bootmem.c 20.7 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2
 *  bootmem - A boot-time physical memory allocator and configurator
L
Linus Torvalds 已提交
3 4
 *
 *  Copyright (C) 1999 Ingo Molnar
5 6
 *                1999 Kanoj Sarcar, SGI
 *                2008 Johannes Weiner
L
Linus Torvalds 已提交
7
 *
8 9
 * Access to this subsystem has to be serialized externally (which is true
 * for the boot process anyway).
L
Linus Torvalds 已提交
10 11
 */
#include <linux/init.h>
12
#include <linux/pfn.h>
L
Linus Torvalds 已提交
13 14
#include <linux/bootmem.h>
#include <linux/module.h>
15 16

#include <asm/bug.h>
L
Linus Torvalds 已提交
17
#include <asm/io.h>
18
#include <asm/processor.h>
19

L
Linus Torvalds 已提交
20 21 22 23 24 25
#include "internal.h"

unsigned long max_low_pfn;
unsigned long min_low_pfn;
unsigned long max_pfn;

26
static LIST_HEAD(bdata_list);
27 28 29 30 31 32 33 34
#ifdef CONFIG_CRASH_DUMP
/*
 * If we have booted due to a crash, max_pfn will be a very low value. We need
 * to know the amount of memory that the previous kernel used.
 */
unsigned long saved_max_pfn;
#endif

35 36
bootmem_data_t bootmem_node_data[MAX_NUMNODES] __initdata;

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
static int bootmem_debug;

static int __init bootmem_debug_setup(char *buf)
{
	bootmem_debug = 1;
	return 0;
}
early_param("bootmem_debug", bootmem_debug_setup);

#define bdebug(fmt, args...) ({				\
	if (unlikely(bootmem_debug))			\
		printk(KERN_INFO			\
			"bootmem::%s " fmt,		\
			__FUNCTION__, ## args);		\
})

53 54 55 56 57 58 59 60 61 62 63 64 65
/*
 * Given an initialised bdata, it returns the size of the boot bitmap
 */
static unsigned long __init get_mapsize(bootmem_data_t *bdata)
{
	unsigned long mapsize;
	unsigned long start = PFN_DOWN(bdata->node_boot_start);
	unsigned long end = bdata->node_low_pfn;

	mapsize = ((end - start) + 7) / 8;
	return ALIGN(mapsize, sizeof(long));
}

66 67 68 69
/**
 * bootmem_bootmap_pages - calculate bitmap size in pages
 * @pages: number of pages the bitmap has to represent
 */
70
unsigned long __init bootmem_bootmap_pages(unsigned long pages)
L
Linus Torvalds 已提交
71 72 73 74 75 76 77 78 79
{
	unsigned long mapsize;

	mapsize = (pages+7)/8;
	mapsize = (mapsize + ~PAGE_MASK) & PAGE_MASK;
	mapsize >>= PAGE_SHIFT;

	return mapsize;
}
80

81 82 83
/*
 * link bdata in order
 */
84
static void __init link_bootmem(bootmem_data_t *bdata)
85 86
{
	bootmem_data_t *ent;
87

88 89 90 91 92 93 94 95 96 97 98 99 100 101
	if (list_empty(&bdata_list)) {
		list_add(&bdata->list, &bdata_list);
		return;
	}
	/* insert in order */
	list_for_each_entry(ent, &bdata_list, list) {
		if (bdata->node_boot_start < ent->node_boot_start) {
			list_add_tail(&bdata->list, &ent->list);
			return;
		}
	}
	list_add_tail(&bdata->list, &bdata_list);
}

L
Linus Torvalds 已提交
102 103 104
/*
 * Called once to set up the allocator itself.
 */
105
static unsigned long __init init_bootmem_core(bootmem_data_t *bdata,
L
Linus Torvalds 已提交
106 107
	unsigned long mapstart, unsigned long start, unsigned long end)
{
108
	unsigned long mapsize;
L
Linus Torvalds 已提交
109

