page_alloc.c 91.1 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>
40 41
#include <linux/sort.h>
#include <linux/pfn.h>
42
#include <linux/backing-dev.h>
43
#include <linux/fault-inject.h>
L
Linus Torvalds 已提交
44 45

#include <asm/tlbflush.h>
46
#include <asm/div64.h>
L
Linus Torvalds 已提交
47 48 49 50 51 52
#include "internal.h"

/*
 * MCD - HACK: Find somewhere to initialize this EARLY, or make this
 * initializer cleaner
 */
53
nodemask_t node_online_map __read_mostly = { { [0] = 1UL } };
54
EXPORT_SYMBOL(node_online_map);
55
nodemask_t node_possible_map __read_mostly = NODE_MASK_ALL;
56
EXPORT_SYMBOL(node_possible_map);
57
unsigned long totalram_pages __read_mostly;
58
unsigned long totalreserve_pages __read_mostly;
L
Linus Torvalds 已提交
59
long nr_swap_pages;
60
int percpu_pagelist_fraction;
L
Linus Torvalds 已提交
61

62
static void __free_pages_ok(struct page *page, unsigned int order);
63

L
Linus Torvalds 已提交
64 65 66 67 68 69 70
/*
 * 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
71 72 73
 *
 * 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
74
 */
75
int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES-1] = {
76
#ifdef CONFIG_ZONE_DMA
77
	 256,
78
#endif
79
#ifdef CONFIG_ZONE_DMA32
80
	 256,
81
#endif
82
#ifdef CONFIG_HIGHMEM
83
	 32
84
#endif
85
};
L
Linus Torvalds 已提交
86 87 88

EXPORT_SYMBOL(totalram_pages);

89
static char * const zone_names[MAX_NR_ZONES] = {
90
#ifdef CONFIG_ZONE_DMA
91
	 "DMA",
92
#endif
93
#ifdef CONFIG_ZONE_DMA32
94
	 "DMA32",
95
#endif
96
	 "Normal",
97
#ifdef CONFIG_HIGHMEM
98
	 "HighMem"
99
#endif
100 101
};

L
Linus Torvalds 已提交
102 103
int min_free_kbytes = 1024;

104 105
unsigned long __meminitdata nr_kernel_pages;
unsigned long __meminitdata nr_all_pages;
106
static unsigned long __meminitdata dma_reserve;
L
Linus Torvalds 已提交
107

108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP
  /*
   * MAX_ACTIVE_REGIONS determines the maxmimum number of distinct
   * ranges of memory (RAM) that may be registered with add_active_range().
   * Ranges passed to add_active_range() will be merged if possible
   * so the number of times add_active_range() can be called is
   * related to the number of nodes and the number of holes
   */
  #ifdef CONFIG_MAX_ACTIVE_REGIONS
    /* Allow an architecture to set MAX_ACTIVE_REGIONS to save memory */
    #define MAX_ACTIVE_REGIONS CONFIG_MAX_ACTIVE_REGIONS
  #else
    #if MAX_NUMNODES >= 32
      /* If there can be many nodes, allow up to 50 holes per node */
      #define MAX_ACTIVE_REGIONS (MAX_NUMNODES*50)
    #else
      /* By default, allow up to 256 distinct regions */
      #define MAX_ACTIVE_REGIONS 256
    #endif
  #endif

129 130 131 132
  struct node_active_region __meminitdata early_node_map[MAX_ACTIVE_REGIONS];
  int __meminitdata nr_nodemap_entries;
  unsigned long __meminitdata arch_zone_lowest_possible_pfn[MAX_NR_ZONES];
  unsigned long __meminitdata arch_zone_highest_possible_pfn[MAX_NR_ZONES];
133 134 135 136
#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE
  unsigned long __initdata node_boundary_start_pfn[MAX_NUMNODES];
  unsigned long __initdata node_boundary_end_pfn[MAX_NUMNODES];
#endif /* CONFIG_MEMORY_HOTPLUG_RESERVE */
137 138
#endif /* CONFIG_ARCH_POPULATES_NODE_MAP */

M
Miklos Szeredi 已提交
139 140 141 142 143
#if MAX_NUMNODES > 1
int nr_node_ids __read_mostly = MAX_NUMNODES;
EXPORT_SYMBOL(nr_node_ids);
#endif

N
Nick Piggin 已提交
144
#ifdef CONFIG_DEBUG_VM
145
static int page_outside_zone_boundaries(struct zone *zone, struct page *page)
L
Linus Torvalds 已提交
146
{
147 148 149
	int ret = 0;
	unsigned seq;
	unsigned long pfn = page_to_pfn(page);
150

151 152 153 154 155 156 157 158 159
	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;
160 161 162 163
}

static int page_is_consistent(struct zone *zone, struct page *page)
{
164
	if (!pfn_valid_within(page_to_pfn(page)))
165
		return 0;
L
Linus Torvalds 已提交
166
	if (zone != page_zone(page))
167 168 169 170 171 172 173 174 175 176
		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
177
		return 1;
178 179 180
	if (!page_is_consistent(zone, page))
		return 1;

L
Linus Torvalds 已提交
181 182
	return 0;
}
N
Nick Piggin 已提交
183 184 185 186 187 188 189
#else
static inline int bad_range(struct zone *zone, struct page *page)
{
	return 0;
}
#endif

N
Nick Piggin 已提交
190
static void bad_page(struct page *page)
L
Linus Torvalds 已提交
191
{
N
Nick Piggin 已提交
192
	printk(KERN_EMERG "Bad page state in process '%s'\n"
193 194 195
		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
196 197 198
		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
199
	dump_stack();
200 201
	page->flags &= ~(1 << PG_lru	|
			1 << PG_private |
L
Linus Torvalds 已提交
202 203 204
			1 << PG_locked	|
			1 << PG_active	|
			1 << PG_dirty	|
205 206
			1 << PG_reclaim |
			1 << PG_slab    |
L
Linus Torvalds 已提交
207
			1 << PG_swapcache |
208 209
			1 << PG_writeback |
			1 << PG_buddy );
L
Linus Torvalds 已提交
210 211 212
	set_page_count(page, 0);
	reset_page_mapcount(page);
	page->mapping = NULL;
213
	add_taint(TAINT_BAD_PAGE);
L
Linus Torvalds 已提交
214 215 216 217 218 219 220 221 222 223 224 225
}

/*
 * 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).
 *
226 227 228
 * 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
229
 */
230 231 232

static void free_compound_page(struct page *page)
{
233
	__free_pages_ok(page, compound_order(page));
234 235
}

L
Linus Torvalds 已提交
236 237 238 239 240
static void prep_compound_page(struct page *page, unsigned long order)
{
	int i;
	int nr_pages = 1 << order;

241
	set_compound_page_dtor(page, free_compound_page);
242
	set_compound_order(page, order);
243
	__SetPageHead(page);
244
	for (i = 1; i < nr_pages; i++) {
L
Linus Torvalds 已提交
245 246
		struct page *p = page + i;

247 248
		__SetPageTail(p);
		p->first_page = page;
L
Linus Torvalds 已提交
249 250 251 252 253 254 255 256
	}
}

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

257
	if (unlikely(compound_order(page) != order))
N
Nick Piggin 已提交
258
		bad_page(page);
L
Linus Torvalds 已提交
259

260
	if (unlikely(!PageHead(page)))
261
			bad_page(page);
262
	__ClearPageHead(page);
263
	for (i = 1; i < nr_pages; i++) {
L
Linus Torvalds 已提交
264 265
		struct page *p = page + i;

266
		if (unlikely(!PageTail(p) |
267
				(p->first_page != page)))
N
Nick Piggin 已提交
268
			bad_page(page);
269
		__ClearPageTail(p);
L
Linus Torvalds 已提交
270 271 272
	}
}

N
Nick Piggin 已提交
273 274 275 276
static inline void prep_zero_page(struct page *page, int order, gfp_t gfp_flags)
{
	int i;

N
Nick Piggin 已提交
277
	VM_BUG_ON((gfp_flags & (__GFP_WAIT | __GFP_HIGHMEM)) == __GFP_HIGHMEM);
278 279 280 281
	/*
	 * 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
282
	VM_BUG_ON((gfp_flags & __GFP_HIGHMEM) && in_interrupt());
N
Nick Piggin 已提交
283 284 285 286
	for (i = 0; i < (1 << order); i++)
		clear_highpage(page + i);
}

L
Linus Torvalds 已提交
287 288 289 290 291
/*
 * 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.
 */
292 293
static inline unsigned long page_order(struct page *page)
{
H
Hugh Dickins 已提交
294
	return page_private(page);
L
Linus Torvalds 已提交
295 296
}

297 298
static inline void set_page_order(struct page *page, int order)
{
H
Hugh Dickins 已提交
299
	set_page_private(page, order);
300
	__SetPageBuddy(page);
L
Linus Torvalds 已提交
301 302 303 304
}

static inline void rmv_page_order(struct page *page)
{
305
	__ClearPageBuddy(page);
H
Hugh Dickins 已提交
306
	set_page_private(page, 0);
L
Linus Torvalds 已提交
307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
}

/*
 * 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
324
 * Assumption: *_mem_map is contiguous at least up to MAX_ORDER
L
Linus Torvalds 已提交
325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342
 */
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
343
 * (a) the buddy is not in a hole &&
344
 * (b) the buddy is in the buddy system &&
345 346
 * (c) a page and its buddy have the same order &&
 * (d) a page and its buddy are in the same zone.
347 348 349
 *
 * 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
350
 *
351
 * For recording page's order, we use page_private(page).
L
Linus Torvalds 已提交
352
 */
353 354
static inline int page_is_buddy(struct page *page, struct page *buddy,
								int order)
L
Linus Torvalds 已提交
355
{
356
	if (!pfn_valid_within(page_to_pfn(buddy)))
N
Nick Piggin 已提交
357 358
		return 0;

359 360 361 362 363
	if (page_zone_id(page) != page_zone_id(buddy))
		return 0;

	if (PageBuddy(buddy) && page_order(buddy) == order) {
		BUG_ON(page_count(buddy) != 0);
364
		return 1;
365
	}
366
	return 0;
L
Linus Torvalds 已提交
367 368 369 370 371 372 373 374 375 376 377 378 379 380 381
}

/*
 * 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
382
 * free pages of length of (1 << order) and marked with PG_buddy. Page's
H
Hugh Dickins 已提交
383
 * order is recorded in page_private(page) field.
L
Linus Torvalds 已提交
384 385 386 387 388 389 390 391 392
 * 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
393
static inline void __free_one_page(struct page *page,
L
Linus Torvalds 已提交
394 395 396 397 398
		struct zone *zone, unsigned int order)
{
	unsigned long page_idx;
	int order_size = 1 << order;

N
Nick Piggin 已提交
399
	if (unlikely(PageCompound(page)))
L
Linus Torvalds 已提交
400 401 402 403
		destroy_compound_page(page, order);

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

N
Nick Piggin 已提交
404 405
	VM_BUG_ON(page_idx & (order_size - 1));
	VM_BUG_ON(bad_range(zone, page));
L
Linus Torvalds 已提交
406

407
	__mod_zone_page_state(zone, NR_FREE_PAGES, order_size);
L
Linus Torvalds 已提交
408 409 410 411 412 413
	while (order < MAX_ORDER-1) {
		unsigned long combined_idx;
		struct free_area *area;
		struct page *buddy;

		buddy = __page_find_buddy(page, page_idx, order);
414
		if (!page_is_buddy(page, buddy, order))
L
Linus Torvalds 已提交
415
			break;		/* Move the buddy up one level. */
N
Nick Piggin 已提交
416

L
Linus Torvalds 已提交
417 418 419 420
		list_del(&buddy->lru);
		area = zone->free_area + order;
		area->nr_free--;
		rmv_page_order(buddy);
N
Nick Piggin 已提交
421
		combined_idx = __find_combined_index(page_idx, order);
L
Linus Torvalds 已提交
422 423 424 425 426 427 428 429 430
		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
431
static inline int free_pages_check(struct page *page)
L
Linus Torvalds 已提交
432
{
N
Nick Piggin 已提交
433 434 435
	if (unlikely(page_mapcount(page) |
		(page->mapping != NULL)  |
		(page_count(page) != 0)  |
L
Linus Torvalds 已提交
436 437 438 439 440 441 442
		(page->flags & (
			1 << PG_lru	|
			1 << PG_private |
			1 << PG_locked	|
			1 << PG_active	|
			1 << PG_slab	|
			1 << PG_swapcache |
N
Nick Piggin 已提交
443
			1 << PG_writeback |
444 445
			1 << PG_reserved |
			1 << PG_buddy ))))
N
Nick Piggin 已提交
446
		bad_page(page);
447 448 449 450 451 452
	/*
	 * PageReclaim == PageTail. It is only an error
	 * for PageReclaim to be set if PageCompound is clear.
	 */
	if (unlikely(!PageCompound(page) && PageReclaim(page)))
		bad_page(page);
L
Linus Torvalds 已提交
453
	if (PageDirty(page))
N
Nick Piggin 已提交
454
		__ClearPageDirty(page);
455 456 457 458 459 460
	/*
	 * 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
461 462 463 464 465
}

