[GitHub/moto-9609/android_kernel_motorola_exynos9610.git] / mm / sparse-vmemmap.c

// SPDX-License-Identifier: GPL-2.0
/*
 * Virtual Memory Map support
 *
 * (C) 2007 sgi. Christoph Lameter.
 *
 * Virtual memory maps allow VM primitives pfn_to_page, page_to_pfn,
 * virt_to_page, page_address() to be implemented as a base offset
 * calculation without memory access.
 *
 * However, virtual mappings need a page table and TLBs. Many Linux
 * architectures already map their physical space using 1-1 mappings
 * via TLBs. For those arches the virtual memory map is essentially
 * for free if we use the same page size as the 1-1 mappings. In that
 * case the overhead consists of a few additional pages that are
 * allocated to create a view of memory for vmemmap.
 *
 * The architecture is expected to provide a vmemmap_populate() function
 * to instantiate the mapping.
 */
#include <linux/mm.h>
#include <linux/mmzone.h>
#include <linux/bootmem.h>
#include <linux/memremap.h>
#include <linux/highmem.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/vmalloc.h>
#include <linux/sched.h>
#include <asm/dma.h>
#include <asm/pgalloc.h>
#include <asm/pgtable.h>

/*
 * Allocate a block of memory to be used to back the virtual memory map
 * or to back the page tables that are used to create the mapping.
 * Uses the main allocators if they are available, else bootmem.
 */

static void * __ref __earlyonly_bootmem_alloc(int node,
				unsigned long size,
				unsigned long align,
				unsigned long goal)
{
	return memblock_virt_alloc_try_nid(size, align, goal,
					    BOOTMEM_ALLOC_ACCESSIBLE, node);
}

static void *vmemmap_buf;
static void *vmemmap_buf_end;

void * __meminit vmemmap_alloc_block(unsigned long size, int node)
{
	/* If the main allocator is up use that, fallback to bootmem. */
	if (slab_is_available()) {
		struct page *page;

		page = alloc_pages_node(node,
			GFP_KERNEL | __GFP_ZERO | __GFP_RETRY_MAYFAIL,
			get_order(size));
		if (page)
			return page_address(page);
		return NULL;
	} else
		return __earlyonly_bootmem_alloc(node, size, size,
				__pa(MAX_DMA_ADDRESS));
}

/* need to make sure size is all the same during early stage */
static void * __meminit alloc_block_buf(unsigned long size, int node)
{
	void *ptr;

	if (!vmemmap_buf)
		return vmemmap_alloc_block(size, node);

	/* take the from buf */
	ptr = (void *)ALIGN((unsigned long)vmemmap_buf, size);
	if (ptr + size > vmemmap_buf_end)
		return vmemmap_alloc_block(size, node);

	vmemmap_buf = ptr + size;

	return ptr;
}

static unsigned long __meminit vmem_altmap_next_pfn(struct vmem_altmap *altmap)
{
	return altmap->base_pfn + altmap->reserve + altmap->alloc
		+ altmap->align;
}

static unsigned long __meminit vmem_altmap_nr_free(struct vmem_altmap *altmap)
{
	unsigned long allocated = altmap->alloc + altmap->align;

	if (altmap->free > allocated)
		return altmap->free - allocated;
	return 0;
}

/**
 * vmem_altmap_alloc - allocate pages from the vmem_altmap reservation
 * @altmap - reserved page pool for the allocation
 * @nr_pfns - size (in pages) of the allocation
 *
 * Allocations are aligned to the size of the request
 */
static unsigned long __meminit vmem_altmap_alloc(struct vmem_altmap *altmap,
		unsigned long nr_pfns)
{
	unsigned long pfn = vmem_altmap_next_pfn(altmap);
	unsigned long nr_align;

	nr_align = 1UL << find_first_bit(&nr_pfns, BITS_PER_LONG);
	nr_align = ALIGN(pfn, nr_align) - pfn;

	if (nr_pfns + nr_align > vmem_altmap_nr_free(altmap))
		return ULONG_MAX;
	altmap->alloc += nr_pfns;
	altmap->align += nr_align;
	return pfn + nr_align;
}

