio / 2 + distress + sc->swappiness;

	/*
	 * If there's huge imbalance between active and inactive
	 * (think active 100 times larger than inactive) we should
	 * become more permissive, or the system will take too much
	 * cpu before it start swapping during memory pressure.
	 * Distress is about avoiding early-oom, this is about
	 * making swappiness graceful despite setting it to low
	 * values.
	 *
	 * Avoid div by zero with nr_inactive+1, and max resulting
	 * value is vm_total_pages.
	 */
	if (scan_global_lru(sc)) {
		imbalance  = zone_page_state(zone, NR_ACTIVE);
		imbalance /= zone_page_state(zone, NR_INACTIVE) + 1;
	} else
		imbalance = mem_cgroup_reclaim_imbalance(sc->mem_cgroup);

	/*
	 * Reduce the effect of imbalance if swappiness is low,
	 * this means for a swappiness very low, the imbalance
	 * must be much higher than 100 for this logic to make
	 * the difference.
	 *
	 * Max temporary value is vm_total_pages*100.
	 */
	imbalance *= (vm_swappiness + 1);
	imbalance /= 100;

	/*
	 * If not much of the ram is mapped, makes the imbalance
	 * less relevant, it's high priority we refill the inactive
	 * list with mapped pages only in presence of high ratio of
	 * mapped pages.
	 *
	 * Max temporary value is vm_total_pages*100.
	 */
	imbalance *= mapped_ratio;
	imbalance /= 100;

	/* apply imbalance feedback to swap_tendency */
	swap_tendency += imbalance;

	/*
	 * Now use this metric to decide whether to start moving mapped
	 * memory onto the inactive list.
	 */
	if (swap_tendency >= 100)
		reclaim_mapped = 1;

	return reclaim_mapped;
}

L
Linus Torvalds 已提交
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057
/*
 * 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.
 */
1058 1059


A
Andrew Morton 已提交
1060
static void shrink_active_list(unsigned long nr_pages, struct zone *zone,
1061
				struct scan_control *sc, int priority)
L
Linus Torvalds 已提交
1062
{
1063
	unsigned long pgmoved;
L
Linus Torvalds 已提交
1064
	int pgdeactivate = 0;
1065
	unsigned long pgscanned;
L
Linus Torvalds 已提交
1066 1067 1068 1069 1070 1071
	LIST_HEAD(l_hold);	/* The pages which were snipped off */
	LIST_HEAD(l_inactive);	/* Pages to go onto the inactive_list */
	LIST_HEAD(l_active);	/* Pages to go onto the active_list */
	struct page *page;
	struct pagevec pvec;
	int reclaim_mapped = 0;
1072

1073 1074
	if (sc->may_swap)
		reclaim_mapped = calc_reclaim_mapped(sc, zone, priority);
L
Linus Torvalds 已提交
1075 1076 1077

	lru_add_drain();
	spin_lock_irq(&zone->lru_lock);
1078 1079 1080
	pgmoved = sc->isolate_pages(nr_pages, &l_hold, &pgscanned, sc->order,
					ISOLATE_ACTIVE, zone,
					sc->mem_cgroup, 1);
1081 1082 1083 1084 1085 1086 1087
	/*
	 * zone->pages_scanned is used for detect zone's oom
	 * mem_cgroup remembers nr_scan by itself.
	 */
	if (scan_global_lru(sc))
		zone->pages_scanned += pgscanned;

1088
	__mod_zone_page_state(zone, NR_ACTIVE, -pgmoved);
L
Linus Torvalds 已提交
1089 1090 1091 1092 1093 1094 1095 1096 1097
	spin_unlock_irq(&zone->lru_lock);

	while (!list_empty(&l_hold)) {
		cond_resched();
		page = lru_to_page(&l_hold);
		list_del(&page->lru);
		if (page_mapped(page)) {
			if (!reclaim_mapped ||
			    (total_swap_pages == 0 && PageAnon(page)) ||
1098
			    page_referenced(page, 0, sc->mem_cgroup)) {
L
Linus Torvalds 已提交
1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111
				list_add(&page->lru, &l_active);
				continue;
			}
		}
		list_add(&page->lru, &l_inactive);
	}

	pagevec_init(&pvec, 1);
	pgmoved = 0;
	spin_lock_irq(&zone->lru_lock);
	while (!list_empty(&l_inactive)) {
		page = lru_to_page(&l_inactive);
		prefetchw_prev_lru_page(page, &l_inactive, flags);
N
Nick Piggin 已提交
1112
		VM_BUG_ON(PageLRU(page));
N
Nick Piggin 已提交
1113
		SetPageLRU(page);
N
Nick Piggin 已提交
1114
		VM_BUG_ON(!PageActive(page));
N
Nick Piggin 已提交
1115 1116
		ClearPageActive(page);

