page_alloc.c 211 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*
 *  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>
22
#include <linux/jiffies.h>
Linus Torvalds's avatar
Linus Torvalds committed
23
#include <linux/bootmem.h>
24
#include <linux/memblock.h>
Linus Torvalds's avatar
Linus Torvalds committed
25
#include <linux/compiler.h>
26
#include <linux/kernel.h>
27
#include <linux/kasan.h>
Linus Torvalds's avatar
Linus Torvalds committed
28 29 30 31 32
#include <linux/module.h>
#include <linux/suspend.h>
#include <linux/pagevec.h>
#include <linux/blkdev.h>
#include <linux/slab.h>
33
#include <linux/ratelimit.h>
34
#include <linux/oom.h>
Linus Torvalds's avatar
Linus Torvalds committed
35 36 37 38 39
#include <linux/notifier.h>
#include <linux/topology.h>
#include <linux/sysctl.h>
#include <linux/cpu.h>
#include <linux/cpuset.h>
40
#include <linux/memory_hotplug.h>
Linus Torvalds's avatar
Linus Torvalds committed
41 42
#include <linux/nodemask.h>
#include <linux/vmalloc.h>
43
#include <linux/vmstat.h>
44
#include <linux/mempolicy.h>
45
#include <linux/memremap.h>
46
#include <linux/stop_machine.h>
47 48
#include <linux/sort.h>
#include <linux/pfn.h>
49
#include <linux/backing-dev.h>
50
#include <linux/fault-inject.h>
51
#include <linux/page-isolation.h>
52
#include <linux/page_ext.h>
53
#include <linux/debugobjects.h>
54
#include <linux/kmemleak.h>
55
#include <linux/compaction.h>
56
#include <trace/events/kmem.h>
57
#include <trace/events/oom.h>
58
#include <linux/prefetch.h>
59
#include <linux/mm_inline.h>
60
#include <linux/migrate.h>
61
#include <linux/hugetlb.h>
62
#include <linux/sched/rt.h>
63
#include <linux/sched/mm.h>
64
#include <linux/page_owner.h>
65
#include <linux/kthread.h>
66
#include <linux/memcontrol.h>
67
#include <linux/ftrace.h>
68
#include <linux/lockdep.h>
69
#include <linux/nmi.h>
Linus Torvalds's avatar
Linus Torvalds committed
70

71
#include <asm/sections.h>
Linus Torvalds's avatar
Linus Torvalds committed
72
#include <asm/tlbflush.h>
73
#include <asm/div64.h>
Linus Torvalds's avatar
Linus Torvalds committed
74 75
#include "internal.h"

76 77
/* prevent >1 _updater_ of zone percpu pageset ->high and ->batch fields */
static DEFINE_MUTEX(pcp_batch_high_lock);
78
#define MIN_PERCPU_PAGELIST_FRACTION	(8)
79

80 81 82 83 84
#ifdef CONFIG_USE_PERCPU_NUMA_NODE_ID
DEFINE_PER_CPU(int, numa_node);
EXPORT_PER_CPU_SYMBOL(numa_node);
#endif

85 86 87 88 89 90 91 92 93
#ifdef CONFIG_HAVE_MEMORYLESS_NODES
/*
 * N.B., Do NOT reference the '_numa_mem_' per cpu variable directly.
 * It will not be defined when CONFIG_HAVE_MEMORYLESS_NODES is not defined.
 * Use the accessor functions set_numa_mem(), numa_mem_id() and cpu_to_mem()
 * defined in <linux/topology.h>.
 */
DEFINE_PER_CPU(int, _numa_mem_);		/* Kernel "local memory" node */
EXPORT_PER_CPU_SYMBOL(_numa_mem_);
94
int _node_numa_mem_[MAX_NUMNODES];
95 96
#endif

97 98 99 100
/* work_structs for global per-cpu drains */
DEFINE_MUTEX(pcpu_drain_mutex);
DEFINE_PER_CPU(struct work_struct, pcpu_drain);

101
#ifdef CONFIG_GCC_PLUGIN_LATENT_ENTROPY
102
volatile unsigned long latent_entropy __latent_entropy;
103 104 105
EXPORT_SYMBOL(latent_entropy);
#endif

Linus Torvalds's avatar
Linus Torvalds committed
106
/*
107
 * Array of node states.
Linus Torvalds's avatar
Linus Torvalds committed
108
 */
109 110 111 112 113 114 115
nodemask_t node_states[NR_NODE_STATES] __read_mostly = {
	[N_POSSIBLE] = NODE_MASK_ALL,
	[N_ONLINE] = { { [0] = 1UL } },
#ifndef CONFIG_NUMA
	[N_NORMAL_MEMORY] = { { [0] = 1UL } },
#ifdef CONFIG_HIGHMEM
	[N_HIGH_MEMORY] = { { [0] = 1UL } },
116 117
#endif
	[N_MEMORY] = { { [0] = 1UL } },
118 119 120 121 122
	[N_CPU] = { { [0] = 1UL } },
#endif	/* NUMA */
};
EXPORT_SYMBOL(node_states);

123 124 125
/* Protect totalram_pages and zone->managed_pages */
static DEFINE_SPINLOCK(managed_page_count_lock);

126
unsigned long totalram_pages __read_mostly;
127
unsigned long totalreserve_pages __read_mostly;
128
unsigned long totalcma_pages __read_mostly;
129

130
int percpu_pagelist_fraction;
131
gfp_t gfp_allowed_mask __read_mostly = GFP_BOOT_MASK;
Linus Torvalds's avatar
Linus Torvalds committed
132

133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
/*
 * A cached value of the page's pageblock's migratetype, used when the page is
 * put on a pcplist. Used to avoid the pageblock migratetype lookup when
 * freeing from pcplists in most cases, at the cost of possibly becoming stale.
 * Also the migratetype set in the page does not necessarily match the pcplist
 * index, e.g. page might have MIGRATE_CMA set but be on a pcplist with any
 * other index - this ensures that it will be put on the correct CMA freelist.
 */
static inline int get_pcppage_migratetype(struct page *page)
{
	return page->index;
}

static inline void set_pcppage_migratetype(struct page *page, int migratetype)
{
	page->index = migratetype;
}