static void * __meminit altmap_alloc_block_buf(unsigned long size,
		struct vmem_altmap *altmap)
{
	unsigned long pfn, nr_pfns;
	void *ptr;

	if (size & ~PAGE_MASK) {
		pr_warn_once("%s: allocations must be multiple of PAGE_SIZE (%ld)\n",
				__func__, size);
		return NULL;
	}

	nr_pfns = size >> PAGE_SHIFT;
	pfn = vmem_altmap_alloc(altmap, nr_pfns);
	if (pfn < ULONG_MAX)
		ptr = __va(__pfn_to_phys(pfn));
	else
		ptr = NULL;
	pr_debug("%s: pfn: %#lx alloc: %ld align: %ld nr: %#lx\n",
			__func__, pfn, altmap->alloc, altmap->align, nr_pfns);

	return ptr;
}

/* need to make sure size is all the same during early stage */
void * __meminit __vmemmap_alloc_block_buf(unsigned long size, int node,
		struct vmem_altmap *altmap)
{
	if (altmap)
		return altmap_alloc_block_buf(size, altmap);
	return alloc_block_buf(size, node);
}

void __meminit vmemmap_verify(pte_t *pte, int node,
				unsigned long start, unsigned long end)
{
	unsigned long pfn = pte_pfn(*pte);
	int actual_node = early_pfn_to_nid(pfn);

	if (node_distance(actual_node, node) > LOCAL_DISTANCE)
		pr_warn("[%lx-%lx] potential offnode page_structs\n",
			start, end - 1);
}

pte_t * __meminit vmemmap_pte_populate(pmd_t *pmd, unsigned long addr, int node)
{
	pte_t *pte = pte_offset_kernel(pmd, addr);
	if (pte_none(*pte)) {
		pte_t entry;
		void *p = alloc_block_buf(PAGE_SIZE, node);
		if (!p)
			return NULL;
		entry = pfn_pte(__pa(p) >> PAGE_SHIFT, PAGE_KERNEL);
		set_pte_at(&init_mm, addr, pte, entry);
	}
	return pte;
}

pmd_t * __meminit vmemmap_pmd_populate(pud_t *pud, unsigned long addr, int node)
{
	pmd_t *pmd = pmd_offset(pud, addr);
	if (pmd_none(*pmd)) {
		void *p = vmemmap_alloc_block(PAGE_SIZE, node);
		if (!p)
			return NULL;
		pmd_populate_kernel(&init_mm, pmd, p);
	}
	return pmd;
}

pud_t * __meminit vmemmap_pud_populate(p4d_t *p4d, unsigned long addr, int node)
{
	pud_t *pud = pud_offset(p4d, addr);
	if (pud_none(*pud)) {
		void *p = vmemmap_alloc_block(PAGE_SIZE, node);
		if (!p)
			return NULL;
		pud_populate(&init_mm, pud, p);
	}
	return pud;
}

p4d_t * __meminit vmemmap_p4d_populate(pgd_t *pgd, unsigned long addr, int node)
{
	p4d_t *p4d = p4d_offset(pgd, addr);
	if (p4d_none(*p4d)) {
		void *p = vmemmap_alloc_block(PAGE_SIZE, node);
		if (!p)
			return NULL;
		p4d_populate(&init_mm, p4d, p);
	}
	return p4d;
}

pgd_t * __meminit vmemmap_pgd_populate(unsigned long addr, int node)
{
	pgd_t *pgd = pgd_offset_k(addr);
	if (pgd_none(*pgd)) {
		void *p = vmemmap_alloc_block(PAGE_SIZE, node);
		if (!p)
			return NULL;
		pgd_populate(&init_mm, pgd, p);
	}
	return pgd;
}

int __meminit vmemmap_populate_basepages(unsigned long start,
					 unsigned long end, int node)
{
	unsigned long addr = start;
	pgd_t *pgd;
	p4d_t *p4d;
	pud_t *pud;
	pmd_t *pmd;
	pte_t *pte;