/*
 * Frees a list of pages. 
 * Assumes all pages on list are in same zone, and of same order.
466
 * count is the number of pages to free.
L
Linus Torvalds 已提交
467 468 469 470 471 472 473
 *
 * 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
474 475
static void free_pages_bulk(struct zone *zone, int count,
					struct list_head *list, int order)
L
Linus Torvalds 已提交
476
{
N
Nick Piggin 已提交
477
	spin_lock(&zone->lock);
L
Linus Torvalds 已提交
478 479
	zone->all_unreclaimable = 0;
	zone->pages_scanned = 0;
N
Nick Piggin 已提交
480 481 482
	while (count--) {
		struct page *page;

N
Nick Piggin 已提交
483
		VM_BUG_ON(list_empty(list));
L
Linus Torvalds 已提交
484
		page = list_entry(list->prev, struct page, lru);
N
Nick Piggin 已提交
485
		/* have to delete it as __free_one_page list manipulates */
L
Linus Torvalds 已提交
486
		list_del(&page->lru);
N
Nick Piggin 已提交
487
		__free_one_page(page, zone, order);
L
Linus Torvalds 已提交
488
	}
N
Nick Piggin 已提交
489
	spin_unlock(&zone->lock);
L
Linus Torvalds 已提交
490 491
}

N
Nick Piggin 已提交
492
static void free_one_page(struct zone *zone, struct page *page, int order)
L
Linus Torvalds 已提交
493
{
494 495 496
	spin_lock(&zone->lock);
	zone->all_unreclaimable = 0;
	zone->pages_scanned = 0;
497
	__free_one_page(page, zone, order);
498
	spin_unlock(&zone->lock);
N
Nick Piggin 已提交
499 500 501 502 503
}

static void __free_pages_ok(struct page *page, unsigned int order)
{
	unsigned long flags;
L
Linus Torvalds 已提交
504
	int i;
505
	int reserved = 0;
L
Linus Torvalds 已提交
506 507

	for (i = 0 ; i < (1 << order) ; ++i)
N
Nick Piggin 已提交
508
		reserved += free_pages_check(page + i);
509 510 511
	if (reserved)
		return;

N
Nick Piggin 已提交
512 513
	if (!PageHighMem(page))
		debug_check_no_locks_freed(page_address(page),PAGE_SIZE<<order);
N
Nick Piggin 已提交
514
	arch_free_page(page, order);
N
Nick Piggin 已提交
515
	kernel_map_pages(page, 1 << order, 0);
N
Nick Piggin 已提交
516

N
Nick Piggin 已提交
517
	local_irq_save(flags);
518
	__count_vm_events(PGFREE, 1 << order);
N
Nick Piggin 已提交
519
	free_one_page(page_zone(page), page, order);
N
Nick Piggin 已提交
520
	local_irq_restore(flags);
L
Linus Torvalds 已提交
521 522
}

523 524 525 526 527 528 529 530
/*
 * 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);
531
		set_page_refcounted(page);
N
Nick Piggin 已提交
532
		__free_page(page);
533 534 535
	} else {
		int loop;

N
Nick Piggin 已提交
536
		prefetchw(page);
537 538 539
		for (loop = 0; loop < BITS_PER_LONG; loop++) {
			struct page *p = &page[loop];

N
Nick Piggin 已提交
540 541
			if (loop + 1 < BITS_PER_LONG)
				prefetchw(p + 1);
542 543 544 545
			__ClearPageReserved(p);
			set_page_count(p, 0);
		}

546
		set_page_refcounted(page);
N
Nick Piggin 已提交
547
		__free_pages(page, order);
548 549 550
	}
}

L
Linus Torvalds 已提交
551 552 553 554 555 556 557 558 559 560 561 562 563 564 565

/*
 * 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
566
static inline void expand(struct zone *zone, struct page *page,
L
Linus Torvalds 已提交
567 568 569 570 571 572 573 574
 	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
575
		VM_BUG_ON(bad_range(zone, &page[size]));
L
Linus Torvalds 已提交
576 577 578 579 580 581 582 583 584
		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
585
static int prep_new_page(struct page *page, int order, gfp_t gfp_flags)
L
Linus Torvalds 已提交
586
{
N
Nick Piggin 已提交
587 588 589
	if (unlikely(page_mapcount(page) |
		(page->mapping != NULL)  |
		(page_count(page) != 0)  |
590 591
		(page->flags & (
			1 << PG_lru	|
L
Linus Torvalds 已提交
592 593 594 595 596
			1 << PG_private	|
			1 << PG_locked	|
			1 << PG_active	|
			1 << PG_dirty	|
			1 << PG_reclaim	|
597
			1 << PG_slab    |
L
Linus Torvalds 已提交
598
			1 << PG_swapcache |
N
Nick Piggin 已提交
599
			1 << PG_writeback |
600 601
			1 << PG_reserved |
			1 << PG_buddy ))))
N
Nick Piggin 已提交
602
		bad_page(page);
L
Linus Torvalds 已提交
603

604 605 606 607 608 609 610
	/*
	 * 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
611 612
	page->flags &= ~(1 << PG_uptodate | 1 << PG_error |
			1 << PG_referenced | 1 << PG_arch_1 |
613
			1 << PG_owner_priv_1 | 1 << PG_mappedtodisk);
H
Hugh Dickins 已提交
614
	set_page_private(page, 0);
615
	set_page_refcounted(page);
N
Nick Piggin 已提交
616 617

	arch_alloc_page(page, order);
L
Linus Torvalds 已提交
618
	kernel_map_pages(page, 1 << order, 1);
N
Nick Piggin 已提交
619 620 621 622 623 624 625

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

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

626
	return 0;
L
Linus Torvalds 已提交
627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647
}

/* 
 * 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--;
648
		__mod_zone_page_state(zone, NR_FREE_PAGES, - (1UL << order));
N
Nick Piggin 已提交
649 650
		expand(zone, page, order, current_order, area);
		return page;
L
Linus Torvalds 已提交
651 652 653 654 655 656 657 658 659 660 661 662 663 664 665
	}

	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
666
	spin_lock(&zone->lock);
L
Linus Torvalds 已提交
667
	for (i = 0; i < count; ++i) {
N
Nick Piggin 已提交
668 669
		struct page *page = __rmqueue(zone, order);
		if (unlikely(page == NULL))
L
Linus Torvalds 已提交
670 671 672
			break;
		list_add_tail(&page->lru, list);
	}
N
Nick Piggin 已提交
673
	spin_unlock(&zone->lock);
N
Nick Piggin 已提交
674
	return i;
L
Linus Torvalds 已提交
675 676
}

677
#ifdef CONFIG_NUMA
678
/*
679 680 681 682
 * Called from the vmstat counter updater to drain pagesets of this
 * currently executing processor on remote nodes after they have
 * expired.
 *
683 684
 * Note that this function must be called with the thread pinned to
 * a single processor.
685
 */
686
void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp)
687 688
{
	unsigned long flags;
689
	int to_drain;
690

691 692 693 694 695 696 697 698
	local_irq_save(flags);
	if (pcp->count >= pcp->batch)
		to_drain = pcp->batch;
	else
		to_drain = pcp->count;
	free_pages_bulk(zone, to_drain, &pcp->list, 0);
	pcp->count -= to_drain;
	local_irq_restore(flags);
699 700 701
}
#endif

L
Linus Torvalds 已提交
702 703
static void __drain_pages(unsigned int cpu)
{
N
Nick Piggin 已提交
704
	unsigned long flags;
L
Linus Torvalds 已提交
705 706 707 708 709 710
	struct zone *zone;
	int i;

	for_each_zone(zone) {
		struct per_cpu_pageset *pset;

711 712 713
		if (!populated_zone(zone))
			continue;

714
		pset = zone_pcp(zone, cpu);
L
Linus Torvalds 已提交
715 716 717 718
		for (i = 0; i < ARRAY_SIZE(pset->pcp); i++) {
			struct per_cpu_pages *pcp;

			pcp = &pset->pcp[i];
N
Nick Piggin 已提交
719
			local_irq_save(flags);
N
Nick Piggin 已提交
720 721
			free_pages_bulk(zone, pcp->count, &pcp->list, 0);
			pcp->count = 0;
N
Nick Piggin 已提交
722
			local_irq_restore(flags);
L
Linus Torvalds 已提交
723 724 725 726 727 728 729 730
		}
	}
}

#ifdef CONFIG_PM

void mark_free_pages(struct zone *zone)
{
731 732
	unsigned long pfn, max_zone_pfn;
	unsigned long flags;
L
Linus Torvalds 已提交
733 734 735 736 737 738 739
	int order;
	struct list_head *curr;

	if (!zone->spanned_pages)
		return;

	spin_lock_irqsave(&zone->lock, flags);
740 741 742 743 744 745

	max_zone_pfn = zone->zone_start_pfn + zone->spanned_pages;
	for (pfn = zone->zone_start_pfn; pfn < max_zone_pfn; pfn++)
		if (pfn_valid(pfn)) {
			struct page *page = pfn_to_page(pfn);

746 747
			if (!swsusp_page_is_forbidden(page))
				swsusp_unset_page_free(page);
748
		}
L
Linus Torvalds 已提交
749 750 751

	for (order = MAX_ORDER - 1; order >= 0; --order)
		list_for_each(curr, &zone->free_area[order].free_list) {
752
			unsigned long i;
L
Linus Torvalds 已提交
753

754 755
			pfn = page_to_pfn(list_entry(curr, struct page, lru));
			for (i = 0; i < (1UL << order); i++)
756
				swsusp_set_page_free(pfn_to_page(pfn + i));
757
		}
L
Linus Torvalds 已提交
758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785

	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;

	if (PageAnon(page))
		page->mapping = NULL;
N
Nick Piggin 已提交
786
	if (free_pages_check(page))
787 788
		return;

N
Nick Piggin 已提交
789 790
	if (!PageHighMem(page))
		debug_check_no_locks_freed(page_address(page), PAGE_SIZE);
N
Nick Piggin 已提交
791
	arch_free_page(page, 0);
792 793
	kernel_map_pages(page, 1, 0);

794
	pcp = &zone_pcp(zone, get_cpu())->pcp[cold];
L
Linus Torvalds 已提交
795
	local_irq_save(flags);
796
	__count_vm_event(PGFREE);
L
Linus Torvalds 已提交
797 798
	list_add(&page->lru, &pcp->list);
	pcp->count++;
N
Nick Piggin 已提交
799 800 801 802
	if (pcp->count >= pcp->high) {
		free_pages_bulk(zone, pcp->batch, &pcp->list, 0);
		pcp->count -= pcp->batch;
	}
L
Linus Torvalds 已提交
803 804 805 806 807 808 809 810 811 812 813 814 815 816
	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
817 818 819 820 821 822 823 824 825 826 827 828
/*
 * 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
829 830
	VM_BUG_ON(PageCompound(page));
	VM_BUG_ON(!page_count(page));
831 832
	for (i = 1; i < (1 << order); i++)
		set_page_refcounted(page + i);
N
Nick Piggin 已提交
833 834
}

L
Linus Torvalds 已提交
835 836 837 838 839
/*
 * 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
840 841
static struct page *buffered_rmqueue(struct zonelist *zonelist,
			struct zone *zone, int order, gfp_t gfp_flags)
L
Linus Torvalds 已提交
842 843
{
	unsigned long flags;
844
	struct page *page;
L
Linus Torvalds 已提交
845
	int cold = !!(gfp_flags & __GFP_COLD);
N
Nick Piggin 已提交
846
	int cpu;
L
Linus Torvalds 已提交
847

848
again:
N
Nick Piggin 已提交
849
	cpu  = get_cpu();
N
Nick Piggin 已提交
850
	if (likely(order == 0)) {
L
Linus Torvalds 已提交
851 852
		struct per_cpu_pages *pcp;