L
Linus Torvalds 已提交
1117
		list_move(&page->lru, &zone->inactive_list);
1118
		mem_cgroup_move_lists(page, false);
L
Linus Torvalds 已提交
1119 1120
		pgmoved++;
		if (!pagevec_add(&pvec, page)) {
1121
			__mod_zone_page_state(zone, NR_INACTIVE, pgmoved);
L
Linus Torvalds 已提交
1122 1123 1124 1125 1126 1127 1128 1129 1130
			spin_unlock_irq(&zone->lru_lock);
			pgdeactivate += pgmoved;
			pgmoved = 0;
			if (buffer_heads_over_limit)
				pagevec_strip(&pvec);
			__pagevec_release(&pvec);
			spin_lock_irq(&zone->lru_lock);
		}
	}
1131
	__mod_zone_page_state(zone, NR_INACTIVE, pgmoved);
L
Linus Torvalds 已提交
1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142
	pgdeactivate += pgmoved;
	if (buffer_heads_over_limit) {
		spin_unlock_irq(&zone->lru_lock);
		pagevec_strip(&pvec);
		spin_lock_irq(&zone->lru_lock);
	}

	pgmoved = 0;
	while (!list_empty(&l_active)) {
		page = lru_to_page(&l_active);
		prefetchw_prev_lru_page(page, &l_active, flags);
N
Nick Piggin 已提交
1143
		VM_BUG_ON(PageLRU(page));
N
Nick Piggin 已提交
1144
		SetPageLRU(page);
N
Nick Piggin 已提交
1145
		VM_BUG_ON(!PageActive(page));
1146

L
Linus Torvalds 已提交
1147
		list_move(&page->lru, &zone->active_list);
1148
		mem_cgroup_move_lists(page, true);
L
Linus Torvalds 已提交
1149 1150
		pgmoved++;
		if (!pagevec_add(&pvec, page)) {
1151
			__mod_zone_page_state(zone, NR_ACTIVE, pgmoved);
L
Linus Torvalds 已提交
1152 1153 1154 1155 1156 1157
			pgmoved = 0;
			spin_unlock_irq(&zone->lru_lock);
			__pagevec_release(&pvec);
			spin_lock_irq(&zone->lru_lock);
		}
	}
1158
	__mod_zone_page_state(zone, NR_ACTIVE, pgmoved);
N
Nick Piggin 已提交
1159

1160 1161 1162
	__count_zone_vm_events(PGREFILL, zone, pgscanned);
	__count_vm_events(PGDEACTIVATE, pgdeactivate);
	spin_unlock_irq(&zone->lru_lock);
L
Linus Torvalds 已提交
1163

N
Nick Piggin 已提交
1164
	pagevec_release(&pvec);
L
Linus Torvalds 已提交
1165 1166 1167 1168 1169
}

/*
 * This is a basic per-zone page freer.  Used by both kswapd and direct reclaim.
 */
1170 1171
static unsigned long shrink_zone(int priority, struct zone *zone,
				struct scan_control *sc)
L
Linus Torvalds 已提交
1172 1173 1174
{
	unsigned long nr_active;
	unsigned long nr_inactive;
1175
	unsigned long nr_to_scan;
1176
	unsigned long nr_reclaimed = 0;
L
Linus Torvalds 已提交
1177

1178 1179 1180 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
	if (scan_global_lru(sc)) {
		/*
		 * Add one to nr_to_scan just to make sure that the kernel
		 * will slowly sift through the active list.
		 */
		zone->nr_scan_active +=
			(zone_page_state(zone, NR_ACTIVE) >> priority) + 1;
		nr_active = zone->nr_scan_active;
		zone->nr_scan_inactive +=
			(zone_page_state(zone, NR_INACTIVE) >> priority) + 1;
		nr_inactive = zone->nr_scan_inactive;
		if (nr_inactive >= sc->swap_cluster_max)
			zone->nr_scan_inactive = 0;
		else
			nr_inactive = 0;

		if (nr_active >= sc->swap_cluster_max)
			zone->nr_scan_active = 0;
		else
			nr_active = 0;
	} else {
		/*
		 * This reclaim occurs not because zone memory shortage but
		 * because memory controller hits its limit.
		 * Then, don't modify zone reclaim related data.
		 */
		nr_active = mem_cgroup_calc_reclaim_active(sc->mem_cgroup,
					zone, priority);

		nr_inactive = mem_cgroup_calc_reclaim_inactive(sc->mem_cgroup,
					zone, priority);
	}
L
Linus Torvalds 已提交
1210 1211 1212 1213


	while (nr_active || nr_inactive) {
		if (nr_active) {
1214
			nr_to_scan = min(nr_active,
L
Linus Torvalds 已提交
1215
					(unsigned long)sc->swap_cluster_max);
1216
			nr_active -= nr_to_scan;
1217
			shrink_active_list(nr_to_scan, zone, sc, priority);
L
Linus Torvalds 已提交
1218 1219 1220
		}

		if (nr_inactive) {
1221
			nr_to_scan = min(nr_inactive,
L
Linus Torvalds 已提交
1222
					(unsigned long)sc->swap_cluster_max);
1223
			nr_inactive -= nr_to_scan;
A
Andrew Morton 已提交
1224 1225
			nr_reclaimed += shrink_inactive_list(nr_to_scan, zone,
								sc);
L
Linus Torvalds 已提交
1226 1227 1228
		}
	}