151 152 153 154 155 156 157 158 159
#ifdef CONFIG_PM_SLEEP
/*
 * The following functions are used by the suspend/hibernate code to temporarily
 * change gfp_allowed_mask in order to avoid using I/O during memory allocations
 * while devices are suspended.  To avoid races with the suspend/hibernate code,
 * they should always be called with pm_mutex held (gfp_allowed_mask also should
 * only be modified with pm_mutex held, unless the suspend/hibernate code is
 * guaranteed not to run in parallel with that modification).
 */
160 161 162 163

static gfp_t saved_gfp_mask;

void pm_restore_gfp_mask(void)
164 165
{
	WARN_ON(!mutex_is_locked(&pm_mutex));
166 167 168 169
	if (saved_gfp_mask) {
		gfp_allowed_mask = saved_gfp_mask;
		saved_gfp_mask = 0;
	}
170 171
}

172
void pm_restrict_gfp_mask(void)
173 174
{
	WARN_ON(!mutex_is_locked(&pm_mutex));
175 176
	WARN_ON(saved_gfp_mask);
	saved_gfp_mask = gfp_allowed_mask;
177
	gfp_allowed_mask &= ~(__GFP_IO | __GFP_FS);
178
}
179 180 181

bool pm_suspended_storage(void)
{
182
	if ((gfp_allowed_mask & (__GFP_IO | __GFP_FS)) == (__GFP_IO | __GFP_FS))
183 184 185
		return false;
	return true;
}
186 187
#endif /* CONFIG_PM_SLEEP */

188
#ifdef CONFIG_HUGETLB_PAGE_SIZE_VARIABLE
189
unsigned int pageblock_order __read_mostly;
190 191
#endif

192
static void __free_pages_ok(struct page *page, unsigned int order);
193

Linus Torvalds's avatar
Linus Torvalds committed
194 195 196 197 198 199
/*
 * 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
Yaowei Bai's avatar
Yaowei Bai committed
200
 *	HIGHMEM allocation will leave (224M+784M)/256 of ram reserved in ZONE_DMA
201 202 203
 *
 * TBD: should special case ZONE_DMA32 machines here - in those we normally
 * don't need any ZONE_NORMAL reservation
Linus Torvalds's avatar
Linus Torvalds committed
204
 */
205
int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES-1] = {
206
#ifdef CONFIG_ZONE_DMA
207
	 256,
208
#endif
209
#ifdef CONFIG_ZONE_DMA32
210
	 256,
211
#endif
212
#ifdef CONFIG_HIGHMEM
213
	 32,
214
#endif
215
	 32,
216
};
Linus Torvalds's avatar
Linus Torvalds committed
217 218 219

EXPORT_SYMBOL(totalram_pages);

220
static char * const zone_names[MAX_NR_ZONES] = {
221
#ifdef CONFIG_ZONE_DMA
222
	 "DMA",
223
#endif
224
#ifdef CONFIG_ZONE_DMA32
225
	 "DMA32",
226
#endif
227
	 "Normal",
228
#ifdef CONFIG_HIGHMEM
229
	 "HighMem",
230
#endif
231
	 "Movable",
232 233 234
#ifdef CONFIG_ZONE_DEVICE
	 "Device",
#endif
235 236
};

237 238 239 240 241 242 243 244 245 246 247 248 249
char * const migratetype_names[MIGRATE_TYPES] = {
	"Unmovable",
	"Movable",
	"Reclaimable",
	"HighAtomic",
#ifdef CONFIG_CMA
	"CMA",
#endif
#ifdef CONFIG_MEMORY_ISOLATION
	"Isolate",
#endif
};

250 251 252 253 254 255
compound_page_dtor * const compound_page_dtors[] = {
	NULL,
	free_compound_page,
#ifdef CONFIG_HUGETLB_PAGE
	free_huge_page,
#endif
256 257 258
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	free_transhuge_page,
#endif
259 260
};

Linus Torvalds's avatar
Linus Torvalds committed
261
int min_free_kbytes = 1024;
262
int user_min_free_kbytes = -1;
263
int watermark_scale_factor = 10;
Linus Torvalds's avatar
Linus Torvalds committed
264

265 266
static unsigned long __meminitdata nr_kernel_pages;
static unsigned long __meminitdata nr_all_pages;
267
static unsigned long __meminitdata dma_reserve;
Linus Torvalds's avatar
Linus Torvalds committed
268

269 270 271 272 273 274
#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP
static unsigned long __meminitdata arch_zone_lowest_possible_pfn[MAX_NR_ZONES];
static unsigned long __meminitdata arch_zone_highest_possible_pfn[MAX_NR_ZONES];
static unsigned long __initdata required_kernelcore;
static unsigned long __initdata required_movablecore;
static unsigned long __meminitdata zone_movable_pfn[MAX_NUMNODES];
275
static bool mirrored_kernelcore;
276 277 278 279 280

/* movable_zone is the "real" zone pages in ZONE_MOVABLE are taken from */
int movable_zone;
EXPORT_SYMBOL(movable_zone);
#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */
281

282 283
#if MAX_NUMNODES > 1
int nr_node_ids __read_mostly = MAX_NUMNODES;
284
int nr_online_nodes __read_mostly = 1;
285
EXPORT_SYMBOL(nr_node_ids);
286
EXPORT_SYMBOL(nr_online_nodes);
287 288
#endif

289 290
int page_group_by_mobility_disabled __read_mostly;

291
#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT
292 293 294 295 296 297 298

/*
 * Determine how many pages need to be initialized durig early boot
 * (non-deferred initialization).
 * The value of first_deferred_pfn will be set later, once non-deferred pages
 * are initialized, but for now set it ULONG_MAX.
 */
299 300
static inline void reset_deferred_meminit(pg_data_t *pgdat)
{
301 302 303
	phys_addr_t start_addr, end_addr;
	unsigned long max_pgcnt;
	unsigned long reserved;
304 305 306 307 308

	/*
	 * Initialise at least 2G of a node but also take into account that
	 * two large system hashes that can take up 1GB for 0.25TB/node.
	 */
309 310
	max_pgcnt = max(2UL << (30 - PAGE_SHIFT),
			(pgdat->node_spanned_pages >> 8));
311 312 313 314 315 316

	/*
	 * Compensate the all the memblock reservations (e.g. crash kernel)
	 * from the initial estimation to make sure we will initialize enough
	 * memory to boot.
	 */
317 318 319 320
	start_addr = PFN_PHYS(pgdat->node_start_pfn);
	end_addr = PFN_PHYS(pgdat->node_start_pfn + max_pgcnt);
	reserved = memblock_reserved_memory_within(start_addr, end_addr);
	max_pgcnt += PHYS_PFN(reserved);
321

322
	pgdat->static_init_pgcnt = min(max_pgcnt, pgdat->node_spanned_pages);
323 324 325 326
	pgdat->first_deferred_pfn = ULONG_MAX;
}

