[linux-2.6-block.git] / mm / sparse.c

/*
 * sparse memory mappings.
 */
#include <linux/mm.h>
#include <linux/mmzone.h>
#include <linux/bootmem.h>
#include <linux/highmem.h>
#include <linux/module.h>
#include <linux/spinlock.h>
#include <linux/vmalloc.h>
#include <asm/dma.h>
#include <asm/pgalloc.h>
#include <asm/pgtable.h>

/*
 * Permanent SPARSEMEM data:
 *
 * 1) mem_section	- memory sections, mem_map's for valid memory
 */
#ifdef CONFIG_SPARSEMEM_EXTREME
struct mem_section *mem_section[NR_SECTION_ROOTS]
	____cacheline_internodealigned_in_smp;
#else
struct mem_section mem_section[NR_SECTION_ROOTS][SECTIONS_PER_ROOT]
	____cacheline_internodealigned_in_smp;
#endif
EXPORT_SYMBOL(mem_section);

#ifdef NODE_NOT_IN_PAGE_FLAGS
/*
 * If we did not store the node number in the page then we have to
 * do a lookup in the section_to_node_table in order to find which
 * node the page belongs to.
 */
#if MAX_NUMNODES <= 256
static u8 section_to_node_table[NR_MEM_SECTIONS] __cacheline_aligned;
#else
static u16 section_to_node_table[NR_MEM_SECTIONS] __cacheline_aligned;
#endif

int page_to_nid(struct page *page)
{
	return section_to_node_table[page_to_section(page)];
}
EXPORT_SYMBOL(page_to_nid);

static void set_section_nid(unsigned long section_nr, int nid)
{
	section_to_node_table[section_nr] = nid;
}
#else /* !NODE_NOT_IN_PAGE_FLAGS */
static inline void set_section_nid(unsigned long section_nr, int nid)
{
}
#endif

#ifdef CONFIG_SPARSEMEM_EXTREME
static struct mem_section noinline __init_refok *sparse_index_alloc(int nid)
{
	struct mem_section *section = NULL;
	unsigned long array_size = SECTIONS_PER_ROOT *
				   sizeof(struct mem_section);

	if (slab_is_available())
		section = kmalloc_node(array_size, GFP_KERNEL, nid);
	else
		section = alloc_bootmem_node(NODE_DATA(nid), array_size);

	if (section)
		memset(section, 0, array_size);

	return section;
}

static int __meminit sparse_index_init(unsigned long section_nr, int nid)
{
	static DEFINE_SPINLOCK(index_init_lock);
	unsigned long root = SECTION_NR_TO_ROOT(section_nr);
	struct mem_section *section;
	int ret = 0;

	if (mem_section[root])
		return -EEXIST;

	section = sparse_index_alloc(nid);
	/*
	 * This lock keeps two different sections from
	 * reallocating for the same index
	 */
	spin_lock(&index_init_lock);

	if (mem_section[root]) {
		ret = -EEXIST;
		goto out;
	}

	mem_section[root] = section;
out:
	spin_unlock(&index_init_lock);
	return ret;
}
#else /* !SPARSEMEM_EXTREME */
static inline int sparse_index_init(unsigned long section_nr, int nid)
{
	return 0;
}
#endif

/*
 * Although written for the SPARSEMEM_EXTREME case, this happens
 * to also work for the flat array case because
 * NR_SECTION_ROOTS==NR_MEM_SECTIONS.
 */
int __section_nr(struct mem_section* ms)
{
	unsigned long root_nr;
	struct mem_section* root;

	for (root_nr = 0; root_nr < NR_SECTION_ROOTS; root_nr++) {
		root = __nr_to_section(root_nr * SECTIONS_PER_ROOT);
		if (!root)
			continue;

		if ((ms >= root) && (ms < (root + SECTIONS_PER_ROOT)))
		     break;
	}

	return (root_nr * SECTIONS_PER_ROOT) + (ms - root);
}

/*
 * During early boot, before section_mem_map is used for an actual
 * mem_map, we use section_mem_map to store the section's NUMA
 * node.  This keeps us from having to use another data structure.  The
 * node information is cleared just before we store the real mem_map.
 */
static inline unsigned long sparse_encode_early_nid(int nid)
{
	return (nid << SECTION_NID_SHIFT);
}

static inline int sparse_early_nid(struct mem_section *section)
{
	return (section->section_mem_map >> SECTION_NID_SHIFT);
}