N
Nick Piggin 已提交
853
		pcp = &zone_pcp(zone, cpu)->pcp[cold];
L
Linus Torvalds 已提交
854
		local_irq_save(flags);
N
Nick Piggin 已提交
855
		if (!pcp->count) {
856
			pcp->count = rmqueue_bulk(zone, 0,
L
Linus Torvalds 已提交
857
						pcp->batch, &pcp->list);
N
Nick Piggin 已提交
858 859
			if (unlikely(!pcp->count))
				goto failed;
L
Linus Torvalds 已提交
860
		}
N
Nick Piggin 已提交
861 862 863
		page = list_entry(pcp->list.next, struct page, lru);
		list_del(&page->lru);
		pcp->count--;
R
Rohit Seth 已提交
864
	} else {
L
Linus Torvalds 已提交
865 866
		spin_lock_irqsave(&zone->lock, flags);
		page = __rmqueue(zone, order);
N
Nick Piggin 已提交
867 868 869
		spin_unlock(&zone->lock);
		if (!page)
			goto failed;
L
Linus Torvalds 已提交
870 871
	}

872
	__count_zone_vm_events(PGALLOC, zone, 1 << order);
873
	zone_statistics(zonelist, zone);
N
Nick Piggin 已提交
874 875
	local_irq_restore(flags);
	put_cpu();
L
Linus Torvalds 已提交
876

N
Nick Piggin 已提交
877
	VM_BUG_ON(bad_range(zone, page));
N
Nick Piggin 已提交
878
	if (prep_new_page(page, order, gfp_flags))
N
Nick Piggin 已提交
879
		goto again;
L
Linus Torvalds 已提交
880
	return page;
N
Nick Piggin 已提交
881 882 883 884 885

failed:
	local_irq_restore(flags);
	put_cpu();
	return NULL;
L
Linus Torvalds 已提交
886 887
}

R
Rohit Seth 已提交
888
#define ALLOC_NO_WATERMARKS	0x01 /* don't check watermarks at all */
889 890 891 892 893 894
#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
895

896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912
#ifdef CONFIG_FAIL_PAGE_ALLOC

static struct fail_page_alloc_attr {
	struct fault_attr attr;

	u32 ignore_gfp_highmem;
	u32 ignore_gfp_wait;

#ifdef CONFIG_FAULT_INJECTION_DEBUG_FS

	struct dentry *ignore_gfp_highmem_file;
	struct dentry *ignore_gfp_wait_file;

#endif /* CONFIG_FAULT_INJECTION_DEBUG_FS */

} fail_page_alloc = {
	.attr = FAULT_ATTR_INITIALIZER,
913 914
	.ignore_gfp_wait = 1,
	.ignore_gfp_highmem = 1,
915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980
};

static int __init setup_fail_page_alloc(char *str)
{
	return setup_fault_attr(&fail_page_alloc.attr, str);
}
__setup("fail_page_alloc=", setup_fail_page_alloc);

static int should_fail_alloc_page(gfp_t gfp_mask, unsigned int order)
{
	if (gfp_mask & __GFP_NOFAIL)
		return 0;
	if (fail_page_alloc.ignore_gfp_highmem && (gfp_mask & __GFP_HIGHMEM))
		return 0;
	if (fail_page_alloc.ignore_gfp_wait && (gfp_mask & __GFP_WAIT))
		return 0;

	return should_fail(&fail_page_alloc.attr, 1 << order);
}

#ifdef CONFIG_FAULT_INJECTION_DEBUG_FS

static int __init fail_page_alloc_debugfs(void)
{
	mode_t mode = S_IFREG | S_IRUSR | S_IWUSR;
	struct dentry *dir;
	int err;

	err = init_fault_attr_dentries(&fail_page_alloc.attr,
				       "fail_page_alloc");
	if (err)
		return err;
	dir = fail_page_alloc.attr.dentries.dir;

	fail_page_alloc.ignore_gfp_wait_file =
		debugfs_create_bool("ignore-gfp-wait", mode, dir,
				      &fail_page_alloc.ignore_gfp_wait);

	fail_page_alloc.ignore_gfp_highmem_file =
		debugfs_create_bool("ignore-gfp-highmem", mode, dir,
				      &fail_page_alloc.ignore_gfp_highmem);

	if (!fail_page_alloc.ignore_gfp_wait_file ||
			!fail_page_alloc.ignore_gfp_highmem_file) {
		err = -ENOMEM;
		debugfs_remove(fail_page_alloc.ignore_gfp_wait_file);
		debugfs_remove(fail_page_alloc.ignore_gfp_highmem_file);
		cleanup_fault_attr_dentries(&fail_page_alloc.attr);
	}

	return err;
}

late_initcall(fail_page_alloc_debugfs);

#endif /* CONFIG_FAULT_INJECTION_DEBUG_FS */

#else /* CONFIG_FAIL_PAGE_ALLOC */

static inline int should_fail_alloc_page(gfp_t gfp_mask, unsigned int order)
{
	return 0;
}

#endif /* CONFIG_FAIL_PAGE_ALLOC */

L
Linus Torvalds 已提交
981 982 983 984 985
/*
 * 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
986
		      int classzone_idx, int alloc_flags)
L
Linus Torvalds 已提交
987 988
{
	/* free_pages my go negative - that's OK */
989 990
	long min = mark;
	long free_pages = zone_page_state(z, NR_FREE_PAGES) - (1 << order) + 1;
L
Linus Torvalds 已提交
991 992
	int o;

R
Rohit Seth 已提交
993
	if (alloc_flags & ALLOC_HIGH)
L
Linus Torvalds 已提交
994
		min -= min / 2;
R
Rohit Seth 已提交
995
	if (alloc_flags & ALLOC_HARDER)
L
Linus Torvalds 已提交
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012
		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;
}

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 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132
#ifdef CONFIG_NUMA
/*
 * zlc_setup - Setup for "zonelist cache".  Uses cached zone data to
 * skip over zones that are not allowed by the cpuset, or that have
 * been recently (in last second) found to be nearly full.  See further
 * comments in mmzone.h.  Reduces cache footprint of zonelist scans
 * that have to skip over alot of full or unallowed zones.
 *
 * If the zonelist cache is present in the passed in zonelist, then
 * returns a pointer to the allowed node mask (either the current
 * tasks mems_allowed, or node_online_map.)
 *
 * If the zonelist cache is not available for this zonelist, does
 * nothing and returns NULL.
 *
 * If the fullzones BITMAP in the zonelist cache is stale (more than
 * a second since last zap'd) then we zap it out (clear its bits.)
 *
 * We hold off even calling zlc_setup, until after we've checked the
 * first zone in the zonelist, on the theory that most allocations will
 * be satisfied from that first zone, so best to examine that zone as
 * quickly as we can.
 */
static nodemask_t *zlc_setup(struct zonelist *zonelist, int alloc_flags)
{
	struct zonelist_cache *zlc;	/* cached zonelist speedup info */
	nodemask_t *allowednodes;	/* zonelist_cache approximation */

	zlc = zonelist->zlcache_ptr;
	if (!zlc)
		return NULL;

	if (jiffies - zlc->last_full_zap > 1 * HZ) {
		bitmap_zero(zlc->fullzones, MAX_ZONES_PER_ZONELIST);
		zlc->last_full_zap = jiffies;
	}

	allowednodes = !in_interrupt() && (alloc_flags & ALLOC_CPUSET) ?
					&cpuset_current_mems_allowed :
					&node_online_map;
	return allowednodes;
}

/*
 * Given 'z' scanning a zonelist, run a couple of quick checks to see
 * if it is worth looking at further for free memory:
 *  1) Check that the zone isn't thought to be full (doesn't have its
 *     bit set in the zonelist_cache fullzones BITMAP).
 *  2) Check that the zones node (obtained from the zonelist_cache
 *     z_to_n[] mapping) is allowed in the passed in allowednodes mask.
 * Return true (non-zero) if zone is worth looking at further, or
 * else return false (zero) if it is not.
 *
 * This check -ignores- the distinction between various watermarks,
 * such as GFP_HIGH, GFP_ATOMIC, PF_MEMALLOC, ...  If a zone is
 * found to be full for any variation of these watermarks, it will
 * be considered full for up to one second by all requests, unless
 * we are so low on memory on all allowed nodes that we are forced
 * into the second scan of the zonelist.
 *
 * In the second scan we ignore this zonelist cache and exactly
 * apply the watermarks to all zones, even it is slower to do so.
 * We are low on memory in the second scan, and should leave no stone
 * unturned looking for a free page.
 */
static int zlc_zone_worth_trying(struct zonelist *zonelist, struct zone **z,
						nodemask_t *allowednodes)
{
	struct zonelist_cache *zlc;	/* cached zonelist speedup info */
	int i;				/* index of *z in zonelist zones */
	int n;				/* node that zone *z is on */

	zlc = zonelist->zlcache_ptr;
	if (!zlc)
		return 1;

	i = z - zonelist->zones;
	n = zlc->z_to_n[i];

	/* This zone is worth trying if it is allowed but not full */
	return node_isset(n, *allowednodes) && !test_bit(i, zlc->fullzones);
}

/*
 * Given 'z' scanning a zonelist, set the corresponding bit in
 * zlc->fullzones, so that subsequent attempts to allocate a page
 * from that zone don't waste time re-examining it.
 */
static void zlc_mark_zone_full(struct zonelist *zonelist, struct zone **z)
{
	struct zonelist_cache *zlc;	/* cached zonelist speedup info */
	int i;				/* index of *z in zonelist zones */

	zlc = zonelist->zlcache_ptr;
	if (!zlc)
		return;

	i = z - zonelist->zones;

	set_bit(i, zlc->fullzones);
}

#else	/* CONFIG_NUMA */

static nodemask_t *zlc_setup(struct zonelist *zonelist, int alloc_flags)
{
	return NULL;
}

static int zlc_zone_worth_trying(struct zonelist *zonelist, struct zone **z,
				nodemask_t *allowednodes)
{
	return 1;
}

static void zlc_mark_zone_full(struct zonelist *zonelist, struct zone **z)
{
}
#endif	/* CONFIG_NUMA */

R
Rohit Seth 已提交
1133
/*
1134
 * get_page_from_freelist goes through the zonelist trying to allocate
R
Rohit Seth 已提交
1135 1136 1137 1138 1139
 * 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
1140
{
1141
	struct zone **z;
R
Rohit Seth 已提交
1142
	struct page *page = NULL;
1143
	int classzone_idx = zone_idx(zonelist->zones[0]);
1144
	struct zone *zone;
1145 1146 1147
	nodemask_t *allowednodes = NULL;/* zonelist_cache approximation */
	int zlc_active = 0;		/* set if using zonelist_cache */
	int did_zlc_setup = 0;		/* just call zlc_setup() one time */
R
Rohit Seth 已提交
1148

1149
zonelist_scan:
R
Rohit Seth 已提交
1150
	/*
1151
	 * Scan zonelist, looking for a zone with enough free.
R
Rohit Seth 已提交
1152 1153
	 * See also cpuset_zone_allowed() comment in kernel/cpuset.c.
	 */
1154 1155
	z = zonelist->zones;

R
Rohit Seth 已提交
1156
	do {
1157 1158 1159
		if (NUMA_BUILD && zlc_active &&
			!zlc_zone_worth_trying(zonelist, z, allowednodes))
				continue;
1160
		zone = *z;
1161
		if (unlikely(NUMA_BUILD && (gfp_mask & __GFP_THISNODE) &&
1162
			zone->zone_pgdat != zonelist->zones[0]->zone_pgdat))
1163
				break;
R
Rohit Seth 已提交
1164
		if ((alloc_flags & ALLOC_CPUSET) &&
1165
			!cpuset_zone_allowed_softwall(zone, gfp_mask))
1166
				goto try_next_zone;
R
Rohit Seth 已提交
1167 1168

		if (!(alloc_flags & ALLOC_NO_WATERMARKS)) {
1169 1170
			unsigned long mark;
			if (alloc_flags & ALLOC_WMARK_MIN)
1171
				mark = zone->pages_min;
1172
			else if (alloc_flags & ALLOC_WMARK_LOW)
1173
				mark = zone->pages_low;
1174
			else
1175
				mark = zone->pages_high;
1176 1177
			if (!zone_watermark_ok(zone, order, mark,
				    classzone_idx, alloc_flags)) {
1178
				if (!zone_reclaim_mode ||
1179
				    !zone_reclaim(zone, gfp_mask, order))
1180
					goto this_zone_full;
1181
			}
R
Rohit Seth 已提交
1182 1183
		}

1184
		page = buffered_rmqueue(zonelist, zone, order, gfp_mask);
1185
		if (page)
R
Rohit Seth 已提交
1186
			break;
1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
this_zone_full:
		if (NUMA_BUILD)
			zlc_mark_zone_full(zonelist, z);
try_next_zone:
		if (NUMA_BUILD && !did_zlc_setup) {
			/* we do zlc_setup after the first zone is tried */
			allowednodes = zlc_setup(zonelist, alloc_flags);
			zlc_active = 1;
			did_zlc_setup = 1;
		}
R
Rohit Seth 已提交
1197
	} while (*(++z) != NULL);
1198 1199 1200 1201 1202 1203

	if (unlikely(NUMA_BUILD && page == NULL && zlc_active)) {
		/* Disable zlc cache for second zonelist scan */
		zlc_active = 0;
		goto zonelist_scan;
	}
R
Rohit Seth 已提交
1204
	return page;
M
Martin Hicks 已提交
1205 1206
}