110
	mminit_validate_memmodel_limits(&start, &end);
111 112
	bdata->node_bootmem_map = phys_to_virt(PFN_PHYS(mapstart));
	bdata->node_boot_start = PFN_PHYS(start);
L
Linus Torvalds 已提交
113
	bdata->node_low_pfn = end;
114
	link_bootmem(bdata);
L
Linus Torvalds 已提交
115 116 117 118 119

	/*
	 * Initially all pages are reserved - setup_arch() has to
	 * register free RAM areas explicitly.
	 */
120
	mapsize = get_mapsize(bdata);
L
Linus Torvalds 已提交
121 122
	memset(bdata->node_bootmem_map, 0xff, mapsize);

123 124 125
	bdebug("nid=%td start=%lx map=%lx end=%lx mapsize=%lx\n",
		bdata - bootmem_node_data, start, mapstart, end, mapsize);

L
Linus Torvalds 已提交
126 127 128
	return mapsize;
}

129 130 131 132 133 134 135 136 137
/**
 * init_bootmem_node - register a node as boot memory
 * @pgdat: node to register
 * @freepfn: pfn where the bitmap for this node is to be placed
 * @startpfn: first pfn on the node
 * @endpfn: first pfn after the node
 *
 * Returns the number of bytes needed to hold the bitmap for this node.
 */
138 139 140 141 142 143
unsigned long __init init_bootmem_node(pg_data_t *pgdat, unsigned long freepfn,
				unsigned long startpfn, unsigned long endpfn)
{
	return init_bootmem_core(pgdat->bdata, freepfn, startpfn, endpfn);
}

144 145 146 147 148 149 150
/**
 * init_bootmem - register boot memory
 * @start: pfn where the bitmap is to be placed
 * @pages: number of available physical pages
 *
 * Returns the number of bytes needed to hold the bitmap.
 */
151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219
unsigned long __init init_bootmem(unsigned long start, unsigned long pages)
{
	max_low_pfn = pages;
	min_low_pfn = start;
	return init_bootmem_core(NODE_DATA(0)->bdata, start, 0, pages);
}

static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata)
{
	struct page *page;
	unsigned long pfn;
	unsigned long i, count;
	unsigned long idx;
	unsigned long *map;
	int gofast = 0;

	BUG_ON(!bdata->node_bootmem_map);

	count = 0;
	/* first extant page of the node */
	pfn = PFN_DOWN(bdata->node_boot_start);
	idx = bdata->node_low_pfn - pfn;
	map = bdata->node_bootmem_map;
	/*
	 * Check if we are aligned to BITS_PER_LONG pages.  If so, we might
	 * be able to free page orders of that size at once.
	 */
	if (!(pfn & (BITS_PER_LONG-1)))
		gofast = 1;

	for (i = 0; i < idx; ) {
		unsigned long v = ~map[i / BITS_PER_LONG];

		if (gofast && v == ~0UL) {
			int order;

			page = pfn_to_page(pfn);
			count += BITS_PER_LONG;
			order = ffs(BITS_PER_LONG) - 1;
			__free_pages_bootmem(page, order);
			i += BITS_PER_LONG;
			page += BITS_PER_LONG;
		} else if (v) {
			unsigned long m;

			page = pfn_to_page(pfn);
			for (m = 1; m && i < idx; m<<=1, page++, i++) {
				if (v & m) {
					count++;
					__free_pages_bootmem(page, 0);
				}
			}
		} else {
			i += BITS_PER_LONG;
		}
		pfn += BITS_PER_LONG;
	}

	/*
	 * Now free the allocator bitmap itself, it's not
	 * needed anymore:
	 */
	page = virt_to_page(bdata->node_bootmem_map);
	idx = (get_mapsize(bdata) + PAGE_SIZE-1) >> PAGE_SHIFT;
	for (i = 0; i < idx; i++, page++)
		__free_pages_bootmem(page, 0);
	count += i;
	bdata->node_bootmem_map = NULL;