/* Record a memory area against a node. */
void __init memory_present(int nid, unsigned long start, unsigned long end)
{
	unsigned long pfn;

	start &= PAGE_SECTION_MASK;
	for (pfn = start; pfn < end; pfn += PAGES_PER_SECTION) {
		unsigned long section = pfn_to_section_nr(pfn);
		struct mem_section *ms;

		sparse_index_init(section, nid);
		set_section_nid(section, nid);

		ms = __nr_to_section(section);
		if (!ms->section_mem_map)
			ms->section_mem_map = sparse_encode_early_nid(nid) |
							SECTION_MARKED_PRESENT;
	}
}

/*
 * Only used by the i386 NUMA architecures, but relatively
 * generic code.
 */
unsigned long __init node_memmap_size_bytes(int nid, unsigned long start_pfn,
						     unsigned long end_pfn)
{
	unsigned long pfn;
	unsigned long nr_pages = 0;

	for (pfn = start_pfn; pfn < end_pfn; pfn += PAGES_PER_SECTION) {
		if (nid != early_pfn_to_nid(pfn))
			continue;

		if (pfn_present(pfn))
			nr_pages += PAGES_PER_SECTION;
	}

	return nr_pages * sizeof(struct page);
}

/*
 * Subtle, we encode the real pfn into the mem_map such that
 * the identity pfn - section_mem_map will return the actual
 * physical page frame number.
 */
static unsigned long sparse_encode_mem_map(struct page *mem_map, unsigned long pnum)
{
	return (unsigned long)(mem_map - (section_nr_to_pfn(pnum)));
}

/*
 * We need this if we ever free the mem_maps.  While not implemented yet,
 * this function is included for parity with its sibling.
 */
static __attribute((unused))
struct page *sparse_decode_mem_map(unsigned long coded_mem_map, unsigned long pnum)
{
	return ((struct page *)coded_mem_map) + section_nr_to_pfn(pnum);
}

static int __meminit sparse_init_one_section(struct mem_section *ms,
		unsigned long pnum, struct page *mem_map,
		unsigned long *pageblock_bitmap)
{
	if (!present_section(ms))
		return -EINVAL;

	ms->section_mem_map &= ~SECTION_MAP_MASK;
	ms->section_mem_map |= sparse_encode_mem_map(mem_map, pnum) |
							SECTION_HAS_MEM_MAP;
 	ms->pageblock_flags = pageblock_bitmap;

	return 1;
}

__attribute__((weak)) __init
void *alloc_bootmem_high_node(pg_data_t *pgdat, unsigned long size)
{
	return NULL;
}

static unsigned long usemap_size(void)
{
	unsigned long size_bytes;
	size_bytes = roundup(SECTION_BLOCKFLAGS_BITS, 8) / 8;
	size_bytes = roundup(size_bytes, sizeof(unsigned long));
	return size_bytes;
}

#ifdef CONFIG_MEMORY_HOTPLUG
static unsigned long *__kmalloc_section_usemap(void)
{
	return kmalloc(usemap_size(), GFP_KERNEL);
}
#endif /* CONFIG_MEMORY_HOTPLUG */

static unsigned long *sparse_early_usemap_alloc(unsigned long pnum)
{
	unsigned long *usemap;
	struct mem_section *ms = __nr_to_section(pnum);
	int nid = sparse_early_nid(ms);

	usemap = alloc_bootmem_node(NODE_DATA(nid), usemap_size());
	if (usemap)
		return usemap;

	/* Stupid: suppress gcc warning for SPARSEMEM && !NUMA */
	nid = 0;

	printk(KERN_WARNING "%s: allocation failed\n", __FUNCTION__);
	return NULL;
}

#ifndef CONFIG_SPARSEMEM_VMEMMAP
struct page __init *sparse_early_mem_map_populate(unsigned long pnum, int nid)
{
	struct page *map;

	map = alloc_remap(nid, sizeof(struct page) * PAGES_PER_SECTION);
	if (map)
		return map;

  	map = alloc_bootmem_high_node(NODE_DATA(nid),
                       sizeof(struct page) * PAGES_PER_SECTION);
	if (map)
		return map;