	for (; addr < end; addr += PAGE_SIZE) {
		pgd = vmemmap_pgd_populate(addr, node);
		if (!pgd)
			return -ENOMEM;
		p4d = vmemmap_p4d_populate(pgd, addr, node);
		if (!p4d)
			return -ENOMEM;
		pud = vmemmap_pud_populate(p4d, addr, node);
		if (!pud)
			return -ENOMEM;
		pmd = vmemmap_pmd_populate(pud, addr, node);
		if (!pmd)
			return -ENOMEM;
		pte = vmemmap_pte_populate(pmd, addr, node);
		if (!pte)
			return -ENOMEM;
		vmemmap_verify(pte, node, addr, addr + PAGE_SIZE);
	}

	return 0;
}

struct page * __meminit sparse_mem_map_populate(unsigned long pnum, int nid)
{
	unsigned long start;
	unsigned long end;
	struct page *map;

	map = pfn_to_page(pnum * PAGES_PER_SECTION);
	start = (unsigned long)map;
	end = (unsigned long)(map + PAGES_PER_SECTION);

	if (vmemmap_populate(start, end, nid))
		return NULL;

	return map;
}

void __init sparse_mem_maps_populate_node(struct page **map_map,
					  unsigned long pnum_begin,
					  unsigned long pnum_end,
					  unsigned long map_count, int nodeid)
{
	unsigned long pnum;
	unsigned long size = sizeof(struct page) * PAGES_PER_SECTION;
	void *vmemmap_buf_start;

	size = ALIGN(size, PMD_SIZE);
	vmemmap_buf_start = __earlyonly_bootmem_alloc(nodeid, size * map_count,
			 PMD_SIZE, __pa(MAX_DMA_ADDRESS));

	if (vmemmap_buf_start) {
		vmemmap_buf = vmemmap_buf_start;
		vmemmap_buf_end = vmemmap_buf_start + size * map_count;
	}

	for (pnum = pnum_begin; pnum < pnum_end; pnum++) {
		struct mem_section *ms;

		if (!present_section_nr(pnum))
			continue;

		map_map[pnum] = sparse_mem_map_populate(pnum, nodeid);
		if (map_map[pnum])
			continue;
		ms = __nr_to_section(pnum);
		pr_err("%s: sparsemem memory map backing failed some memory will not be available\n",
		       __func__);
		ms->section_mem_map = 0;
	}