L
Linus Torvalds 已提交
1207 1208 1209 1210
/*
 * This is the 'heart' of the zoned buddy allocator.
 */
struct page * fastcall
A
Al Viro 已提交
1211
__alloc_pages(gfp_t gfp_mask, unsigned int order,
L
Linus Torvalds 已提交
1212 1213
		struct zonelist *zonelist)
{
A
Al Viro 已提交
1214
	const gfp_t wait = gfp_mask & __GFP_WAIT;
R
Rohit Seth 已提交
1215
	struct zone **z;
L
Linus Torvalds 已提交
1216 1217 1218 1219
	struct page *page;
	struct reclaim_state reclaim_state;
	struct task_struct *p = current;
	int do_retry;
R
Rohit Seth 已提交
1220
	int alloc_flags;
L
Linus Torvalds 已提交
1221 1222 1223 1224
	int did_some_progress;

	might_sleep_if(wait);

1225 1226 1227
	if (should_fail_alloc_page(gfp_mask, order))
		return NULL;

1228
restart:
R
Rohit Seth 已提交
1229
	z = zonelist->zones;  /* the list of zones suitable for gfp_mask */
L
Linus Torvalds 已提交
1230

R
Rohit Seth 已提交
1231
	if (unlikely(*z == NULL)) {
L
Linus Torvalds 已提交
1232 1233 1234
		/* Should this ever happen?? */
		return NULL;
	}
1235

R
Rohit Seth 已提交
1236
	page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, order,
1237
				zonelist, ALLOC_WMARK_LOW|ALLOC_CPUSET);
R
Rohit Seth 已提交
1238 1239
	if (page)
		goto got_pg;
L
Linus Torvalds 已提交
1240

1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
	/*
	 * GFP_THISNODE (meaning __GFP_THISNODE, __GFP_NORETRY and
	 * __GFP_NOWARN set) should not cause reclaim since the subsystem
	 * (f.e. slab) using GFP_THISNODE may choose to trigger reclaim
	 * using a larger set of nodes after it has established that the
	 * allowed per node queues are empty and that nodes are
	 * over allocated.
	 */
	if (NUMA_BUILD && (gfp_mask & GFP_THISNODE) == GFP_THISNODE)
		goto nopage;

1252
	for (z = zonelist->zones; *z; z++)
1253
		wakeup_kswapd(*z, order);
L
Linus Torvalds 已提交
1254

1255
	/*
R
Rohit Seth 已提交
1256 1257 1258 1259 1260 1261
	 * 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
1262 1263
	 * policy or is asking for __GFP_HIGH memory.  GFP_ATOMIC requests will
	 * set both ALLOC_HARDER (!wait) and ALLOC_HIGH (__GFP_HIGH).
1264
	 */
1265
	alloc_flags = ALLOC_WMARK_MIN;
R
Rohit Seth 已提交
1266 1267 1268 1269
	if ((unlikely(rt_task(p)) && !in_interrupt()) || !wait)
		alloc_flags |= ALLOC_HARDER;
	if (gfp_mask & __GFP_HIGH)
		alloc_flags |= ALLOC_HIGH;
1270 1271
	if (wait)
		alloc_flags |= ALLOC_CPUSET;
L
Linus Torvalds 已提交
1272 1273 1274

	/*
	 * Go through the zonelist again. Let __GFP_HIGH and allocations
R
Rohit Seth 已提交
1275
	 * coming from realtime tasks go deeper into reserves.
L
Linus Torvalds 已提交
1276 1277 1278
	 *
	 * This is the last chance, in general, before the goto nopage.
	 * Ignore cpuset if GFP_ATOMIC (!wait) rather than fail alloc.
1279
	 * See also cpuset_zone_allowed() comment in kernel/cpuset.c.
L
Linus Torvalds 已提交
1280
	 */
R
Rohit Seth 已提交
1281 1282 1283
	page = get_page_from_freelist(gfp_mask, order, zonelist, alloc_flags);
	if (page)
		goto got_pg;
L
Linus Torvalds 已提交
1284 1285

	/* This allocation should allow future memory freeing. */
1286

1287
rebalance:
1288 1289 1290
	if (((p->flags & PF_MEMALLOC) || unlikely(test_thread_flag(TIF_MEMDIE)))
			&& !in_interrupt()) {
		if (!(gfp_mask & __GFP_NOMEMALLOC)) {
K
Kirill Korotaev 已提交
1291
nofail_alloc:
1292
			/* go through the zonelist yet again, ignoring mins */
R
Rohit Seth 已提交
1293
			page = get_page_from_freelist(gfp_mask, order,
1294
				zonelist, ALLOC_NO_WATERMARKS);
R
Rohit Seth 已提交
1295 1296
			if (page)
				goto got_pg;
K
Kirill Korotaev 已提交
1297
			if (gfp_mask & __GFP_NOFAIL) {
1298
				congestion_wait(WRITE, HZ/50);
K
Kirill Korotaev 已提交
1299 1300
				goto nofail_alloc;
			}
L
Linus Torvalds 已提交
1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311
		}
		goto nopage;
	}

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

	cond_resched();

	/* We now go into synchronous reclaim */
1312
	cpuset_memory_pressure_bump();
L
Linus Torvalds 已提交
1313 1314 1315 1316
	p->flags |= PF_MEMALLOC;
	reclaim_state.reclaimed_slab = 0;
	p->reclaim_state = &reclaim_state;

R
Rohit Seth 已提交
1317
	did_some_progress = try_to_free_pages(zonelist->zones, gfp_mask);
L
Linus Torvalds 已提交
1318 1319 1320 1321 1322 1323 1324

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

	cond_resched();

	if (likely(did_some_progress)) {
R
Rohit Seth 已提交
1325 1326 1327 1328
		page = get_page_from_freelist(gfp_mask, order,
						zonelist, alloc_flags);
		if (page)
			goto got_pg;
L
Linus Torvalds 已提交
1329 1330 1331 1332 1333 1334 1335
	} 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
1336
		page = get_page_from_freelist(gfp_mask|__GFP_HARDWALL, order,
1337
				zonelist, ALLOC_WMARK_HIGH|ALLOC_CPUSET);
R
Rohit Seth 已提交
1338 1339
		if (page)
			goto got_pg;
L
Linus Torvalds 已提交
1340

1341
		out_of_memory(zonelist, gfp_mask, order);
L
Linus Torvalds 已提交
1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359
		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) {
1360
		congestion_wait(WRITE, HZ/50);
L
Linus Torvalds 已提交
1361 1362 1363 1364 1365 1366 1367 1368 1369
		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
1370
		show_mem();
L
Linus Torvalds 已提交
1371 1372 1373 1374 1375 1376 1377 1378 1379 1380
	}
got_pg:
	return page;
}

EXPORT_SYMBOL(__alloc_pages);

/*
 * Common helper functions.
 */
A
Al Viro 已提交
1381
fastcall unsigned long __get_free_pages(gfp_t gfp_mask, unsigned int order)
L
Linus Torvalds 已提交
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391
{
	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
1392
fastcall unsigned long get_zeroed_page(gfp_t gfp_mask)
L
Linus Torvalds 已提交
1393 1394 1395 1396 1397 1398 1399
{
	struct page * page;

	/*
	 * get_zeroed_page() returns a 32-bit address, which cannot represent
	 * a highmem page
	 */
N
Nick Piggin 已提交
1400
	VM_BUG_ON((gfp_mask & __GFP_HIGHMEM) != 0);
L
Linus Torvalds 已提交
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419

	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
1420
	if (put_page_testzero(page)) {
L
Linus Torvalds 已提交
1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432
		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
1433
		VM_BUG_ON(!virt_addr_valid((void *)addr));
L
Linus Torvalds 已提交
1434 1435 1436 1437 1438 1439 1440 1441
		__free_pages(virt_to_page((void *)addr), order);
	}
}

EXPORT_SYMBOL(free_pages);

static unsigned int nr_free_zone_pages(int offset)
{
1442 1443
	/* Just pick one node, since fallback list is circular */
	pg_data_t *pgdat = NODE_DATA(numa_node_id());
L
Linus Torvalds 已提交
1444 1445
	unsigned int sum = 0;

1446 1447 1448
	struct zonelist *zonelist = pgdat->node_zonelists + offset;
	struct zone **zonep = zonelist->zones;
	struct zone *zone;
L
Linus Torvalds 已提交
1449

1450 1451 1452 1453 1454
	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
1455 1456 1457 1458 1459 1460 1461 1462 1463 1464
	}

	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
1465
	return nr_free_zone_pages(gfp_zone(GFP_USER));
L
Linus Torvalds 已提交
1466 1467 1468 1469 1470 1471 1472
}

/*
 * Amount of free RAM allocatable within all zones
 */
unsigned int nr_free_pagecache_pages(void)
{
A
Al Viro 已提交
1473
	return nr_free_zone_pages(gfp_zone(GFP_HIGHUSER));
L
Linus Torvalds 已提交
1474
}
1475 1476

static inline void show_node(struct zone *zone)
L
Linus Torvalds 已提交
1477
{
1478
	if (NUMA_BUILD)
1479
		printk("Node %d ", zone_to_nid(zone));
L
Linus Torvalds 已提交
1480 1481 1482 1483 1484 1485
}

void si_meminfo(struct sysinfo *val)
{
	val->totalram = totalram_pages;
	val->sharedram = 0;
1486
	val->freeram = global_page_state(NR_FREE_PAGES);
L
Linus Torvalds 已提交
1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500
	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;
1501
	val->freeram = node_page_state(nid, NR_FREE_PAGES);
1502
#ifdef CONFIG_HIGHMEM
L
Linus Torvalds 已提交
1503
	val->totalhigh = pgdat->node_zones[ZONE_HIGHMEM].present_pages;
1504 1505
	val->freehigh = zone_page_state(&pgdat->node_zones[ZONE_HIGHMEM],
			NR_FREE_PAGES);
1506 1507 1508 1509
#else
	val->totalhigh = 0;
	val->freehigh = 0;
#endif
L
Linus Torvalds 已提交
1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522
	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)
{
1523
	int cpu;
L
Linus Torvalds 已提交
1524 1525 1526
	struct zone *zone;

	for_each_zone(zone) {
1527
		if (!populated_zone(zone))
L
Linus Torvalds 已提交
1528
			continue;
1529 1530 1531

		show_node(zone);
		printk("%s per-cpu:\n", zone->name);
L
Linus Torvalds 已提交
1532

1533
		for_each_online_cpu(cpu) {
L
Linus Torvalds 已提交
1534 1535
			struct per_cpu_pageset *pageset;

1536
			pageset = zone_pcp(zone, cpu);
L
Linus Torvalds 已提交
1537

1538 1539 1540 1541 1542 1543
			printk("CPU %4d: Hot: hi:%5d, btch:%4d usd:%4d   "
			       "Cold: hi:%5d, btch:%4d usd:%4d\n",
			       cpu, pageset->pcp[0].high,
			       pageset->pcp[0].batch, pageset->pcp[0].count,
			       pageset->pcp[1].high, pageset->pcp[1].batch,
			       pageset->pcp[1].count);
L
Linus Torvalds 已提交
1544 1545 1546
		}
	}

1547
	printk("Active:%lu inactive:%lu dirty:%lu writeback:%lu unstable:%lu\n"
1548
		" free:%lu slab:%lu mapped:%lu pagetables:%lu bounce:%lu\n",
1549 1550
		global_page_state(NR_ACTIVE),
		global_page_state(NR_INACTIVE),
1551
		global_page_state(NR_FILE_DIRTY),
1552
		global_page_state(NR_WRITEBACK),
1553
		global_page_state(NR_UNSTABLE_NFS),
1554
		global_page_state(NR_FREE_PAGES),
1555 1556
		global_page_state(NR_SLAB_RECLAIMABLE) +
			global_page_state(NR_SLAB_UNRECLAIMABLE),
1557
		global_page_state(NR_FILE_MAPPED),
1558 1559
		global_page_state(NR_PAGETABLE),
		global_page_state(NR_BOUNCE));
L
Linus Torvalds 已提交
1560 1561 1562 1563

	for_each_zone(zone) {
		int i;

1564 1565 1566
		if (!populated_zone(zone))
			continue;