220 221
	bdebug("nid=%td released=%lx\n", bdata - bootmem_node_data, count);

222 223 224
	return count;
}

225 226 227 228 229 230
/**
 * free_all_bootmem_node - release a node's free pages to the buddy allocator
 * @pgdat: node to be released
 *
 * Returns the number of pages actually released.
 */
231 232 233 234 235 236
unsigned long __init free_all_bootmem_node(pg_data_t *pgdat)
{
	register_page_bootmem_info_node(pgdat);
	return free_all_bootmem_core(pgdat->bdata);
}

237 238 239 240 241
/**
 * free_all_bootmem - release free pages to the buddy allocator
 *
 * Returns the number of pages actually released.
 */
242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278
unsigned long __init free_all_bootmem(void)
{
	return free_all_bootmem_core(NODE_DATA(0)->bdata);
}

static void __init free_bootmem_core(bootmem_data_t *bdata, unsigned long addr,
				     unsigned long size)
{
	unsigned long sidx, eidx;
	unsigned long i;

	BUG_ON(!size);

	/* out range */
	if (addr + size < bdata->node_boot_start ||
		PFN_DOWN(addr) > bdata->node_low_pfn)
		return;
	/*
	 * round down end of usable mem, partially free pages are
	 * considered reserved.
	 */

	if (addr >= bdata->node_boot_start && addr < bdata->last_success)
		bdata->last_success = addr;

	/*
	 * Round up to index to the range.
	 */
	if (PFN_UP(addr) > PFN_DOWN(bdata->node_boot_start))
		sidx = PFN_UP(addr) - PFN_DOWN(bdata->node_boot_start);
	else
		sidx = 0;

	eidx = PFN_DOWN(addr + size - bdata->node_boot_start);
	if (eidx > bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start))
		eidx = bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start);

279 280 281 282
	bdebug("nid=%td start=%lx end=%lx\n", bdata - bootmem_node_data,
		sidx + PFN_DOWN(bdata->node_boot_start),
		eidx + PFN_DOWN(bdata->node_boot_start));

283 284 285 286 287 288
	for (i = sidx; i < eidx; i++) {
		if (unlikely(!test_and_clear_bit(i, bdata->node_bootmem_map)))
			BUG();
	}
}

289 290 291 292 293 294 295 296 297 298
/**
 * free_bootmem_node - mark a page range as usable
 * @pgdat: node the range resides on
 * @physaddr: starting address of the range
 * @size: size of the range in bytes
 *
 * Partial pages will be considered reserved and left as they are.
 *
 * Only physical pages that actually reside on @pgdat are marked.
 */
299 300 301 302 303 304
void __init free_bootmem_node(pg_data_t *pgdat, unsigned long physaddr,
			      unsigned long size)
{
	free_bootmem_core(pgdat->bdata, physaddr, size);
}

305 306 307 308 309 310 311 312 313 314
/**
 * free_bootmem - mark a page range as usable
 * @addr: starting address of the range
 * @size: size of the range in bytes
 *
 * Partial pages will be considered reserved and left as they are.
 *
 * All physical pages within the range are marked, no matter what
 * node they reside on.
 */
315 316 317 318 319 320 321
void __init free_bootmem(unsigned long addr, unsigned long size)
{
	bootmem_data_t *bdata;
	list_for_each_entry(bdata, &bdata_list, list)
		free_bootmem_core(bdata, addr, size);
}

L
Linus Torvalds 已提交
322 323 324 325 326
/*
 * Marks a particular physical memory range as unallocatable. Usable RAM
 * might be used for boot-time allocations - or it might get added
 * to the free page pool later on.
 */
327
static int __init can_reserve_bootmem_core(bootmem_data_t *bdata,
328
			unsigned long addr, unsigned long size, int flags)
L
Linus Torvalds 已提交
329
{
330
	unsigned long sidx, eidx;
L
Linus Torvalds 已提交
331
	unsigned long i;
332 333 334 335 336 337 338

	BUG_ON(!size);