1229
	throttle_vm_writeout(sc->gfp_mask);
1230
	return nr_reclaimed;
L
Linus Torvalds 已提交
1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248
}

/*
 * 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.
 *
 * We reclaim from a zone even if that zone is over pages_high.  Because:
 * a) The caller may be trying to free *extra* pages to satisfy a higher-order
 *    allocation or
 * b) The zones may be over pages_high but they must go *over* pages_high to
 *    satisfy the `incremental min' zone defense algorithm.
 *
 * Returns the number of reclaimed pages.
 *
 * If a zone is deemed to be full of pinned pages then just give it a light
 * scan then give up on it.
 */
1249
static unsigned long shrink_zones(int priority, struct zonelist *zonelist,
1250
					struct scan_control *sc)
L
Linus Torvalds 已提交
1251
{
1252
	enum zone_type high_zoneidx = gfp_zone(sc->gfp_mask);
1253
	unsigned long nr_reclaimed = 0;
1254
	struct zoneref *z;
1255
	struct zone *zone;
1256

1257
	sc->all_unreclaimable = 1;
1258
	for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
1259
		if (!populated_zone(zone))
L
Linus Torvalds 已提交
1260
			continue;
1261 1262 1263 1264 1265 1266 1267 1268
		/*
		 * Take care memory controller reclaiming has small influence
		 * to global LRU.
		 */
		if (scan_global_lru(sc)) {
			if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL))
				continue;
			note_zone_scanning_priority(zone, priority);
L
Linus Torvalds 已提交
1269

1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282
			if (zone_is_all_unreclaimable(zone) &&
						priority != DEF_PRIORITY)
				continue;	/* Let kswapd poll it */
			sc->all_unreclaimable = 0;
		} else {
			/*
			 * Ignore cpuset limitation here. We just want to reduce
			 * # of used pages by us regardless of memory shortage.
			 */
			sc->all_unreclaimable = 0;
			mem_cgroup_note_reclaim_priority(sc->mem_cgroup,
							priority);
		}
1283

1284
		nr_reclaimed += shrink_zone(priority, zone, sc);
L
Linus Torvalds 已提交
1285
	}
1286

1287
	return nr_reclaimed;
L
Linus Torvalds 已提交
1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301
}
 
/*
 * 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
 * caller can't do much about.  We kick pdflush 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.
1302 1303 1304
 *
 * returns:	0, if no pages reclaimed
 * 		else, the number of pages reclaimed
L
Linus Torvalds 已提交
1305
 */
1306
static unsigned long do_try_to_free_pages(struct zonelist *zonelist,
1307
					struct scan_control *sc)
L
Linus Torvalds 已提交
1308 1309 1310
{
	int priority;
	int ret = 0;
1311
	unsigned long total_scanned = 0;
1312
	unsigned long nr_reclaimed = 0;
L
Linus Torvalds 已提交
1313 1314
	struct reclaim_state *reclaim_state = current->reclaim_state;
	unsigned long lru_pages = 0;
1315
	struct zoneref *z;
1316
	struct zone *zone;
1317
	enum zone_type high_zoneidx = gfp_zone(sc->gfp_mask);
L
Linus Torvalds 已提交
1318

1319 1320 1321 1322 1323 1324
	if (scan_global_lru(sc))
		count_vm_event(ALLOCSTALL);
	/*
	 * mem_cgroup will not do shrink_slab.
	 */
	if (scan_global_lru(sc)) {
1325
		for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
L
Linus Torvalds 已提交
1326

1327 1328
			if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL))
				continue;
L
Linus Torvalds 已提交
1329

1330 1331 1332
			lru_pages += zone_page_state(zone, NR_ACTIVE)
					+ zone_page_state(zone, NR_INACTIVE);
		}
L
Linus Torvalds 已提交
1333 1334 1335
	}

	for (priority = DEF_PRIORITY; priority >= 0; priority--) {
1336
		sc->nr_scanned = 0;
1337 1338
		if (!priority)
			disable_swap_token();
1339
		nr_reclaimed += shrink_zones(priority, zonelist, sc);
1340 1341 1342 1343
		/*
		 * Don't shrink slabs when reclaiming memory from
		 * over limit cgroups
		 */
1344
		if (scan_global_lru(sc)) {
1345
			shrink_slab(sc->nr_scanned, sc->gfp_mask, lru_pages);
1346 1347 1348 1349
			if (reclaim_state) {
				nr_reclaimed += reclaim_state->reclaimed_slab;
				reclaim_state->reclaimed_slab = 0;
			}
L
Linus Torvalds 已提交
1350
		}
1351 1352
		total_scanned += sc->nr_scanned;
		if (nr_reclaimed >= sc->swap_cluster_max) {
1353
			ret = nr_reclaimed;
L
Linus Torvalds 已提交
1354 1355 1356 1357 1358 1359 1360 1361 1362 1363
			goto out;
		}