L
Linus Torvalds 已提交
1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579
		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,
1580
			K(zone_page_state(zone, NR_FREE_PAGES)),
L
Linus Torvalds 已提交
1581 1582 1583
			K(zone->pages_min),
			K(zone->pages_low),
			K(zone->pages_high),
1584 1585
			K(zone_page_state(zone, NR_ACTIVE)),
			K(zone_page_state(zone, NR_INACTIVE)),
L
Linus Torvalds 已提交
1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596
			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) {
1597
 		unsigned long nr[MAX_ORDER], flags, order, total = 0;
L
Linus Torvalds 已提交
1598

1599 1600 1601
		if (!populated_zone(zone))
			continue;

L
Linus Torvalds 已提交
1602 1603 1604 1605 1606
		show_node(zone);
		printk("%s: ", zone->name);

		spin_lock_irqsave(&zone->lock, flags);
		for (order = 0; order < MAX_ORDER; order++) {
1607 1608
			nr[order] = zone->free_area[order].nr_free;
			total += nr[order] << order;
L
Linus Torvalds 已提交
1609 1610
		}
		spin_unlock_irqrestore(&zone->lock, flags);
1611 1612
		for (order = 0; order < MAX_ORDER; order++)
			printk("%lu*%lukB ", nr[order], K(1UL) << order);
L
Linus Torvalds 已提交
1613 1614 1615 1616 1617 1618 1619 1620
		printk("= %lukB\n", K(total));
	}

	show_swap_cache_info();
}

/*
 * Builds allocation fallback zone lists.
1621 1622
 *
 * Add all populated zones of a node to the zonelist.
L
Linus Torvalds 已提交
1623
 */
1624
static int __meminit build_zonelists_node(pg_data_t *pgdat,
1625
			struct zonelist *zonelist, int nr_zones, enum zone_type zone_type)
L
Linus Torvalds 已提交
1626
{
1627 1628
	struct zone *zone;

1629
	BUG_ON(zone_type >= MAX_NR_ZONES);
1630
	zone_type++;
1631 1632

	do {
1633
		zone_type--;
1634
		zone = pgdat->node_zones + zone_type;
1635
		if (populated_zone(zone)) {
1636 1637
			zonelist->zones[nr_zones++] = zone;
			check_highest_zone(zone_type);
L
Linus Torvalds 已提交
1638
		}
1639

1640
	} while (zone_type);
1641
	return nr_zones;
L
Linus Torvalds 已提交
1642 1643 1644 1645
}

#ifdef CONFIG_NUMA
#define MAX_NODE_LOAD (num_online_nodes())
1646
static int __meminitdata node_load[MAX_NUMNODES];
L
Linus Torvalds 已提交
1647
/**
1648
 * find_next_best_node - find the next node that should appear in a given node's fallback list
L
Linus Torvalds 已提交
1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660
 * @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.
 */
1661
static int __meminit find_next_best_node(int node, nodemask_t *used_node_mask)
L
Linus Torvalds 已提交
1662
{
1663
	int n, val;
L
Linus Torvalds 已提交
1664 1665 1666
	int min_val = INT_MAX;
	int best_node = -1;

1667 1668 1669 1670 1671
	/* 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
1672

1673 1674
	for_each_online_node(n) {
		cpumask_t tmp;
L
Linus Torvalds 已提交
1675 1676 1677 1678 1679 1680 1681 1682

		/* 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);

1683 1684 1685
		/* Penalize nodes under us ("prefer the next node") */
		val += (n < node);

L
Linus Torvalds 已提交
1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706
		/* 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;
}

1707
static void __meminit build_zonelists(pg_data_t *pgdat)
L
Linus Torvalds 已提交
1708
{
1709 1710
	int j, node, local_node;
	enum zone_type i;
L
Linus Torvalds 已提交
1711 1712 1713 1714 1715
	int prev_node, load;
	struct zonelist *zonelist;
	nodemask_t used_mask;

	/* initialize zonelists */
1716
	for (i = 0; i < MAX_NR_ZONES; i++) {
L
Linus Torvalds 已提交
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726
		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) {
1727 1728 1729 1730 1731 1732 1733 1734 1735
		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
1736 1737 1738 1739 1740
		/*
		 * 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.
		 */
1741 1742

		if (distance != node_distance(local_node, prev_node))
L
Linus Torvalds 已提交
1743 1744 1745
			node_load[node] += load;
		prev_node = node;
		load--;
1746
		for (i = 0; i < MAX_NR_ZONES; i++) {
L
Linus Torvalds 已提交
1747 1748 1749
			zonelist = pgdat->node_zonelists + i;
			for (j = 0; zonelist->zones[j] != NULL; j++);

1750
	 		j = build_zonelists_node(NODE_DATA(node), zonelist, j, i);
L
Linus Torvalds 已提交
1751 1752 1753 1754 1755
			zonelist->zones[j] = NULL;
		}
	}
}

1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773
/* Construct the zonelist performance cache - see further mmzone.h */
static void __meminit build_zonelist_cache(pg_data_t *pgdat)
{
	int i;

	for (i = 0; i < MAX_NR_ZONES; i++) {
		struct zonelist *zonelist;
		struct zonelist_cache *zlc;
		struct zone **z;

		zonelist = pgdat->node_zonelists + i;
		zonelist->zlcache_ptr = zlc = &zonelist->zlcache;
		bitmap_zero(zlc->fullzones, MAX_ZONES_PER_ZONELIST);
		for (z = zonelist->zones; *z; z++)
			zlc->z_to_n[z - zonelist->zones] = zone_to_nid(*z);
	}
}

L
Linus Torvalds 已提交
1774 1775
#else	/* CONFIG_NUMA */

1776
static void __meminit build_zonelists(pg_data_t *pgdat)
L
Linus Torvalds 已提交
1777
{
1778 1779
	int node, local_node;
	enum zone_type i,j;
L
Linus Torvalds 已提交
1780 1781

	local_node = pgdat->node_id;
1782
	for (i = 0; i < MAX_NR_ZONES; i++) {
L
Linus Torvalds 已提交
1783 1784 1785 1786
		struct zonelist *zonelist;

		zonelist = pgdat->node_zonelists + i;

1787
 		j = build_zonelists_node(pgdat, zonelist, 0, i);
L
Linus Torvalds 已提交
1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798
 		/*
 		 * 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;
1799
			j = build_zonelists_node(NODE_DATA(node), zonelist, j, i);
L
Linus Torvalds 已提交
1800 1801 1802 1803
		}
		for (node = 0; node < local_node; node++) {
			if (!node_online(node))
				continue;
1804
			j = build_zonelists_node(NODE_DATA(node), zonelist, j, i);
L
Linus Torvalds 已提交
1805 1806 1807 1808 1809 1810
		}

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

1811 1812 1813 1814 1815 1816 1817 1818 1819
/* non-NUMA variant of zonelist performance cache - just NULL zlcache_ptr */
static void __meminit build_zonelist_cache(pg_data_t *pgdat)
{
	int i;

	for (i = 0; i < MAX_NR_ZONES; i++)
		pgdat->node_zonelists[i].zlcache_ptr = NULL;
}

L
Linus Torvalds 已提交
1820 1821
#endif	/* CONFIG_NUMA */

1822 1823
/* return values int ....just for stop_machine_run() */
static int __meminit __build_all_zonelists(void *dummy)
L
Linus Torvalds 已提交
1824
{
1825
	int nid;
1826 1827

	for_each_online_node(nid) {
1828
		build_zonelists(NODE_DATA(nid));
1829 1830
		build_zonelist_cache(NODE_DATA(nid));
	}
1831 1832 1833 1834 1835 1836
	return 0;
}

void __meminit build_all_zonelists(void)
{
	if (system_state == SYSTEM_BOOTING) {
1837
		__build_all_zonelists(NULL);
1838 1839 1840 1841 1842 1843 1844
		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 */
	}
1845 1846 1847
	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
1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862
}

/*
 * 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

1863
#ifndef CONFIG_MEMORY_HOTPLUG
1864
static inline unsigned long wait_table_hash_nr_entries(unsigned long pages)
L
Linus Torvalds 已提交
1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
{
	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);
}
1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904
#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
1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922

/*
 * 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))

/*
 * 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.
 */
1923
void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,
D
Dave Hansen 已提交
1924
		unsigned long start_pfn, enum memmap_context context)
L
Linus Torvalds 已提交
1925 1926
{
	struct page *page;
A
Andy Whitcroft 已提交
1927 1928
	unsigned long end_pfn = start_pfn + size;
	unsigned long pfn;
L
Linus Torvalds 已提交
1929

1930
	for (pfn = start_pfn; pfn < end_pfn; pfn++) {
D
Dave Hansen 已提交
1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941
		/*
		 * There can be holes in boot-time mem_map[]s
		 * handed to this function.  They do not
		 * exist on hotplugged memory.
		 */
		if (context == MEMMAP_EARLY) {
			if (!early_pfn_valid(pfn))
				continue;
			if (!early_pfn_in_nid(pfn, nid))
				continue;
		}
A
Andy Whitcroft 已提交
1942 1943
		page = pfn_to_page(pfn);
		set_page_links(page, zone, nid, pfn);
1944
		init_page_count(page);
L
Linus Torvalds 已提交
1945 1946 1947 1948 1949 1950
		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))
1951
			set_page_address(page, __va(pfn << PAGE_SHIFT));
L
Linus Torvalds 已提交
1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967
#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;
	}
}

#ifndef __HAVE_ARCH_MEMMAP_INIT
#define memmap_init(size, nid, zone, start_pfn) \
D
Dave Hansen 已提交
1968
	memmap_init_zone((size), (nid), (zone), (start_pfn), MEMMAP_EARLY)
L
Linus Torvalds 已提交
1969 1970
#endif

1971
static int __cpuinit zone_batchsize(struct zone *zone)
1972 1973 1974 1975 1976
{
	int batch;

	/*
	 * The per-cpu-pages pools are set to around 1000th of the
1977
	 * size of the zone.  But no more than 1/2 of a meg.
1978 1979 1980 1981
	 *
	 * OK, so we don't know how big the cache is.  So guess.
	 */
	batch = zone->present_pages / 1024;
1982 1983
	if (batch * PAGE_SIZE > 512 * 1024)
		batch = (512 * 1024) / PAGE_SIZE;
1984 1985 1986 1987 1988
	batch /= 4;		/* We effectively *= 4 below */
	if (batch < 1)
		batch = 1;

	/*
1989 1990 1991
	 * 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.
1992
	 *
1993 1994 1995 1996
	 * 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.
1997
	 */
1998
	batch = (1 << (fls(batch + batch/2)-1)) - 1;
1999

2000 2001 2002
	return batch;
}

2003 2004 2005 2006
inline void setup_pageset(struct per_cpu_pageset *p, unsigned long batch)
{
	struct per_cpu_pages *pcp;

2007 2008
	memset(p, 0, sizeof(*p));

2009 2010 2011 2012 2013 2014 2015 2016 2017
	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;
2018
	pcp->batch = max(1UL, batch/2);
2019 2020 2021
	INIT_LIST_HEAD(&pcp->list);
}

2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039
/*
 * 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;
}


2040 2041
#ifdef CONFIG_NUMA
/*
2042 2043 2044 2045 2046 2047 2048
 * 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.
2049 2050 2051 2052 2053 2054 2055 2056
 *
 * 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.
2057
 */
2058
static struct per_cpu_pageset boot_pageset[NR_CPUS];
2059 2060 2061

/*
 * Dynamically allocate memory for the
2062 2063
 * per cpu pageset array in struct zone.
 */
2064
static int __cpuinit process_zones(int cpu)
2065 2066 2067 2068 2069
{
	struct zone *zone, *dzone;

	for_each_zone(zone) {

2070 2071 2072
		if (!populated_zone(zone))
			continue;

N
Nick Piggin 已提交
2073
		zone_pcp(zone, cpu) = kmalloc_node(sizeof(struct per_cpu_pageset),
2074
					 GFP_KERNEL, cpu_to_node(cpu));
N
Nick Piggin 已提交
2075
		if (!zone_pcp(zone, cpu))
2076 2077
			goto bad;

N
Nick Piggin 已提交
2078
		setup_pageset(zone_pcp(zone, cpu), zone_batchsize(zone));
2079 2080 2081 2082

		if (percpu_pagelist_fraction)
			setup_pagelist_highmark(zone_pcp(zone, cpu),
			 	(zone->present_pages / percpu_pagelist_fraction));
2083 2084 2085 2086 2087 2088 2089
	}

	return 0;
bad:
	for_each_zone(dzone) {
		if (dzone == zone)
			break;
N
Nick Piggin 已提交
2090 2091
		kfree(zone_pcp(dzone, cpu));
		zone_pcp(dzone, cpu) = NULL;
2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102
	}
	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);