	map = alloc_bootmem_node(NODE_DATA(nid),
			sizeof(struct page) * PAGES_PER_SECTION);
	return map;
}
#endif /* !CONFIG_SPARSEMEM_VMEMMAP */

struct page __init *sparse_early_mem_map_alloc(unsigned long pnum)
{
	struct page *map;
	struct mem_section *ms = __nr_to_section(pnum);
	int nid = sparse_early_nid(ms);

	map = sparse_early_mem_map_populate(pnum, nid);
	if (map)
		return map;

	printk(KERN_ERR "%s: sparsemem memory map backing failed "
			"some memory will not be available.\n", __FUNCTION__);
	ms->section_mem_map = 0;
	return NULL;
}

/*
 * Allocate the accumulated non-linear sections, allocate a mem_map
 * for each and record the physical to section mapping.
 */
void __init sparse_init(void)
{
	unsigned long pnum;
	struct page *map;
	unsigned long *usemap;

	for (pnum = 0; pnum < NR_MEM_SECTIONS; pnum++) {
		if (!present_section_nr(pnum))
			continue;

		map = sparse_early_mem_map_alloc(pnum);
		if (!map)
			continue;

		usemap = sparse_early_usemap_alloc(pnum);
		if (!usemap)
			continue;

		sparse_init_one_section(__nr_to_section(pnum), pnum, map,
								usemap);
	}
}

#ifdef CONFIG_MEMORY_HOTPLUG
static struct page *__kmalloc_section_memmap(unsigned long nr_pages)
{
	struct page *page, *ret;
	unsigned long memmap_size = sizeof(struct page) * nr_pages;

	page = alloc_pages(GFP_KERNEL|__GFP_NOWARN, get_order(memmap_size));
	if (page)
		goto got_map_page;

	ret = vmalloc(memmap_size);
	if (ret)
		goto got_map_ptr;

	return NULL;
got_map_page:
	ret = (struct page *)pfn_to_kaddr(page_to_pfn(page));
got_map_ptr:
	memset(ret, 0, memmap_size);

	return ret;
}

static int vaddr_in_vmalloc_area(void *addr)
{
	if (addr >= (void *)VMALLOC_START &&
	    addr < (void *)VMALLOC_END)
		return 1;
	return 0;
}

static void __kfree_section_memmap(struct page *memmap, unsigned long nr_pages)
{
	if (vaddr_in_vmalloc_area(memmap))
		vfree(memmap);
	else
		free_pages((unsigned long)memmap,
			   get_order(sizeof(struct page) * nr_pages));
}

/*
 * returns the number of sections whose mem_maps were properly
 * set.  If this is <=0, then that means that the passed-in
 * map was not consumed and must be freed.
 */
int sparse_add_one_section(struct zone *zone, unsigned long start_pfn,
			   int nr_pages)
{
	unsigned long section_nr = pfn_to_section_nr(start_pfn);
	struct pglist_data *pgdat = zone->zone_pgdat;
	struct mem_section *ms;
	struct page *memmap;
	unsigned long *usemap;
	unsigned long flags;
	int ret;

	/*
	 * no locking for this, because it does its own
	 * plus, it does a kmalloc
	 */
	sparse_index_init(section_nr, pgdat->node_id);
	memmap = __kmalloc_section_memmap(nr_pages);
	usemap = __kmalloc_section_usemap();

	pgdat_resize_lock(pgdat, &flags);

	ms = __pfn_to_section(start_pfn);
	if (ms->section_mem_map & SECTION_MARKED_PRESENT) {
		ret = -EEXIST;
		goto out;
	}

	if (!usemap) {
		ret = -ENOMEM;
		goto out;
	}
	ms->section_mem_map |= SECTION_MARKED_PRESENT;

	ret = sparse_init_one_section(ms, section_nr, memmap, usemap);