/* Returns true if the struct page for the pfn is uninitialised */
327
static inline bool __meminit early_page_uninitialised(unsigned long pfn)
328
{
329 330 331
	int nid = early_pfn_to_nid(pfn);

	if (node_online(nid) && pfn >= NODE_DATA(nid)->first_deferred_pfn)
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348
		return true;

	return false;
}

/*
 * Returns false when the remaining initialisation should be deferred until
 * later in the boot cycle when it can be parallelised.
 */
static inline bool update_defer_init(pg_data_t *pgdat,
				unsigned long pfn, unsigned long zone_end,
				unsigned long *nr_initialised)
{
	/* Always populate low zones for address-contrained allocations */
	if (zone_end < pgdat_end_pfn(pgdat))
		return true;
	(*nr_initialised)++;
349
	if ((*nr_initialised > pgdat->static_init_pgcnt) &&
350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
	    (pfn & (PAGES_PER_SECTION - 1)) == 0) {
		pgdat->first_deferred_pfn = pfn;
		return false;
	}

	return true;
}
#else
static inline void reset_deferred_meminit(pg_data_t *pgdat)
{
}

static inline bool early_page_uninitialised(unsigned long pfn)
{
	return false;
}

static inline bool update_defer_init(pg_data_t *pgdat,
				unsigned long pfn, unsigned long zone_end,
				unsigned long *nr_initialised)
{
	return true;
}
#endif

375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474
/* Return a pointer to the bitmap storing bits affecting a block of pages */
static inline unsigned long *get_pageblock_bitmap(struct page *page,
							unsigned long pfn)
{
#ifdef CONFIG_SPARSEMEM
	return __pfn_to_section(pfn)->pageblock_flags;
#else
	return page_zone(page)->pageblock_flags;
#endif /* CONFIG_SPARSEMEM */
}

static inline int pfn_to_bitidx(struct page *page, unsigned long pfn)
{
#ifdef CONFIG_SPARSEMEM
	pfn &= (PAGES_PER_SECTION-1);
	return (pfn >> pageblock_order) * NR_PAGEBLOCK_BITS;
#else
	pfn = pfn - round_down(page_zone(page)->zone_start_pfn, pageblock_nr_pages);
	return (pfn >> pageblock_order) * NR_PAGEBLOCK_BITS;
#endif /* CONFIG_SPARSEMEM */
}

/**
 * get_pfnblock_flags_mask - Return the requested group of flags for the pageblock_nr_pages block of pages
 * @page: The page within the block of interest
 * @pfn: The target page frame number
 * @end_bitidx: The last bit of interest to retrieve
 * @mask: mask of bits that the caller is interested in
 *
 * Return: pageblock_bits flags
 */
static __always_inline unsigned long __get_pfnblock_flags_mask(struct page *page,
					unsigned long pfn,
					unsigned long end_bitidx,
					unsigned long mask)
{
	unsigned long *bitmap;
	unsigned long bitidx, word_bitidx;
	unsigned long word;

	bitmap = get_pageblock_bitmap(page, pfn);
	bitidx = pfn_to_bitidx(page, pfn);
	word_bitidx = bitidx / BITS_PER_LONG;
	bitidx &= (BITS_PER_LONG-1);

	word = bitmap[word_bitidx];
	bitidx += end_bitidx;
	return (word >> (BITS_PER_LONG - bitidx - 1)) & mask;
}

unsigned long get_pfnblock_flags_mask(struct page *page, unsigned long pfn,
					unsigned long end_bitidx,
					unsigned long mask)
{
	return __get_pfnblock_flags_mask(page, pfn, end_bitidx, mask);
}

static __always_inline int get_pfnblock_migratetype(struct page *page, unsigned long pfn)
{
	return __get_pfnblock_flags_mask(page, pfn, PB_migrate_end, MIGRATETYPE_MASK);
}

/**
 * set_pfnblock_flags_mask - Set the requested group of flags for a pageblock_nr_pages block of pages
 * @page: The page within the block of interest
 * @flags: The flags to set
 * @pfn: The target page frame number
 * @end_bitidx: The last bit of interest
 * @mask: mask of bits that the caller is interested in
 */
void set_pfnblock_flags_mask(struct page *page, unsigned long flags,
					unsigned long pfn,
					unsigned long end_bitidx,
					unsigned long mask)
{
	unsigned long *bitmap;
	unsigned long bitidx, word_bitidx;
	unsigned long old_word, word;

	BUILD_BUG_ON(NR_PAGEBLOCK_BITS != 4);

	bitmap = get_pageblock_bitmap(page, pfn);
	bitidx = pfn_to_bitidx(page, pfn);
	word_bitidx = bitidx / BITS_PER_LONG;
	bitidx &= (BITS_PER_LONG-1);

	VM_BUG_ON_PAGE(!zone_spans_pfn(page_zone(page), pfn), page);

	bitidx += end_bitidx;
	mask <<= (BITS_PER_LONG - bitidx - 1);
	flags <<= (BITS_PER_LONG - bitidx - 1);

	word = READ_ONCE(bitmap[word_bitidx]);
	for (;;) {
		old_word = cmpxchg(&bitmap[word_bitidx], word, (word & ~mask) | flags);
		if (word == old_word)
			break;
		word = old_word;
	}
}
475

476
void set_pageblock_migratetype(struct page *page, int migratetype)
477
{
478 479
	if (unlikely(page_group_by_mobility_disabled &&
		     migratetype < MIGRATE_PCPTYPES))
480 481
		migratetype = MIGRATE_UNMOVABLE;

482 483 484 485
	set_pageblock_flags_group(page, (unsigned long)migratetype,
					PB_migrate, PB_migrate_end);
}

486
#ifdef CONFIG_DEBUG_VM
487
static int page_outside_zone_boundaries(struct zone *zone, struct page *page)
Linus Torvalds's avatar
Linus Torvalds committed
488
{
489 490 491
	int ret = 0;
	unsigned seq;
	unsigned long pfn = page_to_pfn(page);
492
	unsigned long sp, start_pfn;
493

494 495
	do {
		seq = zone_span_seqbegin(zone);
496 497
		start_pfn = zone->zone_start_pfn;
		sp = zone->spanned_pages;
498
		if (!zone_spans_pfn(zone, pfn))
499 500 501
			ret = 1;
	} while (zone_span_seqretry(zone, seq));

502
	if (ret)
503 504 505
		pr_err("page 0x%lx outside node %d zone %s [ 0x%lx - 0x%lx ]\n",
			pfn, zone_to_nid(zone), zone->name,
			start_pfn, start_pfn + sp);
506