		/*
		 * Try to write back as many pages as we just scanned.  This
		 * tends to cause slow streaming writers to write data to the
		 * disk smoothly, at the dirtying rate, which is nice.   But
		 * that's undesirable in laptop mode, where we *want* lumpy
		 * writeout.  So in laptop mode, write out the whole world.
		 */
1364 1365
		if (total_scanned > sc->swap_cluster_max +
					sc->swap_cluster_max / 2) {
1366
			wakeup_pdflush(laptop_mode ? 0 : total_scanned);
1367
			sc->may_writepage = 1;
L
Linus Torvalds 已提交
1368 1369 1370
		}

		/* Take a nap, wait for some writeback to complete */
1371
		if (sc->nr_scanned && priority < DEF_PRIORITY - 2)
1372
			congestion_wait(WRITE, HZ/10);
L
Linus Torvalds 已提交
1373
	}
1374
	/* top priority shrink_caches still had more to do? don't OOM, then */
1375
	if (!sc->all_unreclaimable && scan_global_lru(sc))
1376
		ret = nr_reclaimed;
L
Linus Torvalds 已提交
1377
out:
1378 1379 1380 1381 1382 1383 1384 1385 1386
	/*
	 * 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;
L
Linus Torvalds 已提交
1387

1388
	if (scan_global_lru(sc)) {
1389
		for_each_zone_zonelist(zone, z, zonelist, high_zoneidx) {
1390 1391 1392 1393 1394 1395 1396 1397

			if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL))
				continue;

			zone->prev_priority = priority;
		}
	} else
		mem_cgroup_record_reclaim_priority(sc->mem_cgroup, priority);
L
Linus Torvalds 已提交
1398 1399 1400 1401

	return ret;
}

1402 1403
unsigned long try_to_free_pages(struct zonelist *zonelist, int order,
								gfp_t gfp_mask)
1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415
{
	struct scan_control sc = {
		.gfp_mask = gfp_mask,
		.may_writepage = !laptop_mode,
		.swap_cluster_max = SWAP_CLUSTER_MAX,
		.may_swap = 1,
		.swappiness = vm_swappiness,
		.order = order,
		.mem_cgroup = NULL,
		.isolate_pages = isolate_pages_global,
	};

1416
	return do_try_to_free_pages(zonelist, &sc);
1417 1418
}

1419
#ifdef CONFIG_CGROUP_MEM_RES_CTLR
1420

1421 1422
unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
						gfp_t gfp_mask)
1423 1424 1425 1426 1427 1428 1429 1430 1431 1432
{
	struct scan_control sc = {
		.may_writepage = !laptop_mode,
		.may_swap = 1,
		.swap_cluster_max = SWAP_CLUSTER_MAX,
		.swappiness = vm_swappiness,
		.order = 0,
		.mem_cgroup = mem_cont,
		.isolate_pages = mem_cgroup_isolate_pages,
	};
1433
	struct zonelist *zonelist;
1434

1435 1436 1437 1438
	sc.gfp_mask = (gfp_mask & GFP_RECLAIM_MASK) |
			(GFP_HIGHUSER_MOVABLE & ~GFP_RECLAIM_MASK);
	zonelist = NODE_DATA(numa_node_id())->node_zonelists;
	return do_try_to_free_pages(zonelist, &sc);
1439 1440 1441
}
#endif

L
Linus Torvalds 已提交
1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462
/*
 * For kswapd, balance_pgdat() will work across all this node's zones until
 * they are all at pages_high.
 *
 * 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
 * zones which have free_pages > pages_high, but once a zone is found to have
 * free_pages <= pages_high, 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.
 */
1463
static unsigned long balance_pgdat(pg_data_t *pgdat, int order)
L
Linus Torvalds 已提交
1464 1465 1466 1467
{
	int all_zones_ok;
	int priority;
	int i;
1468
	unsigned long total_scanned;
1469
	unsigned long nr_reclaimed;
L
Linus Torvalds 已提交
1470
	struct reclaim_state *reclaim_state = current->reclaim_state;
1471 1472 1473
	struct scan_control sc = {
		.gfp_mask = GFP_KERNEL,
		.may_swap = 1,
1474 1475
		.swap_cluster_max = SWAP_CLUSTER_MAX,
		.swappiness = vm_swappiness,
A
Andy Whitcroft 已提交
1476
		.order = order,
1477 1478
		.mem_cgroup = NULL,
		.isolate_pages = isolate_pages_global,
1479
	};
1480 1481 1482 1483 1484
	/*
	 * temp_priority is used to remember the scanning priority at which
	 * this zone was successfully refilled to free_pages == pages_high.
	 */
	int temp_priority[MAX_NR_ZONES];
L
Linus Torvalds 已提交
1485 1486 1487

loop_again:
	total_scanned = 0;
1488
	nr_reclaimed = 0;
C
Christoph Lameter 已提交
1489
	sc.may_writepage = !laptop_mode;
1490
	count_vm_event(PAGEOUTRUN);
L
Linus Torvalds 已提交
1491

1492 1493
	for (i = 0; i < pgdat->nr_zones; i++)
		temp_priority[i] = DEF_PRIORITY;
L
Linus Torvalds 已提交
1494 1495 1496 1497 1498

	for (priority = DEF_PRIORITY; priority >= 0; priority--) {
		int end_zone = 0;	/* Inclusive.  0 = ZONE_DMA */
		unsigned long lru_pages = 0;