	/* out of range, don't hold other */
	if (addr + size < bdata->node_boot_start ||
		PFN_DOWN(addr) > bdata->node_low_pfn)
		return 0;
339

L
Linus Torvalds 已提交
340
	/*
341
	 * Round up to index to the range.
L
Linus Torvalds 已提交
342
	 */
343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368
	if (addr > bdata->node_boot_start)
		sidx= PFN_DOWN(addr - bdata->node_boot_start);
	else
		sidx = 0;

	eidx = PFN_UP(addr + size - bdata->node_boot_start);
	if (eidx > bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start))
		eidx = bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start);

	for (i = sidx; i < eidx; i++) {
		if (test_bit(i, bdata->node_bootmem_map)) {
			if (flags & BOOTMEM_EXCLUSIVE)
				return -EBUSY;
		}
	}

	return 0;

}

static void __init reserve_bootmem_core(bootmem_data_t *bdata,
			unsigned long addr, unsigned long size, int flags)
{
	unsigned long sidx, eidx;
	unsigned long i;

L
Linus Torvalds 已提交
369
	BUG_ON(!size);
370

371 372 373 374 375 376 377 378 379 380 381 382 383
	/* out of range */
	if (addr + size < bdata->node_boot_start ||
		PFN_DOWN(addr) > bdata->node_low_pfn)
		return;

	/*
	 * Round up to index to the range.
	 */
	if (addr > bdata->node_boot_start)
		sidx= PFN_DOWN(addr - bdata->node_boot_start);
	else
		sidx = 0;

384
	eidx = PFN_UP(addr + size - bdata->node_boot_start);
385 386
	if (eidx > bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start))
		eidx = bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start);
L
Linus Torvalds 已提交
387

388 389 390 391 392 393 394 395 396 397
	bdebug("nid=%td start=%lx end=%lx flags=%x\n",
		bdata - bootmem_node_data,
		sidx + PFN_DOWN(bdata->node_boot_start),
		eidx + PFN_DOWN(bdata->node_boot_start),
		flags);

	for (i = sidx; i < eidx; i++)
		if (test_and_set_bit(i, bdata->node_bootmem_map))
			bdebug("hm, page %lx reserved twice.\n",
				PFN_DOWN(bdata->node_boot_start) + i);
L
Linus Torvalds 已提交
398 399
}

400 401 402 403 404 405 406 407 408 409 410
/**
 * reserve_bootmem_node - mark a page range as reserved
 * @pgdat: node the range resides on
 * @physaddr: starting address of the range
 * @size: size of the range in bytes
 * @flags: reservation flags (see linux/bootmem.h)
 *
 * Partial pages will be reserved.
 *
 * Only physical pages that actually reside on @pgdat are marked.
 */
411 412
int __init reserve_bootmem_node(pg_data_t *pgdat, unsigned long physaddr,
				 unsigned long size, int flags)
L
Linus Torvalds 已提交
413
{
414
	int ret;
L
Linus Torvalds 已提交
415

416 417 418 419 420 421
	ret = can_reserve_bootmem_core(pgdat->bdata, physaddr, size, flags);
	if (ret < 0)
		return -ENOMEM;
	reserve_bootmem_core(pgdat->bdata, physaddr, size, flags);
	return 0;
}
422

423
#ifndef CONFIG_HAVE_ARCH_BOOTMEM_NODE
424 425 426 427 428 429 430 431 432 433 434
/**
 * reserve_bootmem - mark a page range as usable
 * @addr: starting address of the range
 * @size: size of the range in bytes
 * @flags: reservation flags (see linux/bootmem.h)
 *
 * Partial pages will be reserved.
 *
 * All physical pages within the range are marked, no matter what
 * node they reside on.
 */
435 436 437 438 439
int __init reserve_bootmem(unsigned long addr, unsigned long size,
			    int flags)
{
	bootmem_data_t *bdata;
	int ret;
L
Linus Torvalds 已提交
440

