page_alloc.c 62.3 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 22 23
/*
 *  linux/mm/page_alloc.c
 *
 *  Manages the free list, the system allocates free pages here.
 *  Note that kmalloc() lives in slab.c
 *
 *  Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
 *  Swap reorganised 29.12.95, Stephen Tweedie
 *  Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
 *  Reshaped it to be a zoned allocator, Ingo Molnar, Red Hat, 1999
 *  Discontiguous memory support, Kanoj Sarcar, SGI, Nov 1999
 *  Zone balancing, Kanoj Sarcar, SGI, Jan 2000
 *  Per cpu hot/cold page lists, bulk allocation, Martin J. Bligh, Sept 2002
 *          (lots of bits borrowed from Ingo Molnar & Andrew Morton)
 */

#include <linux/stddef.h>
#include <linux/mm.h>
#include <linux/swap.h>
#include <linux/interrupt.h>
#include <linux/pagemap.h>
#include <linux/bootmem.h>
#include <linux/compiler.h>
24
#include <linux/kernel.h>
L
Linus Torvalds 已提交
25 26 27 28 29 30 31 32 33 34
#include <linux/module.h>
#include <linux/suspend.h>
#include <linux/pagevec.h>
#include <linux/blkdev.h>
#include <linux/slab.h>
#include <linux/notifier.h>
#include <linux/topology.h>
#include <linux/sysctl.h>
#include <linux/cpu.h>
#include <linux/cpuset.h>
35
#include <linux/memory_hotplug.h>
L
Linus Torvalds 已提交
36 37
#include <linux/nodemask.h>
#include <linux/vmalloc.h>
38
#include <linux/mempolicy.h>
39
#include <linux/stop_machine.h>
L
Linus Torvalds 已提交
40 41

#include <asm/tlbflush.h>
42
#include <asm/div64.h>
L
Linus Torvalds 已提交
43 44 45 46 47 48
#include "internal.h"

/*
 * MCD - HACK: Find somewhere to initialize this EARLY, or make this
 * initializer cleaner
 */
49
nodemask_t node_online_map __read_mostly = { { [0] = 1UL } };
50
EXPORT_SYMBOL(node_online_map);
51
nodemask_t node_possible_map __read_mostly = NODE_MASK_ALL;
52
EXPORT_SYMBOL(node_possible_map);
53
unsigned long totalram_pages __read_mostly;
54
unsigned long totalreserve_pages __read_mostly;
L
Linus Torvalds 已提交
55
long nr_swap_pages;
56
int percpu_pagelist_fraction;
L
Linus Torvalds 已提交
57

58
static void __free_pages_ok(struct page *page, unsigned int order);
59

L
Linus Torvalds 已提交
60 61 62 63 64 65 66
/*
 * results with 256, 32 in the lowmem_reserve sysctl:
 *	1G machine -> (16M dma, 800M-16M normal, 1G-800M high)
 *	1G machine -> (16M dma, 784M normal, 224M high)
 *	NORMAL allocation will leave 784M/256 of ram reserved in the ZONE_DMA
 *	HIGHMEM allocation will leave 224M/32 of ram reserved in ZONE_NORMAL
 *	HIGHMEM allocation will (224M+784M)/256 of ram reserved in ZONE_DMA
A
Andi Kleen 已提交
67 68 69
 *
 * TBD: should special case ZONE_DMA32 machines here - in those we normally
 * don't need any ZONE_NORMAL reservation
L
Linus Torvalds 已提交
70
 */
A
Andi Kleen 已提交
71
int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES-1] = { 256, 256, 32 };
L
Linus Torvalds 已提交
72 73 74 75 76 77 78

EXPORT_SYMBOL(totalram_pages);

/*
 * Used by page_zone() to look up the address of the struct zone whose
 * id is encoded in the upper bits of page->flags
 */
79
struct zone *zone_table[1 << ZONETABLE_SHIFT] __read_mostly;
L
Linus Torvalds 已提交
80 81
EXPORT_SYMBOL(zone_table);

A
Andi Kleen 已提交
82
static char *zone_names[MAX_NR_ZONES] = { "DMA", "DMA32", "Normal", "HighMem" };
L
Linus Torvalds 已提交
83 84
int min_free_kbytes = 1024;

85 86
unsigned long __meminitdata nr_kernel_pages;
unsigned long __meminitdata nr_all_pages;
L
Linus Torvalds 已提交
87

N
Nick Piggin 已提交
88
#ifdef CONFIG_DEBUG_VM
89
static int page_outside_zone_boundaries(struct zone *zone, struct page *page)
L
Linus Torvalds 已提交
90
{
91 92 93
	int ret = 0;
	unsigned seq;
	unsigned long pfn = page_to_pfn(page);
94

95 96 97 98 99 100 101 102 103
	do {
		seq = zone_span_seqbegin(zone);
		if (pfn >= zone->zone_start_pfn + zone->spanned_pages)
			ret = 1;
		else if (pfn < zone->zone_start_pfn)
			ret = 1;
	} while (zone_span_seqretry(zone, seq));

	return ret;
104 105 106 107
}

static int page_is_consistent(struct zone *zone, struct page *page)
{
L
Linus Torvalds 已提交
108 109
#ifdef CONFIG_HOLES_IN_ZONE
	if (!pfn_valid(page_to_pfn(page)))
110
		return 0;
L
Linus Torvalds 已提交
111 112
#endif
	if (zone != page_zone(page))
113 114 115 116 117 118 119 120 121 122
		return 0;

	return 1;
}
/*
 * Temporary debugging check for pages not lying within a given zone.
 */
static int bad_range(struct zone *zone, struct page *page)
{
	if (page_outside_zone_boundaries(zone, page))
L
Linus Torvalds 已提交
123
		return 1;
124 125 126
	if (!page_is_consistent(zone, page))
		return 1;

L
Linus Torvalds 已提交
127 128
	return 0;
}
N
Nick Piggin 已提交
129 130 131 132 133 134 135
#else
static inline int bad_range(struct zone *zone, struct page *page)
{
	return 0;
}
#endif

N
Nick Piggin 已提交
136
static void bad_page(struct page *page)
L
Linus Torvalds 已提交
137
{
N
Nick Piggin 已提交
138
	printk(KERN_EMERG "Bad page state in process '%s'\n"
139 140 141
		KERN_EMERG "page:%p flags:0x%0*lx mapping:%p mapcount:%d count:%d\n"
		KERN_EMERG "Trying to fix it up, but a reboot is needed\n"
		KERN_EMERG "Backtrace:\n",
N
Nick Piggin 已提交
142 143 144
		current->comm, page, (int)(2*sizeof(unsigned long)),
		(unsigned long)page->flags, page->mapping,
		page_mapcount(page), page_count(page));
L
Linus Torvalds 已提交
145
	dump_stack();
146 147
	page->flags &= ~(1 << PG_lru	|
			1 << PG_private |
L
Linus Torvalds 已提交
148 149 150
			1 << PG_locked	|
			1 << PG_active	|
			1 << PG_dirty	|
151 152
			1 << PG_reclaim |
			1 << PG_slab    |
L
Linus Torvalds 已提交
153
			1 << PG_swapcache |
154 155
			1 << PG_writeback |
			1 << PG_buddy );
L
Linus Torvalds 已提交
156 157 158
	set_page_count(page, 0);
	reset_page_mapcount(page);
	page->mapping = NULL;
159
	add_taint(TAINT_BAD_PAGE);
L
Linus Torvalds 已提交
160 161 162 163 164 165 166 167 168 169 170 171
}

/*
 * Higher-order pages are called "compound pages".  They are structured thusly:
 *
 * The first PAGE_SIZE page is called the "head page".
 *
 * The remaining PAGE_SIZE pages are called "tail pages".
 *
 * All pages have PG_compound set.  All pages have their ->private pointing at
 * the head page (even the head page has this).
 *
172 173 174
 * The first tail page's ->lru.next holds the address of the compound page's
 * put_page() function.  Its ->lru.prev holds the order of allocation.
 * This usage means that zero-order pages may not be compound.
L
Linus Torvalds 已提交
175
 */
176 177 178 179 180 181

static void free_compound_page(struct page *page)
{
	__free_pages_ok(page, (unsigned long)page[1].lru.prev);
}

L
Linus Torvalds 已提交
182 183 184 185 186
static void prep_compound_page(struct page *page, unsigned long order)
{
	int i;
	int nr_pages = 1 << order;

187
	page[1].lru.next = (void *)free_compound_page;	/* set dtor */
188
	page[1].lru.prev = (void *)order;
L
Linus Torvalds 已提交
189 190 191
	for (i = 0; i < nr_pages; i++) {
		struct page *p = page + i;

192
		__SetPageCompound(p);
H
Hugh Dickins 已提交
193
		set_page_private(p, (unsigned long)page);
L
Linus Torvalds 已提交
194 195 196 197 198 199 200 201
	}
}

static void destroy_compound_page(struct page *page, unsigned long order)
{
	int i;
	int nr_pages = 1 << order;

202
	if (unlikely((unsigned long)page[1].lru.prev != order))
N
Nick Piggin 已提交
203
		bad_page(page);
L
Linus Torvalds 已提交
204 205 206 207

	for (i = 0; i < nr_pages; i++) {
		struct page *p = page + i;

N
Nick Piggin 已提交
208 209 210
		if (unlikely(!PageCompound(p) |
				(page_private(p) != (unsigned long)page)))
			bad_page(page);
211
		__ClearPageCompound(p);
L
Linus Torvalds 已提交
212 213 214
	}
}

N
Nick Piggin 已提交
215 216 217 218
static inline void prep_zero_page(struct page *page, int order, gfp_t gfp_flags)
{
	int i;

N
Nick Piggin 已提交
219
	VM_BUG_ON((gfp_flags & (__GFP_WAIT | __GFP_HIGHMEM)) == __GFP_HIGHMEM);
220 221 222 223
	/*
	 * clear_highpage() will use KM_USER0, so it's a bug to use __GFP_ZERO
	 * and __GFP_HIGHMEM from hard or soft interrupt context.
	 */
N
Nick Piggin 已提交
224
	VM_BUG_ON((gfp_flags & __GFP_HIGHMEM) && in_interrupt());
N
Nick Piggin 已提交
225 226 227 228
	for (i = 0; i < (1 << order); i++)
		clear_highpage(page + i);
}

L
Linus Torvalds 已提交
229 230 231 232 233
/*
 * function for dealing with page's order in buddy system.
 * zone->lock is already acquired when we use these.
 * So, we don't need atomic page->flags operations here.
 */
234 235
static inline unsigned long page_order(struct page *page)
{
H
Hugh Dickins 已提交
236
	return page_private(page);
L
Linus Torvalds 已提交
237 238
}

239 240
static inline void set_page_order(struct page *page, int order)
{
H
Hugh Dickins 已提交
241
	set_page_private(page, order);
242
	__SetPageBuddy(page);
L
Linus Torvalds 已提交
243 244 245 246
}

static inline void rmv_page_order(struct page *page)
{
247
	__ClearPageBuddy(page);
H
Hugh Dickins 已提交
248
	set_page_private(page, 0);
L
Linus Torvalds 已提交
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
}

/*
 * Locate the struct page for both the matching buddy in our
 * pair (buddy1) and the combined O(n+1) page they form (page).
 *
 * 1) Any buddy B1 will have an order O twin B2 which satisfies
 * the following equation:
 *     B2 = B1 ^ (1 << O)
 * For example, if the starting buddy (buddy2) is #8 its order
 * 1 buddy is #10:
 *     B2 = 8 ^ (1 << 1) = 8 ^ 2 = 10
 *
 * 2) Any buddy B will have an order O+1 parent P which
 * satisfies the following equation:
 *     P = B & ~(1 << O)
 *
A
Andreas Mohr 已提交
266
 * Assumption: *_mem_map is contiguous at least up to MAX_ORDER
L
Linus Torvalds 已提交
267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284
 */
static inline struct page *
__page_find_buddy(struct page *page, unsigned long page_idx, unsigned int order)
{
	unsigned long buddy_idx = page_idx ^ (1 << order);

	return page + (buddy_idx - page_idx);
}

static inline unsigned long
__find_combined_index(unsigned long page_idx, unsigned int order)
{
	return (page_idx & ~(1 << order));
}