	if (vmemmap_buf_start) {
		/* need to free left buf */
		memblock_free_early(__pa(vmemmap_buf),
				    vmemmap_buf_end - vmemmap_buf);
		vmemmap_buf = NULL;
		vmemmap_buf_end = NULL;
	}
}
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
8f6aac41 CL	2	/*
	3	* Virtual Memory Map support
	4	*
cde53535	5	* (C) 2007 sgi. Christoph Lameter.
8f6aac41 CL	6	*
	7	* Virtual memory maps allow VM primitives pfn_to_page, page_to_pfn,
	8	* virt_to_page, page_address() to be implemented as a base offset
	9	* calculation without memory access.
	10	*
	11	* However, virtual mappings need a page table and TLBs. Many Linux
	12	* architectures already map their physical space using 1-1 mappings
b595076a	13	* via TLBs. For those arches the virtual memory map is essentially
8f6aac41 CL	14	* for free if we use the same page size as the 1-1 mappings. In that
	15	* case the overhead consists of a few additional pages that are
	16	* allocated to create a view of memory for vmemmap.
	17	*
29c71111 AW	18	* The architecture is expected to provide a vmemmap_populate() function
29c71111 AW	19	* to instantiate the mapping.
8f6aac41 CL	20	*/
	21	#include <linux/mm.h>
	22	#include <linux/mmzone.h>
	23	#include <linux/bootmem.h>
4b94ffdc	24	#include <linux/memremap.h>
8f6aac41	25	#include <linux/highmem.h>
5a0e3ad6	26	#include <linux/slab.h>
8f6aac41 CL	27	#include <linux/spinlock.h>
8f6aac41 CL	28	#include <linux/vmalloc.h>
8bca44bb	29	#include <linux/sched.h>
8f6aac41 CL	30	#include <asm/dma.h>
	31	#include <asm/pgalloc.h>
	32	#include <asm/pgtable.h>
	33
	34	/*
	35	* Allocate a block of memory to be used to back the virtual memory map
	36	* or to back the page tables that are used to create the mapping.
	37	* Uses the main allocators if they are available, else bootmem.
	38	*/
e0dc3a53	39
bd721ea7	40	static void * __ref __earlyonly_bootmem_alloc(int node,
e0dc3a53 KH	41	unsigned long size,
	42	unsigned long align,
	43	unsigned long goal)
	44	{
bb016b84 SS	45	return memblock_virt_alloc_try_nid(size, align, goal,
bb016b84 SS	46	BOOTMEM_ALLOC_ACCESSIBLE, node);
e0dc3a53 KH	47	}
e0dc3a53 KH	48
9bdac914 YL	49	static void *vmemmap_buf;
9bdac914 YL	50	static void *vmemmap_buf_end;
e0dc3a53	51
8f6aac41 CL	52	void * __meminit vmemmap_alloc_block(unsigned long size, int node)
	53	{
	54	/* If the main allocator is up use that, fallback to bootmem. */
	55	if (slab_is_available()) {
f52407ce SL	56	struct page *page;
f52407ce SL	57
b95046b0 MH	58	page = alloc_pages_node(node,
	59	GFP_KERNEL \| __GFP_ZERO \| __GFP_RETRY_MAYFAIL,
	60	get_order(size));
8f6aac41 CL	61	if (page)
	62	return page_address(page);
	63	return NULL;
	64	} else
e0dc3a53	65	return __earlyonly_bootmem_alloc(node, size, size,
8f6aac41 CL	66	__pa(MAX_DMA_ADDRESS));
	67	}
	68
9bdac914	69	/* need to make sure size is all the same during early stage */
4b94ffdc	70	static void * __meminit alloc_block_buf(unsigned long size, int node)
9bdac914 YL	71	{
	72	void *ptr;
	73
	74	if (!vmemmap_buf)
	75	return vmemmap_alloc_block(size, node);
	76
	77	/* take the from buf */
	78	ptr = (void *)ALIGN((unsigned long)vmemmap_buf, size);
	79	if (ptr + size > vmemmap_buf_end)
	80	return vmemmap_alloc_block(size, node);
	81
	82	vmemmap_buf = ptr + size;
	83
	84	return ptr;
	85	}
	86
4b94ffdc DW	87	static unsigned long __meminit vmem_altmap_next_pfn(struct vmem_altmap *altmap)
	88	{
	89	return altmap->base_pfn + altmap->reserve + altmap->alloc
	90	+ altmap->align;
	91	}
	92
	93	static unsigned long __meminit vmem_altmap_nr_free(struct vmem_altmap *altmap)