441 442 443 444
	list_for_each_entry(bdata, &bdata_list, list) {
		ret = can_reserve_bootmem_core(bdata, addr, size, flags);
		if (ret < 0)
			return ret;
L
Linus Torvalds 已提交
445
	}
446 447 448 449
	list_for_each_entry(bdata, &bdata_list, list)
		reserve_bootmem_core(bdata, addr, size, flags);

	return 0;
L
Linus Torvalds 已提交
450
}
451
#endif /* !CONFIG_HAVE_ARCH_BOOTMEM_NODE */
L
Linus Torvalds 已提交
452 453 454 455 456 457 458 459 460 461 462 463 464 465

/*
 * We 'merge' subsequent allocations to save space. We might 'lose'
 * some fraction of a page if allocations cannot be satisfied due to
 * size constraints on boxes where there is physical RAM space
 * fragmentation - in these cases (mostly large memory boxes) this
 * is not a problem.
 *
 * On low memory boxes we get it right in 100% of the cases.
 *
 * alignment has to be a power of 2 value.
 *
 * NOTE:  This function is _not_ reentrant.
 */
466 467 468
static void * __init
alloc_bootmem_core(struct bootmem_data *bdata, unsigned long size,
		unsigned long align, unsigned long goal, unsigned long limit)
L
Linus Torvalds 已提交
469
{
Y
Yinghai Lu 已提交
470
	unsigned long areasize, preferred;
471
	unsigned long i, start = 0, incr, eidx, end_pfn;
L
Linus Torvalds 已提交
472
	void *ret;
Y
Yinghai Lu 已提交
473 474
	unsigned long node_boot_start;
	void *node_bootmem_map;
L
Linus Torvalds 已提交
475

476
	if (!size) {
477
		printk("alloc_bootmem_core(): zero-sized request\n");
L
Linus Torvalds 已提交
478 479 480 481
		BUG();
	}
	BUG_ON(align & (align-1));

482 483 484 485
	/* on nodes without memory - bootmem_map is NULL */
	if (!bdata->node_bootmem_map)
		return NULL;

486 487 488 489
	bdebug("nid=%td size=%lx [%lu pages] align=%lx goal=%lx limit=%lx\n",
		bdata - bootmem_node_data, size, PAGE_ALIGN(size) >> PAGE_SHIFT,
		align, goal, limit);

Y
Yinghai Lu 已提交
490 491 492 493 494 495 496 497 498 499 500 501 502
	/* bdata->node_boot_start is supposed to be (12+6)bits alignment on x86_64 ? */
	node_boot_start = bdata->node_boot_start;
	node_bootmem_map = bdata->node_bootmem_map;
	if (align) {
		node_boot_start = ALIGN(bdata->node_boot_start, align);
		if (node_boot_start > bdata->node_boot_start)
			node_bootmem_map = (unsigned long *)bdata->node_bootmem_map +
			    PFN_DOWN(node_boot_start - bdata->node_boot_start)/BITS_PER_LONG;
	}

	if (limit && node_boot_start >= limit)
		return NULL;

503 504
	end_pfn = bdata->node_low_pfn;
	limit = PFN_DOWN(limit);
505 506 507
	if (limit && end_pfn > limit)
		end_pfn = limit;

Y
Yinghai Lu 已提交
508
	eidx = end_pfn - PFN_DOWN(node_boot_start);
L
Linus Torvalds 已提交
509 510 511 512 513

	/*
	 * We try to allocate bootmem pages above 'goal'
	 * first, then we try to allocate lower pages.
	 */
514 515
	preferred = 0;
	if (goal && PFN_DOWN(goal) < end_pfn) {
Y
Yinghai Lu 已提交
516 517
		if (goal > node_boot_start)
			preferred = goal - node_boot_start;
L
Linus Torvalds 已提交
518

Y
Yinghai Lu 已提交
519 520
		if (bdata->last_success > node_boot_start &&
			bdata->last_success - node_boot_start >= preferred)
521
			if (!limit || (limit && limit > bdata->last_success))
Y
Yinghai Lu 已提交
522
				preferred = bdata->last_success - node_boot_start;
523
	}
L
Linus Torvalds 已提交
524