/*
 * This function checks whether a page is free && is the buddy
 * we can do coalesce a page and its buddy if
N
Nick Piggin 已提交
285
 * (a) the buddy is not in a hole &&
286
 * (b) the buddy is in the buddy system &&
287 288
 * (c) a page and its buddy have the same order &&
 * (d) a page and its buddy are in the same zone.
289 290 291
 *
 * For recording whether a page is in the buddy system, we use PG_buddy.
 * Setting, clearing, and testing PG_buddy is serialized by zone->lock.
L
Linus Torvalds 已提交
292
 *
293
 * For recording page's order, we use page_private(page).
L
Linus Torvalds 已提交
294
 */
295 296
static inline int page_is_buddy(struct page *page, struct page *buddy,
								int order)
L
Linus Torvalds 已提交
297
{
N
Nick Piggin 已提交
298
#ifdef CONFIG_HOLES_IN_ZONE
299
	if (!pfn_valid(page_to_pfn(buddy)))
N
Nick Piggin 已提交
300 301 302
		return 0;
#endif

303 304 305 306 307
	if (page_zone_id(page) != page_zone_id(buddy))
		return 0;

	if (PageBuddy(buddy) && page_order(buddy) == order) {
		BUG_ON(page_count(buddy) != 0);
308
		return 1;
309
	}
310
	return 0;
L
Linus Torvalds 已提交
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
}

/*
 * Freeing function for a buddy system allocator.
 *
 * The concept of a buddy system is to maintain direct-mapped table
 * (containing bit values) for memory blocks of various "orders".
 * The bottom level table contains the map for the smallest allocatable
 * units of memory (here, pages), and each level above it describes
 * pairs of units from the levels below, hence, "buddies".
 * At a high level, all that happens here is marking the table entry
 * at the bottom level available, and propagating the changes upward
 * as necessary, plus some accounting needed to play nicely with other
 * parts of the VM system.
 * At each level, we keep a list of pages, which are heads of continuous
326
 * free pages of length of (1 << order) and marked with PG_buddy. Page's
H
Hugh Dickins 已提交
327
 * order is recorded in page_private(page) field.
L
Linus Torvalds 已提交
328 329 330 331 332 333 334 335 336
 * So when we are allocating or freeing one, we can derive the state of the
 * other.  That is, if we allocate a small block, and both were   
 * free, the remainder of the region must be split into blocks.   
 * If a block is freed, and its buddy is also free, then this
 * triggers coalescing into a block of larger size.            
 *
 * -- wli
 */

N
Nick Piggin 已提交
337
static inline void __free_one_page(struct page *page,
L
Linus Torvalds 已提交
338 339 340 341 342
		struct zone *zone, unsigned int order)
{
	unsigned long page_idx;
	int order_size = 1 << order;

N
Nick Piggin 已提交
343
	if (unlikely(PageCompound(page)))
L
Linus Torvalds 已提交
344 345 346 347
		destroy_compound_page(page, order);

	page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);

N
Nick Piggin 已提交
348 349
	VM_BUG_ON(page_idx & (order_size - 1));
	VM_BUG_ON(bad_range(zone, page));
L
Linus Torvalds 已提交
350 351 352 353 354 355 356 357

	zone->free_pages += order_size;
	while (order < MAX_ORDER-1) {
		unsigned long combined_idx;
		struct free_area *area;
		struct page *buddy;

		buddy = __page_find_buddy(page, page_idx, order);
358
		if (!page_is_buddy(page, buddy, order))
L
Linus Torvalds 已提交
359
			break;		/* Move the buddy up one level. */
N
Nick Piggin 已提交
360

L
Linus Torvalds 已提交
361 362 363 364
		list_del(&buddy->lru);
		area = zone->free_area + order;
		area->nr_free--;
		rmv_page_order(buddy);
N
Nick Piggin 已提交
365
		combined_idx = __find_combined_index(page_idx, order);
L
Linus Torvalds 已提交
366 367 368 369 370 371 372 373 374
		page = page + (combined_idx - page_idx);
		page_idx = combined_idx;
		order++;
	}
	set_page_order(page, order);
	list_add(&page->lru, &zone->free_area[order].free_list);
	zone->free_area[order].nr_free++;
}

N
Nick Piggin 已提交
375
static inline int free_pages_check(struct page *page)
L
Linus Torvalds 已提交
376
{
N
Nick Piggin 已提交
377 378 379
	if (unlikely(page_mapcount(page) |
		(page->mapping != NULL)  |
		(page_count(page) != 0)  |
L
Linus Torvalds 已提交
380 381 382 383 384 385 386 387
		(page->flags & (
			1 << PG_lru	|
			1 << PG_private |
			1 << PG_locked	|
			1 << PG_active	|
			1 << PG_reclaim	|
			1 << PG_slab	|
			1 << PG_swapcache |
N
Nick Piggin 已提交
388
			1 << PG_writeback |
389 390
			1 << PG_reserved |
			1 << PG_buddy ))))
N
Nick Piggin 已提交
391
		bad_page(page);
L
Linus Torvalds 已提交
392
	if (PageDirty(page))
N
Nick Piggin 已提交
393
		__ClearPageDirty(page);
394 395 396 397 398 399
	/*
	 * For now, we report if PG_reserved was found set, but do not
	 * clear it, and do not free the page.  But we shall soon need
	 * to do more, for when the ZERO_PAGE count wraps negative.
	 */
	return PageReserved(page);
L
Linus Torvalds 已提交
400 401 402 403 404
}

/*
 * Frees a list of pages. 
 * Assumes all pages on list are in same zone, and of same order.
405
 * count is the number of pages to free.
L
Linus Torvalds 已提交
406 407 408 409 410 411 412
 *
 * If the zone was previously in an "all pages pinned" state then look to
 * see if this freeing clears that state.
 *
 * And clear the zone's pages_scanned counter, to hold off the "all pages are
 * pinned" detection logic.
 */
N
Nick Piggin 已提交
413 414
static void free_pages_bulk(struct zone *zone, int count,
					struct list_head *list, int order)
L
Linus Torvalds 已提交
415
{
N
Nick Piggin 已提交
416
	spin_lock(&zone->lock);
L
Linus Torvalds 已提交
417 418
	zone->all_unreclaimable = 0;
	zone->pages_scanned = 0;
N
Nick Piggin 已提交
419 420 421
	while (count--) {
		struct page *page;

N
Nick Piggin 已提交
422
		VM_BUG_ON(list_empty(list));
L
Linus Torvalds 已提交
423
		page = list_entry(list->prev, struct page, lru);
N
Nick Piggin 已提交
424
		/* have to delete it as __free_one_page list manipulates */
L
Linus Torvalds 已提交
425
		list_del(&page->lru);
N
Nick Piggin 已提交
426
		__free_one_page(page, zone, order);
L
Linus Torvalds 已提交
427
	}
N
Nick Piggin 已提交
428
	spin_unlock(&zone->lock);
L
Linus Torvalds 已提交
429 430
}

N
Nick Piggin 已提交
431
static void free_one_page(struct zone *zone, struct page *page, int order)
L
Linus Torvalds 已提交
432 433
{
	LIST_HEAD(list);
N
Nick Piggin 已提交
434 435 436 437 438 439 440
	list_add(&page->lru, &list);
	free_pages_bulk(zone, 1, &list, order);
}

static void __free_pages_ok(struct page *page, unsigned int order)
{
	unsigned long flags;
L
Linus Torvalds 已提交
441
	int i;
442
	int reserved = 0;
L
Linus Torvalds 已提交
443 444

	arch_free_page(page, order);
445
	if (!PageHighMem(page))
446 447
		debug_check_no_locks_freed(page_address(page),
					   PAGE_SIZE<<order);
L
Linus Torvalds 已提交
448 449

	for (i = 0 ; i < (1 << order) ; ++i)
N
Nick Piggin 已提交
450
		reserved += free_pages_check(page + i);
451 452 453
	if (reserved)
		return;

N
Nick Piggin 已提交
454
	kernel_map_pages(page, 1 << order, 0);
N
Nick Piggin 已提交
455
	local_irq_save(flags);
456
	__count_vm_events(PGFREE, 1 << order);
N
Nick Piggin 已提交
457
	free_one_page(page_zone(page), page, order);
N
Nick Piggin 已提交
458
	local_irq_restore(flags);
L
Linus Torvalds 已提交
459 460
}

461 462 463 464 465 466 467 468
/*
 * permit the bootmem allocator to evade page validation on high-order frees
 */
void fastcall __init __free_pages_bootmem(struct page *page, unsigned int order)
{
	if (order == 0) {
		__ClearPageReserved(page);
		set_page_count(page, 0);
469
		set_page_refcounted(page);
N
Nick Piggin 已提交
470
		__free_page(page);
471 472 473
	} else {
		int loop;

N
Nick Piggin 已提交
474
		prefetchw(page);
475 476 477
		for (loop = 0; loop < BITS_PER_LONG; loop++) {
			struct page *p = &page[loop];

N
Nick Piggin 已提交
478 479
			if (loop + 1 < BITS_PER_LONG)
				prefetchw(p + 1);
480 481 482 483
			__ClearPageReserved(p);
			set_page_count(p, 0);
		}

484
		set_page_refcounted(page);
N
Nick Piggin 已提交
485
		__free_pages(page, order);
486 487 488
	}
}

L
Linus Torvalds 已提交
489 490 491 492 493 494 495 496 497 498 499 500 501 502 503

/*
 * The order of subdivision here is critical for the IO subsystem.
 * Please do not alter this order without good reasons and regression
 * testing. Specifically, as large blocks of memory are subdivided,
 * the order in which smaller blocks are delivered depends on the order
 * they're subdivided in this function. This is the primary factor
 * influencing the order in which pages are delivered to the IO
 * subsystem according to empirical testing, and this is also justified
 * by considering the behavior of a buddy system containing a single
 * large block of memory acted on by a series of small allocations.
 * This behavior is a critical factor in sglist merging's success.
 *
 * -- wli
 */
N
Nick Piggin 已提交
504
static inline void expand(struct zone *zone, struct page *page,
L
Linus Torvalds 已提交
505 506 507 508 509 510 511 512
 	int low, int high, struct free_area *area)
{
	unsigned long size = 1 << high;

	while (high > low) {
		area--;
		high--;
		size >>= 1;
N
Nick Piggin 已提交
513
		VM_BUG_ON(bad_range(zone, &page[size]));
L
Linus Torvalds 已提交
514 515 516 517 518 519 520 521 522
		list_add(&page[size].lru, &area->free_list);
		area->nr_free++;
		set_page_order(&page[size], high);
	}
}

/*
 * This page is about to be returned from the page allocator
 */
N
Nick Piggin 已提交
523
static int prep_new_page(struct page *page, int order, gfp_t gfp_flags)
L
Linus Torvalds 已提交
524
{
N
Nick Piggin 已提交
525 526 527
	if (unlikely(page_mapcount(page) |
		(page->mapping != NULL)  |
		(page_count(page) != 0)  |
528 529
		(page->flags & (
			1 << PG_lru	|
L
Linus Torvalds 已提交
530 531 532 533 534
			1 << PG_private	|
			1 << PG_locked	|
			1 << PG_active	|
			1 << PG_dirty	|
			1 << PG_reclaim	|
535
			1 << PG_slab    |
L
Linus Torvalds 已提交
536
			1 << PG_swapcache |
N
Nick Piggin 已提交
537
			1 << PG_writeback |
538 539
			1 << PG_reserved |
			1 << PG_buddy ))))
N
Nick Piggin 已提交
540
		bad_page(page);
L
Linus Torvalds 已提交
541

542 543 544 545 546 547 548
	/*
	 * For now, we report if PG_reserved was found set, but do not
	 * clear it, and do not allocate the page: as a safety net.
	 */
	if (PageReserved(page))
		return 1;

L
Linus Torvalds 已提交
549 550 551
	page->flags &= ~(1 << PG_uptodate | 1 << PG_error |
			1 << PG_referenced | 1 << PG_arch_1 |
			1 << PG_checked | 1 << PG_mappedtodisk);
H
Hugh Dickins 已提交
552
	set_page_private(page, 0);
553
	set_page_refcounted(page);
L
Linus Torvalds 已提交
554
	kernel_map_pages(page, 1 << order, 1);
N
Nick Piggin 已提交
555 556 557 558 559 560 561

	if (gfp_flags & __GFP_ZERO)
		prep_zero_page(page, order, gfp_flags);

	if (order && (gfp_flags & __GFP_COMP))
		prep_compound_page(page, order);