507
	return ret;
508 509 510 511
}

static int page_is_consistent(struct zone *zone, struct page *page)
{
512
	if (!pfn_valid_within(page_to_pfn(page)))
513
		return 0;
Linus Torvalds's avatar
Linus Torvalds committed
514
	if (zone != page_zone(page))
515 516 517 518 519 520 521
		return 0;

	return 1;
}
/*
 * Temporary debugging check for pages not lying within a given zone.
 */
522
static int __maybe_unused bad_range(struct zone *zone, struct page *page)
523 524
{
	if (page_outside_zone_boundaries(zone, page))
Linus Torvalds's avatar
Linus Torvalds committed
525
		return 1;
526 527 528
	if (!page_is_consistent(zone, page))
		return 1;

Linus Torvalds's avatar
Linus Torvalds committed
529 530
	return 0;
}
531
#else
532
static inline int __maybe_unused bad_range(struct zone *zone, struct page *page)
533 534 535 536 537
{
	return 0;
}
#endif

538 539
static void bad_page(struct page *page, const char *reason,
		unsigned long bad_flags)
Linus Torvalds's avatar
Linus Torvalds committed
540
{
541 542 543 544 545 546 547 548 549 550 551 552 553 554
	static unsigned long resume;
	static unsigned long nr_shown;
	static unsigned long nr_unshown;

	/*
	 * Allow a burst of 60 reports, then keep quiet for that minute;
	 * or allow a steady drip of one report per second.
	 */
	if (nr_shown == 60) {
		if (time_before(jiffies, resume)) {
			nr_unshown++;
			goto out;
		}
		if (nr_unshown) {
555
			pr_alert(
556
			      "BUG: Bad page state: %lu messages suppressed\n",
557 558 559 560 561 562 563 564
				nr_unshown);
			nr_unshown = 0;
		}
		nr_shown = 0;
	}
	if (nr_shown++ == 0)
		resume = jiffies + 60 * HZ;

565
	pr_alert("BUG: Bad page state in process %s  pfn:%05lx\n",
566
		current->comm, page_to_pfn(page));
567 568 569 570 571
	__dump_page(page, reason);
	bad_flags &= page->flags;
	if (bad_flags)
		pr_alert("bad because of flags: %#lx(%pGp)\n",
						bad_flags, &bad_flags);
572
	dump_page_owner(page);
573

574
	print_modules();
Linus Torvalds's avatar
Linus Torvalds committed
575
	dump_stack();
576
out:
577
	/* Leave bad fields for debug, except PageBuddy could make trouble */
578
	page_mapcount_reset(page); /* remove PageBuddy */
579
	add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE);
Linus Torvalds's avatar
Linus Torvalds committed
580 581 582 583 584
}

/*
 * Higher-order pages are called "compound pages".  They are structured thusly:
 *
585
 * The first PAGE_SIZE page is called the "head page" and have PG_head set.
Linus Torvalds's avatar
Linus Torvalds committed
586
 *
587 588
 * The remaining PAGE_SIZE pages are called "tail pages". PageTail() is encoded
 * in bit 0 of page->compound_head. The rest of bits is pointer to head page.
Linus Torvalds's avatar
Linus Torvalds committed
589
 *
590 591
 * The first tail page's ->compound_dtor holds the offset in array of compound
 * page destructors. See compound_page_dtors.
Linus Torvalds's avatar
Linus Torvalds committed
592
 *
593
 * The first tail page's ->compound_order holds the order of allocation.
594
 * This usage means that zero-order pages may not be compound.
Linus Torvalds's avatar
Linus Torvalds committed
595
 */
596

597
void free_compound_page(struct page *page)
598
{
599
	__free_pages_ok(page, compound_order(page));
600 601
}

602
void prep_compound_page(struct page *page, unsigned int order)
603 604 605 606
{
	int i;
	int nr_pages = 1 << order;

607
	set_compound_page_dtor(page, COMPOUND_PAGE_DTOR);
608 609 610 611
	set_compound_order(page, order);
	__SetPageHead(page);
	for (i = 1; i < nr_pages; i++) {
		struct page *p = page + i;
612
		set_page_count(p, 0);
613
		p->mapping = TAIL_MAPPING;
614
		set_compound_head(p, page);
615
	}
616
	atomic_set(compound_mapcount_ptr(page), -1);
617 618
}

619 620
#ifdef CONFIG_DEBUG_PAGEALLOC
unsigned int _debug_guardpage_minorder;
621 622
bool _debug_pagealloc_enabled __read_mostly
			= IS_ENABLED(CONFIG_DEBUG_PAGEALLOC_ENABLE_DEFAULT);
623
EXPORT_SYMBOL(_debug_pagealloc_enabled);
624 625
bool _debug_guardpage_enabled __read_mostly;

626 627 628 629
static int __init early_debug_pagealloc(char *buf)
{
	if (!buf)
		return -EINVAL;
630
	return kstrtobool(buf, &_debug_pagealloc_enabled);
631 632 633
}
early_param("debug_pagealloc", early_debug_pagealloc);

634 635
static bool need_debug_guardpage(void)
{
636 637 638 639
	/* If we don't use debug_pagealloc, we don't need guard page */
	if (!debug_pagealloc_enabled())
		return false;

640 641 642
	if (!debug_guardpage_minorder())
		return false;

643 644 645 646 647
	return true;
}

static void init_debug_guardpage(void)
{
648 649 650
	if (!debug_pagealloc_enabled())
		return;

651 652 653
	if (!debug_guardpage_minorder())
		return;

654 655 656 657 658 659 660
	_debug_guardpage_enabled = true;
}

struct page_ext_operations debug_guardpage_ops = {
	.need = need_debug_guardpage,
	.init = init_debug_guardpage,
};
661 662 663 664 665 666

static int __init debug_guardpage_minorder_setup(char *buf)
{
	unsigned long res;

	if (kstrtoul(buf, 10, &res) < 0 ||  res > MAX_ORDER / 2) {
667
		pr_err("Bad debug_guardpage_minorder value\n");
668 669 670
		return 0;
	}
	_debug_guardpage_minorder = res;
671
	pr_info("Setting debug_guardpage_minorder to %lu\n", res);
672 673
	return 0;
}
674
early_param("debug_guardpage_minorder", debug_guardpage_minorder_setup);
675