2103 2104 2105
		/* Free per_cpu_pageset if it is slab allocated */
		if (pset != &boot_pageset[cpu])
			kfree(pset);
2106 2107 2108 2109
		zone_pcp(zone, cpu) = NULL;
	}
}

2110
static int __cpuinit pageset_cpuup_callback(struct notifier_block *nfb,
2111 2112 2113 2114 2115 2116 2117
		unsigned long action,
		void *hcpu)
{
	int cpu = (long)hcpu;
	int ret = NOTIFY_OK;

	switch (action) {
2118
	case CPU_UP_PREPARE:
2119
	case CPU_UP_PREPARE_FROZEN:
2120 2121 2122 2123
		if (process_zones(cpu))
			ret = NOTIFY_BAD;
		break;
	case CPU_UP_CANCELED:
2124
	case CPU_UP_CANCELED_FROZEN:
2125
	case CPU_DEAD:
2126
	case CPU_DEAD_FROZEN:
2127 2128 2129 2130
		free_zone_pagesets(cpu);
		break;
	default:
		break;
2131 2132 2133 2134
	}
	return ret;
}

2135
static struct notifier_block __cpuinitdata pageset_notifier =
2136 2137
	{ &pageset_cpuup_callback, NULL, 0 };

2138
void __init setup_per_cpu_pageset(void)
2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152
{
	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

S
Sam Ravnborg 已提交
2153
static noinline __init_refok
2154
int zone_wait_table_init(struct zone *zone, unsigned long zone_size_pages)
2155 2156 2157
{
	int i;
	struct pglist_data *pgdat = zone->zone_pgdat;
2158
	size_t alloc_size;
2159 2160 2161 2162 2163

	/*
	 * The per-page waitqueue mechanism uses hashed waitqueues
	 * per zone.
	 */
2164 2165 2166 2167
	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);
2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188
	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;
2189

2190
	for(i = 0; i < zone->wait_table_hash_nr_entries; ++i)
2191
		init_waitqueue_head(zone->wait_table + i);
2192 2193

	return 0;
2194 2195
}

2196
static __meminit void zone_pcp_init(struct zone *zone)
2197 2198 2199 2200 2201 2202 2203
{
	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
2204
		zone_pcp(zone, cpu) = &boot_pageset[cpu];
2205 2206 2207 2208 2209
		setup_pageset(&boot_pageset[cpu],0);
#else
		setup_pageset(zone_pcp(zone,cpu), batch);
#endif
	}
A
Anton Blanchard 已提交
2210 2211 2212
	if (zone->present_pages)
		printk(KERN_DEBUG "  %s zone: %lu pages, LIFO batch:%lu\n",
			zone->name, zone->present_pages, batch);
2213 2214
}

2215 2216
__meminit int init_currently_empty_zone(struct zone *zone,
					unsigned long zone_start_pfn,
D
Dave Hansen 已提交
2217 2218
					unsigned long size,
					enum memmap_context context)
2219 2220
{
	struct pglist_data *pgdat = zone->zone_pgdat;
2221 2222 2223 2224
	int ret;
	ret = zone_wait_table_init(zone, size);
	if (ret)
		return ret;
2225 2226 2227 2228 2229 2230 2231
	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);
2232 2233

	return 0;
2234 2235
}

2236 2237 2238 2239 2240
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP
/*
 * Basic iterator support. Return the first range of PFNs for a node
 * Note: nid == MAX_NUMNODES returns first region regardless of node
 */
2241
static int __meminit first_active_region_index_in_nid(int nid)
2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255
{
	int i;

	for (i = 0; i < nr_nodemap_entries; i++)
		if (nid == MAX_NUMNODES || early_node_map[i].nid == nid)
			return i;

	return -1;
}

/*
 * Basic iterator support. Return the next active range of PFNs for a node
 * Note: nid == MAX_NUMNODES returns next region regardles of node
 */
2256
static int __meminit next_active_region_index_in_nid(int index, int nid)
2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271
{
	for (index = index + 1; index < nr_nodemap_entries; index++)
		if (nid == MAX_NUMNODES || early_node_map[index].nid == nid)
			return index;

	return -1;
}

#ifndef CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID
/*
 * Required by SPARSEMEM. Given a PFN, return what node the PFN is on.
 * Architectures may implement their own version but if add_active_range()
 * was used and there are no special requirements, this is a convenient
 * alternative
 */
2272
int __meminit early_pfn_to_nid(unsigned long pfn)
2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294
{
	int i;

	for (i = 0; i < nr_nodemap_entries; i++) {
		unsigned long start_pfn = early_node_map[i].start_pfn;
		unsigned long end_pfn = early_node_map[i].end_pfn;

		if (start_pfn <= pfn && pfn < end_pfn)
			return early_node_map[i].nid;
	}

	return 0;
}
#endif /* CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID */

/* Basic iterator support to walk early_node_map[] */
#define for_each_active_range_index_in_nid(i, nid) \
	for (i = first_active_region_index_in_nid(nid); i != -1; \
				i = next_active_region_index_in_nid(i, nid))

/**
 * free_bootmem_with_active_regions - Call free_bootmem_node for each active range
2295 2296
 * @nid: The node to free memory on. If MAX_NUMNODES, all nodes are freed.
 * @max_low_pfn: The highest PFN that will be passed to free_bootmem_node
2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325
 *
 * If an architecture guarantees that all ranges registered with
 * add_active_ranges() contain no holes and may be freed, this
 * this function may be used instead of calling free_bootmem() manually.
 */
void __init free_bootmem_with_active_regions(int nid,
						unsigned long max_low_pfn)
{
	int i;

	for_each_active_range_index_in_nid(i, nid) {
		unsigned long size_pages = 0;
		unsigned long end_pfn = early_node_map[i].end_pfn;

		if (early_node_map[i].start_pfn >= max_low_pfn)
			continue;

		if (end_pfn > max_low_pfn)
			end_pfn = max_low_pfn;

		size_pages = end_pfn - early_node_map[i].start_pfn;
		free_bootmem_node(NODE_DATA(early_node_map[i].nid),
				PFN_PHYS(early_node_map[i].start_pfn),
				size_pages << PAGE_SHIFT);
	}
}

/**
 * sparse_memory_present_with_active_regions - Call memory_present for each active range
2326
 * @nid: The node to call memory_present for. If MAX_NUMNODES, all nodes will be used.
2327 2328 2329
 *
 * If an architecture guarantees that all ranges registered with
 * add_active_ranges() contain no holes and may be freed, this
2330
 * function may be used instead of calling memory_present() manually.
2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341
 */
void __init sparse_memory_present_with_active_regions(int nid)
{
	int i;

	for_each_active_range_index_in_nid(i, nid)
		memory_present(early_node_map[i].nid,
				early_node_map[i].start_pfn,
				early_node_map[i].end_pfn);
}

2342 2343 2344 2345 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 2391 2392 2393 2394 2395 2396 2397
/**
 * push_node_boundaries - Push node boundaries to at least the requested boundary
 * @nid: The nid of the node to push the boundary for
 * @start_pfn: The start pfn of the node
 * @end_pfn: The end pfn of the node
 *
 * In reserve-based hot-add, mem_map is allocated that is unused until hotadd
 * time. Specifically, on x86_64, SRAT will report ranges that can potentially
 * be hotplugged even though no physical memory exists. This function allows
 * an arch to push out the node boundaries so mem_map is allocated that can
 * be used later.
 */
#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE
void __init push_node_boundaries(unsigned int nid,
		unsigned long start_pfn, unsigned long end_pfn)
{
	printk(KERN_DEBUG "Entering push_node_boundaries(%u, %lu, %lu)\n",
			nid, start_pfn, end_pfn);

	/* Initialise the boundary for this node if necessary */
	if (node_boundary_end_pfn[nid] == 0)
		node_boundary_start_pfn[nid] = -1UL;

	/* Update the boundaries */
	if (node_boundary_start_pfn[nid] > start_pfn)
		node_boundary_start_pfn[nid] = start_pfn;
	if (node_boundary_end_pfn[nid] < end_pfn)
		node_boundary_end_pfn[nid] = end_pfn;
}

/* If necessary, push the node boundary out for reserve hotadd */
static void __init account_node_boundary(unsigned int nid,
		unsigned long *start_pfn, unsigned long *end_pfn)
{
	printk(KERN_DEBUG "Entering account_node_boundary(%u, %lu, %lu)\n",
			nid, *start_pfn, *end_pfn);

	/* Return if boundary information has not been provided */
	if (node_boundary_end_pfn[nid] == 0)
		return;

	/* Check the boundaries and update if necessary */
	if (node_boundary_start_pfn[nid] < *start_pfn)
		*start_pfn = node_boundary_start_pfn[nid];
	if (node_boundary_end_pfn[nid] > *end_pfn)
		*end_pfn = node_boundary_end_pfn[nid];
}
#else
void __init push_node_boundaries(unsigned int nid,
		unsigned long start_pfn, unsigned long end_pfn) {}

static void __init account_node_boundary(unsigned int nid,
		unsigned long *start_pfn, unsigned long *end_pfn) {}
#endif


2398 2399
/**
 * get_pfn_range_for_nid - Return the start and end page frames for a node
2400 2401 2402
 * @nid: The nid to return the range for. If MAX_NUMNODES, the min and max PFN are returned.
 * @start_pfn: Passed by reference. On return, it will have the node start_pfn.
 * @end_pfn: Passed by reference. On return, it will have the node end_pfn.
2403 2404 2405 2406
 *
 * It returns the start and end page frame of a node based on information
 * provided by an arch calling add_active_range(). If called for a node
 * with no available memory, a warning is printed and the start and end
2407
 * PFNs will be 0.
2408
 */
2409
void __meminit get_pfn_range_for_nid(unsigned int nid,
2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424
			unsigned long *start_pfn, unsigned long *end_pfn)
{
	int i;
	*start_pfn = -1UL;
	*end_pfn = 0;

	for_each_active_range_index_in_nid(i, nid) {
		*start_pfn = min(*start_pfn, early_node_map[i].start_pfn);
		*end_pfn = max(*end_pfn, early_node_map[i].end_pfn);
	}

	if (*start_pfn == -1UL) {
		printk(KERN_WARNING "Node %u active with no memory\n", nid);
		*start_pfn = 0;
	}
2425 2426 2427

	/* Push the node boundaries out if requested */
	account_node_boundary(nid, start_pfn, end_pfn);
2428 2429 2430 2431 2432 2433
}

/*
 * Return the number of pages a zone spans in a node, including holes
 * present_pages = zone_spanned_pages_in_node() - zone_absent_pages_in_node()
 */
2434
unsigned long __meminit zone_spanned_pages_in_node(int nid,
2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459
					unsigned long zone_type,
					unsigned long *ignored)
{
	unsigned long node_start_pfn, node_end_pfn;
	unsigned long zone_start_pfn, zone_end_pfn;

	/* Get the start and end of the node and zone */
	get_pfn_range_for_nid(nid, &node_start_pfn, &node_end_pfn);
	zone_start_pfn = arch_zone_lowest_possible_pfn[zone_type];
	zone_end_pfn = arch_zone_highest_possible_pfn[zone_type];

	/* Check that this node has pages within the zone's required range */
	if (zone_end_pfn < node_start_pfn || zone_start_pfn > node_end_pfn)
		return 0;

	/* Move the zone boundaries inside the node if necessary */
	zone_end_pfn = min(zone_end_pfn, node_end_pfn);
	zone_start_pfn = max(zone_start_pfn, node_start_pfn);

	/* Return the spanned pages */
	return zone_end_pfn - zone_start_pfn;
}

/*
 * Return the number of holes in a range on a node. If nid is MAX_NUMNODES,
2460
 * then all holes in the requested range will be accounted for.
2461
 */
2462
unsigned long __meminit __absent_pages_in_range(int nid,
2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474
				unsigned long range_start_pfn,
				unsigned long range_end_pfn)
{
	int i = 0;
	unsigned long prev_end_pfn = 0, hole_pages = 0;
	unsigned long start_pfn;

	/* Find the end_pfn of the first active range of pfns in the node */
	i = first_active_region_index_in_nid(nid);
	if (i == -1)
		return 0;

2475 2476 2477 2478
	/* Account for ranges before physical memory on this node */
	if (early_node_map[i].start_pfn > range_start_pfn)
		hole_pages = early_node_map[i].start_pfn - range_start_pfn;

2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499
	prev_end_pfn = early_node_map[i].start_pfn;

	/* Find all holes for the zone within the node */
	for (; i != -1; i = next_active_region_index_in_nid(i, nid)) {

		/* No need to continue if prev_end_pfn is outside the zone */
		if (prev_end_pfn >= range_end_pfn)
			break;