562
	return 0;
L
Linus Torvalds 已提交
563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584
}

/* 
 * Do the hard work of removing an element from the buddy allocator.
 * Call me with the zone->lock already held.
 */
static struct page *__rmqueue(struct zone *zone, unsigned int order)
{
	struct free_area * area;
	unsigned int current_order;
	struct page *page;

	for (current_order = order; current_order < MAX_ORDER; ++current_order) {
		area = zone->free_area + current_order;
		if (list_empty(&area->free_list))
			continue;

		page = list_entry(area->free_list.next, struct page, lru);
		list_del(&page->lru);
		rmv_page_order(page);
		area->nr_free--;
		zone->free_pages -= 1UL << order;
N
Nick Piggin 已提交
585 586
		expand(zone, page, order, current_order, area);
		return page;
L
Linus Torvalds 已提交
587 588 589 590 591 592 593 594 595 596 597 598 599 600 601
	}

	return NULL;
}

/* 
 * Obtain a specified number of elements from the buddy allocator, all under
 * a single hold of the lock, for efficiency.  Add them to the supplied list.
 * Returns the number of new pages which were placed at *list.
 */
static int rmqueue_bulk(struct zone *zone, unsigned int order, 
			unsigned long count, struct list_head *list)
{
	int i;
	
N
Nick Piggin 已提交
602
	spin_lock(&zone->lock);
L
Linus Torvalds 已提交
603
	for (i = 0; i < count; ++i) {
N
Nick Piggin 已提交
604 605
		struct page *page = __rmqueue(zone, order);
		if (unlikely(page == NULL))
L
Linus Torvalds 已提交
606 607 608
			break;
		list_add_tail(&page->lru, list);
	}
N
Nick Piggin 已提交
609
	spin_unlock(&zone->lock);
N
Nick Piggin 已提交
610
	return i;
L
Linus Torvalds 已提交
611 612
}

613
#ifdef CONFIG_NUMA
614 615 616
/*
 * Called from the slab reaper to drain pagesets on a particular node that
 * belong to the currently executing processor.
617 618
 * Note that this function must be called with the thread pinned to
 * a single processor.
619 620
 */
void drain_node_pages(int nodeid)
621
{
622
	int i, z;
623 624
	unsigned long flags;

625 626
	for (z = 0; z < MAX_NR_ZONES; z++) {
		struct zone *zone = NODE_DATA(nodeid)->node_zones + z;
627 628
		struct per_cpu_pageset *pset;

N
Nick Piggin 已提交
629
		pset = zone_pcp(zone, smp_processor_id());
630 631 632 633
		for (i = 0; i < ARRAY_SIZE(pset->pcp); i++) {
			struct per_cpu_pages *pcp;

			pcp = &pset->pcp[i];
634 635 636 637 638 639
			if (pcp->count) {
				local_irq_save(flags);
				free_pages_bulk(zone, pcp->count, &pcp->list, 0);
				pcp->count = 0;
				local_irq_restore(flags);
			}
640 641 642 643 644
		}
	}
}
#endif

L
Linus Torvalds 已提交
645 646 647
#if defined(CONFIG_PM) || defined(CONFIG_HOTPLUG_CPU)
static void __drain_pages(unsigned int cpu)
{
N
Nick Piggin 已提交
648
	unsigned long flags;
L
Linus Torvalds 已提交
649 650 651 652 653 654
	struct zone *zone;
	int i;

	for_each_zone(zone) {
		struct per_cpu_pageset *pset;

655
		pset = zone_pcp(zone, cpu);
L
Linus Torvalds 已提交
656 657 658 659
		for (i = 0; i < ARRAY_SIZE(pset->pcp); i++) {
			struct per_cpu_pages *pcp;

			pcp = &pset->pcp[i];
N
Nick Piggin 已提交
660
			local_irq_save(flags);
N
Nick Piggin 已提交
661 662
			free_pages_bulk(zone, pcp->count, &pcp->list, 0);
			pcp->count = 0;
N
Nick Piggin 已提交
663
			local_irq_restore(flags);
L
Linus Torvalds 已提交
664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721
		}
	}
}
#endif /* CONFIG_PM || CONFIG_HOTPLUG_CPU */

#ifdef CONFIG_PM

void mark_free_pages(struct zone *zone)
{
	unsigned long zone_pfn, flags;
	int order;
	struct list_head *curr;

	if (!zone->spanned_pages)
		return;

	spin_lock_irqsave(&zone->lock, flags);
	for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn)
		ClearPageNosaveFree(pfn_to_page(zone_pfn + zone->zone_start_pfn));

	for (order = MAX_ORDER - 1; order >= 0; --order)
		list_for_each(curr, &zone->free_area[order].free_list) {
			unsigned long start_pfn, i;

			start_pfn = page_to_pfn(list_entry(curr, struct page, lru));

			for (i=0; i < (1<<order); i++)
				SetPageNosaveFree(pfn_to_page(start_pfn+i));
	}
	spin_unlock_irqrestore(&zone->lock, flags);
}

/*
 * Spill all of this CPU's per-cpu pages back into the buddy allocator.
 */
void drain_local_pages(void)
{
	unsigned long flags;

	local_irq_save(flags);	
	__drain_pages(smp_processor_id());
	local_irq_restore(flags);	
}
#endif /* CONFIG_PM */

/*
 * Free a 0-order page
 */
static void fastcall free_hot_cold_page(struct page *page, int cold)
{
	struct zone *zone = page_zone(page);
	struct per_cpu_pages *pcp;
	unsigned long flags;

	arch_free_page(page, 0);

	if (PageAnon(page))
		page->mapping = NULL;
N
Nick Piggin 已提交
722
	if (free_pages_check(page))
723 724 725 726
		return;

	kernel_map_pages(page, 1, 0);

727
	pcp = &zone_pcp(zone, get_cpu())->pcp[cold];
L
Linus Torvalds 已提交
728
	local_irq_save(flags);
729
	__count_vm_event(PGFREE);
L
Linus Torvalds 已提交
730 731
	list_add(&page->lru, &pcp->list);
	pcp->count++;
N
Nick Piggin 已提交
732 733 734 735
	if (pcp->count >= pcp->high) {
		free_pages_bulk(zone, pcp->batch, &pcp->list, 0);
		pcp->count -= pcp->batch;
	}
L
Linus Torvalds 已提交
736 737 738 739 740 741 742 743 744 745 746 747 748 749
	local_irq_restore(flags);
	put_cpu();
}

void fastcall free_hot_page(struct page *page)
{
	free_hot_cold_page(page, 0);
}
	
void fastcall free_cold_page(struct page *page)
{
	free_hot_cold_page(page, 1);
}

N
Nick Piggin 已提交
750 751 752 753 754 755 756 757 758 759 760 761
/*
 * split_page takes a non-compound higher-order page, and splits it into
 * n (1<<order) sub-pages: page[0..n]
 * Each sub-page must be freed individually.
 *
 * Note: this is probably too low level an operation for use in drivers.
 * Please consult with lkml before using this in your driver.
 */
void split_page(struct page *page, unsigned int order)
{
	int i;

N
Nick Piggin 已提交
762 763
	VM_BUG_ON(PageCompound(page));
	VM_BUG_ON(!page_count(page));
764 765
	for (i = 1; i < (1 << order); i++)
		set_page_refcounted(page + i);
N
Nick Piggin 已提交
766 767
}

L
Linus Torvalds 已提交
768 769 770 771 772
/*
 * Really, prep_compound_page() should be called from __rmqueue_bulk().  But
 * we cheat by calling it from here, in the order > 0 path.  Saves a branch
 * or two.
 */
N
Nick Piggin 已提交
773 774
static struct page *buffered_rmqueue(struct zonelist *zonelist,
			struct zone *zone, int order, gfp_t gfp_flags)
L
Linus Torvalds 已提交
775 776
{
	unsigned long flags;
777
	struct page *page;
L
Linus Torvalds 已提交
778
	int cold = !!(gfp_flags & __GFP_COLD);
N
Nick Piggin 已提交
779
	int cpu;
L
Linus Torvalds 已提交
780

781
again:
N
Nick Piggin 已提交
782
	cpu  = get_cpu();
N
Nick Piggin 已提交
783
	if (likely(order == 0)) {
L
Linus Torvalds 已提交
784 785
		struct per_cpu_pages *pcp;

N
Nick Piggin 已提交
786
		pcp = &zone_pcp(zone, cpu)->pcp[cold];
L
Linus Torvalds 已提交
787
		local_irq_save(flags);
N
Nick Piggin 已提交
788
		if (!pcp->count) {
L
Linus Torvalds 已提交
789 790
			pcp->count += rmqueue_bulk(zone, 0,
						pcp->batch, &pcp->list);
N
Nick Piggin 已提交
791 792
			if (unlikely(!pcp->count))
				goto failed;
L
Linus Torvalds 已提交
793
		}
N
Nick Piggin 已提交
794 795 796
		page = list_entry(pcp->list.next, struct page, lru);
		list_del(&page->lru);
		pcp->count--;
R
Rohit Seth 已提交
797
	} else {
L
Linus Torvalds 已提交
798 799
		spin_lock_irqsave(&zone->lock, flags);
		page = __rmqueue(zone, order);
N
Nick Piggin 已提交
800 801 802
		spin_unlock(&zone->lock);
		if (!page)
			goto failed;
L
Linus Torvalds 已提交
803 804
	}

805
	__count_zone_vm_events(PGALLOC, zone, 1 << order);
806
	zone_statistics(zonelist, zone);
N
Nick Piggin 已提交
807 808
	local_irq_restore(flags);
	put_cpu();
L
Linus Torvalds 已提交
809

N
Nick Piggin 已提交
810
	VM_BUG_ON(bad_range(zone, page));
N
Nick Piggin 已提交
811
	if (prep_new_page(page, order, gfp_flags))
N
Nick Piggin 已提交
812
		goto again;
L
Linus Torvalds 已提交
813
	return page;
N
Nick Piggin 已提交
814 815 816 817 818

failed:
	local_irq_restore(flags);
	put_cpu();
	return NULL;
L
Linus Torvalds 已提交
819 820
}

R
Rohit Seth 已提交
821
#define ALLOC_NO_WATERMARKS	0x01 /* don't check watermarks at all */
822 823 824 825 826 827
#define ALLOC_WMARK_MIN		0x02 /* use pages_min watermark */
#define ALLOC_WMARK_LOW		0x04 /* use pages_low watermark */
#define ALLOC_WMARK_HIGH	0x08 /* use pages_high watermark */
#define ALLOC_HARDER		0x10 /* try to alloc harder */
#define ALLOC_HIGH		0x20 /* __GFP_HIGH set */
#define ALLOC_CPUSET		0x40 /* check for correct cpuset */
R
Rohit Seth 已提交
828

L
Linus Torvalds 已提交
829 830 831 832 833
/*
 * Return 1 if free pages are above 'mark'. This takes into account the order
 * of the allocation.
 */
int zone_watermark_ok(struct zone *z, int order, unsigned long mark,
R
Rohit Seth 已提交
834
		      int classzone_idx, int alloc_flags)
L
Linus Torvalds 已提交
835 836 837 838 839
{
	/* free_pages my go negative - that's OK */
	long min = mark, free_pages = z->free_pages - (1 << order) + 1;
	int o;

R
Rohit Seth 已提交
840
	if (alloc_flags & ALLOC_HIGH)
L
Linus Torvalds 已提交
841
		min -= min / 2;
R
Rohit Seth 已提交
842
	if (alloc_flags & ALLOC_HARDER)
L
Linus Torvalds 已提交
843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859
		min -= min / 4;

	if (free_pages <= min + z->lowmem_reserve[classzone_idx])
		return 0;
	for (o = 0; o < order; o++) {
		/* At the next order, this order's pages become unavailable */
		free_pages -= z->free_area[o].nr_free << o;