676
static inline bool set_page_guard(struct zone *zone, struct page *page,
677
				unsigned int order, int migratetype)
678
{
679 680 681
	struct page_ext *page_ext;

	if (!debug_guardpage_enabled())
682 683 684 685
		return false;

	if (order >= debug_guardpage_minorder())
		return false;
686 687

	page_ext = lookup_page_ext(page);
688
	if (unlikely(!page_ext))
689
		return false;
690

691 692
	__set_bit(PAGE_EXT_DEBUG_GUARD, &page_ext->flags);

693 694 695 696
	INIT_LIST_HEAD(&page->lru);
	set_page_private(page, order);
	/* Guard pages are not available for any usage */
	__mod_zone_freepage_state(zone, -(1 << order), migratetype);
697 698

	return true;
699 700
}

701 702
static inline void clear_page_guard(struct zone *zone, struct page *page,
				unsigned int order, int migratetype)
703
{
704 705 706 707 708 709
	struct page_ext *page_ext;

	if (!debug_guardpage_enabled())
		return;

	page_ext = lookup_page_ext(page);
710 711 712
	if (unlikely(!page_ext))
		return;

713 714
	__clear_bit(PAGE_EXT_DEBUG_GUARD, &page_ext->flags);

715 716 717
	set_page_private(page, 0);
	if (!is_migrate_isolate(migratetype))
		__mod_zone_freepage_state(zone, (1 << order), migratetype);
718 719
}
#else
720
struct page_ext_operations debug_guardpage_ops;
721 722
static inline bool set_page_guard(struct zone *zone, struct page *page,
			unsigned int order, int migratetype) { return false; }
723 724
static inline void clear_page_guard(struct zone *zone, struct page *page,
				unsigned int order, int migratetype) {}
725 726
#endif

727
static inline void set_page_order(struct page *page, unsigned int order)
728
{
729
	set_page_private(page, order);
730
	__SetPageBuddy(page);
Linus Torvalds's avatar
Linus Torvalds committed
731 732 733 734
}

static inline void rmv_page_order(struct page *page)
{
735
	__ClearPageBuddy(page);
736
	set_page_private(page, 0);
Linus Torvalds's avatar
Linus Torvalds committed
737 738 739 740 741
}

/*
 * This function checks whether a page is free && is the buddy
 * we can do coalesce a page and its buddy if
742
 * (a) the buddy is not in a hole (check before calling!) &&
743
 * (b) the buddy is in the buddy system &&
744 745
 * (c) a page and its buddy have the same order &&
 * (d) a page and its buddy are in the same zone.
746
 *
747 748 749 750
 * For recording whether a page is in the buddy system, we set ->_mapcount
 * PAGE_BUDDY_MAPCOUNT_VALUE.
 * Setting, clearing, and testing _mapcount PAGE_BUDDY_MAPCOUNT_VALUE is
 * serialized by zone->lock.
Linus Torvalds's avatar
Linus Torvalds committed
751
 *
752
 * For recording page's order, we use page_private(page).
Linus Torvalds's avatar
Linus Torvalds committed
753
 */
754
static inline int page_is_buddy(struct page *page, struct page *buddy,
755
							unsigned int order)
Linus Torvalds's avatar
Linus Torvalds committed
756
{
757
	if (page_is_guard(buddy) && page_order(buddy) == order) {
758 759 760
		if (page_zone_id(page) != page_zone_id(buddy))
			return 0;

761 762
		VM_BUG_ON_PAGE(page_count(buddy) != 0, buddy);

763 764 765
		return 1;
	}

766
	if (PageBuddy(buddy) && page_order(buddy) == order) {
767 768 769 770 771 772 773 774
		/*
		 * zone check is done late to avoid uselessly
		 * calculating zone/node ids for pages that could
		 * never merge.
		 */
		if (page_zone_id(page) != page_zone_id(buddy))
			return 0;

775 776
		VM_BUG_ON_PAGE(page_count(buddy) != 0, buddy);

777
		return 1;
778
	}
779
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
780 781 782 783 784 785 786 787 788 789 790 791 792 793 794
}

/*
 * 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
795 796 797
 * free pages of length of (1 << order) and marked with _mapcount
 * PAGE_BUDDY_MAPCOUNT_VALUE. Page's order is recorded in page_private(page)
 * field.
Linus Torvalds's avatar
Linus Torvalds committed
798
 * So when we are allocating or freeing one, we can derive the state of the
799 800
 * other.  That is, if we allocate a small block, and both were
 * free, the remainder of the region must be split into blocks.
Linus Torvalds's avatar
Linus Torvalds committed
801
 * If a block is freed, and its buddy is also free, then this
802
 * triggers coalescing into a block of larger size.
Linus Torvalds's avatar
Linus Torvalds committed
803
 *
804
 * -- nyc
Linus Torvalds's avatar
Linus Torvalds committed
805 806
 */

Nick Piggin's avatar
Nick Piggin committed
807
static inline void __free_one_page(struct page *page,
808
		unsigned long pfn,
809 810
		struct zone *zone, unsigned int order,
		int migratetype)
Linus Torvalds's avatar
Linus Torvalds committed
811
{
812 813
	unsigned long combined_pfn;
	unsigned long uninitialized_var(buddy_pfn);
814
	struct page *buddy;
815 816 817
	unsigned int max_order;

	max_order = min_t(unsigned int, MAX_ORDER, pageblock_order + 1);
Linus Torvalds's avatar
Linus Torvalds committed
818

819
	VM_BUG_ON(!zone_is_initialized(zone));
820
	VM_BUG_ON_PAGE(page->flags & PAGE_FLAGS_CHECK_AT_PREP, page);
Linus Torvalds's avatar
Linus Torvalds committed
821

822
	VM_BUG_ON(migratetype == -1);
823
	if (likely(!is_migrate_isolate(migratetype)))
824
		__mod_zone_freepage_state(zone, 1 << order, migratetype);
825

826
	VM_BUG_ON_PAGE(pfn & ((1 << order) - 1), page);
827
	VM_BUG_ON_PAGE(bad_range(zone, page), page);
Linus Torvalds's avatar
Linus Torvalds committed
828