out:
	pgdat_resize_unlock(pgdat, &flags);
	if (ret <= 0)
		__kfree_section_memmap(memmap, nr_pages);
	return ret;
}
#endif
Commit	Line	Data
d41dee36 AW	1	/*
	2	* sparse memory mappings.
	3	*/
d41dee36 AW	4	#include <linux/mm.h>
	5	#include <linux/mmzone.h>
	6	#include <linux/bootmem.h>
0b0acbec	7	#include <linux/highmem.h>
d41dee36	8	#include <linux/module.h>
28ae55c9	9	#include <linux/spinlock.h>
0b0acbec	10	#include <linux/vmalloc.h>
d41dee36	11	#include <asm/dma.h>
8f6aac41 CL	12	#include <asm/pgalloc.h>
8f6aac41 CL	13	#include <asm/pgtable.h>
d41dee36 AW	14
	15	/*
	16	* Permanent SPARSEMEM data:
	17	*
	18	* 1) mem_section - memory sections, mem_map's for valid memory
	19	*/
3e347261	20	#ifdef CONFIG_SPARSEMEM_EXTREME
802f192e	21	struct mem_section *mem_section[NR_SECTION_ROOTS]
22fc6ecc	22	____cacheline_internodealigned_in_smp;
3e347261 BP	23	#else
3e347261 BP	24	struct mem_section mem_section[NR_SECTION_ROOTS][SECTIONS_PER_ROOT]
22fc6ecc	25	____cacheline_internodealigned_in_smp;
3e347261 BP	26	#endif
	27	EXPORT_SYMBOL(mem_section);
	28
89689ae7 CL	29	#ifdef NODE_NOT_IN_PAGE_FLAGS
	30	/*
	31	* If we did not store the node number in the page then we have to
	32	* do a lookup in the section_to_node_table in order to find which
	33	* node the page belongs to.
	34	*/
	35	#if MAX_NUMNODES <= 256
	36	static u8 section_to_node_table[NR_MEM_SECTIONS] __cacheline_aligned;
	37	#else
	38	static u16 section_to_node_table[NR_MEM_SECTIONS] __cacheline_aligned;
	39	#endif
	40
25ba77c1	41	int page_to_nid(struct page *page)
89689ae7 CL	42	{
	43	return section_to_node_table[page_to_section(page)];
	44	}
	45	EXPORT_SYMBOL(page_to_nid);
85770ffe AW	46
	47	static void set_section_nid(unsigned long section_nr, int nid)
	48	{
	49	section_to_node_table[section_nr] = nid;
	50	}
	51	#else /* !NODE_NOT_IN_PAGE_FLAGS */
	52	static inline void set_section_nid(unsigned long section_nr, int nid)
	53	{
	54	}
89689ae7 CL	55	#endif
89689ae7 CL	56
3e347261	57	#ifdef CONFIG_SPARSEMEM_EXTREME
577a32f6	58	static struct mem_section noinline __init_refok *sparse_index_alloc(int nid)
28ae55c9 DH	59	{
	60	struct mem_section *section = NULL;
	61	unsigned long array_size = SECTIONS_PER_ROOT *
	62	sizeof(struct mem_section);
	63
39d24e64	64	if (slab_is_available())
46a66eec MK	65	section = kmalloc_node(array_size, GFP_KERNEL, nid);
	66	else
	67	section = alloc_bootmem_node(NODE_DATA(nid), array_size);
28ae55c9 DH	68
	69	if (section)
	70	memset(section, 0, array_size);
	71
	72	return section;
3e347261	73	}
802f192e	74
a3142c8e	75	static int __meminit sparse_index_init(unsigned long section_nr, int nid)
802f192e	76	{
34af946a	77	static DEFINE_SPINLOCK(index_init_lock);
28ae55c9 DH	78	unsigned long root = SECTION_NR_TO_ROOT(section_nr);
	79	struct mem_section *section;
	80	int ret = 0;
802f192e BP	81
802f192e BP	82	if (mem_section[root])
28ae55c9	83	return -EEXIST;
3e347261	84
28ae55c9 DH	85	section = sparse_index_alloc(nid);
	86	/*
	87	* This lock keeps two different sections from
	88	* reallocating for the same index
	89	*/
	90	spin_lock(&index_init_lock);
3e347261	91
28ae55c9 DH	92	if (mem_section[root]) {
	93	ret = -EEXIST;
	94	goto out;
	95	}
	96
	97	mem_section[root] = section;
	98	out:
	99	spin_unlock(&index_init_lock);