		/* Require fewer higher order pages to be free */
		min >>= 1;

		if (free_pages <= min)
			return 0;
	}
	return 1;
}

R
Rohit Seth 已提交
860 861 862 863 864 865 866
/*
 * get_page_from_freeliest goes through the zonelist trying to allocate
 * a page.
 */
static struct page *
get_page_from_freelist(gfp_t gfp_mask, unsigned int order,
		struct zonelist *zonelist, int alloc_flags)
M
Martin Hicks 已提交
867
{
R
Rohit Seth 已提交
868 869 870 871 872 873 874 875 876 877 878 879 880 881
	struct zone **z = zonelist->zones;
	struct page *page = NULL;
	int classzone_idx = zone_idx(*z);

	/*
	 * Go through the zonelist once, looking for a zone with enough free.
	 * See also cpuset_zone_allowed() comment in kernel/cpuset.c.
	 */
	do {
		if ((alloc_flags & ALLOC_CPUSET) &&
				!cpuset_zone_allowed(*z, gfp_mask))
			continue;

		if (!(alloc_flags & ALLOC_NO_WATERMARKS)) {
882 883 884 885 886 887 888 889
			unsigned long mark;
			if (alloc_flags & ALLOC_WMARK_MIN)
				mark = (*z)->pages_min;
			else if (alloc_flags & ALLOC_WMARK_LOW)
				mark = (*z)->pages_low;
			else
				mark = (*z)->pages_high;
			if (!zone_watermark_ok(*z, order, mark,
R
Rohit Seth 已提交
890
				    classzone_idx, alloc_flags))
891 892 893
				if (!zone_reclaim_mode ||
				    !zone_reclaim(*z, gfp_mask, order))
					continue;
R
Rohit Seth 已提交
894 895
		}

N
Nick Piggin 已提交
896
		page = buffered_rmqueue(zonelist, *z, order, gfp_mask);
R
Rohit Seth 已提交
897 898 899 900 901
		if (page) {
			break;
		}
	} while (*(++z) != NULL);
	return page;
M
Martin Hicks 已提交
902 903
}

L
Linus Torvalds 已提交
904 905 906 907
/*
 * This is the 'heart' of the zoned buddy allocator.
 */
struct page * fastcall
A
Al Viro 已提交
908
__alloc_pages(gfp_t gfp_mask, unsigned int order,
L
Linus Torvalds 已提交
909 910
		struct zonelist *zonelist)
{
A
Al Viro 已提交
911
	const gfp_t wait = gfp_mask & __GFP_WAIT;
R
Rohit Seth 已提交
912
	struct zone **z;
L
Linus Torvalds 已提交
913 914 915 916
	struct page *page;
	struct reclaim_state reclaim_state;
	struct task_struct *p = current;
	int do_retry;
R
Rohit Seth 已提交
917
	int alloc_flags;
L
Linus Torvalds 已提交
918 919 920 921
	int did_some_progress;

	might_sleep_if(wait);

922
restart:
R
Rohit Seth 已提交
923
	z = zonelist->zones;  /* the list of zones suitable for gfp_mask */
L
Linus Torvalds 已提交
924

R
Rohit Seth 已提交
925
	if (unlikely(*z == NULL)) {
L
Linus Torvalds 已提交
926 927 928
		/* Should this ever happen?? */
		return NULL;
	}
929

R
Rohit Seth 已提交
930
	page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, order,
931
				zonelist, ALLOC_WMARK_LOW|ALLOC_CPUSET);
R
Rohit Seth 已提交
932 933
	if (page)
		goto got_pg;
L
Linus Torvalds 已提交
934

935
	do {
936
		wakeup_kswapd(*z, order);
937
	} while (*(++z));
L
Linus Torvalds 已提交
938

939
	/*
R
Rohit Seth 已提交
940 941 942 943 944 945
	 * OK, we're below the kswapd watermark and have kicked background
	 * reclaim. Now things get more complex, so set up alloc_flags according
	 * to how we want to proceed.
	 *
	 * The caller may dip into page reserves a bit more if the caller
	 * cannot run direct reclaim, or if the caller has realtime scheduling
P
Paul Jackson 已提交
946 947
	 * policy or is asking for __GFP_HIGH memory.  GFP_ATOMIC requests will
	 * set both ALLOC_HARDER (!wait) and ALLOC_HIGH (__GFP_HIGH).
948
	 */
949
	alloc_flags = ALLOC_WMARK_MIN;
R
Rohit Seth 已提交
950 951 952 953
	if ((unlikely(rt_task(p)) && !in_interrupt()) || !wait)
		alloc_flags |= ALLOC_HARDER;
	if (gfp_mask & __GFP_HIGH)
		alloc_flags |= ALLOC_HIGH;
954 955
	if (wait)
		alloc_flags |= ALLOC_CPUSET;
L
Linus Torvalds 已提交
956 957 958

	/*
	 * Go through the zonelist again. Let __GFP_HIGH and allocations
R
Rohit Seth 已提交
959
	 * coming from realtime tasks go deeper into reserves.
L
Linus Torvalds 已提交
960 961 962
	 *
	 * This is the last chance, in general, before the goto nopage.
	 * Ignore cpuset if GFP_ATOMIC (!wait) rather than fail alloc.
963
	 * See also cpuset_zone_allowed() comment in kernel/cpuset.c.
L
Linus Torvalds 已提交
964
	 */
R
Rohit Seth 已提交
965 966 967
	page = get_page_from_freelist(gfp_mask, order, zonelist, alloc_flags);
	if (page)
		goto got_pg;
L
Linus Torvalds 已提交
968 969

	/* This allocation should allow future memory freeing. */
970 971 972 973

	if (((p->flags & PF_MEMALLOC) || unlikely(test_thread_flag(TIF_MEMDIE)))
			&& !in_interrupt()) {
		if (!(gfp_mask & __GFP_NOMEMALLOC)) {
K
Kirill Korotaev 已提交
974
nofail_alloc:
975
			/* go through the zonelist yet again, ignoring mins */
R
Rohit Seth 已提交
976
			page = get_page_from_freelist(gfp_mask, order,
977
				zonelist, ALLOC_NO_WATERMARKS);
R
Rohit Seth 已提交
978 979
			if (page)
				goto got_pg;
K
Kirill Korotaev 已提交
980 981 982 983
			if (gfp_mask & __GFP_NOFAIL) {
				blk_congestion_wait(WRITE, HZ/50);
				goto nofail_alloc;
			}
L
Linus Torvalds 已提交
984 985 986 987 988 989 990 991 992 993 994 995
		}
		goto nopage;
	}

	/* Atomic allocations - we can't balance anything */
	if (!wait)
		goto nopage;

rebalance:
	cond_resched();

	/* We now go into synchronous reclaim */
996
	cpuset_memory_pressure_bump();
L
Linus Torvalds 已提交
997 998 999 1000
	p->flags |= PF_MEMALLOC;
	reclaim_state.reclaimed_slab = 0;
	p->reclaim_state = &reclaim_state;

R
Rohit Seth 已提交
1001
	did_some_progress = try_to_free_pages(zonelist->zones, gfp_mask);
L
Linus Torvalds 已提交
1002 1003 1004 1005 1006 1007 1008

	p->reclaim_state = NULL;
	p->flags &= ~PF_MEMALLOC;

	cond_resched();

	if (likely(did_some_progress)) {
R
Rohit Seth 已提交
1009 1010 1011 1012
		page = get_page_from_freelist(gfp_mask, order,
						zonelist, alloc_flags);
		if (page)
			goto got_pg;
L
Linus Torvalds 已提交
1013 1014 1015 1016 1017 1018 1019
	} else if ((gfp_mask & __GFP_FS) && !(gfp_mask & __GFP_NORETRY)) {
		/*
		 * Go through the zonelist yet one more time, keep
		 * very high watermark here, this is only to catch
		 * a parallel oom killing, we must fail if we're still
		 * under heavy pressure.
		 */
R
Rohit Seth 已提交
1020
		page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, order,
1021
				zonelist, ALLOC_WMARK_HIGH|ALLOC_CPUSET);
R
Rohit Seth 已提交
1022 1023
		if (page)
			goto got_pg;
L
Linus Torvalds 已提交
1024

1025
		out_of_memory(zonelist, gfp_mask, order);
L
Linus Torvalds 已提交
1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053
		goto restart;
	}

	/*
	 * Don't let big-order allocations loop unless the caller explicitly
	 * requests that.  Wait for some write requests to complete then retry.
	 *
	 * In this implementation, __GFP_REPEAT means __GFP_NOFAIL for order
	 * <= 3, but that may not be true in other implementations.
	 */
	do_retry = 0;
	if (!(gfp_mask & __GFP_NORETRY)) {
		if ((order <= 3) || (gfp_mask & __GFP_REPEAT))
			do_retry = 1;
		if (gfp_mask & __GFP_NOFAIL)
			do_retry = 1;
	}
	if (do_retry) {
		blk_congestion_wait(WRITE, HZ/50);
		goto rebalance;
	}

nopage:
	if (!(gfp_mask & __GFP_NOWARN) && printk_ratelimit()) {
		printk(KERN_WARNING "%s: page allocation failure."
			" order:%d, mode:0x%x\n",
			p->comm, order, gfp_mask);
		dump_stack();
J
Janet Morgan 已提交
1054
		show_mem();
L
Linus Torvalds 已提交
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
	}
got_pg:
	return page;
}

EXPORT_SYMBOL(__alloc_pages);

/*
 * Common helper functions.
 */
A
Al Viro 已提交
1065
fastcall unsigned long __get_free_pages(gfp_t gfp_mask, unsigned int order)
L
Linus Torvalds 已提交
1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
{
	struct page * page;
	page = alloc_pages(gfp_mask, order);
	if (!page)
		return 0;
	return (unsigned long) page_address(page);
}

EXPORT_SYMBOL(__get_free_pages);

A
Al Viro 已提交
1076
fastcall unsigned long get_zeroed_page(gfp_t gfp_mask)
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081 1082 1083
{
	struct page * page;

	/*
	 * get_zeroed_page() returns a 32-bit address, which cannot represent
	 * a highmem page
	 */
N
Nick Piggin 已提交
1084
	VM_BUG_ON((gfp_mask & __GFP_HIGHMEM) != 0);
L
Linus Torvalds 已提交
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103

	page = alloc_pages(gfp_mask | __GFP_ZERO, 0);
	if (page)
		return (unsigned long) page_address(page);
	return 0;
}

EXPORT_SYMBOL(get_zeroed_page);

void __pagevec_free(struct pagevec *pvec)
{
	int i = pagevec_count(pvec);

	while (--i >= 0)
		free_hot_cold_page(pvec->pages[i], pvec->cold);
}

fastcall void __free_pages(struct page *page, unsigned int order)
{
N
Nick Piggin 已提交
1104
	if (put_page_testzero(page)) {
L
Linus Torvalds 已提交
1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116
		if (order == 0)
			free_hot_page(page);
		else
			__free_pages_ok(page, order);
	}
}

EXPORT_SYMBOL(__free_pages);

fastcall void free_pages(unsigned long addr, unsigned int order)
{
	if (addr != 0) {
N
Nick Piggin 已提交
1117
		VM_BUG_ON(!virt_addr_valid((void *)addr));
L
Linus Torvalds 已提交
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 1150 1151 1152 1153
		__free_pages(virt_to_page((void *)addr), order);
	}
}

EXPORT_SYMBOL(free_pages);

/*
 * Total amount of free (allocatable) RAM:
 */
unsigned int nr_free_pages(void)
{
	unsigned int sum = 0;
	struct zone *zone;

	for_each_zone(zone)
		sum += zone->free_pages;

	return sum;
}

EXPORT_SYMBOL(nr_free_pages);

#ifdef CONFIG_NUMA
unsigned int nr_free_pages_pgdat(pg_data_t *pgdat)
{
	unsigned int i, sum = 0;

	for (i = 0; i < MAX_NR_ZONES; i++)
		sum += pgdat->node_zones[i].free_pages;

	return sum;
}
#endif

static unsigned int nr_free_zone_pages(int offset)
{
1154 1155
	/* Just pick one node, since fallback list is circular */
	pg_data_t *pgdat = NODE_DATA(numa_node_id());
L
Linus Torvalds 已提交
1156 1157
	unsigned int sum = 0;

1158 1159 1160
	struct zonelist *zonelist = pgdat->node_zonelists + offset;
	struct zone **zonep = zonelist->zones;
	struct zone *zone;
L
Linus Torvalds 已提交
1161