829
continue_merging:
830
	while (order < max_order - 1) {
831 832
		buddy_pfn = __find_buddy_pfn(pfn, order);
		buddy = page + (buddy_pfn - pfn);
833 834 835

		if (!pfn_valid_within(buddy_pfn))
			goto done_merging;
836
		if (!page_is_buddy(page, buddy, order))
837
			goto done_merging;
838 839 840 841 842
		/*
		 * Our buddy is free or it is CONFIG_DEBUG_PAGEALLOC guard page,
		 * merge with it and move up one order.
		 */
		if (page_is_guard(buddy)) {
843
			clear_page_guard(zone, buddy, order, migratetype);
844 845 846 847 848
		} else {
			list_del(&buddy->lru);
			zone->free_area[order].nr_free--;
			rmv_page_order(buddy);
		}
849 850 851
		combined_pfn = buddy_pfn & pfn;
		page = page + (combined_pfn - pfn);
		pfn = combined_pfn;
Linus Torvalds's avatar
Linus Torvalds committed
852 853
		order++;
	}
854 855 856 857 858 859 860 861 862 863 864 865
	if (max_order < MAX_ORDER) {
		/* If we are here, it means order is >= pageblock_order.
		 * We want to prevent merge between freepages on isolate
		 * pageblock and normal pageblock. Without this, pageblock
		 * isolation could cause incorrect freepage or CMA accounting.
		 *
		 * We don't want to hit this code for the more frequent
		 * low-order merging.
		 */
		if (unlikely(has_isolate_pageblock(zone))) {
			int buddy_mt;

866 867
			buddy_pfn = __find_buddy_pfn(pfn, order);
			buddy = page + (buddy_pfn - pfn);
868 869 870 871 872 873 874 875 876 877 878 879
			buddy_mt = get_pageblock_migratetype(buddy);

			if (migratetype != buddy_mt
					&& (is_migrate_isolate(migratetype) ||
						is_migrate_isolate(buddy_mt)))
				goto done_merging;
		}
		max_order++;
		goto continue_merging;
	}

done_merging:
Linus Torvalds's avatar
Linus Torvalds committed
880
	set_page_order(page, order);
881 882 883 884 885 886 887 888 889

	/*
	 * If this is not the largest possible page, check if the buddy
	 * of the next-highest order is free. If it is, it's possible
	 * that pages are being freed that will coalesce soon. In case,
	 * that is happening, add the free page to the tail of the list
	 * so it's less likely to be used soon and more likely to be merged
	 * as a higher order page
	 */
890
	if ((order < MAX_ORDER-2) && pfn_valid_within(buddy_pfn)) {
891
		struct page *higher_page, *higher_buddy;
892 893 894 895
		combined_pfn = buddy_pfn & pfn;
		higher_page = page + (combined_pfn - pfn);
		buddy_pfn = __find_buddy_pfn(combined_pfn, order + 1);
		higher_buddy = higher_page + (buddy_pfn - combined_pfn);
896 897
		if (pfn_valid_within(buddy_pfn) &&
		    page_is_buddy(higher_page, higher_buddy, order + 1)) {
898 899 900 901 902 903 904 905
			list_add_tail(&page->lru,
				&zone->free_area[order].free_list[migratetype]);
			goto out;
		}
	}

	list_add(&page->lru, &zone->free_area[order].free_list[migratetype]);
out:
Linus Torvalds's avatar
Linus Torvalds committed
906 907 908
	zone->free_area[order].nr_free++;
}

909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930
/*
 * A bad page could be due to a number of fields. Instead of multiple branches,
 * try and check multiple fields with one check. The caller must do a detailed
 * check if necessary.
 */
static inline bool page_expected_state(struct page *page,
					unsigned long check_flags)
{
	if (unlikely(atomic_read(&page->_mapcount) != -1))
		return false;

	if (unlikely((unsigned long)page->mapping |
			page_ref_count(page) |
#ifdef CONFIG_MEMCG
			(unsigned long)page->mem_cgroup |
#endif
			(page->flags & check_flags)))
		return false;

	return true;
}

931
static void free_pages_check_bad(struct page *page)
Linus Torvalds's avatar
Linus Torvalds committed
932
{
933 934 935 936 937
	const char *bad_reason;
	unsigned long bad_flags;

	bad_reason = NULL;
	bad_flags = 0;
938

939
	if (unlikely(atomic_read(&page->_mapcount) != -1))
940 941 942
		bad_reason = "nonzero mapcount";
	if (unlikely(page->mapping != NULL))
		bad_reason = "non-NULL mapping";
943
	if (unlikely(page_ref_count(page) != 0))
944
		bad_reason = "nonzero _refcount";
945 946 947 948
	if (unlikely(page->flags & PAGE_FLAGS_CHECK_AT_FREE)) {
		bad_reason = "PAGE_FLAGS_CHECK_AT_FREE flag(s) set";
		bad_flags = PAGE_FLAGS_CHECK_AT_FREE;
	}
949 950 951 952
#ifdef CONFIG_MEMCG
	if (unlikely(page->mem_cgroup))
		bad_reason = "page still charged to cgroup";
#endif
953
	bad_page(page, bad_reason, bad_flags);
954 955 956 957
}

static inline int free_pages_check(struct page *page)
{
958
	if (likely(page_expected_state(page, PAGE_FLAGS_CHECK_AT_FREE)))
959 960 961 962
		return 0;

	/* Something has gone sideways, find it */
	free_pages_check_bad(page);
963
	return 1;
Linus Torvalds's avatar
Linus Torvalds committed
964 965
}

966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
static int free_tail_pages_check(struct page *head_page, struct page *page)
{
	int ret = 1;

	/*
	 * We rely page->lru.next never has bit 0 set, unless the page
	 * is PageTail(). Let's make sure that's true even for poisoned ->lru.
	 */
	BUILD_BUG_ON((unsigned long)LIST_POISON1 & 1);

	if (!IS_ENABLED(CONFIG_DEBUG_VM)) {
		ret = 0;
		goto out;
	}
	switch (page - head_page) {
	case 1:
		/* the first tail page: ->mapping is compound_mapcount() */
		if (unlikely(compound_mapcount(page))) {
			bad_page(page, "nonzero compound_mapcount", 0);
			goto out;
		}
		break;
	case 2:
		/*
		 * the second tail page: ->mapping is
		 * page_deferred_list().next -- ignore value.
		 */
		break;
	default:
		if (page->mapping != TAIL_MAPPING) {
			bad_page(page, "corrupted mapping in tail page", 0);
			goto out;
		}
		break;
	}
	if (unlikely(!PageTail(page))) {
		bad_page(page, "PageTail not set", 0);
		goto out;
	}
	if (unlikely(compound_head(page) != head_page)) {
		bad_page(page, "compound_head not consistent", 0);
		goto out;
	}
	ret = 0;
out:
	page->mapping = NULL;
	clear_compound_head(page);
	return ret;
}

1016 1017
static __always_inline bool free_pages_prepare(struct page *page,
					unsigned int order, bool check_free)
1018
{
1019
	int bad = 0;
1020 1021 1022