	100	return ret;
	101	}
	102	#else /* !SPARSEMEM_EXTREME */
	103	static inline int sparse_index_init(unsigned long section_nr, int nid)
	104	{
	105	return 0;
802f192e	106	}
28ae55c9 DH	107	#endif
28ae55c9 DH	108
4ca644d9 DH	109	/*
4ca644d9 DH	110	* Although written for the SPARSEMEM_EXTREME case, this happens
cd881a6b	111	* to also work for the flat array case because
4ca644d9 DH	112	* NR_SECTION_ROOTS==NR_MEM_SECTIONS.
	113	*/
	114	int __section_nr(struct mem_section* ms)
	115	{
	116	unsigned long root_nr;
	117	struct mem_section* root;
	118
12783b00 MK	119	for (root_nr = 0; root_nr < NR_SECTION_ROOTS; root_nr++) {
12783b00 MK	120	root = __nr_to_section(root_nr * SECTIONS_PER_ROOT);
4ca644d9 DH	121	if (!root)
	122	continue;
	123
	124	if ((ms >= root) && (ms < (root + SECTIONS_PER_ROOT)))
	125	break;
	126	}
	127
	128	return (root_nr * SECTIONS_PER_ROOT) + (ms - root);
	129	}
	130
30c253e6 AW	131	/*
	132	* During early boot, before section_mem_map is used for an actual
	133	* mem_map, we use section_mem_map to store the section's NUMA
	134	* node. This keeps us from having to use another data structure. The
	135	* node information is cleared just before we store the real mem_map.
	136	*/
	137	static inline unsigned long sparse_encode_early_nid(int nid)
	138	{
	139	return (nid << SECTION_NID_SHIFT);
	140	}
	141
	142	static inline int sparse_early_nid(struct mem_section *section)
	143	{
	144	return (section->section_mem_map >> SECTION_NID_SHIFT);
	145	}
	146
d41dee36	147	/* Record a memory area against a node. */
a3142c8e	148	void __init memory_present(int nid, unsigned long start, unsigned long end)
d41dee36 AW	149	{
	150	unsigned long pfn;
	151
	152	start &= PAGE_SECTION_MASK;
	153	for (pfn = start; pfn < end; pfn += PAGES_PER_SECTION) {
	154	unsigned long section = pfn_to_section_nr(pfn);
802f192e BP	155	struct mem_section *ms;
	156
	157	sparse_index_init(section, nid);
85770ffe	158	set_section_nid(section, nid);
802f192e BP	159
	160	ms = __nr_to_section(section);
	161	if (!ms->section_mem_map)
30c253e6 AW	162	ms->section_mem_map = sparse_encode_early_nid(nid) \|
30c253e6 AW	163	SECTION_MARKED_PRESENT;
d41dee36 AW	164	}
	165	}
	166
	167	/*
	168	* Only used by the i386 NUMA architecures, but relatively
	169	* generic code.
	170	*/
	171	unsigned long __init node_memmap_size_bytes(int nid, unsigned long start_pfn,
	172	unsigned long end_pfn)
	173	{
	174	unsigned long pfn;
	175	unsigned long nr_pages = 0;
	176
	177	for (pfn = start_pfn; pfn < end_pfn; pfn += PAGES_PER_SECTION) {
	178	if (nid != early_pfn_to_nid(pfn))
	179	continue;
	180
540557b9	181	if (pfn_present(pfn))
d41dee36 AW	182	nr_pages += PAGES_PER_SECTION;
	183	}
	184
	185	return nr_pages * sizeof(struct page);
	186	}
	187
29751f69 AW	188	/*
	189	* Subtle, we encode the real pfn into the mem_map such that
	190	* the identity pfn - section_mem_map will return the actual
	191	* physical page frame number.
	192	*/
	193	static unsigned long sparse_encode_mem_map(struct page *mem_map, unsigned long pnum)
	194	{
	195	return (unsigned long)(mem_map - (section_nr_to_pfn(pnum)));
	196	}
	197
	198	/*
	199	* We need this if we ever free the mem_maps. While not implemented yet,