1162 1163 1164 1165 1166
	for (zone = *zonep++; zone; zone = *zonep++) {
		unsigned long size = zone->present_pages;
		unsigned long high = zone->pages_high;
		if (size > high)
			sum += size - high;
L
Linus Torvalds 已提交
1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
	}

	return sum;
}

/*
 * Amount of free RAM allocatable within ZONE_DMA and ZONE_NORMAL
 */
unsigned int nr_free_buffer_pages(void)
{
A
Al Viro 已提交
1177
	return nr_free_zone_pages(gfp_zone(GFP_USER));
L
Linus Torvalds 已提交
1178 1179 1180 1181 1182 1183 1184
}

/*
 * Amount of free RAM allocatable within all zones
 */
unsigned int nr_free_pagecache_pages(void)
{
A
Al Viro 已提交
1185
	return nr_free_zone_pages(gfp_zone(GFP_HIGHUSER));
L
Linus Torvalds 已提交
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
}
#ifdef CONFIG_NUMA
static void show_node(struct zone *zone)
{
	printk("Node %d ", zone->zone_pgdat->node_id);
}
#else
#define show_node(zone)	do { } while (0)
#endif

void si_meminfo(struct sysinfo *val)
{
	val->totalram = totalram_pages;
	val->sharedram = 0;
	val->freeram = nr_free_pages();
	val->bufferram = nr_blockdev_pages();
	val->totalhigh = totalhigh_pages;
	val->freehigh = nr_free_highpages();
	val->mem_unit = PAGE_SIZE;
}

EXPORT_SYMBOL(si_meminfo);

#ifdef CONFIG_NUMA
void si_meminfo_node(struct sysinfo *val, int nid)
{
	pg_data_t *pgdat = NODE_DATA(nid);

	val->totalram = pgdat->node_present_pages;
	val->freeram = nr_free_pages_pgdat(pgdat);
1216
#ifdef CONFIG_HIGHMEM
L
Linus Torvalds 已提交
1217 1218
	val->totalhigh = pgdat->node_zones[ZONE_HIGHMEM].present_pages;
	val->freehigh = pgdat->node_zones[ZONE_HIGHMEM].free_pages;
1219 1220 1221 1222
#else
	val->totalhigh = 0;
	val->freehigh = 0;
#endif
L
Linus Torvalds 已提交
1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245
	val->mem_unit = PAGE_SIZE;
}
#endif

#define K(x) ((x) << (PAGE_SHIFT-10))

/*
 * Show free area list (used inside shift_scroll-lock stuff)
 * We also calculate the percentage fragmentation. We do this by counting the
 * memory on each free list with the exception of the first item on the list.
 */
void show_free_areas(void)
{
	int cpu, temperature;
	unsigned long active;
	unsigned long inactive;
	unsigned long free;
	struct zone *zone;

	for_each_zone(zone) {
		show_node(zone);
		printk("%s per-cpu:", zone->name);

1246
		if (!populated_zone(zone)) {
L
Linus Torvalds 已提交
1247 1248 1249 1250 1251
			printk(" empty\n");
			continue;
		} else
			printk("\n");

1252
		for_each_online_cpu(cpu) {
L
Linus Torvalds 已提交
1253 1254
			struct per_cpu_pageset *pageset;

1255
			pageset = zone_pcp(zone, cpu);
L
Linus Torvalds 已提交
1256 1257

			for (temperature = 0; temperature < 2; temperature++)
N
Nick Piggin 已提交
1258
				printk("cpu %d %s: high %d, batch %d used:%d\n",
L
Linus Torvalds 已提交
1259 1260 1261
					cpu,
					temperature ? "cold" : "hot",
					pageset->pcp[temperature].high,
1262 1263
					pageset->pcp[temperature].batch,
					pageset->pcp[temperature].count);
L
Linus Torvalds 已提交
1264 1265 1266 1267 1268 1269 1270 1271 1272
		}
	}

	get_zone_counts(&active, &inactive, &free);

	printk("Active:%lu inactive:%lu dirty:%lu writeback:%lu "
		"unstable:%lu free:%u slab:%lu mapped:%lu pagetables:%lu\n",
		active,
		inactive,
1273
		global_page_state(NR_FILE_DIRTY),
1274
		global_page_state(NR_WRITEBACK),
1275
		global_page_state(NR_UNSTABLE_NFS),
L
Linus Torvalds 已提交
1276
		nr_free_pages(),
1277
		global_page_state(NR_SLAB),
1278
		global_page_state(NR_FILE_MAPPED),
1279
		global_page_state(NR_PAGETABLE));
L
Linus Torvalds 已提交
1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313

	for_each_zone(zone) {
		int i;

		show_node(zone);
		printk("%s"
			" free:%lukB"
			" min:%lukB"
			" low:%lukB"
			" high:%lukB"
			" active:%lukB"
			" inactive:%lukB"
			" present:%lukB"
			" pages_scanned:%lu"
			" all_unreclaimable? %s"
			"\n",
			zone->name,
			K(zone->free_pages),
			K(zone->pages_min),
			K(zone->pages_low),
			K(zone->pages_high),
			K(zone->nr_active),
			K(zone->nr_inactive),
			K(zone->present_pages),
			zone->pages_scanned,
			(zone->all_unreclaimable ? "yes" : "no")
			);
		printk("lowmem_reserve[]:");
		for (i = 0; i < MAX_NR_ZONES; i++)
			printk(" %lu", zone->lowmem_reserve[i]);
		printk("\n");
	}

	for_each_zone(zone) {
1314
 		unsigned long nr[MAX_ORDER], flags, order, total = 0;
L
Linus Torvalds 已提交
1315 1316 1317

		show_node(zone);
		printk("%s: ", zone->name);
1318
		if (!populated_zone(zone)) {
L
Linus Torvalds 已提交
1319 1320 1321 1322 1323 1324
			printk("empty\n");
			continue;
		}

		spin_lock_irqsave(&zone->lock, flags);
		for (order = 0; order < MAX_ORDER; order++) {
1325 1326
			nr[order] = zone->free_area[order].nr_free;
			total += nr[order] << order;
L
Linus Torvalds 已提交
1327 1328
		}
		spin_unlock_irqrestore(&zone->lock, flags);
1329 1330
		for (order = 0; order < MAX_ORDER; order++)
			printk("%lu*%lukB ", nr[order], K(1UL) << order);
L
Linus Torvalds 已提交
1331 1332 1333 1334 1335 1336 1337 1338
		printk("= %lukB\n", K(total));
	}

	show_swap_cache_info();
}

/*
 * Builds allocation fallback zone lists.
1339 1340
 *
 * Add all populated zones of a node to the zonelist.
L
Linus Torvalds 已提交
1341
 */
1342
static int __meminit build_zonelists_node(pg_data_t *pgdat,
1343
			struct zonelist *zonelist, int nr_zones, int zone_type)
L
Linus Torvalds 已提交
1344
{
1345 1346
	struct zone *zone;

1347
	BUG_ON(zone_type >= MAX_NR_ZONES);
1348 1349

	do {
1350
		zone = pgdat->node_zones + zone_type;
1351
		if (populated_zone(zone)) {
1352 1353
			zonelist->zones[nr_zones++] = zone;
			check_highest_zone(zone_type);
L
Linus Torvalds 已提交
1354
		}
1355
		zone_type--;
1356

1357 1358
	} while (zone_type >= 0);
	return nr_zones;
L
Linus Torvalds 已提交
1359 1360
}

A
Al Viro 已提交
1361 1362 1363 1364 1365
static inline int highest_zone(int zone_bits)
{
	int res = ZONE_NORMAL;
	if (zone_bits & (__force int)__GFP_HIGHMEM)
		res = ZONE_HIGHMEM;
A
Andi Kleen 已提交
1366 1367
	if (zone_bits & (__force int)__GFP_DMA32)
		res = ZONE_DMA32;
A
Al Viro 已提交
1368 1369 1370 1371 1372
	if (zone_bits & (__force int)__GFP_DMA)
		res = ZONE_DMA;
	return res;
}

L
Linus Torvalds 已提交
1373 1374
#ifdef CONFIG_NUMA
#define MAX_NODE_LOAD (num_online_nodes())
1375
static int __meminitdata node_load[MAX_NUMNODES];
L
Linus Torvalds 已提交
1376
/**
1377
 * find_next_best_node - find the next node that should appear in a given node's fallback list
L
Linus Torvalds 已提交
1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389
 * @node: node whose fallback list we're appending
 * @used_node_mask: nodemask_t of already used nodes
 *
 * We use a number of factors to determine which is the next node that should
 * appear on a given node's fallback list.  The node should not have appeared
 * already in @node's fallback list, and it should be the next closest node
 * according to the distance array (which contains arbitrary distance values
 * from each node to each node in the system), and should also prefer nodes
 * with no CPUs, since presumably they'll have very little allocation pressure
 * on them otherwise.
 * It returns -1 if no node is found.
 */
1390
static int __meminit find_next_best_node(int node, nodemask_t *used_node_mask)
L
Linus Torvalds 已提交
1391
{
1392
	int n, val;
L
Linus Torvalds 已提交
1393 1394 1395
	int min_val = INT_MAX;
	int best_node = -1;

1396 1397 1398 1399 1400
	/* Use the local node if we haven't already */
	if (!node_isset(node, *used_node_mask)) {
		node_set(node, *used_node_mask);
		return node;
	}
L
Linus Torvalds 已提交
1401

1402 1403
	for_each_online_node(n) {
		cpumask_t tmp;
L
Linus Torvalds 已提交
1404 1405 1406 1407 1408 1409 1410 1411

		/* Don't want a node to appear more than once */
		if (node_isset(n, *used_node_mask))
			continue;

		/* Use the distance array to find the distance */
		val = node_distance(node, n);

1412 1413 1414
		/* Penalize nodes under us ("prefer the next node") */
		val += (n < node);

L
Linus Torvalds 已提交
1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435
		/* Give preference to headless and unused nodes */
		tmp = node_to_cpumask(n);
		if (!cpus_empty(tmp))
			val += PENALTY_FOR_NODE_WITH_CPUS;

		/* Slight preference for less loaded node */
		val *= (MAX_NODE_LOAD*MAX_NUMNODES);
		val += node_load[n];

		if (val < min_val) {
			min_val = val;
			best_node = n;
		}
	}

	if (best_node >= 0)
		node_set(best_node, *used_node_mask);

	return best_node;
}

1436
static void __meminit build_zonelists(pg_data_t *pgdat)
L
Linus Torvalds 已提交
1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454
{
	int i, j, k, node, local_node;
	int prev_node, load;
	struct zonelist *zonelist;
	nodemask_t used_mask;

	/* initialize zonelists */
	for (i = 0; i < GFP_ZONETYPES; i++) {
		zonelist = pgdat->node_zonelists + i;
		zonelist->zones[0] = NULL;
	}

	/* NUMA-aware ordering of nodes */
	local_node = pgdat->node_id;
	load = num_online_nodes();
	prev_node = local_node;
	nodes_clear(used_mask);
	while ((node = find_next_best_node(local_node, &used_mask)) >= 0) {
1455 1456 1457 1458 1459 1460 1461 1462 1463
		int distance = node_distance(local_node, node);

		/*
		 * If another node is sufficiently far away then it is better
		 * to reclaim pages in a zone before going off node.
		 */
		if (distance > RECLAIM_DISTANCE)
			zone_reclaim_mode = 1;

L
Linus Torvalds 已提交
1464 1465 1466 1467 1468
		/*
		 * We don't want to pressure a particular node.
		 * So adding penalty to the first node in same
		 * distance group to make it round-robin.
		 */
1469 1470

		if (distance != node_distance(local_node, prev_node))
L
Linus Torvalds 已提交
1471 1472 1473 1474 1475 1476 1477
			node_load[node] += load;
		prev_node = node;
		load--;
		for (i = 0; i < GFP_ZONETYPES; i++) {
			zonelist = pgdat->node_zonelists + i;
			for (j = 0; zonelist->zones[j] != NULL; j++);

A
Al Viro 已提交
1478
			k = highest_zone(i);
L
Linus Torvalds 已提交
1479 1480 1481 1482 1483 1484 1485 1486 1487

	 		j = build_zonelists_node(NODE_DATA(node), zonelist, j, k);
			zonelist->zones[j] = NULL;
		}
	}
}

