[net-next-2.6.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);
	if (!section)
		return -ENOMEM;
	/*
	 * 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;
}

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_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_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_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
#ifdef CONFIG_SPARSEMEM_VMEMMAP
static inline struct page *kmalloc_section_memmap(unsigned long pnum, int nid,
						 unsigned long nr_pages)
{
	/* This will make the necessary allocations eventually. */
	return sparse_mem_map_populate(pnum, nid);
}
static void __kfree_section_memmap(struct page *memmap, unsigned long nr_pages)
{
	return; /* XXX: Not implemented yet */
}
#else
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 inline struct page *kmalloc_section_memmap(unsigned long pnum, int nid,
						  unsigned long nr_pages)
{
	return __kmalloc_section_memmap(nr_pages);
}

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));
}
#endif /* CONFIG_SPARSEMEM_VMEMMAP */

/*
 * 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
	 */
	ret = sparse_index_init(section_nr, pgdat->node_id);
	if (ret < 0 && ret != -EEXIST)
		return ret;
	memmap = kmalloc_section_memmap(section_nr, pgdat->node_id, nr_pages);
	if (!memmap)
		return -ENOMEM;
	usemap = __kmalloc_section_usemap();
	if (!usemap) {
		__kfree_section_memmap(memmap, nr_pages);
		return -ENOMEM;
	}

	pgdat_resize_lock(pgdat, &flags);

	ms = __pfn_to_section(start_pfn);
	if (ms->section_mem_map & SECTION_MARKED_PRESENT) {
		ret = -EEXIST;
		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(usemap);
		__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	85	section = sparse_index_alloc(nid);
af0cd5a7 WC	86	if (!section)
af0cd5a7 WC	87	return -ENOMEM;
28ae55c9 DH	88	/*
	89	* This lock keeps two different sections from
	90	* reallocating for the same index
	91	*/
	92	spin_lock(&index_init_lock);
3e347261	93
28ae55c9 DH	94	if (mem_section[root]) {
	95	ret = -EEXIST;
	96	goto out;
	97	}
	98
	99	mem_section[root] = section;
	100	out:
	101	spin_unlock(&index_init_lock);
	102	return ret;
	103	}
	104	#else /* !SPARSEMEM_EXTREME */
	105	static inline int sparse_index_init(unsigned long section_nr, int nid)
	106	{
	107	return 0;
802f192e	108	}
28ae55c9 DH	109	#endif
28ae55c9 DH	110
4ca644d9 DH	111	/*
4ca644d9 DH	112	* Although written for the SPARSEMEM_EXTREME case, this happens
cd881a6b	113	* to also work for the flat array case because
4ca644d9 DH	114	* NR_SECTION_ROOTS==NR_MEM_SECTIONS.
	115	*/
	116	int __section_nr(struct mem_section* ms)
	117	{
	118	unsigned long root_nr;
	119	struct mem_section* root;
	120
12783b00 MK	121	for (root_nr = 0; root_nr < NR_SECTION_ROOTS; root_nr++) {
12783b00 MK	122	root = __nr_to_section(root_nr * SECTIONS_PER_ROOT);
4ca644d9 DH	123	if (!root)
	124	continue;
	125
	126	if ((ms >= root) && (ms < (root + SECTIONS_PER_ROOT)))
	127	break;
	128	}
	129
	130	return (root_nr * SECTIONS_PER_ROOT) + (ms - root);
	131	}
	132
30c253e6 AW	133	/*
	134	* During early boot, before section_mem_map is used for an actual
	135	* mem_map, we use section_mem_map to store the section's NUMA
	136	* node. This keeps us from having to use another data structure. The
	137	* node information is cleared just before we store the real mem_map.