1499 1500 1501 1502
		/* The swap token gets in the way of swapout... */
		if (!priority)
			disable_swap_token();

L
Linus Torvalds 已提交
1503 1504
		all_zones_ok = 1;

1505 1506 1507 1508 1509 1510
		/*
		 * 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;
L
Linus Torvalds 已提交
1511

1512 1513
			if (!populated_zone(zone))
				continue;
L
Linus Torvalds 已提交
1514

1515 1516
			if (zone_is_all_unreclaimable(zone) &&
			    priority != DEF_PRIORITY)
1517
				continue;
L
Linus Torvalds 已提交
1518

1519 1520 1521
			if (!zone_watermark_ok(zone, order, zone->pages_high,
					       0, 0)) {
				end_zone = i;
A
Andrew Morton 已提交
1522
				break;
L
Linus Torvalds 已提交
1523 1524
			}
		}
A
Andrew Morton 已提交
1525 1526 1527
		if (i < 0)
			goto out;

L
Linus Torvalds 已提交
1528 1529 1530
		for (i = 0; i <= end_zone; i++) {
			struct zone *zone = pgdat->node_zones + i;

1531 1532
			lru_pages += zone_page_state(zone, NR_ACTIVE)
					+ zone_page_state(zone, NR_INACTIVE);
L
Linus Torvalds 已提交
1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
		}

		/*
		 * 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;
1546
			int nr_slab;
L
Linus Torvalds 已提交
1547

1548
			if (!populated_zone(zone))
L
Linus Torvalds 已提交
1549 1550
				continue;

1551 1552
			if (zone_is_all_unreclaimable(zone) &&
					priority != DEF_PRIORITY)
L
Linus Torvalds 已提交
1553 1554
				continue;

1555 1556 1557
			if (!zone_watermark_ok(zone, order, zone->pages_high,
					       end_zone, 0))
				all_zones_ok = 0;
1558
			temp_priority[i] = priority;
L
Linus Torvalds 已提交
1559
			sc.nr_scanned = 0;
1560
			note_zone_scanning_priority(zone, priority);
1561 1562 1563 1564 1565 1566 1567
			/*
			 * We put equal pressure on every zone, unless one
			 * zone has way too many pages free already.
			 */
			if (!zone_watermark_ok(zone, order, 8*zone->pages_high,
						end_zone, 0))
				nr_reclaimed += shrink_zone(priority, zone, &sc);
L
Linus Torvalds 已提交
1568
			reclaim_state->reclaimed_slab = 0;
1569 1570
			nr_slab = shrink_slab(sc.nr_scanned, GFP_KERNEL,
						lru_pages);
1571
			nr_reclaimed += reclaim_state->reclaimed_slab;
L
Linus Torvalds 已提交
1572
			total_scanned += sc.nr_scanned;
1573
			if (zone_is_all_unreclaimable(zone))
L
Linus Torvalds 已提交
1574
				continue;
1575
			if (nr_slab == 0 && zone->pages_scanned >=
1576 1577
				(zone_page_state(zone, NR_ACTIVE)
				+ zone_page_state(zone, NR_INACTIVE)) * 6)
1578 1579
					zone_set_flag(zone,
						      ZONE_ALL_UNRECLAIMABLE);
L
Linus Torvalds 已提交
1580 1581 1582 1583 1584 1585
			/*
			 * 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 &&
1586
			    total_scanned > nr_reclaimed + nr_reclaimed / 2)
L
Linus Torvalds 已提交
1587 1588 1589 1590 1591 1592 1593 1594
				sc.may_writepage = 1;
		}
		if (all_zones_ok)
			break;		/* kswapd: all done */
		/*
		 * OK, kswapd is getting into trouble.  Take a nap, then take
		 * another pass across the zones.
		 */
1595
		if (total_scanned && priority < DEF_PRIORITY - 2)
1596
			congestion_wait(WRITE, HZ/10);
L
Linus Torvalds 已提交
1597 1598 1599 1600 1601 1602 1603

		/*
		 * 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.
		 */
1604
		if (nr_reclaimed >= SWAP_CLUSTER_MAX)
L
Linus Torvalds 已提交
1605 1606 1607
			break;
	}
out:
1608 1609 1610 1611 1612
	/*
	 * Note within each zone the priority level at which this zone was
	 * brought into a happy state.  So that the next thread which scans this
	 * zone will start out at that priority level.
	 */
L
Linus Torvalds 已提交
1613 1614 1615
	for (i = 0; i < pgdat->nr_zones; i++) {
		struct zone *zone = pgdat->node_zones + i;

1616
		zone->prev_priority = temp_priority[i];
L
Linus Torvalds 已提交
1617 1618 1619
	}
	if (!all_zones_ok) {
		cond_resched();
1620 1621 1622

		try_to_freeze();

L
Linus Torvalds 已提交
1623 1624 1625
		goto loop_again;
	}

1626
	return nr_reclaimed;
L
Linus Torvalds 已提交
1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650
}