		/* Make sure the end of the zone is not within the hole */
		start_pfn = min(early_node_map[i].start_pfn, range_end_pfn);
		prev_end_pfn = max(prev_end_pfn, range_start_pfn);

		/* Update the hole size cound and move on */
		if (start_pfn > range_start_pfn) {
			BUG_ON(prev_end_pfn > start_pfn);
			hole_pages += start_pfn - prev_end_pfn;
		}
		prev_end_pfn = early_node_map[i].end_pfn;
	}

2500 2501
	/* Account for ranges past physical memory on this node */
	if (range_end_pfn > prev_end_pfn)
2502
		hole_pages += range_end_pfn -
2503 2504
				max(range_start_pfn, prev_end_pfn);

2505 2506 2507 2508 2509 2510 2511 2512
	return hole_pages;
}

/**
 * absent_pages_in_range - Return number of page frames in holes within a range
 * @start_pfn: The start PFN to start searching for holes
 * @end_pfn: The end PFN to stop searching for holes
 *
2513
 * It returns the number of pages frames in memory holes within a range.
2514 2515 2516 2517 2518 2519 2520 2521
 */
unsigned long __init absent_pages_in_range(unsigned long start_pfn,
							unsigned long end_pfn)
{
	return __absent_pages_in_range(MAX_NUMNODES, start_pfn, end_pfn);
}

/* Return the number of page frames in holes in a zone on a node */
2522
unsigned long __meminit zone_absent_pages_in_node(int nid,
2523 2524 2525
					unsigned long zone_type,
					unsigned long *ignored)
{
2526 2527 2528 2529 2530 2531 2532 2533 2534 2535
	unsigned long node_start_pfn, node_end_pfn;
	unsigned long zone_start_pfn, zone_end_pfn;

	get_pfn_range_for_nid(nid, &node_start_pfn, &node_end_pfn);
	zone_start_pfn = max(arch_zone_lowest_possible_pfn[zone_type],
							node_start_pfn);
	zone_end_pfn = min(arch_zone_highest_possible_pfn[zone_type],
							node_end_pfn);

	return __absent_pages_in_range(nid, zone_start_pfn, zone_end_pfn);
2536
}
2537

2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554
#else
static inline unsigned long zone_spanned_pages_in_node(int nid,
					unsigned long zone_type,
					unsigned long *zones_size)
{
	return zones_size[zone_type];
}

static inline unsigned long zone_absent_pages_in_node(int nid,
						unsigned long zone_type,
						unsigned long *zholes_size)
{
	if (!zholes_size)
		return 0;

	return zholes_size[zone_type];
}
2555

2556 2557
#endif

2558
static void __meminit calculate_node_totalpages(struct pglist_data *pgdat,
2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578
		unsigned long *zones_size, unsigned long *zholes_size)
{
	unsigned long realtotalpages, totalpages = 0;
	enum zone_type i;

	for (i = 0; i < MAX_NR_ZONES; i++)
		totalpages += zone_spanned_pages_in_node(pgdat->node_id, i,
								zones_size);
	pgdat->node_spanned_pages = totalpages;

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

L
Linus Torvalds 已提交
2579 2580 2581 2582 2583 2584
/*
 * Set up the zone data structures:
 *   - mark all pages reserved
 *   - mark all memory queues empty
 *   - clear the memory bitmaps
 */
2585
static void __meminit free_area_init_core(struct pglist_data *pgdat,
L
Linus Torvalds 已提交
2586 2587
		unsigned long *zones_size, unsigned long *zholes_size)
{
2588
	enum zone_type j;
2589
	int nid = pgdat->node_id;
L
Linus Torvalds 已提交
2590
	unsigned long zone_start_pfn = pgdat->node_start_pfn;
2591
	int ret;
L
Linus Torvalds 已提交
2592

2593
	pgdat_resize_init(pgdat);
L
Linus Torvalds 已提交
2594 2595 2596 2597 2598 2599
	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;
2600
		unsigned long size, realsize, memmap_pages;
L
Linus Torvalds 已提交
2601

2602 2603 2604
		size = zone_spanned_pages_in_node(nid, j, zones_size);
		realsize = size - zone_absent_pages_in_node(nid, j,
								zholes_size);
L
Linus Torvalds 已提交
2605

2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621
		/*
		 * Adjust realsize so that it accounts for how much memory
		 * is used by this zone for memmap. This affects the watermark
		 * and per-cpu initialisations
		 */
		memmap_pages = (size * sizeof(struct page)) >> PAGE_SHIFT;
		if (realsize >= memmap_pages) {
			realsize -= memmap_pages;
			printk(KERN_DEBUG
				"  %s zone: %lu pages used for memmap\n",
				zone_names[j], memmap_pages);
		} else
			printk(KERN_WARNING
				"  %s zone: %lu pages exceeds realsize %lu\n",
				zone_names[j], memmap_pages, realsize);

2622 2623
		/* Account for reserved pages */
		if (j == 0 && realsize > dma_reserve) {
2624
			realsize -= dma_reserve;
2625 2626
			printk(KERN_DEBUG "  %s zone: %lu pages reserved\n",
					zone_names[0], dma_reserve);
2627 2628
		}

2629
		if (!is_highmem_idx(j))
L
Linus Torvalds 已提交
2630 2631 2632 2633 2634
			nr_kernel_pages += realsize;
		nr_all_pages += realsize;

		zone->spanned_pages = size;
		zone->present_pages = realsize;
2635
#ifdef CONFIG_NUMA
2636
		zone->node = nid;
2637
		zone->min_unmapped_pages = (realsize*sysctl_min_unmapped_ratio)
2638
						/ 100;
2639
		zone->min_slab_pages = (realsize * sysctl_min_slab_ratio) / 100;
2640
#endif
L
Linus Torvalds 已提交
2641 2642 2643
		zone->name = zone_names[j];
		spin_lock_init(&zone->lock);
		spin_lock_init(&zone->lru_lock);
2644
		zone_seqlock_init(zone);
L
Linus Torvalds 已提交
2645 2646
		zone->zone_pgdat = pgdat;

2647
		zone->prev_priority = DEF_PRIORITY;
L
Linus Torvalds 已提交
2648

2649
		zone_pcp_init(zone);
L
Linus Torvalds 已提交
2650 2651 2652 2653
		INIT_LIST_HEAD(&zone->active_list);
		INIT_LIST_HEAD(&zone->inactive_list);
		zone->nr_scan_active = 0;
		zone->nr_scan_inactive = 0;
2654
		zap_zone_vm_stats(zone);
2655
		atomic_set(&zone->reclaim_in_progress, 0);
L
Linus Torvalds 已提交
2656 2657 2658
		if (!size)
			continue;

D
Dave Hansen 已提交
2659 2660
		ret = init_currently_empty_zone(zone, zone_start_pfn,
						size, MEMMAP_EARLY);
2661
		BUG_ON(ret);
L
Linus Torvalds 已提交
2662 2663 2664 2665
		zone_start_pfn += size;
	}
}

S
Sam Ravnborg 已提交
2666
static void __init_refok alloc_node_mem_map(struct pglist_data *pgdat)
L
Linus Torvalds 已提交
2667 2668 2669 2670 2671
{
	/* Skip empty nodes */
	if (!pgdat->node_spanned_pages)
		return;

A
Andy Whitcroft 已提交
2672
#ifdef CONFIG_FLAT_NODE_MEM_MAP
L
Linus Torvalds 已提交
2673 2674
	/* ia64 gets its own node_mem_map, before this, without bootmem */
	if (!pgdat->node_mem_map) {
2675
		unsigned long size, start, end;
A
Andy Whitcroft 已提交
2676 2677
		struct page *map;

2678 2679 2680 2681 2682 2683 2684 2685 2686
		/*
		 * 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);
2687 2688 2689
		map = alloc_remap(pgdat->node_id, size);
		if (!map)
			map = alloc_bootmem_node(pgdat, size);
2690
		pgdat->node_mem_map = map + (pgdat->node_start_pfn - start);
L
Linus Torvalds 已提交
2691
	}
A
Andy Whitcroft 已提交
2692
#ifdef CONFIG_FLATMEM
L
Linus Torvalds 已提交
2693 2694 2695
	/*
	 * With no DISCONTIG, the global mem_map is just set as node 0's
	 */
2696
	if (pgdat == NODE_DATA(0)) {
L
Linus Torvalds 已提交
2697
		mem_map = NODE_DATA(0)->node_mem_map;
2698 2699 2700 2701 2702
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP
		if (page_to_pfn(mem_map) != pgdat->node_start_pfn)
			mem_map -= pgdat->node_start_pfn;
#endif /* CONFIG_ARCH_POPULATES_NODE_MAP */
	}
L
Linus Torvalds 已提交
2703
#endif
A
Andy Whitcroft 已提交
2704
#endif /* CONFIG_FLAT_NODE_MEM_MAP */
L
Linus Torvalds 已提交
2705 2706
}

2707
void __meminit free_area_init_node(int nid, struct pglist_data *pgdat,
L
Linus Torvalds 已提交
2708 2709 2710 2711 2712
		unsigned long *zones_size, unsigned long node_start_pfn,
		unsigned long *zholes_size)
{
	pgdat->node_id = nid;
	pgdat->node_start_pfn = node_start_pfn;
2713
	calculate_node_totalpages(pgdat, zones_size, zholes_size);
L
Linus Torvalds 已提交
2714 2715 2716 2717 2718 2719

	alloc_node_mem_map(pgdat);

	free_area_init_core(pgdat, zones_size, zholes_size);
}

2720
#ifdef CONFIG_ARCH_POPULATES_NODE_MAP
M
Miklos Szeredi 已提交
2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740

#if MAX_NUMNODES > 1
/*
 * Figure out the number of possible node ids.
 */
static void __init setup_nr_node_ids(void)
{
	unsigned int node;
	unsigned int highest = 0;

	for_each_node_mask(node, node_possible_map)
		highest = node;
	nr_node_ids = highest + 1;
}
#else
static inline void setup_nr_node_ids(void)
{
}
#endif

2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826
/**
 * add_active_range - Register a range of PFNs backed by physical memory
 * @nid: The node ID the range resides on
 * @start_pfn: The start PFN of the available physical memory
 * @end_pfn: The end PFN of the available physical memory
 *
 * These ranges are stored in an early_node_map[] and later used by
 * free_area_init_nodes() to calculate zone sizes and holes. If the
 * range spans a memory hole, it is up to the architecture to ensure
 * the memory is not freed by the bootmem allocator. If possible
 * the range being registered will be merged with existing ranges.
 */
void __init add_active_range(unsigned int nid, unsigned long start_pfn,
						unsigned long end_pfn)
{
	int i;

	printk(KERN_DEBUG "Entering add_active_range(%d, %lu, %lu) "
			  "%d entries of %d used\n",
			  nid, start_pfn, end_pfn,
			  nr_nodemap_entries, MAX_ACTIVE_REGIONS);

	/* Merge with existing active regions if possible */
	for (i = 0; i < nr_nodemap_entries; i++) {
		if (early_node_map[i].nid != nid)
			continue;

		/* Skip if an existing region covers this new one */
		if (start_pfn >= early_node_map[i].start_pfn &&
				end_pfn <= early_node_map[i].end_pfn)
			return;

		/* Merge forward if suitable */
		if (start_pfn <= early_node_map[i].end_pfn &&
				end_pfn > early_node_map[i].end_pfn) {
			early_node_map[i].end_pfn = end_pfn;
			return;
		}

		/* Merge backward if suitable */
		if (start_pfn < early_node_map[i].end_pfn &&
				end_pfn >= early_node_map[i].start_pfn) {
			early_node_map[i].start_pfn = start_pfn;
			return;
		}
	}

	/* Check that early_node_map is large enough */
	if (i >= MAX_ACTIVE_REGIONS) {
		printk(KERN_CRIT "More than %d memory regions, truncating\n",
							MAX_ACTIVE_REGIONS);
		return;
	}

	early_node_map[i].nid = nid;
	early_node_map[i].start_pfn = start_pfn;
	early_node_map[i].end_pfn = end_pfn;
	nr_nodemap_entries = i + 1;
}

/**
 * shrink_active_range - Shrink an existing registered range of PFNs
 * @nid: The node id the range is on that should be shrunk
 * @old_end_pfn: The old end PFN of the range
 * @new_end_pfn: The new PFN of the range
 *
 * i386 with NUMA use alloc_remap() to store a node_mem_map on a local node.
 * The map is kept at the end physical page range that has already been
 * registered with add_active_range(). This function allows an arch to shrink
 * an existing registered range.
 */
void __init shrink_active_range(unsigned int nid, unsigned long old_end_pfn,
						unsigned long new_end_pfn)
{
	int i;