#else	/* CONFIG_NUMA */

1488
static void __meminit build_zonelists(pg_data_t *pgdat)
L
Linus Torvalds 已提交
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
{
	int i, j, k, node, local_node;

	local_node = pgdat->node_id;
	for (i = 0; i < GFP_ZONETYPES; i++) {
		struct zonelist *zonelist;

		zonelist = pgdat->node_zonelists + i;

		j = 0;
A
Al Viro 已提交
1499
		k = highest_zone(i);
L
Linus Torvalds 已提交
1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525
 		j = build_zonelists_node(pgdat, zonelist, j, k);
 		/*
 		 * Now we build the zonelist so that it contains the zones
 		 * of all the other nodes.
 		 * We don't want to pressure a particular node, so when
 		 * building the zones for node N, we make sure that the
 		 * zones coming right after the local ones are those from
 		 * node N+1 (modulo N)
 		 */
		for (node = local_node + 1; node < MAX_NUMNODES; node++) {
			if (!node_online(node))
				continue;
			j = build_zonelists_node(NODE_DATA(node), zonelist, j, k);
		}
		for (node = 0; node < local_node; node++) {
			if (!node_online(node))
				continue;
			j = build_zonelists_node(NODE_DATA(node), zonelist, j, k);
		}

		zonelist->zones[j] = NULL;
	}
}

#endif	/* CONFIG_NUMA */

1526 1527
/* return values int ....just for stop_machine_run() */
static int __meminit __build_all_zonelists(void *dummy)
L
Linus Torvalds 已提交
1528
{
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
	int nid;
	for_each_online_node(nid)
		build_zonelists(NODE_DATA(nid));
	return 0;
}

void __meminit build_all_zonelists(void)
{
	if (system_state == SYSTEM_BOOTING) {
		__build_all_zonelists(0);
		cpuset_init_current_mems_allowed();
	} else {
		/* we have to stop all cpus to guaranntee there is no user
		   of zonelist */
		stop_machine_run(__build_all_zonelists, NULL, NR_CPUS);
		/* cpuset refresh routine should be here */
	}
1546 1547 1548
	vm_total_pages = nr_free_pagecache_pages();
	printk("Built %i zonelists.  Total pages: %ld\n",
			num_online_nodes(), vm_total_pages);
L
Linus Torvalds 已提交
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563
}

/*
 * Helper functions to size the waitqueue hash table.
 * Essentially these want to choose hash table sizes sufficiently
 * large so that collisions trying to wait on pages are rare.
 * But in fact, the number of active page waitqueues on typical
 * systems is ridiculously low, less than 200. So this is even
 * conservative, even though it seems large.
 *
 * The constant PAGES_PER_WAITQUEUE specifies the ratio of pages to
 * waitqueues, i.e. the size of the waitq table given the number of pages.
 */
#define PAGES_PER_WAITQUEUE	256

1564
#ifndef CONFIG_MEMORY_HOTPLUG
1565
static inline unsigned long wait_table_hash_nr_entries(unsigned long pages)
L
Linus Torvalds 已提交
1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582
{
	unsigned long size = 1;

	pages /= PAGES_PER_WAITQUEUE;

	while (size < pages)
		size <<= 1;

	/*
	 * Once we have dozens or even hundreds of threads sleeping
	 * on IO we've got bigger problems than wait queue collision.
	 * Limit the size of the wait table to a reasonable size.
	 */
	size = min(size, 4096UL);

	return max(size, 4UL);
}
1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605
#else
/*
 * A zone's size might be changed by hot-add, so it is not possible to determine
 * a suitable size for its wait_table.  So we use the maximum size now.
 *
 * The max wait table size = 4096 x sizeof(wait_queue_head_t).   ie:
 *
 *    i386 (preemption config)    : 4096 x 16 = 64Kbyte.
 *    ia64, x86-64 (no preemption): 4096 x 20 = 80Kbyte.
 *    ia64, x86-64 (preemption)   : 4096 x 24 = 96Kbyte.
 *
 * The maximum entries are prepared when a zone's memory is (512K + 256) pages
 * or more by the traditional way. (See above).  It equals:
 *
 *    i386, x86-64, powerpc(4K page size) : =  ( 2G + 1M)byte.
 *    ia64(16K page size)                 : =  ( 8G + 4M)byte.
 *    powerpc (64K page size)             : =  (32G +16M)byte.
 */
static inline unsigned long wait_table_hash_nr_entries(unsigned long pages)
{
	return 4096UL;
}
#endif
L
Linus Torvalds 已提交
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642

/*
 * This is an integer logarithm so that shifts can be used later
 * to extract the more random high bits from the multiplicative
 * hash function before the remainder is taken.
 */
static inline unsigned long wait_table_bits(unsigned long size)
{
	return ffz(~size);
}

#define LONG_ALIGN(x) (((x)+(sizeof(long))-1)&~((sizeof(long))-1))

static void __init calculate_zone_totalpages(struct pglist_data *pgdat,
		unsigned long *zones_size, unsigned long *zholes_size)
{
	unsigned long realtotalpages, totalpages = 0;
	int i;

	for (i = 0; i < MAX_NR_ZONES; i++)
		totalpages += zones_size[i];
	pgdat->node_spanned_pages = totalpages;

	realtotalpages = totalpages;
	if (zholes_size)
		for (i = 0; i < MAX_NR_ZONES; i++)
			realtotalpages -= zholes_size[i];
	pgdat->node_present_pages = realtotalpages;
	printk(KERN_DEBUG "On node %d totalpages: %lu\n", pgdat->node_id, realtotalpages);
}


/*
 * Initially all pages are reserved - free ones are freed
 * up by free_all_bootmem() once the early boot process is
 * done. Non-atomic initialization, single-pass.
 */
1643
void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,
L
Linus Torvalds 已提交
1644 1645 1646
		unsigned long start_pfn)
{
	struct page *page;
A
Andy Whitcroft 已提交
1647 1648
	unsigned long end_pfn = start_pfn + size;
	unsigned long pfn;
L
Linus Torvalds 已提交
1649

1650
	for (pfn = start_pfn; pfn < end_pfn; pfn++) {
A
Andy Whitcroft 已提交
1651 1652 1653 1654
		if (!early_pfn_valid(pfn))
			continue;
		page = pfn_to_page(pfn);
		set_page_links(page, zone, nid, pfn);
1655
		init_page_count(page);
L
Linus Torvalds 已提交
1656 1657 1658 1659 1660 1661
		reset_page_mapcount(page);
		SetPageReserved(page);
		INIT_LIST_HEAD(&page->lru);
#ifdef WANT_PAGE_VIRTUAL
		/* The shift won't overflow because ZONE_NORMAL is below 4G. */
		if (!is_highmem_idx(zone))
1662
			set_page_address(page, __va(pfn << PAGE_SHIFT));
L
Linus Torvalds 已提交
1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676
#endif
	}
}

void zone_init_free_lists(struct pglist_data *pgdat, struct zone *zone,
				unsigned long size)
{
	int order;
	for (order = 0; order < MAX_ORDER ; order++) {
		INIT_LIST_HEAD(&zone->free_area[order].free_list);
		zone->free_area[order].nr_free = 0;
	}
}

A
Andy Whitcroft 已提交
1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690
#define ZONETABLE_INDEX(x, zone_nr)	((x << ZONES_SHIFT) | zone_nr)
void zonetable_add(struct zone *zone, int nid, int zid, unsigned long pfn,
		unsigned long size)
{
	unsigned long snum = pfn_to_section_nr(pfn);
	unsigned long end = pfn_to_section_nr(pfn + size);

	if (FLAGS_HAS_NODE)
		zone_table[ZONETABLE_INDEX(nid, zid)] = zone;
	else
		for (; snum <= end; snum++)
			zone_table[ZONETABLE_INDEX(snum, zid)] = zone;
}

L
Linus Torvalds 已提交
1691 1692 1693 1694 1695
#ifndef __HAVE_ARCH_MEMMAP_INIT
#define memmap_init(size, nid, zone, start_pfn) \
	memmap_init_zone((size), (nid), (zone), (start_pfn))
#endif

1696
static int __cpuinit zone_batchsize(struct zone *zone)
1697 1698 1699 1700 1701
{
	int batch;

	/*
	 * The per-cpu-pages pools are set to around 1000th of the
1702
	 * size of the zone.  But no more than 1/2 of a meg.
1703 1704 1705 1706
	 *
	 * OK, so we don't know how big the cache is.  So guess.
	 */
	batch = zone->present_pages / 1024;
1707 1708
	if (batch * PAGE_SIZE > 512 * 1024)
		batch = (512 * 1024) / PAGE_SIZE;
1709 1710 1711 1712 1713
	batch /= 4;		/* We effectively *= 4 below */
	if (batch < 1)
		batch = 1;

	/*
1714 1715 1716
	 * Clamp the batch to a 2^n - 1 value. Having a power
	 * of 2 value was found to be more likely to have
	 * suboptimal cache aliasing properties in some cases.
1717
	 *
1718 1719 1720 1721
	 * For example if 2 tasks are alternately allocating
	 * batches of pages, one task can end up with a lot
	 * of pages of one half of the possible page colors
	 * and the other with pages of the other colors.
1722
	 */
1723
	batch = (1 << (fls(batch + batch/2)-1)) - 1;
1724

1725 1726 1727
	return batch;
}

1728 1729 1730 1731
inline void setup_pageset(struct per_cpu_pageset *p, unsigned long batch)
{
	struct per_cpu_pages *pcp;

1732 1733
	memset(p, 0, sizeof(*p));

1734 1735 1736 1737 1738 1739 1740 1741 1742
	pcp = &p->pcp[0];		/* hot */
	pcp->count = 0;
	pcp->high = 6 * batch;
	pcp->batch = max(1UL, 1 * batch);
	INIT_LIST_HEAD(&pcp->list);

	pcp = &p->pcp[1];		/* cold*/
	pcp->count = 0;
	pcp->high = 2 * batch;
1743
	pcp->batch = max(1UL, batch/2);
1744 1745 1746
	INIT_LIST_HEAD(&pcp->list);
}

1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764
/*
 * setup_pagelist_highmark() sets the high water mark for hot per_cpu_pagelist
 * to the value high for the pageset p.
 */

static void setup_pagelist_highmark(struct per_cpu_pageset *p,
				unsigned long high)
{
	struct per_cpu_pages *pcp;

	pcp = &p->pcp[0]; /* hot list */
	pcp->high = high;
	pcp->batch = max(1UL, high/4);
	if ((high/4) > (PAGE_SHIFT * 8))
		pcp->batch = PAGE_SHIFT * 8;
}


1765 1766
#ifdef CONFIG_NUMA
/*
1767 1768 1769 1770 1771 1772 1773
 * Boot pageset table. One per cpu which is going to be used for all
 * zones and all nodes. The parameters will be set in such a way
 * that an item put on a list will immediately be handed over to
 * the buddy list. This is safe since pageset manipulation is done
 * with interrupts disabled.
 *
 * Some NUMA counter updates may also be caught by the boot pagesets.
1774 1775 1776 1777 1778 1779 1780 1781
 *
 * The boot_pagesets must be kept even after bootup is complete for
 * unused processors and/or zones. They do play a role for bootstrapping
 * hotplugged processors.
 *
 * zoneinfo_show() and maybe other functions do
 * not check if the processor is online before following the pageset pointer.
 * Other parts of the kernel may not check if the zone is available.
1782
 */
1783
static struct per_cpu_pageset boot_pageset[NR_CPUS];
1784 1785 1786

/*
 * Dynamically allocate memory for the
1787 1788
 * per cpu pageset array in struct zone.
 */
1789
static int __cpuinit process_zones(int cpu)
1790 1791 1792 1793 1794
{
	struct zone *zone, *dzone;

	for_each_zone(zone) {

N
Nick Piggin 已提交
1795
		zone_pcp(zone, cpu) = kmalloc_node(sizeof(struct per_cpu_pageset),
1796
					 GFP_KERNEL, cpu_to_node(cpu));
N
Nick Piggin 已提交
1797
		if (!zone_pcp(zone, cpu))
1798 1799
			goto bad;