	138	*/
	139	static inline unsigned long sparse_encode_early_nid(int nid)
	140	{
	141	return (nid << SECTION_NID_SHIFT);
	142	}
	143
	144	static inline int sparse_early_nid(struct mem_section *section)
	145	{
	146	return (section->section_mem_map >> SECTION_NID_SHIFT);
	147	}
	148
d41dee36	149	/* Record a memory area against a node. */
a3142c8e	150	void __init memory_present(int nid, unsigned long start, unsigned long end)
d41dee36 AW	151	{
	152	unsigned long pfn;
	153
	154	start &= PAGE_SECTION_MASK;
	155	for (pfn = start; pfn < end; pfn += PAGES_PER_SECTION) {
	156	unsigned long section = pfn_to_section_nr(pfn);
802f192e BP	157	struct mem_section *ms;
	158
	159	sparse_index_init(section, nid);
85770ffe	160	set_section_nid(section, nid);
802f192e BP	161
	162	ms = __nr_to_section(section);
	163	if (!ms->section_mem_map)
30c253e6 AW	164	ms->section_mem_map = sparse_encode_early_nid(nid) \|
30c253e6 AW	165	SECTION_MARKED_PRESENT;
d41dee36 AW	166	}
	167	}
	168
	169	/*
	170	* Only used by the i386 NUMA architecures, but relatively
	171	* generic code.
	172	*/
	173	unsigned long __init node_memmap_size_bytes(int nid, unsigned long start_pfn,
	174	unsigned long end_pfn)
	175	{
	176	unsigned long pfn;
	177	unsigned long nr_pages = 0;
	178
	179	for (pfn = start_pfn; pfn < end_pfn; pfn += PAGES_PER_SECTION) {
	180	if (nid != early_pfn_to_nid(pfn))
	181	continue;
	182
540557b9	183	if (pfn_present(pfn))
d41dee36 AW	184	nr_pages += PAGES_PER_SECTION;
	185	}
	186
	187	return nr_pages * sizeof(struct page);
	188	}
	189
29751f69 AW	190	/*
	191	* Subtle, we encode the real pfn into the mem_map such that
	192	* the identity pfn - section_mem_map will return the actual
	193	* physical page frame number.
	194	*/
	195	static unsigned long sparse_encode_mem_map(struct page *mem_map, unsigned long pnum)
	196	{
	197	return (unsigned long)(mem_map - (section_nr_to_pfn(pnum)));
	198	}
	199
	200	/*
	201	* We need this if we ever free the mem_maps. While not implemented yet,
	202	* this function is included for parity with its sibling.
	203	*/
	204	static __attribute((unused))
	205	struct page *sparse_decode_mem_map(unsigned long coded_mem_map, unsigned long pnum)
	206	{
	207	return ((struct page *)coded_mem_map) + section_nr_to_pfn(pnum);
	208	}
	209
a3142c8e	210	static int __meminit sparse_init_one_section(struct mem_section *ms,
5c0e3066 MG	211	unsigned long pnum, struct page *mem_map,
5c0e3066 MG	212	unsigned long *pageblock_bitmap)
29751f69	213	{
540557b9	214	if (!present_section(ms))
29751f69 AW	215	return -EINVAL;
29751f69 AW	216
30c253e6	217	ms->section_mem_map &= ~SECTION_MAP_MASK;
540557b9 AW	218	ms->section_mem_map \|= sparse_encode_mem_map(mem_map, pnum) \|
540557b9 AW	219	SECTION_HAS_MEM_MAP;
5c0e3066	220	ms->pageblock_flags = pageblock_bitmap;
29751f69 AW	221
	222	return 1;
	223	}
	224
5c0e3066 MG	225	static unsigned long usemap_size(void)
	226	{
	227	unsigned long size_bytes;
	228	size_bytes = roundup(SECTION_BLOCKFLAGS_BITS, 8) / 8;
	229	size_bytes = roundup(size_bytes, sizeof(unsigned long));