	/* Find the old active region end and shrink */
	for_each_active_range_index_in_nid(i, nid)
		if (early_node_map[i].end_pfn == old_end_pfn) {
			early_node_map[i].end_pfn = new_end_pfn;
			break;
		}
}

/**
 * remove_all_active_ranges - Remove all currently registered regions
2827
 *
2828 2829 2830 2831
 * During discovery, it may be found that a table like SRAT is invalid
 * and an alternative discovery method must be used. This function removes
 * all currently registered regions.
 */
2832
void __init remove_all_active_ranges(void)
2833 2834 2835
{
	memset(early_node_map, 0, sizeof(early_node_map));
	nr_nodemap_entries = 0;
2836 2837 2838 2839
#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE
	memset(node_boundary_start_pfn, 0, sizeof(node_boundary_start_pfn));
	memset(node_boundary_end_pfn, 0, sizeof(node_boundary_end_pfn));
#endif /* CONFIG_MEMORY_HOTPLUG_RESERVE */
2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864
}

/* Compare two active node_active_regions */
static int __init cmp_node_active_region(const void *a, const void *b)
{
	struct node_active_region *arange = (struct node_active_region *)a;
	struct node_active_region *brange = (struct node_active_region *)b;

	/* Done this way to avoid overflows */
	if (arange->start_pfn > brange->start_pfn)
		return 1;
	if (arange->start_pfn < brange->start_pfn)
		return -1;

	return 0;
}

/* sort the node_map by start_pfn */
static void __init sort_node_map(void)
{
	sort(early_node_map, (size_t)nr_nodemap_entries,
			sizeof(struct node_active_region),
			cmp_node_active_region, NULL);
}

2865
/* Find the lowest pfn for a node */
2866 2867 2868
unsigned long __init find_min_pfn_for_node(unsigned long nid)
{
	int i;
2869
	unsigned long min_pfn = ULONG_MAX;
2870

2871 2872
	/* Assuming a sorted map, the first range found has the starting pfn */
	for_each_active_range_index_in_nid(i, nid)
2873
		min_pfn = min(min_pfn, early_node_map[i].start_pfn);
2874

2875 2876 2877 2878 2879 2880 2881
	if (min_pfn == ULONG_MAX) {
		printk(KERN_WARNING
			"Could not find start_pfn for node %lu\n", nid);
		return 0;
	}

	return min_pfn;
2882 2883 2884 2885 2886 2887
}

/**
 * find_min_pfn_with_active_regions - Find the minimum PFN registered
 *
 * It returns the minimum PFN based on information provided via
2888
 * add_active_range().
2889 2890 2891 2892 2893 2894 2895 2896 2897 2898
 */
unsigned long __init find_min_pfn_with_active_regions(void)
{
	return find_min_pfn_for_node(MAX_NUMNODES);
}

/**
 * find_max_pfn_with_active_regions - Find the maximum PFN registered
 *
 * It returns the maximum PFN based on information provided via
2899
 * add_active_range().
2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913
 */
unsigned long __init find_max_pfn_with_active_regions(void)
{
	int i;
	unsigned long max_pfn = 0;

	for (i = 0; i < nr_nodemap_entries; i++)
		max_pfn = max(max_pfn, early_node_map[i].end_pfn);

	return max_pfn;
}

/**
 * free_area_init_nodes - Initialise all pg_data_t and zone data
2914
 * @max_zone_pfn: an array of max PFNs for each zone
2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929
 *
 * This will call free_area_init_node() for each active node in the system.
 * Using the page ranges provided by add_active_range(), the size of each
 * zone in each node and their holes is calculated. If the maximum PFN
 * between two adjacent zones match, it is assumed that the zone is empty.
 * For example, if arch_max_dma_pfn == arch_max_dma32_pfn, it is assumed
 * that arch_max_dma32_pfn has no pages. It is also assumed that a zone
 * starts where the previous one ended. For example, ZONE_DMA32 starts
 * at arch_max_dma_pfn.
 */
void __init free_area_init_nodes(unsigned long *max_zone_pfn)
{
	unsigned long nid;
	enum zone_type i;

2930 2931 2932
	/* Sort early_node_map as initialisation assumes it is sorted */
	sort_node_map();

2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962
	/* Record where the zone boundaries are */
	memset(arch_zone_lowest_possible_pfn, 0,
				sizeof(arch_zone_lowest_possible_pfn));
	memset(arch_zone_highest_possible_pfn, 0,
				sizeof(arch_zone_highest_possible_pfn));
	arch_zone_lowest_possible_pfn[0] = find_min_pfn_with_active_regions();
	arch_zone_highest_possible_pfn[0] = max_zone_pfn[0];
	for (i = 1; i < MAX_NR_ZONES; i++) {
		arch_zone_lowest_possible_pfn[i] =
			arch_zone_highest_possible_pfn[i-1];
		arch_zone_highest_possible_pfn[i] =
			max(max_zone_pfn[i], arch_zone_lowest_possible_pfn[i]);
	}

	/* Print out the zone ranges */
	printk("Zone PFN ranges:\n");
	for (i = 0; i < MAX_NR_ZONES; i++)
		printk("  %-8s %8lu -> %8lu\n",
				zone_names[i],
				arch_zone_lowest_possible_pfn[i],
				arch_zone_highest_possible_pfn[i]);

	/* Print out the early_node_map[] */
	printk("early_node_map[%d] active PFN ranges\n", nr_nodemap_entries);
	for (i = 0; i < nr_nodemap_entries; i++)
		printk("  %3d: %8lu -> %8lu\n", early_node_map[i].nid,
						early_node_map[i].start_pfn,
						early_node_map[i].end_pfn);

	/* Initialise every node */
2963
	setup_nr_node_ids();
2964 2965 2966 2967 2968 2969 2970 2971
	for_each_online_node(nid) {
		pg_data_t *pgdat = NODE_DATA(nid);
		free_area_init_node(nid, pgdat, NULL,
				find_min_pfn_for_node(nid), NULL);
	}
}
#endif /* CONFIG_ARCH_POPULATES_NODE_MAP */

2972
/**
2973 2974
 * set_dma_reserve - set the specified number of pages reserved in the first zone
 * @new_dma_reserve: The number of pages to mark reserved
2975 2976 2977 2978
 *
 * The per-cpu batchsize and zone watermarks are determined by present_pages.
 * In the DMA zone, a significant percentage may be consumed by kernel image
 * and other unfreeable allocations which can skew the watermarks badly. This
2979 2980 2981
 * function may optionally be used to account for unfreeable pages in the
 * first zone (e.g., ZONE_DMA). The effect will be lower watermarks and
 * smaller per-cpu batchsize.
2982 2983 2984 2985 2986 2987
 */
void __init set_dma_reserve(unsigned long new_dma_reserve)
{
	dma_reserve = new_dma_reserve;
}

2988
#ifndef CONFIG_NEED_MULTIPLE_NODES
L
Linus Torvalds 已提交
2989 2990 2991 2992
static bootmem_data_t contig_bootmem_data;
struct pglist_data contig_page_data = { .bdata = &contig_bootmem_data };

EXPORT_SYMBOL(contig_page_data);
2993
#endif
L
Linus Torvalds 已提交
2994 2995 2996

void __init free_area_init(unsigned long *zones_size)
{
2997
	free_area_init_node(0, NODE_DATA(0), zones_size,
L
Linus Torvalds 已提交
2998 2999 3000 3001 3002 3003 3004 3005
			__pa(PAGE_OFFSET) >> PAGE_SHIFT, NULL);
}

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

3006
	if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) {
L
Linus Torvalds 已提交
3007 3008
		local_irq_disable();
		__drain_pages(cpu);
3009
		vm_events_fold_cpu(cpu);
L
Linus Torvalds 已提交
3010
		local_irq_enable();
3011
		refresh_cpu_vm_stats(cpu);
L
Linus Torvalds 已提交
3012 3013 3014 3015 3016 3017 3018 3019 3020
	}
	return NOTIFY_OK;
}

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

3021 3022 3023 3024 3025 3026 3027 3028
/*
 * 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;
3029
	enum zone_type i, j;
3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052

	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
3053 3054 3055 3056 3057 3058 3059 3060 3061
/*
 * 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;
3062
	enum zone_type j, idx;
L
Linus Torvalds 已提交
3063

3064
	for_each_online_pgdat(pgdat) {
L
Linus Torvalds 已提交
3065 3066 3067 3068 3069 3070
		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;

3071 3072
			idx = j;
			while (idx) {
L
Linus Torvalds 已提交
3073 3074
				struct zone *lower_zone;

3075 3076
				idx--;

L
Linus Torvalds 已提交
3077 3078 3079 3080 3081 3082 3083 3084 3085 3086
				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;
			}
		}
	}
3087 3088 3089

	/* update totalreserve_pages */
	calculate_totalreserve_pages();
L
Linus Torvalds 已提交
3090 3091
}

3092 3093 3094 3095 3096
/**
 * 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.
L
Linus Torvalds 已提交
3097
 */
3098
void setup_per_zone_pages_min(void)
L
Linus Torvalds 已提交
3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111
{
	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) {
3112 3113
		u64 tmp;

L
Linus Torvalds 已提交
3114
		spin_lock_irqsave(&zone->lru_lock, flags);
3115 3116
		tmp = (u64)pages_min * zone->present_pages;
		do_div(tmp, lowmem_pages);
L
Linus Torvalds 已提交
3117 3118
		if (is_highmem(zone)) {
			/*
N
Nick Piggin 已提交
3119 3120 3121 3122 3123 3124 3125
			 * __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
3126 3127 3128 3129 3130 3131 3132 3133 3134 3135
			 */
			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
3136 3137
			/*
			 * If it's a lowmem zone, reserve a number of pages
L
Linus Torvalds 已提交
3138 3139
			 * proportionate to the zone's size.
			 */
N
Nick Piggin 已提交
3140
			zone->pages_min = tmp;
L
Linus Torvalds 已提交
3141 3142
		}

3143 3144
		zone->pages_low   = zone->pages_min + (tmp >> 2);
		zone->pages_high  = zone->pages_min + (tmp >> 1);
L
Linus Torvalds 已提交
3145 3146
		spin_unlock_irqrestore(&zone->lru_lock, flags);
	}
3147 3148 3149

	/* update totalreserve_pages */
	calculate_totalreserve_pages();
L
Linus Torvalds 已提交
3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201
}

/*
 * 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);
3202 3203
	if (write)
		setup_per_zone_pages_min();
L
Linus Torvalds 已提交
3204 3205 3206
	return 0;
}

3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218
#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)
3219
		zone->min_unmapped_pages = (zone->present_pages *
3220 3221 3222
				sysctl_min_unmapped_ratio) / 100;
	return 0;
}
3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238

int sysctl_min_slab_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_slab_pages = (zone->present_pages *
				sysctl_min_slab_ratio) / 100;
	return 0;
}
3239 3240
#endif

L
Linus Torvalds 已提交
3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257
/*
 * 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;
}

3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283
/*
 * 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;
}

3284
int hashdist = HASHDIST_DEFAULT;
L
Linus Torvalds 已提交
3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318

#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 */
A
Andrew Morton 已提交
3319
		numentries = nr_kernel_pages;
L
Linus Torvalds 已提交
3320 3321 3322 3323 3324 3325 3326 3327 3328
		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);
3329 3330 3331 3332

		/* Make sure we've got at least a 0-order allocation.. */
		if (unlikely((numentries * bucketsize) < PAGE_SIZE))
			numentries = PAGE_SIZE / bucketsize;
L
Linus Torvalds 已提交
3333
	}
3334
	numentries = roundup_pow_of_two(numentries);
L
Linus Torvalds 已提交
3335 3336 3337 3338 3339 3340 3341 3342 3343 3344

	/* 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;

3345
	log2qty = ilog2(numentries);
L
Linus Torvalds 已提交
3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366

	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),
3367
	       ilog2(size) - PAGE_SHIFT,
L
Linus Torvalds 已提交
3368 3369 3370 3371 3372 3373 3374 3375 3376
	       size);

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

	return table;
}
3377 3378 3379 3380

#ifdef CONFIG_OUT_OF_LINE_PFN_TO_PAGE
struct page *pfn_to_page(unsigned long pfn)
{
3381
	return __pfn_to_page(pfn);
3382 3383 3384
}
unsigned long page_to_pfn(struct page *page)
{
3385
	return __page_to_pfn(page);
3386 3387 3388 3389
}
EXPORT_SYMBOL(pfn_to_page);
EXPORT_SYMBOL(page_to_pfn);
#endif /* CONFIG_OUT_OF_LINE_PFN_TO_PAGE */
3390 3391