	VM_BUG_ON_PAGE(PageTail(page), page);

1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033
	trace_mm_page_free(page, order);

	/*
	 * Check tail pages before head page information is cleared to
	 * avoid checking PageCompound for order-0 pages.
	 */
	if (unlikely(order)) {
		bool compound = PageCompound(page);
		int i;

		VM_BUG_ON_PAGE(compound && compound_order(page) != order, page);
1034

1035 1036
		if (compound)
			ClearPageDoubleMap(page);
1037 1038 1039 1040 1041 1042 1043 1044 1045 1046
		for (i = 1; i < (1 << order); i++) {
			if (compound)
				bad += free_tail_pages_check(page, page + i);
			if (unlikely(free_pages_check(page + i))) {
				bad++;
				continue;
			}
			(page + i)->flags &= ~PAGE_FLAGS_CHECK_AT_PREP;
		}
	}
1047
	if (PageMappingFlags(page))
1048
		page->mapping = NULL;
1049
	if (memcg_kmem_enabled() && PageKmemcg(page))
1050
		memcg_kmem_uncharge(page, order);
1051 1052 1053 1054
	if (check_free)
		bad += free_pages_check(page);
	if (bad)
		return false;
1055

1056 1057 1058
	page_cpupid_reset_last(page);
	page->flags &= ~PAGE_FLAGS_CHECK_AT_PREP;
	reset_page_owner(page, order);
1059 1060 1061

	if (!PageHighMem(page)) {
		debug_check_no_locks_freed(page_address(page),
1062
					   PAGE_SIZE << order);
1063
		debug_check_no_obj_freed(page_address(page),
1064
					   PAGE_SIZE << order);
1065
	}
1066 1067 1068
	arch_free_page(page, order);
	kernel_poison_pages(page, 1 << order, 0);
	kernel_map_pages(page, 1 << order, 0);
1069
	kasan_free_pages(page, order);
1070 1071 1072 1073

	return true;
}

1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089
#ifdef CONFIG_DEBUG_VM
static inline bool free_pcp_prepare(struct page *page)
{
	return free_pages_prepare(page, 0, true);
}

static inline bool bulkfree_pcp_prepare(struct page *page)
{
	return false;
}
#else
static bool free_pcp_prepare(struct page *page)
{
	return free_pages_prepare(page, 0, false);
}

1090 1091 1092 1093 1094 1095
static bool bulkfree_pcp_prepare(struct page *page)
{
	return free_pages_check(page);
}
#endif /* CONFIG_DEBUG_VM */

Linus Torvalds's avatar
Linus Torvalds committed
1096
/*
1097
 * Frees a number of pages from the PCP lists
Linus Torvalds's avatar
Linus Torvalds committed
1098
 * Assumes all pages on list are in same zone, and of same order.
1099
 * count is the number of pages to free.
Linus Torvalds's avatar
Linus Torvalds committed
1100 1101 1102 1103 1104 1105 1106
 *
 * 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.
 */
1107 1108
static void free_pcppages_bulk(struct zone *zone, int count,
					struct per_cpu_pages *pcp)
Linus Torvalds's avatar
Linus Torvalds committed
1109
{
1110
	int migratetype = 0;
1111
	int batch_free = 0;
1112
	bool isolated_pageblocks;
1113

1114
	spin_lock(&zone->lock);
1115
	isolated_pageblocks = has_isolate_pageblock(zone);
1116

1117
	while (count) {
Nick Piggin's avatar
Nick Piggin committed
1118
		struct page *page;
1119 1120 1121
		struct list_head *list;

		/*
1122 1123 1124 1125 1126
		 * Remove pages from lists in a round-robin fashion. A
		 * batch_free count is maintained that is incremented when an
		 * empty list is encountered.  This is so more pages are freed
		 * off fuller lists instead of spinning excessively around empty
		 * lists
1127 1128
		 */
		do {
1129
			batch_free++;
1130 1131 1132 1133
			if (++migratetype == MIGRATE_PCPTYPES)
				migratetype = 0;
			list = &pcp->lists[migratetype];
		} while (list_empty(list));
Nick Piggin's avatar
Nick Piggin committed
1134

1135 1136
		/* This is the only non-empty list. Free them all. */
		if (batch_free == MIGRATE_PCPTYPES)
1137
			batch_free = count;
1138

1139
		do {
1140 1141
			int mt;	/* migratetype of the to-be-freed page */

1142
			page = list_last_entry(list, struct page, lru);
1143 1144
			/* must delete as __free_one_page list manipulates */
			list_del(&page->lru);
1145

1146
			mt = get_pcppage_migratetype(page);
1147 1148 1149
			/* MIGRATE_ISOLATE page should not go to pcplists */
			VM_BUG_ON_PAGE(is_migrate_isolate(mt), page);
			/* Pageblock could have been isolated meanwhile */
1150
			if (unlikely(isolated_pageblocks))
1151 1152
				mt = get_pageblock_migratetype(page);

1153 1154 1155
			if (bulkfree_pcp_prepare(page))
				continue;

1156
			__free_one_page(page, page_to_pfn(page), zone, 0, mt);
1157
			trace_mm_page_pcpu_drain(page, 0, mt);
1158
		} while (--count && --batch_free && !list_empty(list));
Linus Torvalds's avatar
Linus Torvalds committed
1159
	}
1160
	spin_unlock(&zone->lock);
Linus Torvalds's avatar
Linus Torvalds committed
1161 1162
}

1163 1164
static void free_one_page(struct zone *zone,
				struct page *page, unsigned long pfn,
1165
				unsigned int order,
1166
				int migratetype)
Linus Torvalds's avatar
Linus Torvalds committed
1167
{
1168
	spin_lock(&zone->lock);
1169 1170 1171 1172
	if (unlikely(has_isolate_pageblock(zone) ||
		is_migrate_isolate(migratetype))) {
		migratetype = get_pfnblock_migratetype(page, pfn);
	}
1173
	__free_one_page(page, pfn, zone, order, migratetype);
1174
	spin_unlock(&zone->lock);
Nick Piggin's avatar
Nick Piggin committed
1175 1176
}

1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198
static void __meminit __init_single_page(struct page *page, unsigned long pfn,
				unsigned long zone, int nid)
{
	set_page_links(page, zone, nid, pfn);
	init_page_count(page);
	page_mapcount_reset(page);
	page_cpupid_reset_last(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))
		set_page_address(page, __va(pfn << PAGE_SHIFT));
#endif
}

static void __meminit __init_single_pfn(unsigned long pfn, unsigned long zone,
					int nid)
{
	return __init_single_page(pfn_to_page(pfn), pfn, zone, nid);
}