N
Nick Piggin 已提交
1800
		setup_pageset(zone_pcp(zone, cpu), zone_batchsize(zone));
1801 1802 1803 1804

		if (percpu_pagelist_fraction)
			setup_pagelist_highmark(zone_pcp(zone, cpu),
			 	(zone->present_pages / percpu_pagelist_fraction));
1805 1806 1807 1808 1809 1810 1811
	}

	return 0;
bad:
	for_each_zone(dzone) {
		if (dzone == zone)
			break;
N
Nick Piggin 已提交
1812 1813
		kfree(zone_pcp(dzone, cpu));
		zone_pcp(dzone, cpu) = NULL;
1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824
	}
	return -ENOMEM;
}

static inline void free_zone_pagesets(int cpu)
{
	struct zone *zone;

	for_each_zone(zone) {
		struct per_cpu_pageset *pset = zone_pcp(zone, cpu);

1825 1826 1827
		/* Free per_cpu_pageset if it is slab allocated */
		if (pset != &boot_pageset[cpu])
			kfree(pset);
1828 1829 1830 1831
		zone_pcp(zone, cpu) = NULL;
	}
}

1832
static int __cpuinit pageset_cpuup_callback(struct notifier_block *nfb,
1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843
		unsigned long action,
		void *hcpu)
{
	int cpu = (long)hcpu;
	int ret = NOTIFY_OK;

	switch (action) {
		case CPU_UP_PREPARE:
			if (process_zones(cpu))
				ret = NOTIFY_BAD;
			break;
1844
		case CPU_UP_CANCELED:
1845 1846 1847 1848 1849 1850 1851 1852 1853
		case CPU_DEAD:
			free_zone_pagesets(cpu);
			break;
		default:
			break;
	}
	return ret;
}

1854
static struct notifier_block __cpuinitdata pageset_notifier =
1855 1856
	{ &pageset_cpuup_callback, NULL, 0 };

1857
void __init setup_per_cpu_pageset(void)
1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871
{
	int err;

	/* Initialize per_cpu_pageset for cpu 0.
	 * A cpuup callback will do this for every cpu
	 * as it comes online
	 */
	err = process_zones(smp_processor_id());
	BUG_ON(err);
	register_cpu_notifier(&pageset_notifier);
}

#endif

1872
static __meminit
1873
int zone_wait_table_init(struct zone *zone, unsigned long zone_size_pages)
1874 1875 1876
{
	int i;
	struct pglist_data *pgdat = zone->zone_pgdat;
1877
	size_t alloc_size;
1878 1879 1880 1881 1882

	/*
	 * The per-page waitqueue mechanism uses hashed waitqueues
	 * per zone.
	 */
1883 1884 1885 1886
	zone->wait_table_hash_nr_entries =
		 wait_table_hash_nr_entries(zone_size_pages);
	zone->wait_table_bits =
		wait_table_bits(zone->wait_table_hash_nr_entries);
1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907
	alloc_size = zone->wait_table_hash_nr_entries
					* sizeof(wait_queue_head_t);

 	if (system_state == SYSTEM_BOOTING) {
		zone->wait_table = (wait_queue_head_t *)
			alloc_bootmem_node(pgdat, alloc_size);
	} else {
		/*
		 * This case means that a zone whose size was 0 gets new memory
		 * via memory hot-add.
		 * But it may be the case that a new node was hot-added.  In
		 * this case vmalloc() will not be able to use this new node's
		 * memory - this wait_table must be initialized to use this new
		 * node itself as well.
		 * To use this new node's memory, further consideration will be
		 * necessary.
		 */
		zone->wait_table = (wait_queue_head_t *)vmalloc(alloc_size);
	}
	if (!zone->wait_table)
		return -ENOMEM;
1908

1909
	for(i = 0; i < zone->wait_table_hash_nr_entries; ++i)
1910
		init_waitqueue_head(zone->wait_table + i);
1911 1912

	return 0;
1913 1914
}

1915
static __meminit void zone_pcp_init(struct zone *zone)
1916 1917 1918 1919 1920 1921 1922
{
	int cpu;
	unsigned long batch = zone_batchsize(zone);

	for (cpu = 0; cpu < NR_CPUS; cpu++) {
#ifdef CONFIG_NUMA
		/* Early boot. Slab allocator not functional yet */
N
Nick Piggin 已提交
1923
		zone_pcp(zone, cpu) = &boot_pageset[cpu];
1924 1925 1926 1927 1928
		setup_pageset(&boot_pageset[cpu],0);
#else
		setup_pageset(zone_pcp(zone,cpu), batch);
#endif
	}
A
Anton Blanchard 已提交
1929 1930 1931
	if (zone->present_pages)
		printk(KERN_DEBUG "  %s zone: %lu pages, LIFO batch:%lu\n",
			zone->name, zone->present_pages, batch);
1932 1933
}

1934 1935 1936
__meminit int init_currently_empty_zone(struct zone *zone,
					unsigned long zone_start_pfn,
					unsigned long size)
1937 1938
{
	struct pglist_data *pgdat = zone->zone_pgdat;
1939 1940 1941 1942
	int ret;
	ret = zone_wait_table_init(zone, size);
	if (ret)
		return ret;
1943 1944 1945 1946 1947 1948 1949
	pgdat->nr_zones = zone_idx(zone) + 1;

	zone->zone_start_pfn = zone_start_pfn;

	memmap_init(size, pgdat->node_id, zone_idx(zone), zone_start_pfn);

	zone_init_free_lists(pgdat, zone, zone->spanned_pages);
1950 1951

	return 0;
1952 1953
}

L
Linus Torvalds 已提交
1954 1955 1956 1957 1958 1959
/*
 * Set up the zone data structures:
 *   - mark all pages reserved
 *   - mark all memory queues empty
 *   - clear the memory bitmaps
 */
1960
static void __meminit free_area_init_core(struct pglist_data *pgdat,
L
Linus Torvalds 已提交
1961 1962
		unsigned long *zones_size, unsigned long *zholes_size)
{
1963 1964
	unsigned long j;
	int nid = pgdat->node_id;
L
Linus Torvalds 已提交
1965
	unsigned long zone_start_pfn = pgdat->node_start_pfn;
1966
	int ret;
L
Linus Torvalds 已提交
1967

1968
	pgdat_resize_init(pgdat);
L
Linus Torvalds 已提交
1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980
	pgdat->nr_zones = 0;
	init_waitqueue_head(&pgdat->kswapd_wait);
	pgdat->kswapd_max_order = 0;
	
	for (j = 0; j < MAX_NR_ZONES; j++) {
		struct zone *zone = pgdat->node_zones + j;
		unsigned long size, realsize;

		realsize = size = zones_size[j];
		if (zholes_size)
			realsize -= zholes_size[j];

1981
		if (!is_highmem_idx(j))
L
Linus Torvalds 已提交
1982 1983 1984 1985 1986
			nr_kernel_pages += realsize;
		nr_all_pages += realsize;

		zone->spanned_pages = size;
		zone->present_pages = realsize;
1987 1988 1989 1990
#ifdef CONFIG_NUMA
		zone->min_unmapped_ratio = (realsize*sysctl_min_unmapped_ratio)
						/ 100;
#endif
L
Linus Torvalds 已提交
1991 1992 1993
		zone->name = zone_names[j];
		spin_lock_init(&zone->lock);
		spin_lock_init(&zone->lru_lock);
1994
		zone_seqlock_init(zone);
L
Linus Torvalds 已提交
1995 1996 1997 1998 1999
		zone->zone_pgdat = pgdat;
		zone->free_pages = 0;

		zone->temp_priority = zone->prev_priority = DEF_PRIORITY;

2000
		zone_pcp_init(zone);
L
Linus Torvalds 已提交
2001 2002 2003 2004 2005 2006
		INIT_LIST_HEAD(&zone->active_list);
		INIT_LIST_HEAD(&zone->inactive_list);
		zone->nr_scan_active = 0;
		zone->nr_scan_inactive = 0;
		zone->nr_active = 0;
		zone->nr_inactive = 0;
2007
		zap_zone_vm_stats(zone);
2008
		atomic_set(&zone->reclaim_in_progress, 0);
L
Linus Torvalds 已提交
2009 2010 2011
		if (!size)
			continue;

A
Andy Whitcroft 已提交
2012
		zonetable_add(zone, nid, j, zone_start_pfn, size);
2013 2014
		ret = init_currently_empty_zone(zone, zone_start_pfn, size);
		BUG_ON(ret);
L
Linus Torvalds 已提交
2015 2016 2017 2018 2019 2020 2021 2022 2023 2024
		zone_start_pfn += size;
	}
}

static void __init alloc_node_mem_map(struct pglist_data *pgdat)
{
	/* Skip empty nodes */
	if (!pgdat->node_spanned_pages)
		return;

A
Andy Whitcroft 已提交
2025
#ifdef CONFIG_FLAT_NODE_MEM_MAP
L
Linus Torvalds 已提交
2026 2027
	/* ia64 gets its own node_mem_map, before this, without bootmem */
	if (!pgdat->node_mem_map) {
2028
		unsigned long size, start, end;
A
Andy Whitcroft 已提交
2029 2030
		struct page *map;

2031 2032 2033 2034 2035 2036 2037 2038 2039
		/*
		 * The zone's endpoints aren't required to be MAX_ORDER
		 * aligned but the node_mem_map endpoints must be in order
		 * for the buddy allocator to function correctly.
		 */
		start = pgdat->node_start_pfn & ~(MAX_ORDER_NR_PAGES - 1);
		end = pgdat->node_start_pfn + pgdat->node_spanned_pages;
		end = ALIGN(end, MAX_ORDER_NR_PAGES);
		size =  (end - start) * sizeof(struct page);
2040 2041 2042
		map = alloc_remap(pgdat->node_id, size);
		if (!map)
			map = alloc_bootmem_node(pgdat, size);
2043
		pgdat->node_mem_map = map + (pgdat->node_start_pfn - start);
L
Linus Torvalds 已提交
2044
	}
A
Andy Whitcroft 已提交
2045
#ifdef CONFIG_FLATMEM
L
Linus Torvalds 已提交
2046 2047 2048 2049 2050 2051
	/*
	 * With no DISCONTIG, the global mem_map is just set as node 0's
	 */
	if (pgdat == NODE_DATA(0))
		mem_map = NODE_DATA(0)->node_mem_map;
#endif
A
Andy Whitcroft 已提交
2052
#endif /* CONFIG_FLAT_NODE_MEM_MAP */
L
Linus Torvalds 已提交
2053 2054
}

2055
void __meminit free_area_init_node(int nid, struct pglist_data *pgdat,
L
Linus Torvalds 已提交
2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067
		unsigned long *zones_size, unsigned long node_start_pfn,
		unsigned long *zholes_size)
{
	pgdat->node_id = nid;
	pgdat->node_start_pfn = node_start_pfn;
	calculate_zone_totalpages(pgdat, zones_size, zholes_size);

	alloc_node_mem_map(pgdat);

	free_area_init_core(pgdat, zones_size, zholes_size);
}

2068
#ifndef CONFIG_NEED_MULTIPLE_NODES
L
Linus Torvalds 已提交
2069 2070 2071 2072
static bootmem_data_t contig_bootmem_data;
struct pglist_data contig_page_data = { .bdata = &contig_bootmem_data };

EXPORT_SYMBOL(contig_page_data);
2073
#endif
L
Linus Torvalds 已提交
2074 2075 2076

void __init free_area_init(unsigned long *zones_size)
{
2077
	free_area_init_node(0, NODE_DATA(0), zones_size,
L
Linus Torvalds 已提交
2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089
			__pa(PAGE_OFFSET) >> PAGE_SHIFT, NULL);
}

#ifdef CONFIG_HOTPLUG_CPU
static int page_alloc_cpu_notify(struct notifier_block *self,
				 unsigned long action, void *hcpu)
{
	int cpu = (unsigned long)hcpu;

	if (action == CPU_DEAD) {
		local_irq_disable();
		__drain_pages(cpu);
2090
		vm_events_fold_cpu(cpu);
L
Linus Torvalds 已提交
2091
		local_irq_enable();
2092
		refresh_cpu_vm_stats(cpu);
L
Linus Torvalds 已提交
2093 2094 2095 2096 2097 2098 2099 2100 2101 2102
	}
	return NOTIFY_OK;
}
#endif /* CONFIG_HOTPLUG_CPU */

void __init page_alloc_init(void)
{
	hotcpu_notifier(page_alloc_cpu_notify, 0);
}