	94	{
	95	unsigned long allocated = altmap->alloc + altmap->align;
	96
	97	if (altmap->free > allocated)
	98	return altmap->free - allocated;
	99	return 0;
	100	}
	101
	102	/**
	103	* vmem_altmap_alloc - allocate pages from the vmem_altmap reservation
	104	* @altmap - reserved page pool for the allocation
	105	* @nr_pfns - size (in pages) of the allocation
	106	*
	107	* Allocations are aligned to the size of the request
	108	*/
	109	static unsigned long __meminit vmem_altmap_alloc(struct vmem_altmap *altmap,
	110	unsigned long nr_pfns)
	111	{
	112	unsigned long pfn = vmem_altmap_next_pfn(altmap);
	113	unsigned long nr_align;
	114
	115	nr_align = 1UL << find_first_bit(&nr_pfns, BITS_PER_LONG);
	116	nr_align = ALIGN(pfn, nr_align) - pfn;
	117
	118	if (nr_pfns + nr_align > vmem_altmap_nr_free(altmap))
	119	return ULONG_MAX;
	120	altmap->alloc += nr_pfns;
	121	altmap->align += nr_align;
	122	return pfn + nr_align;
	123	}
	124
	125	static void * __meminit altmap_alloc_block_buf(unsigned long size,
	126	struct vmem_altmap *altmap)
	127	{
	128	unsigned long pfn, nr_pfns;
	129	void *ptr;
	130
	131	if (size & ~PAGE_MASK) {
	132	pr_warn_once("%s: allocations must be multiple of PAGE_SIZE (%ld)\n",
	133	__func__, size);
	134	return NULL;
	135	}
	136
	137	nr_pfns = size >> PAGE_SHIFT;
	138	pfn = vmem_altmap_alloc(altmap, nr_pfns);
	139	if (pfn < ULONG_MAX)
	140	ptr = __va(__pfn_to_phys(pfn));
	141	else
	142	ptr = NULL;
	143	pr_debug("%s: pfn: %#lx alloc: %ld align: %ld nr: %#lx\n",
	144	__func__, pfn, altmap->alloc, altmap->align, nr_pfns);
	145
	146	return ptr;
	147	}
	148
	149	/* need to make sure size is all the same during early stage */
	150	void * __meminit __vmemmap_alloc_block_buf(unsigned long size, int node,
151	struct vmem_altmap *altmap)
152	{
153	if (altmap)
154	return altmap_alloc_block_buf(size, altmap);
155	return alloc_block_buf(size, node);
156	}
157
8f6aac41 CL	158	void __meminit vmemmap_verify(pte_t *pte, int node,
	159	unsigned long start, unsigned long end)
	160	{
	161	unsigned long pfn = pte_pfn(*pte);
	162	int actual_node = early_pfn_to_nid(pfn);
	163
b41ad14c	164	if (node_distance(actual_node, node) > LOCAL_DISTANCE)
1170532b JP	165	pr_warn("[%lx-%lx] potential offnode page_structs\n",
1170532b JP	166	start, end - 1);
8f6aac41 CL	167	}
8f6aac41 CL	168
29c71111	169	pte_t * __meminit vmemmap_pte_populate(pmd_t *pmd, unsigned long addr, int node)
8f6aac41	170	{
29c71111 AW	171	pte_t *pte = pte_offset_kernel(pmd, addr);
	172	if (pte_none(*pte)) {
	173	pte_t entry;
4b94ffdc	174	void *p = alloc_block_buf(PAGE_SIZE, node);
29c71111	175	if (!p)
9dce07f1	176	return NULL;
29c71111 AW	177	entry = pfn_pte(__pa(p) >> PAGE_SHIFT, PAGE_KERNEL);
	178	set_pte_at(&init_mm, addr, pte, entry);
	179	}
	180	return pte;
8f6aac41 CL	181	}
8f6aac41 CL	182
29c71111	183	pmd_t * __meminit vmemmap_pmd_populate(pud_t *pud, unsigned long addr, int node)
8f6aac41	184	{
29c71111 AW	185	pmd_t *pmd = pmd_offset(pud, addr);
	186	if (pmd_none(*pmd)) {
	187	void *p = vmemmap_alloc_block(PAGE_SIZE, node);
	188	if (!p)
9dce07f1	189	return NULL;
29c71111	190	pmd_populate_kernel(&init_mm, pmd, p);
8f6aac41	191	}
29c71111	192	return pmd;
8f6aac41	193	}
8f6aac41	194
c2febafc	195	pud_t * __meminit vmemmap_pud_populate(p4d_t *p4d, unsigned long addr, int node)
8f6aac41	196	{
c2febafc	197	pud_t *pud = pud_offset(p4d, addr);
29c71111 AW	198	if (pud_none(*pud)) {
	199	void *p = vmemmap_alloc_block(PAGE_SIZE, node);
	200	if (!p)
9dce07f1	201	return NULL;