/*
 * The background pageout daemon, started as a kernel thread
 * from the init process. 
 *
 * 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,
	};
1651
	node_to_cpumask_ptr(cpumask, pgdat->node_id);
L
Linus Torvalds 已提交
1652

1653 1654
	if (!cpus_empty(*cpumask))
		set_cpus_allowed_ptr(tsk, cpumask);
L
Linus Torvalds 已提交
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668
	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).
	 */
1669
	tsk->flags |= PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD;
1670
	set_freezable();
L
Linus Torvalds 已提交
1671 1672 1673 1674

	order = 0;
	for ( ; ; ) {
		unsigned long new_order;
1675

L
Linus Torvalds 已提交
1676 1677 1678 1679 1680 1681 1682 1683 1684 1685
		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 {
1686 1687 1688
			if (!freezing(current))
				schedule();

L
Linus Torvalds 已提交
1689 1690 1691 1692
			order = pgdat->kswapd_max_order;
		}
		finish_wait(&pgdat->kswapd_wait, &wait);

1693 1694 1695 1696 1697 1698
		if (!try_to_freeze()) {
			/* We can speed up thawing tasks if we don't call
			 * balance_pgdat after returning from the refrigerator
			 */
			balance_pgdat(pgdat, order);
		}
L
Linus Torvalds 已提交
1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
	}
	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;

1710
	if (!populated_zone(zone))
L
Linus Torvalds 已提交
1711 1712 1713
		return;

	pgdat = zone->zone_pgdat;
R
Rohit Seth 已提交
1714
	if (zone_watermark_ok(zone, order, zone->pages_low, 0, 0))
L
Linus Torvalds 已提交
1715 1716 1717
		return;
	if (pgdat->kswapd_max_order < order)
		pgdat->kswapd_max_order = order;
1718
	if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL))
L
Linus Torvalds 已提交
1719
		return;
1720
	if (!waitqueue_active(&pgdat->kswapd_wait))
L
Linus Torvalds 已提交
1721
		return;
1722
	wake_up_interruptible(&pgdat->kswapd_wait);
L
Linus Torvalds 已提交
1723 1724 1725 1726
}

#ifdef CONFIG_PM
/*
1727 1728 1729 1730 1731 1732
 * Helper function for shrink_all_memory().  Tries to reclaim 'nr_pages' pages
 * from LRU lists system-wide, for given pass and priority, and returns the
 * number of reclaimed pages
 *
 * For pass > 3 we also try to shrink the LRU lists that contain a few pages
 */
1733 1734
static unsigned long shrink_all_zones(unsigned long nr_pages, int prio,
				      int pass, struct scan_control *sc)
1735 1736 1737 1738 1739 1740 1741 1742 1743
{
	struct zone *zone;
	unsigned long nr_to_scan, ret = 0;

	for_each_zone(zone) {

		if (!populated_zone(zone))
			continue;

1744
		if (zone_is_all_unreclaimable(zone) && prio != DEF_PRIORITY)
1745 1746 1747 1748
			continue;

		/* For pass = 0 we don't shrink the active list */
		if (pass > 0) {
1749 1750
			zone->nr_scan_active +=
				(zone_page_state(zone, NR_ACTIVE) >> prio) + 1;
1751 1752
			if (zone->nr_scan_active >= nr_pages || pass > 3) {
				zone->nr_scan_active = 0;
1753 1754
				nr_to_scan = min(nr_pages,
					zone_page_state(zone, NR_ACTIVE));
1755
				shrink_active_list(nr_to_scan, zone, sc, prio);
1756 1757 1758
			}
		}

1759 1760
		zone->nr_scan_inactive +=
			(zone_page_state(zone, NR_INACTIVE) >> prio) + 1;
1761 1762
		if (zone->nr_scan_inactive >= nr_pages || pass > 3) {
			zone->nr_scan_inactive = 0;
1763 1764
			nr_to_scan = min(nr_pages,
				zone_page_state(zone, NR_INACTIVE));
1765 1766 1767 1768 1769 1770 1771 1772 1773
			ret += shrink_inactive_list(nr_to_scan, zone, sc);
			if (ret >= nr_pages)
				return ret;
		}
	}

	return ret;
}

1774 1775
static unsigned long count_lru_pages(void)
{
1776
	return global_page_state(NR_ACTIVE) + global_page_state(NR_INACTIVE);
1777 1778
}

1779 1780 1781 1782 1783 1784 1785
/*
 * Try to free `nr_pages' of memory, system-wide, and return the number of
 * 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
L
Linus Torvalds 已提交
1786
 */
1787
unsigned long shrink_all_memory(unsigned long nr_pages)
L
Linus Torvalds 已提交
1788
{
1789
	unsigned long lru_pages, nr_slab;
1790
	unsigned long ret = 0;
1791 1792 1793 1794 1795 1796 1797 1798
	int pass;
	struct reclaim_state reclaim_state;
	struct scan_control sc = {
		.gfp_mask = GFP_KERNEL,
		.may_swap = 0,
		.swap_cluster_max = nr_pages,
		.may_writepage = 1,
		.swappiness = vm_swappiness,
1799
		.isolate_pages = isolate_pages_global,
L
Linus Torvalds 已提交
1800 1801 1802
	};