	230	return size_bytes;
	231	}
	232
	233	#ifdef CONFIG_MEMORY_HOTPLUG
	234	static unsigned long *__kmalloc_section_usemap(void)
	235	{
	236	return kmalloc(usemap_size(), GFP_KERNEL);
	237	}
	238	#endif /* CONFIG_MEMORY_HOTPLUG */
	239
	240	static unsigned long *sparse_early_usemap_alloc(unsigned long pnum)
	241	{
	242	unsigned long *usemap;
	243	struct mem_section *ms = __nr_to_section(pnum);
	244	int nid = sparse_early_nid(ms);
	245
	246	usemap = alloc_bootmem_node(NODE_DATA(nid), usemap_size());
	247	if (usemap)
	248	return usemap;
	249
	250	/* Stupid: suppress gcc warning for SPARSEMEM && !NUMA */
	251	nid = 0;
	252
	253	printk(KERN_WARNING "%s: allocation failed\n", __FUNCTION__);
	254	return NULL;
	255	}
	256
8f6aac41	257	#ifndef CONFIG_SPARSEMEM_VMEMMAP
98f3cfc1	258	struct page __init *sparse_mem_map_populate(unsigned long pnum, int nid)
29751f69 AW	259	{
29751f69 AW	260	struct page *map;
29751f69 AW	261
	262	map = alloc_remap(nid, sizeof(struct page) * PAGES_PER_SECTION);
	263	if (map)
	264	return map;
	265
	266	map = alloc_bootmem_node(NODE_DATA(nid),
	267	sizeof(struct page) * PAGES_PER_SECTION);
8f6aac41 CL	268	return map;
	269	}
	270	#endif /* !CONFIG_SPARSEMEM_VMEMMAP */
	271
	272	struct page __init *sparse_early_mem_map_alloc(unsigned long pnum)
	273	{
	274	struct page *map;
	275	struct mem_section *ms = __nr_to_section(pnum);
	276	int nid = sparse_early_nid(ms);
	277
98f3cfc1	278	map = sparse_mem_map_populate(pnum, nid);
29751f69 AW	279	if (map)
	280	return map;
	281
8f6aac41 CL	282	printk(KERN_ERR "%s: sparsemem memory map backing failed "
8f6aac41 CL	283	"some memory will not be available.\n", __FUNCTION__);
802f192e	284	ms->section_mem_map = 0;
29751f69 AW	285	return NULL;
	286	}
	287
193faea9 SR	288	/*
	289	* Allocate the accumulated non-linear sections, allocate a mem_map
	290	* for each and record the physical to section mapping.
	291	*/
	292	void __init sparse_init(void)
	293	{
	294	unsigned long pnum;
	295	struct page *map;
5c0e3066	296	unsigned long *usemap;
193faea9 SR	297
193faea9 SR	298	for (pnum = 0; pnum < NR_MEM_SECTIONS; pnum++) {
540557b9	299	if (!present_section_nr(pnum))
193faea9 SR	300	continue;
	301
	302	map = sparse_early_mem_map_alloc(pnum);
	303	if (!map)
	304	continue;
5c0e3066 MG	305
	306	usemap = sparse_early_usemap_alloc(pnum);
	307	if (!usemap)
	308	continue;
	309
	310	sparse_init_one_section(__nr_to_section(pnum), pnum, map,
	311	usemap);
193faea9 SR	312	}
	313	}
	314
	315	#ifdef CONFIG_MEMORY_HOTPLUG
98f3cfc1 YG	316	#ifdef CONFIG_SPARSEMEM_VMEMMAP
	317	static inline struct page *kmalloc_section_memmap(unsigned long pnum, int nid,
	318	unsigned long nr_pages)
	319	{
	320	/* This will make the necessary allocations eventually. */
	321	return sparse_mem_map_populate(pnum, nid);
	322	}
	323	static void __kfree_section_memmap(struct page *memmap, unsigned long nr_pages)
	324	{
	325	return; /* XXX: Not implemented yet */
	326	}
	327	#else
0b0acbec DH	328	static struct page *__kmalloc_section_memmap(unsigned long nr_pages)