	200	* this function is included for parity with its sibling.
	201	*/
	202	static __attribute((unused))
	203	struct page *sparse_decode_mem_map(unsigned long coded_mem_map, unsigned long pnum)
	204	{
	205	return ((struct page *)coded_mem_map) + section_nr_to_pfn(pnum);
	206	}
	207
a3142c8e	208	static int __meminit sparse_init_one_section(struct mem_section *ms,
5c0e3066 MG	209	unsigned long pnum, struct page *mem_map,
5c0e3066 MG	210	unsigned long *pageblock_bitmap)
29751f69	211	{
540557b9	212	if (!present_section(ms))
29751f69 AW	213	return -EINVAL;
29751f69 AW	214
30c253e6	215	ms->section_mem_map &= ~SECTION_MAP_MASK;
540557b9 AW	216	ms->section_mem_map \|= sparse_encode_mem_map(mem_map, pnum) \|
540557b9 AW	217	SECTION_HAS_MEM_MAP;
5c0e3066	218	ms->pageblock_flags = pageblock_bitmap;
29751f69 AW	219
	220	return 1;
	221	}
	222
dec2e6b7	223	__attribute__((weak)) __init
2e1c49db ZN	224	void alloc_bootmem_high_node(pg_data_t pgdat, unsigned long size)
	225	{
	226	return NULL;
	227	}
	228
5c0e3066 MG	229	static unsigned long usemap_size(void)
	230	{
	231	unsigned long size_bytes;
	232	size_bytes = roundup(SECTION_BLOCKFLAGS_BITS, 8) / 8;
	233	size_bytes = roundup(size_bytes, sizeof(unsigned long));
	234	return size_bytes;
	235	}
	236
	237	#ifdef CONFIG_MEMORY_HOTPLUG
	238	static unsigned long *__kmalloc_section_usemap(void)
	239	{
	240	return kmalloc(usemap_size(), GFP_KERNEL);
	241	}
	242	#endif /* CONFIG_MEMORY_HOTPLUG */
	243
	244	static unsigned long *sparse_early_usemap_alloc(unsigned long pnum)
	245	{
	246	unsigned long *usemap;
	247	struct mem_section *ms = __nr_to_section(pnum);
	248	int nid = sparse_early_nid(ms);
	249
	250	usemap = alloc_bootmem_node(NODE_DATA(nid), usemap_size());
	251	if (usemap)
	252	return usemap;
	253
	254	/* Stupid: suppress gcc warning for SPARSEMEM && !NUMA */
	255	nid = 0;
	256
	257	printk(KERN_WARNING "%s: allocation failed\n", __FUNCTION__);
	258	return NULL;
	259	}
	260
8f6aac41 CL	261	#ifndef CONFIG_SPARSEMEM_VMEMMAP
8f6aac41 CL	262	struct page __init *sparse_early_mem_map_populate(unsigned long pnum, int nid)
29751f69 AW	263	{
29751f69 AW	264	struct page *map;
29751f69 AW	265
	266	map = alloc_remap(nid, sizeof(struct page) * PAGES_PER_SECTION);
	267	if (map)
	268	return map;
	269
2e1c49db ZN	270	map = alloc_bootmem_high_node(NODE_DATA(nid),
	271	sizeof(struct page) * PAGES_PER_SECTION);
	272	if (map)
	273	return map;
	274
29751f69 AW	275	map = alloc_bootmem_node(NODE_DATA(nid),
29751f69 AW	276	sizeof(struct page) * PAGES_PER_SECTION);
8f6aac41 CL	277	return map;
	278	}
	279	#endif /* !CONFIG_SPARSEMEM_VMEMMAP */
	280
	281	struct page __init *sparse_early_mem_map_alloc(unsigned long pnum)
	282	{
	283	struct page *map;
	284	struct mem_section *ms = __nr_to_section(pnum);
	285	int nid = sparse_early_nid(ms);
	286
	287	map = sparse_early_mem_map_populate(pnum, nid);
29751f69 AW	288	if (map)
	289	return map;
	290
8f6aac41 CL	291	printk(KERN_ERR "%s: sparsemem memory map backing failed "
8f6aac41 CL	292	"some memory will not be available.\n", __FUNCTION__);
802f192e	293	ms->section_mem_map = 0;
29751f69 AW	294	return NULL;
	295	}
	296
193faea9 SR	297	/*
	298	* Allocate the accumulated non-linear sections, allocate a mem_map
	299	* for each and record the physical to section mapping.