Y
Yinghai Lu 已提交
525
	preferred = PFN_DOWN(ALIGN(preferred, align));
526
	areasize = (size + PAGE_SIZE-1) / PAGE_SIZE;
L
Linus Torvalds 已提交
527 528 529
	incr = align >> PAGE_SHIFT ? : 1;

restart_scan:
530
	for (i = preferred; i < eidx;) {
L
Linus Torvalds 已提交
531
		unsigned long j;
532

Y
Yinghai Lu 已提交
533
		i = find_next_zero_bit(node_bootmem_map, eidx, i);
L
Linus Torvalds 已提交
534
		i = ALIGN(i, incr);
535 536
		if (i >= eidx)
			break;
Y
Yinghai Lu 已提交
537
		if (test_bit(i, node_bootmem_map)) {
538
			i += incr;
L
Linus Torvalds 已提交
539
			continue;
540
		}
L
Linus Torvalds 已提交
541 542 543
		for (j = i + 1; j < i + areasize; ++j) {
			if (j >= eidx)
				goto fail_block;
Y
Yinghai Lu 已提交
544
			if (test_bit(j, node_bootmem_map))
L
Linus Torvalds 已提交
545 546 547 548 549 550
				goto fail_block;
		}
		start = i;
		goto found;
	fail_block:
		i = ALIGN(j, incr);
551 552
		if (i == j)
			i += incr;
L
Linus Torvalds 已提交
553 554
	}

Y
Yinghai Lu 已提交
555 556
	if (preferred > 0) {
		preferred = 0;
L
Linus Torvalds 已提交
557 558 559 560 561
		goto restart_scan;
	}
	return NULL;

found:
Y
Yinghai Lu 已提交
562
	bdata->last_success = PFN_PHYS(start) + node_boot_start;
L
Linus Torvalds 已提交
563 564 565 566 567 568 569 570 571
	BUG_ON(start >= eidx);

	/*
	 * Is the next page of the previous allocation-end the start
	 * of this allocation's buffer? If yes then we can 'merge'
	 * the previous partial page with this allocation.
	 */
	if (align < PAGE_SIZE &&
	    bdata->last_offset && bdata->last_pos+1 == start) {
Y
Yinghai Lu 已提交
572
		unsigned long offset, remaining_size;
573
		offset = ALIGN(bdata->last_offset, align);
L
Linus Torvalds 已提交
574
		BUG_ON(offset > PAGE_SIZE);
575
		remaining_size = PAGE_SIZE - offset;
L
Linus Torvalds 已提交
576 577 578
		if (size < remaining_size) {
			areasize = 0;
			/* last_pos unchanged */
579 580
			bdata->last_offset = offset + size;
			ret = phys_to_virt(bdata->last_pos * PAGE_SIZE +
Y
Yinghai Lu 已提交
581
					   offset + node_boot_start);
L
Linus Torvalds 已提交
582 583
		} else {
			remaining_size = size - remaining_size;
584 585
			areasize = (remaining_size + PAGE_SIZE-1) / PAGE_SIZE;
			ret = phys_to_virt(bdata->last_pos * PAGE_SIZE +
Y
Yinghai Lu 已提交
586
					   offset + node_boot_start);
587
			bdata->last_pos = start + areasize - 1;
L
Linus Torvalds 已提交
588 589 590 591 592 593
			bdata->last_offset = remaining_size;
		}
		bdata->last_offset &= ~PAGE_MASK;
	} else {
		bdata->last_pos = start + areasize - 1;
		bdata->last_offset = size & ~PAGE_MASK;
Y
Yinghai Lu 已提交
594
		ret = phys_to_virt(start * PAGE_SIZE + node_boot_start);
L
Linus Torvalds 已提交
595 596
	}