	329	{
	330	struct page page, ret;
	331	unsigned long memmap_size = sizeof(struct page) * nr_pages;
	332
f2d0aa5b	333	page = alloc_pages(GFP_KERNEL\|__GFP_NOWARN, get_order(memmap_size));
0b0acbec DH	334	if (page)
	335	goto got_map_page;
	336
	337	ret = vmalloc(memmap_size);
	338	if (ret)
	339	goto got_map_ptr;
	340
	341	return NULL;
	342	got_map_page:
	343	ret = (struct page *)pfn_to_kaddr(page_to_pfn(page));
	344	got_map_ptr:
	345	memset(ret, 0, memmap_size);
	346
	347	return ret;
	348	}
	349
98f3cfc1 YG	350	static inline struct page *kmalloc_section_memmap(unsigned long pnum, int nid,
	351	unsigned long nr_pages)
	352	{
	353	return __kmalloc_section_memmap(nr_pages);
	354	}
	355
0b0acbec DH	356	static int vaddr_in_vmalloc_area(void *addr)
	357	{
	358	if (addr >= (void *)VMALLOC_START &&
	359	addr < (void *)VMALLOC_END)
	360	return 1;
	361	return 0;
	362	}
	363
	364	static void __kfree_section_memmap(struct page *memmap, unsigned long nr_pages)
	365	{
	366	if (vaddr_in_vmalloc_area(memmap))
	367	vfree(memmap);
	368	else
	369	free_pages((unsigned long)memmap,
	370	get_order(sizeof(struct page) * nr_pages));
	371	}
98f3cfc1	372	#endif /* CONFIG_SPARSEMEM_VMEMMAP */
0b0acbec	373
29751f69 AW	374	/*
	375	* returns the number of sections whose mem_maps were properly
	376	* set. If this is <=0, then that means that the passed-in
	377	* map was not consumed and must be freed.
	378	*/
0b0acbec DH	379	int sparse_add_one_section(struct zone *zone, unsigned long start_pfn,
0b0acbec DH	380	int nr_pages)
29751f69	381	{
0b0acbec DH	382	unsigned long section_nr = pfn_to_section_nr(start_pfn);
	383	struct pglist_data *pgdat = zone->zone_pgdat;
	384	struct mem_section *ms;
	385	struct page *memmap;
5c0e3066	386	unsigned long *usemap;
0b0acbec DH	387	unsigned long flags;
0b0acbec DH	388	int ret;
29751f69	389
0b0acbec DH	390	/*
	391	* no locking for this, because it does its own
	392	* plus, it does a kmalloc
	393	*/
bbd06825 WC	394	ret = sparse_index_init(section_nr, pgdat->node_id);
	395	if (ret < 0 && ret != -EEXIST)
	396	return ret;
98f3cfc1	397	memmap = kmalloc_section_memmap(section_nr, pgdat->node_id, nr_pages);
bbd06825 WC	398	if (!memmap)
bbd06825 WC	399	return -ENOMEM;
5c0e3066	400	usemap = __kmalloc_section_usemap();
bbd06825 WC	401	if (!usemap) {
	402	__kfree_section_memmap(memmap, nr_pages);
	403	return -ENOMEM;
	404	}
0b0acbec DH	405
0b0acbec DH	406	pgdat_resize_lock(pgdat, &flags);
29751f69	407
0b0acbec DH	408	ms = __pfn_to_section(start_pfn);
	409	if (ms->section_mem_map & SECTION_MARKED_PRESENT) {
	410	ret = -EEXIST;
	411	goto out;
	412	}
5c0e3066	413
29751f69 AW	414	ms->section_mem_map \|= SECTION_MARKED_PRESENT;
29751f69 AW	415
5c0e3066	416	ret = sparse_init_one_section(ms, section_nr, memmap, usemap);
0b0acbec	417
0b0acbec DH	418	out:
0b0acbec DH	419	pgdat_resize_unlock(pgdat, &flags);
bbd06825 WC	420	if (ret <= 0) {
bbd06825 WC	421	kfree(usemap);
46a66eec	422	__kfree_section_memmap(memmap, nr_pages);
bbd06825	423	}
0b0acbec	424	return ret;
29751f69	425	}
a3142c8e	426	#endif