29c71111 AW	202	pud_populate(&init_mm, pud, p);
	203	}
	204	return pud;
	205	}
8f6aac41	206
c2febafc KS	207	p4d_t * __meminit vmemmap_p4d_populate(pgd_t *pgd, unsigned long addr, int node)
	208	{
	209	p4d_t *p4d = p4d_offset(pgd, addr);
	210	if (p4d_none(*p4d)) {
	211	void *p = vmemmap_alloc_block(PAGE_SIZE, node);
	212	if (!p)
	213	return NULL;
	214	p4d_populate(&init_mm, p4d, p);
	215	}
	216	return p4d;
	217	}
	218
29c71111 AW	219	pgd_t * __meminit vmemmap_pgd_populate(unsigned long addr, int node)
	220	{
	221	pgd_t *pgd = pgd_offset_k(addr);
	222	if (pgd_none(*pgd)) {
	223	void *p = vmemmap_alloc_block(PAGE_SIZE, node);
	224	if (!p)
9dce07f1	225	return NULL;
29c71111	226	pgd_populate(&init_mm, pgd, p);
8f6aac41	227	}
29c71111	228	return pgd;
8f6aac41 CL	229	}
8f6aac41 CL	230
0aad818b JW	231	int __meminit vmemmap_populate_basepages(unsigned long start,
0aad818b JW	232	unsigned long end, int node)
8f6aac41	233	{
0aad818b	234	unsigned long addr = start;
29c71111	235	pgd_t *pgd;
c2febafc	236	p4d_t *p4d;
29c71111 AW	237	pud_t *pud;
	238	pmd_t *pmd;
	239	pte_t *pte;
8f6aac41	240
29c71111 AW	241	for (; addr < end; addr += PAGE_SIZE) {
	242	pgd = vmemmap_pgd_populate(addr, node);
	243	if (!pgd)
	244	return -ENOMEM;
c2febafc KS	245	p4d = vmemmap_p4d_populate(pgd, addr, node);
	246	if (!p4d)
	247	return -ENOMEM;
	248	pud = vmemmap_pud_populate(p4d, addr, node);
29c71111 AW	249	if (!pud)
	250	return -ENOMEM;
	251	pmd = vmemmap_pmd_populate(pud, addr, node);
	252	if (!pmd)
	253	return -ENOMEM;
	254	pte = vmemmap_pte_populate(pmd, addr, node);
	255	if (!pte)
	256	return -ENOMEM;
	257	vmemmap_verify(pte, node, addr, addr + PAGE_SIZE);
8f6aac41	258	}
29c71111 AW	259
29c71111 AW	260	return 0;
8f6aac41	261	}
8f6aac41	262
98f3cfc1	263	struct page * __meminit sparse_mem_map_populate(unsigned long pnum, int nid)
8f6aac41	264	{
0aad818b JW	265	unsigned long start;
	266	unsigned long end;
	267	struct page *map;
	268
	269	map = pfn_to_page(pnum * PAGES_PER_SECTION);
	270	start = (unsigned long)map;
	271	end = (unsigned long)(map + PAGES_PER_SECTION);
	272
	273	if (vmemmap_populate(start, end, nid))
8f6aac41 CL	274	return NULL;
	275
	276	return map;
	277	}
9bdac914 YL	278
	279	void __init sparse_mem_maps_populate_node(struct page **map_map,
	280	unsigned long pnum_begin,
	281	unsigned long pnum_end,
	282	unsigned long map_count, int nodeid)
	283	{
	284	unsigned long pnum;
	285	unsigned long size = sizeof(struct page) * PAGES_PER_SECTION;
	286	void *vmemmap_buf_start;
	287
	288	size = ALIGN(size, PMD_SIZE);
	289	vmemmap_buf_start = __earlyonly_bootmem_alloc(nodeid, size * map_count,
	290	PMD_SIZE, __pa(MAX_DMA_ADDRESS));
	291
	292	if (vmemmap_buf_start) {
	293	vmemmap_buf = vmemmap_buf_start;
	294	vmemmap_buf_end = vmemmap_buf_start + size * map_count;
	295	}
	296
	297	for (pnum = pnum_begin; pnum < pnum_end; pnum++) {
	298	struct mem_section *ms;
	299
	300	if (!present_section_nr(pnum))
	301	continue;
	302
	303	map_map[pnum] = sparse_mem_map_populate(pnum, nodeid);
	304	if (map_map[pnum])
	305	continue;
	306	ms = __nr_to_section(pnum);
1170532b	307	pr_err("%s: sparsemem memory map backing failed some memory will not be available\n",
756a025f	308	__func__);
9bdac914 YL	309	ms->section_mem_map = 0;
	310	}
	311
	312	if (vmemmap_buf_start) {
	313	/* need to free left buf */
bb016b84 SS	314	memblock_free_early(__pa(vmemmap_buf),
bb016b84 SS	315	vmemmap_buf_end - vmemmap_buf);
9bdac914 YL	316	vmemmap_buf = NULL;
	317	vmemmap_buf_end = NULL;
	318	}
	319	}