597 598 599 600 601
	bdebug("nid=%td start=%lx end=%lx\n",
		bdata - bootmem_node_data,
		start + PFN_DOWN(bdata->node_boot_start),
		start + areasize + PFN_DOWN(bdata->node_boot_start));

L
Linus Torvalds 已提交
602 603 604
	/*
	 * Reserve the area now:
	 */
605
	for (i = start; i < start + areasize; i++)
Y
Yinghai Lu 已提交
606
		if (unlikely(test_and_set_bit(i, node_bootmem_map)))
L
Linus Torvalds 已提交
607 608 609 610 611
			BUG();
	memset(ret, 0, size);
	return ret;
}

612 613 614 615 616 617 618 619 620 621 622 623 624
/**
 * __alloc_bootmem_nopanic - allocate boot memory without panicking
 * @size: size of the request in bytes
 * @align: alignment of the region
 * @goal: preferred starting address of the region
 *
 * The goal is dropped if it can not be satisfied and the allocation will
 * fall back to memory below @goal.
 *
 * Allocation may happen on any node in the system.
 *
 * Returns NULL on failure.
 */
625 626
void * __init __alloc_bootmem_nopanic(unsigned long size, unsigned long align,
				      unsigned long goal)
L
Linus Torvalds 已提交
627
{
628
	bootmem_data_t *bdata;
L
Linus Torvalds 已提交
629 630
	void *ptr;

631
	list_for_each_entry(bdata, &bdata_list, list) {
632
		ptr = alloc_bootmem_core(bdata, size, align, goal, 0);
633 634 635
		if (ptr)
			return ptr;
	}
636 637
	return NULL;
}
L
Linus Torvalds 已提交
638

639 640 641 642 643 644 645 646 647 648 649 650 651
/**
 * __alloc_bootmem - allocate boot memory
 * @size: size of the request in bytes
 * @align: alignment of the region
 * @goal: preferred starting address of the region
 *
 * The goal is dropped if it can not be satisfied and the allocation will
 * fall back to memory below @goal.
 *
 * Allocation may happen on any node in the system.
 *
 * The function panics if the request can not be satisfied.
 */
652 653
void * __init __alloc_bootmem(unsigned long size, unsigned long align,
			      unsigned long goal)
654 655
{
	void *mem = __alloc_bootmem_nopanic(size,align,goal);
656

657 658
	if (mem)
		return mem;
L
Linus Torvalds 已提交
659 660 661 662 663 664 665 666
	/*
	 * Whoops, we cannot satisfy the allocation request.
	 */
	printk(KERN_ALERT "bootmem alloc of %lu bytes failed!\n", size);
	panic("Out of memory");
	return NULL;
}

667 668 669 670 671 672 673 674 675 676 677 678 679 680 681
/**
 * __alloc_bootmem_node - allocate boot memory from a specific node
 * @pgdat: node to allocate from
 * @size: size of the request in bytes
 * @align: alignment of the region
 * @goal: preferred starting address of the region
 *
 * The goal is dropped if it can not be satisfied and the allocation will
 * fall back to memory below @goal.
 *
 * Allocation may fall back to any node in the system if the specified node
 * can not hold the requested memory.
 *
 * The function panics if the request can not be satisfied.
 */
682 683
void * __init __alloc_bootmem_node(pg_data_t *pgdat, unsigned long size,
				   unsigned long align, unsigned long goal)
L
Linus Torvalds 已提交
684 685 686
{
	void *ptr;

687
	ptr = alloc_bootmem_core(pgdat->bdata, size, align, goal, 0);
L
Linus Torvalds 已提交
688
	if (ptr)
689
		return ptr;
L
Linus Torvalds 已提交
690