	300	*/
	301	void __init sparse_init(void)
	302	{
	303	unsigned long pnum;
	304	struct page *map;
5c0e3066	305	unsigned long *usemap;
193faea9 SR	306
193faea9 SR	307	for (pnum = 0; pnum < NR_MEM_SECTIONS; pnum++) {
540557b9	308	if (!present_section_nr(pnum))
193faea9 SR	309	continue;
	310
	311	map = sparse_early_mem_map_alloc(pnum);
	312	if (!map)
	313	continue;
5c0e3066 MG	314
	315	usemap = sparse_early_usemap_alloc(pnum);
	316	if (!usemap)
	317	continue;
	318
	319	sparse_init_one_section(__nr_to_section(pnum), pnum, map,
	320	usemap);
193faea9 SR	321	}
	322	}
	323
	324	#ifdef CONFIG_MEMORY_HOTPLUG
0b0acbec DH	325	static struct page *__kmalloc_section_memmap(unsigned long nr_pages)
	326	{
	327	struct page page, ret;
	328	unsigned long memmap_size = sizeof(struct page) * nr_pages;
	329
f2d0aa5b	330	page = alloc_pages(GFP_KERNEL\|__GFP_NOWARN, get_order(memmap_size));
0b0acbec DH	331	if (page)
	332	goto got_map_page;
	333
	334	ret = vmalloc(memmap_size);
	335	if (ret)
	336	goto got_map_ptr;
	337
	338	return NULL;
	339	got_map_page:
	340	ret = (struct page *)pfn_to_kaddr(page_to_pfn(page));
	341	got_map_ptr:
	342	memset(ret, 0, memmap_size);
	343
	344	return ret;
	345	}
	346
	347	static int vaddr_in_vmalloc_area(void *addr)
	348	{
	349	if (addr >= (void *)VMALLOC_START &&
	350	addr < (void *)VMALLOC_END)
	351	return 1;
	352	return 0;
	353	}
	354
	355	static void __kfree_section_memmap(struct page *memmap, unsigned long nr_pages)
	356	{
	357	if (vaddr_in_vmalloc_area(memmap))
	358	vfree(memmap);
	359	else
	360	free_pages((unsigned long)memmap,
	361	get_order(sizeof(struct page) * nr_pages));
	362	}
	363
29751f69 AW	364	/*
	365	* returns the number of sections whose mem_maps were properly
	366	* set. If this is <=0, then that means that the passed-in
	367	* map was not consumed and must be freed.
	368	*/
0b0acbec DH	369	int sparse_add_one_section(struct zone *zone, unsigned long start_pfn,
0b0acbec DH	370	int nr_pages)
29751f69	371	{
0b0acbec DH	372	unsigned long section_nr = pfn_to_section_nr(start_pfn);
	373	struct pglist_data *pgdat = zone->zone_pgdat;
	374	struct mem_section *ms;
	375	struct page *memmap;
5c0e3066	376	unsigned long *usemap;
0b0acbec DH	377	unsigned long flags;
0b0acbec DH	378	int ret;
29751f69	379
0b0acbec DH	380	/*
	381	* no locking for this, because it does its own
	382	* plus, it does a kmalloc
	383	*/
	384	sparse_index_init(section_nr, pgdat->node_id);
	385	memmap = __kmalloc_section_memmap(nr_pages);
5c0e3066	386	usemap = __kmalloc_section_usemap();
0b0acbec DH	387
0b0acbec DH	388	pgdat_resize_lock(pgdat, &flags);
29751f69	389
0b0acbec DH	390	ms = __pfn_to_section(start_pfn);
	391	if (ms->section_mem_map & SECTION_MARKED_PRESENT) {
	392	ret = -EEXIST;
	393	goto out;
	394	}
5c0e3066 MG	395
	396	if (!usemap) {
	397	ret = -ENOMEM;
	398	goto out;
	399	}
29751f69 AW	400	ms->section_mem_map \|= SECTION_MARKED_PRESENT;
29751f69 AW	401
5c0e3066	402	ret = sparse_init_one_section(ms, section_nr, memmap, usemap);
0b0acbec	403
0b0acbec DH	404	out:
0b0acbec DH	405	pgdat_resize_unlock(pgdat, &flags);
46a66eec MK	406	if (ret <= 0)
46a66eec MK	407	__kfree_section_memmap(memmap, nr_pages);
0b0acbec	408	return ret;
29751f69	409	}
a3142c8e	410	#endif