	current->reclaim_state = &reclaim_state;
1803

1804
	lru_pages = count_lru_pages();
1805
	nr_slab = global_page_state(NR_SLAB_RECLAIMABLE);
1806 1807 1808 1809 1810
	/* If slab caches are huge, it's better to hit them first */
	while (nr_slab >= lru_pages) {
		reclaim_state.reclaimed_slab = 0;
		shrink_slab(nr_pages, sc.gfp_mask, lru_pages);
		if (!reclaim_state.reclaimed_slab)
L
Linus Torvalds 已提交
1811
			break;
1812 1813 1814 1815 1816 1817

		ret += reclaim_state.reclaimed_slab;
		if (ret >= nr_pages)
			goto out;

		nr_slab -= reclaim_state.reclaimed_slab;
L
Linus Torvalds 已提交
1818
	}
1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845

	/*
	 * We try to shrink LRUs in 5 passes:
	 * 0 = Reclaim from inactive_list only
	 * 1 = Reclaim from active list but don't reclaim mapped
	 * 2 = 2nd pass of type 1
	 * 3 = Reclaim mapped (normal reclaim)
	 * 4 = 2nd pass of type 3
	 */
	for (pass = 0; pass < 5; pass++) {
		int prio;

		/* Force reclaiming mapped pages in the passes #3 and #4 */
		if (pass > 2) {
			sc.may_swap = 1;
			sc.swappiness = 100;
		}

		for (prio = DEF_PRIORITY; prio >= 0; prio--) {
			unsigned long nr_to_scan = nr_pages - ret;

			sc.nr_scanned = 0;
			ret += shrink_all_zones(nr_to_scan, prio, pass, &sc);
			if (ret >= nr_pages)
				goto out;

			reclaim_state.reclaimed_slab = 0;
1846 1847
			shrink_slab(sc.nr_scanned, sc.gfp_mask,
					count_lru_pages());
1848 1849 1850 1851 1852
			ret += reclaim_state.reclaimed_slab;
			if (ret >= nr_pages)
				goto out;

			if (sc.nr_scanned && prio < DEF_PRIORITY - 2)
1853
				congestion_wait(WRITE, HZ / 10);
1854
		}
1855
	}
1856 1857 1858 1859 1860

	/*
	 * If ret = 0, we could not shrink LRUs, but there may be something
	 * in slab caches
	 */
1861
	if (!ret) {
1862 1863
		do {
			reclaim_state.reclaimed_slab = 0;
1864
			shrink_slab(nr_pages, sc.gfp_mask, count_lru_pages());
1865 1866
			ret += reclaim_state.reclaimed_slab;
		} while (ret < nr_pages && reclaim_state.reclaimed_slab > 0);
1867
	}
1868 1869

out:
L
Linus Torvalds 已提交
1870
	current->reclaim_state = NULL;
1871

L
Linus Torvalds 已提交
1872 1873 1874 1875 1876 1877 1878 1879
	return ret;
}
#endif

/* 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. */
1880
static int __devinit cpu_callback(struct notifier_block *nfb,
1881
				  unsigned long action, void *hcpu)
L
Linus Torvalds 已提交
1882
{
1883
	int nid;
L
Linus Torvalds 已提交
1884

1885
	if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN) {
1886
		for_each_node_state(nid, N_HIGH_MEMORY) {
1887 1888 1889 1890
			pg_data_t *pgdat = NODE_DATA(nid);
			node_to_cpumask_ptr(mask, pgdat->node_id);

			if (any_online_cpu(*mask) < nr_cpu_ids)
L
Linus Torvalds 已提交
1891
				/* One of our CPUs online: restore mask */
1892
				set_cpus_allowed_ptr(pgdat->kswapd, mask);
L
Linus Torvalds 已提交
1893 1894 1895 1896 1897
		}
	}
	return NOTIFY_OK;
}

1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919
/*
 * 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\n",nid);
		ret = -1;
	}
	return ret;
}

L
Linus Torvalds 已提交
1920 1921
static int __init kswapd_init(void)
{
1922
	int nid;
1923

L
Linus Torvalds 已提交
1924
	swap_setup();
1925
	for_each_node_state(nid, N_HIGH_MEMORY)
1926
 		kswapd_run(nid);
L
Linus Torvalds 已提交
1927 1928 1929 1930 1931
	hotcpu_notifier(cpu_callback, 0);
	return 0;
}

module_init(kswapd_init)
1932 1933 1934 1935 1936 1937 1938 1939 1940 1941

#ifdef CONFIG_NUMA
/*
 * Zone reclaim mode
 *
 * If non-zero call zone_reclaim when the number of free pages falls below
 * the watermarks.
 */
int zone_reclaim_mode __read_mostly;

1942 1943 1944 1945 1946
#define RECLAIM_OFF 0
#define RECLAIM_ZONE (1<<0)	/* Run shrink_cache on the zone */
#define RECLAIM_WRITE (1<<1)	/* Writeout pages during reclaim */
#define RECLAIM_SWAP (1<<2)	/* Swap pages out during reclaim */