691
	return __alloc_bootmem(size, align, goal);
L
Linus Torvalds 已提交
692 693
}

694
#ifdef CONFIG_SPARSEMEM
695 696 697 698 699 700 701
/**
 * alloc_bootmem_section - allocate boot memory from a specific section
 * @size: size of the request in bytes
 * @section_nr: sparse map section to allocate from
 *
 * Return NULL on failure.
 */
702 703 704 705 706 707 708 709 710 711 712
void * __init alloc_bootmem_section(unsigned long size,
				    unsigned long section_nr)
{
	void *ptr;
	unsigned long limit, goal, start_nr, end_nr, pfn;
	struct pglist_data *pgdat;

	pfn = section_nr_to_pfn(section_nr);
	goal = PFN_PHYS(pfn);
	limit = PFN_PHYS(section_nr_to_pfn(section_nr + 1)) - 1;
	pgdat = NODE_DATA(early_pfn_to_nid(pfn));
713 714
	ptr = alloc_bootmem_core(pgdat->bdata, size, SMP_CACHE_BYTES, goal,
				limit);
715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731

	if (!ptr)
		return NULL;

	start_nr = pfn_to_section_nr(PFN_DOWN(__pa(ptr)));
	end_nr = pfn_to_section_nr(PFN_DOWN(__pa(ptr) + size));
	if (start_nr != section_nr || end_nr != section_nr) {
		printk(KERN_WARNING "alloc_bootmem failed on section %ld.\n",
		       section_nr);
		free_bootmem_core(pgdat->bdata, __pa(ptr), size);
		ptr = NULL;
	}

	return ptr;
}
#endif

732 733 734 735 736 737 738 739 740 741 742 743
void * __init __alloc_bootmem_node_nopanic(pg_data_t *pgdat, unsigned long size,
				   unsigned long align, unsigned long goal)
{
	void *ptr;

	ptr = alloc_bootmem_core(pgdat->bdata, size, align, goal, 0);
	if (ptr)
		return ptr;

	return __alloc_bootmem_nopanic(size, align, goal);
}

744 745 746
#ifndef ARCH_LOW_ADDRESS_LIMIT
#define ARCH_LOW_ADDRESS_LIMIT	0xffffffffUL
#endif
747

748 749 750 751 752 753 754 755 756 757 758 759 760
/**
 * __alloc_bootmem_low - allocate low boot memory
 * @size: size of the request in bytes
 * @align: alignment of the region
 * @goal: preferred starting address of the region
 *
 * The goal is dropped if it can not be satisfied and the allocation will
 * fall back to memory below @goal.
 *
 * Allocation may happen on any node in the system.
 *
 * The function panics if the request can not be satisfied.
 */
761 762
void * __init __alloc_bootmem_low(unsigned long size, unsigned long align,
				  unsigned long goal)
763
{
764
	bootmem_data_t *bdata;
765 766
	void *ptr;

767
	list_for_each_entry(bdata, &bdata_list, list) {
768 769
		ptr = alloc_bootmem_core(bdata, size, align, goal,
					ARCH_LOW_ADDRESS_LIMIT);
770 771 772
		if (ptr)
			return ptr;
	}
773 774 775 776 777 778 779 780 781

	/*
	 * Whoops, we cannot satisfy the allocation request.
	 */
	printk(KERN_ALERT "low bootmem alloc of %lu bytes failed!\n", size);
	panic("Out of low memory");
	return NULL;
}

782 783 784 785 786 787 788 789 790 791 792 793 794 795 796
/**
 * __alloc_bootmem_low_node - allocate low boot memory from a specific node
 * @pgdat: node to allocate from
 * @size: size of the request in bytes
 * @align: alignment of the region
 * @goal: preferred starting address of the region
 *
 * The goal is dropped if it can not be satisfied and the allocation will
 * fall back to memory below @goal.
 *
 * Allocation may fall back to any node in the system if the specified node
 * can not hold the requested memory.
 *
 * The function panics if the request can not be satisfied.
 */
797 798 799
void * __init __alloc_bootmem_low_node(pg_data_t *pgdat, unsigned long size,
				       unsigned long align, unsigned long goal)
{
800 801
	return alloc_bootmem_core(pgdat->bdata, size, align, goal,
				ARCH_LOW_ADDRESS_LIMIT);
802
}