2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134
/*
 * calculate_totalreserve_pages - called when sysctl_lower_zone_reserve_ratio
 *	or min_free_kbytes changes.
 */
static void calculate_totalreserve_pages(void)
{
	struct pglist_data *pgdat;
	unsigned long reserve_pages = 0;
	int i, j;

	for_each_online_pgdat(pgdat) {
		for (i = 0; i < MAX_NR_ZONES; i++) {
			struct zone *zone = pgdat->node_zones + i;
			unsigned long max = 0;

			/* Find valid and maximum lowmem_reserve in the zone */
			for (j = i; j < MAX_NR_ZONES; j++) {
				if (zone->lowmem_reserve[j] > max)
					max = zone->lowmem_reserve[j];
			}

			/* we treat pages_high as reserved pages. */
			max += zone->pages_high;

			if (max > zone->present_pages)
				max = zone->present_pages;
			reserve_pages += max;
		}
	}
	totalreserve_pages = reserve_pages;
}

L
Linus Torvalds 已提交
2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145
/*
 * setup_per_zone_lowmem_reserve - called whenever
 *	sysctl_lower_zone_reserve_ratio changes.  Ensures that each zone
 *	has a correct pages reserved value, so an adequate number of
 *	pages are left in the zone after a successful __alloc_pages().
 */
static void setup_per_zone_lowmem_reserve(void)
{
	struct pglist_data *pgdat;
	int j, idx;

2146
	for_each_online_pgdat(pgdat) {
L
Linus Torvalds 已提交
2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165
		for (j = 0; j < MAX_NR_ZONES; j++) {
			struct zone *zone = pgdat->node_zones + j;
			unsigned long present_pages = zone->present_pages;

			zone->lowmem_reserve[j] = 0;

			for (idx = j-1; idx >= 0; idx--) {
				struct zone *lower_zone;

				if (sysctl_lowmem_reserve_ratio[idx] < 1)
					sysctl_lowmem_reserve_ratio[idx] = 1;

				lower_zone = pgdat->node_zones + idx;
				lower_zone->lowmem_reserve[j] = present_pages /
					sysctl_lowmem_reserve_ratio[idx];
				present_pages += lower_zone->present_pages;
			}
		}
	}
2166 2167 2168

	/* update totalreserve_pages */
	calculate_totalreserve_pages();
L
Linus Torvalds 已提交
2169 2170 2171 2172 2173 2174 2175
}

/*
 * setup_per_zone_pages_min - called when min_free_kbytes changes.  Ensures 
 *	that the pages_{min,low,high} values for each zone are set correctly 
 *	with respect to min_free_kbytes.
 */
2176
void setup_per_zone_pages_min(void)
L
Linus Torvalds 已提交
2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189
{
	unsigned long pages_min = min_free_kbytes >> (PAGE_SHIFT - 10);
	unsigned long lowmem_pages = 0;
	struct zone *zone;
	unsigned long flags;

	/* Calculate total number of !ZONE_HIGHMEM pages */
	for_each_zone(zone) {
		if (!is_highmem(zone))
			lowmem_pages += zone->present_pages;
	}

	for_each_zone(zone) {
2190 2191
		u64 tmp;

L
Linus Torvalds 已提交
2192
		spin_lock_irqsave(&zone->lru_lock, flags);
2193 2194
		tmp = (u64)pages_min * zone->present_pages;
		do_div(tmp, lowmem_pages);
L
Linus Torvalds 已提交
2195 2196
		if (is_highmem(zone)) {
			/*
N
Nick Piggin 已提交
2197 2198 2199 2200 2201 2202 2203
			 * __GFP_HIGH and PF_MEMALLOC allocations usually don't
			 * need highmem pages, so cap pages_min to a small
			 * value here.
			 *
			 * The (pages_high-pages_low) and (pages_low-pages_min)
			 * deltas controls asynch page reclaim, and so should
			 * not be capped for highmem.
L
Linus Torvalds 已提交
2204 2205 2206 2207 2208 2209 2210 2211 2212 2213
			 */
			int min_pages;

			min_pages = zone->present_pages / 1024;
			if (min_pages < SWAP_CLUSTER_MAX)
				min_pages = SWAP_CLUSTER_MAX;
			if (min_pages > 128)
				min_pages = 128;
			zone->pages_min = min_pages;
		} else {
N
Nick Piggin 已提交
2214 2215
			/*
			 * If it's a lowmem zone, reserve a number of pages
L
Linus Torvalds 已提交
2216 2217
			 * proportionate to the zone's size.
			 */
N
Nick Piggin 已提交
2218
			zone->pages_min = tmp;
L
Linus Torvalds 已提交
2219 2220
		}

2221 2222
		zone->pages_low   = zone->pages_min + (tmp >> 2);
		zone->pages_high  = zone->pages_min + (tmp >> 1);
L
Linus Torvalds 已提交
2223 2224
		spin_unlock_irqrestore(&zone->lru_lock, flags);
	}
2225 2226 2227

	/* update totalreserve_pages */
	calculate_totalreserve_pages();
L
Linus Torvalds 已提交
2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283
}

/*
 * Initialise min_free_kbytes.
 *
 * For small machines we want it small (128k min).  For large machines
 * we want it large (64MB max).  But it is not linear, because network
 * bandwidth does not increase linearly with machine size.  We use
 *
 * 	min_free_kbytes = 4 * sqrt(lowmem_kbytes), for better accuracy:
 *	min_free_kbytes = sqrt(lowmem_kbytes * 16)
 *
 * which yields
 *
 * 16MB:	512k
 * 32MB:	724k
 * 64MB:	1024k
 * 128MB:	1448k
 * 256MB:	2048k
 * 512MB:	2896k
 * 1024MB:	4096k
 * 2048MB:	5792k
 * 4096MB:	8192k
 * 8192MB:	11584k
 * 16384MB:	16384k
 */
static int __init init_per_zone_pages_min(void)
{
	unsigned long lowmem_kbytes;

	lowmem_kbytes = nr_free_buffer_pages() * (PAGE_SIZE >> 10);

	min_free_kbytes = int_sqrt(lowmem_kbytes * 16);
	if (min_free_kbytes < 128)
		min_free_kbytes = 128;
	if (min_free_kbytes > 65536)
		min_free_kbytes = 65536;
	setup_per_zone_pages_min();
	setup_per_zone_lowmem_reserve();
	return 0;
}
module_init(init_per_zone_pages_min)

/*
 * min_free_kbytes_sysctl_handler - just a wrapper around proc_dointvec() so 
 *	that we can call two helper functions whenever min_free_kbytes
 *	changes.
 */
int min_free_kbytes_sysctl_handler(ctl_table *table, int write, 
	struct file *file, void __user *buffer, size_t *length, loff_t *ppos)
{
	proc_dointvec(table, write, file, buffer, length, ppos);
	setup_per_zone_pages_min();
	return 0;
}

2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301
#ifdef CONFIG_NUMA
int sysctl_min_unmapped_ratio_sysctl_handler(ctl_table *table, int write,
	struct file *file, void __user *buffer, size_t *length, loff_t *ppos)
{
	struct zone *zone;
	int rc;

	rc = proc_dointvec_minmax(table, write, file, buffer, length, ppos);
	if (rc)
		return rc;

	for_each_zone(zone)
		zone->min_unmapped_ratio = (zone->present_pages *
				sysctl_min_unmapped_ratio) / 100;
	return 0;
}
#endif

L
Linus Torvalds 已提交
2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318
/*
 * lowmem_reserve_ratio_sysctl_handler - just a wrapper around
 *	proc_dointvec() so that we can call setup_per_zone_lowmem_reserve()
 *	whenever sysctl_lowmem_reserve_ratio changes.
 *
 * The reserve ratio obviously has absolutely no relation with the
 * pages_min watermarks. The lowmem reserve ratio can only make sense
 * if in function of the boot time zone sizes.
 */
int lowmem_reserve_ratio_sysctl_handler(ctl_table *table, int write,
	struct file *file, void __user *buffer, size_t *length, loff_t *ppos)
{
	proc_dointvec_minmax(table, write, file, buffer, length, ppos);
	setup_per_zone_lowmem_reserve();
	return 0;
}

2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344
/*
 * percpu_pagelist_fraction - changes the pcp->high for each zone on each
 * cpu.  It is the fraction of total pages in each zone that a hot per cpu pagelist
 * can have before it gets flushed back to buddy allocator.
 */

int percpu_pagelist_fraction_sysctl_handler(ctl_table *table, int write,
	struct file *file, void __user *buffer, size_t *length, loff_t *ppos)
{
	struct zone *zone;
	unsigned int cpu;
	int ret;

	ret = proc_dointvec_minmax(table, write, file, buffer, length, ppos);
	if (!write || (ret == -EINVAL))
		return ret;
	for_each_zone(zone) {
		for_each_online_cpu(cpu) {
			unsigned long  high;
			high = zone->present_pages / percpu_pagelist_fraction;
			setup_pagelist_highmark(zone_pcp(zone, cpu), high);
		}
	}
	return 0;
}

2345
int hashdist = HASHDIST_DEFAULT;
L
Linus Torvalds 已提交
2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390

#ifdef CONFIG_NUMA
static int __init set_hashdist(char *str)
{
	if (!str)
		return 0;
	hashdist = simple_strtoul(str, &str, 0);
	return 1;
}
__setup("hashdist=", set_hashdist);
#endif

/*
 * allocate a large system hash table from bootmem
 * - it is assumed that the hash table must contain an exact power-of-2
 *   quantity of entries
 * - limit is the number of hash buckets, not the total allocation size
 */
void *__init alloc_large_system_hash(const char *tablename,
				     unsigned long bucketsize,
				     unsigned long numentries,
				     int scale,
				     int flags,
				     unsigned int *_hash_shift,
				     unsigned int *_hash_mask,
				     unsigned long limit)
{
	unsigned long long max = limit;
	unsigned long log2qty, size;
	void *table = NULL;

	/* allow the kernel cmdline to have a say */
	if (!numentries) {
		/* round applicable memory size up to nearest megabyte */
		numentries = (flags & HASH_HIGHMEM) ? nr_all_pages : nr_kernel_pages;
		numentries += (1UL << (20 - PAGE_SHIFT)) - 1;
		numentries >>= 20 - PAGE_SHIFT;
		numentries <<= 20 - PAGE_SHIFT;

		/* limit to 1 bucket per 2^scale bytes of low memory */
		if (scale > PAGE_SHIFT)
			numentries >>= (scale - PAGE_SHIFT);
		else
			numentries <<= (PAGE_SHIFT - scale);
	}
2391
	numentries = roundup_pow_of_two(numentries);
L
Linus Torvalds 已提交
2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433

	/* limit allocation size to 1/16 total memory by default */
	if (max == 0) {
		max = ((unsigned long long)nr_all_pages << PAGE_SHIFT) >> 4;
		do_div(max, bucketsize);
	}

	if (numentries > max)
		numentries = max;

	log2qty = long_log2(numentries);

	do {
		size = bucketsize << log2qty;
		if (flags & HASH_EARLY)
			table = alloc_bootmem(size);
		else if (hashdist)
			table = __vmalloc(size, GFP_ATOMIC, PAGE_KERNEL);
		else {
			unsigned long order;
			for (order = 0; ((1UL << order) << PAGE_SHIFT) < size; order++)
				;
			table = (void*) __get_free_pages(GFP_ATOMIC, order);
		}
	} while (!table && size > PAGE_SIZE && --log2qty);

	if (!table)
		panic("Failed to allocate %s hash table\n", tablename);

	printk("%s hash table entries: %d (order: %d, %lu bytes)\n",
	       tablename,
	       (1U << log2qty),
	       long_log2(size) - PAGE_SHIFT,
	       size);

	if (_hash_shift)
		*_hash_shift = log2qty;
	if (_hash_mask)
		*_hash_mask = (1 << log2qty) - 1;

	return table;
}
2434 2435 2436 2437

#ifdef CONFIG_OUT_OF_LINE_PFN_TO_PAGE
struct page *pfn_to_page(unsigned long pfn)
{
2438
	return __pfn_to_page(pfn);
2439 2440 2441
}
unsigned long page_to_pfn(struct page *page)
{
2442
	return __page_to_pfn(page);
2443 2444 2445 2446
}
EXPORT_SYMBOL(pfn_to_page);
EXPORT_SYMBOL(page_to_pfn);
#endif /* CONFIG_OUT_OF_LINE_PFN_TO_PAGE */