1947 1948 1949 1950 1951 1952 1953
/*
 * 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

1954 1955 1956 1957 1958 1959
/*
 * Percentage of pages in a zone that must be unmapped for zone_reclaim to
 * occur.
 */
int sysctl_min_unmapped_ratio = 1;

1960 1961 1962 1963 1964 1965
/*
 * 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;

1966 1967 1968
/*
 * Try to free up some pages from this zone through reclaim.
 */
1969
static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
1970
{
1971
	/* Minimum pages needed in order to stay on node */
1972
	const unsigned long nr_pages = 1 << order;
1973 1974
	struct task_struct *p = current;
	struct reclaim_state reclaim_state;
1975
	int priority;
1976
	unsigned long nr_reclaimed = 0;
1977 1978 1979
	struct scan_control sc = {
		.may_writepage = !!(zone_reclaim_mode & RECLAIM_WRITE),
		.may_swap = !!(zone_reclaim_mode & RECLAIM_SWAP),
1980 1981
		.swap_cluster_max = max_t(unsigned long, nr_pages,
					SWAP_CLUSTER_MAX),
1982
		.gfp_mask = gfp_mask,
1983
		.swappiness = vm_swappiness,
1984
		.isolate_pages = isolate_pages_global,
1985
	};
1986
	unsigned long slab_reclaimable;
1987 1988 1989

	disable_swap_token();
	cond_resched();
1990 1991 1992 1993 1994 1995
	/*
	 * 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;
1996 1997
	reclaim_state.reclaimed_slab = 0;
	p->reclaim_state = &reclaim_state;
1998

1999 2000 2001 2002 2003 2004 2005 2006 2007
	if (zone_page_state(zone, NR_FILE_PAGES) -
		zone_page_state(zone, NR_FILE_MAPPED) >
		zone->min_unmapped_pages) {
		/*
		 * Free memory by calling shrink zone with increasing
		 * priorities until we have enough memory freed.
		 */
		priority = ZONE_RECLAIM_PRIORITY;
		do {
2008
			note_zone_scanning_priority(zone, priority);
2009 2010 2011 2012
			nr_reclaimed += shrink_zone(priority, zone, &sc);
			priority--;
		} while (priority >= 0 && nr_reclaimed < nr_pages);
	}
2013

2014 2015
	slab_reclaimable = zone_page_state(zone, NR_SLAB_RECLAIMABLE);
	if (slab_reclaimable > zone->min_slab_pages) {
2016
		/*
2017
		 * shrink_slab() does not currently allow us to determine how
2018 2019 2020 2021
		 * 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.
2022
		 *
2023 2024
		 * Note that shrink_slab will free memory on all zones and may
		 * take a long time.
2025
		 */
2026
		while (shrink_slab(sc.nr_scanned, gfp_mask, order) &&
2027 2028
			zone_page_state(zone, NR_SLAB_RECLAIMABLE) >
				slab_reclaimable - nr_pages)
2029
			;
2030 2031 2032 2033 2034 2035 2036

		/*
		 * Update nr_reclaimed by the number of slab pages we
		 * reclaimed from this zone.
		 */
		nr_reclaimed += slab_reclaimable -
			zone_page_state(zone, NR_SLAB_RECLAIMABLE);
2037 2038
	}

2039
	p->reclaim_state = NULL;
2040
	current->flags &= ~(PF_MEMALLOC | PF_SWAPWRITE);
2041
	return nr_reclaimed >= nr_pages;
2042
}
2043 2044 2045 2046

int zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order)
{
	int node_id;
2047
	int ret;
2048 2049

	/*
2050 2051
	 * Zone reclaim reclaims unmapped file backed pages and
	 * slab pages if we are over the defined limits.
2052
	 *
2053 2054 2055 2056 2057
	 * 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.
2058
	 */
2059
	if (zone_page_state(zone, NR_FILE_PAGES) -
2060 2061 2062
	    zone_page_state(zone, NR_FILE_MAPPED) <= zone->min_unmapped_pages
	    && zone_page_state(zone, NR_SLAB_RECLAIMABLE)
			<= zone->min_slab_pages)
2063
		return 0;
2064

2065 2066 2067
	if (zone_is_all_unreclaimable(zone))
		return 0;

2068
	/*
2069
	 * Do not scan if the allocation should not be delayed.
2070
	 */
2071
	if (!(gfp_mask & __GFP_WAIT) || (current->flags & PF_MEMALLOC))
2072 2073 2074 2075 2076 2077 2078 2079
			return 0;

	/*
	 * 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.
	 */
2080
	node_id = zone_to_nid(zone);
2081
	if (node_state(node_id, N_CPU) && node_id != numa_node_id())
2082
		return 0;
2083 2084 2085 2086 2087 2088 2089

	if (zone_test_and_set_flag(zone, ZONE_RECLAIM_LOCKED))
		return 0;
	ret = __zone_reclaim(zone, gfp_mask, order);
	zone_clear_flag(zone, ZONE_RECLAIM_LOCKED);

	return ret;
2090
}
2091
#endif