1199
#ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT
1200
static void __meminit init_reserved_page(unsigned long pfn)
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
{
	pg_data_t *pgdat;
	int nid, zid;

	if (!early_page_uninitialised(pfn))
		return;

	nid = early_pfn_to_nid(pfn);
	pgdat = NODE_DATA(nid);

	for (zid = 0; zid < MAX_NR_ZONES; zid++) {
		struct zone *zone = &pgdat->node_zones[zid];

		if (pfn >= zone->zone_start_pfn && pfn < zone_end_pfn(zone))
			break;
	}
	__init_single_pfn(pfn, zid, nid);
}
#else
static inline void init_reserved_page(unsigned long pfn)
{
}
#endif /* CONFIG_DEFERRED_STRUCT_PAGE_INIT */

1225 1226 1227 1228 1229 1230
/*
 * Initialised pages do not have PageReserved set. This function is
 * called for each range allocated by the bootmem allocator and
 * marks the pages PageReserved. The remaining valid pages are later
 * sent to the buddy page allocator.
 */
1231
void __meminit reserve_bootmem_region(phys_addr_t start, phys_addr_t end)
1232 1233 1234 1235
{
	unsigned long start_pfn = PFN_DOWN(start);
	unsigned long end_pfn = PFN_UP(end);

1236 1237 1238 1239 1240
	for (; start_pfn < end_pfn; start_pfn++) {
		if (pfn_valid(start_pfn)) {
			struct page *page = pfn_to_page(start_pfn);

			init_reserved_page(start_pfn);
1241 1242 1243 1244

			/* Avoid false-positive PageTail() */
			INIT_LIST_HEAD(&page->lru);

1245 1246 1247
			SetPageReserved(page);
		}
	}
1248 1249
}

1250 1251
static void __free_pages_ok(struct page *page, unsigned int order)
{
1252
	unsigned long flags;
1253
	int migratetype;
1254
	unsigned long pfn = page_to_pfn(page);
1255

1256
	if (!free_pages_prepare(page, order, true))
1257 1258
		return;

1259
	migratetype = get_pfnblock_migratetype(page, pfn);
1260 1261
	local_irq_save(flags);
	__count_vm_events(PGFREE, 1 << order);
1262
	free_one_page(page_zone(page), page, pfn, order, migratetype);
1263
	local_irq_restore(flags);
Linus Torvalds's avatar
Linus Torvalds committed
1264 1265
}

1266
static void __init __free_pages_boot_core(struct page *page, unsigned int order)
1267
{
1268
	unsigned int nr_pages = 1 << order;
1269
	struct page *p = page;
1270
	unsigned int loop;
1271

1272 1273 1274
	prefetchw(p);
	for (loop = 0; loop < (nr_pages - 1); loop++, p++) {
		prefetchw(p + 1);
1275 1276
		__ClearPageReserved(p);
		set_page_count(p, 0);
1277
	}
1278 1279
	__ClearPageReserved(p);
	set_page_count(p, 0);
1280

1281
	page_zone(page)->managed_pages += nr_pages;
1282 1283
	set_page_refcounted(page);
	__free_pages(page, order);
1284 1285
}

1286 1287
#if defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID) || \
	defined(CONFIG_HAVE_MEMBLOCK_NODE_MAP)
1288

1289 1290 1291 1292
static struct mminit_pfnnid_cache early_pfnnid_cache __meminitdata;

int __meminit early_pfn_to_nid(unsigned long pfn)
{
1293
	static DEFINE_SPINLOCK(early_pfn_lock);
1294 1295
	int nid;

1296
	spin_lock(&early_pfn_lock);
1297
	nid = __early_pfn_to_nid(pfn, &early_pfnnid_cache);
1298
	if (nid < 0)
1299
		nid = first_online_node;
1300 1301 1302
	spin_unlock(&early_pfn_lock);

	return nid;
1303 1304 1305 1306
}
#endif

#ifdef CONFIG_NODES_SPAN_OTHER_NODES
1307 1308 1309
static inline bool __meminit __maybe_unused
meminit_pfn_in_nid(unsigned long pfn, int node,
		   struct mminit_pfnnid_cache *state)
1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330
{
	int nid;

	nid = __early_pfn_to_nid(pfn, state);
	if (nid >= 0 && nid != node)
		return false;
	return true;
}

/* Only safe to use early in boot when initialisation is single-threaded */
static inline bool __meminit early_pfn_in_nid(unsigned long pfn, int node)
{
	return meminit_pfn_in_nid(pfn, node, &early_pfnnid_cache);
}

#else

static inline bool __meminit early_pfn_in_nid(unsigned long pfn, int node)
{
	return true;
}
1331 1332 1333
static inline bool __meminit  __maybe_unused
meminit_pfn_in_nid(unsigned long pfn, int node,
		   struct mminit_pfnnid_cache *state)
1334 1335 1336 1337 1338 1339
{
	return true;
}
#endif


1340
void __init __free_pages_bootmem(struct page *page, unsigned long pfn,
1341 1342 1343 1344
							unsigned int order)
{
	if (early_page_uninitialised(pfn))
		return;
1345
	return __free_pages_boot_core(page, order);
1346 1347
}

1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
/*
 * Check that the whole (or subset of) a pageblock given by the interval of
 * [start_pfn, end_pfn) is valid and within the same zone, before scanning it
 * with the migration of free compaction scanner. The scanners then need to
 * use only pfn_valid_within() check for arches that allow holes within
 * pageblocks.
 *
 * Return struct page pointer of start_pfn, or NULL if checks were not passed.
 *
 * It's possible on some configurations to have a setup like node0 node1 node0
 * i.e. it's possible that all pages within a zones range of pages do not
 * belong to a single zone. We assume that a border between node0 and node1
 * can occur within a single pageblock, but not a node0 node1 node0
 * interleaving within a single pageblock. It is therefore sufficient to check
 * the first and last page of a pageblock and avoid checking each individual
 * page in a pageblock.
 */
struct page *__pageblock_pfn_to_page(unsigned long start_pfn,
				     unsigned long end_pfn, struct zone *zone)
{
	struct page *start_page;
	struct page *end_page;

	/* end_pfn is one past the range we are checking */
	end_pfn--;

	if (!pfn_valid(start_pfn) || !pfn_valid(end_pfn))
		return NULL;

1377 1378 1379
	start_page = pfn_to_online_page(start_pfn);
	if (!start_page)
		return NULL;
1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418