hugetlb.h 14 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3
#ifndef _LINUX_HUGETLB_H
#define _LINUX_HUGETLB_H

Linus Torvalds's avatar
Linus Torvalds committed
4
#include <linux/mm_types.h>
5
#include <linux/mmdebug.h>
Alexey Dobriyan's avatar
Alexey Dobriyan committed
6
#include <linux/fs.h>
7
#include <linux/hugetlb_inline.h>
8
#include <linux/cgroup.h>
9 10
#include <linux/list.h>
#include <linux/kref.h>
Alexey Dobriyan's avatar
Alexey Dobriyan committed
11

12 13
struct ctl_table;
struct user_struct;
14
struct mmu_gather;
15

Linus Torvalds's avatar
Linus Torvalds committed
16 17 18
#ifdef CONFIG_HUGETLB_PAGE

#include <linux/mempolicy.h>
19
#include <linux/shm.h>
David Gibson's avatar
David Gibson committed
20
#include <asm/tlbflush.h>
Linus Torvalds's avatar
Linus Torvalds committed
21

22 23 24
struct hugepage_subpool {
	spinlock_t lock;
	long count;
25 26 27 28 29 30 31
	long max_hpages;	/* Maximum huge pages or -1 if no maximum. */
	long used_hpages;	/* Used count against maximum, includes */
				/* both alloced and reserved pages. */
	struct hstate *hstate;
	long min_hpages;	/* Minimum huge pages or -1 if no minimum. */
	long rsv_hpages;	/* Pages reserved against global pool to */
				/* sasitfy minimum size. */
32 33
};

34 35
struct resv_map {
	struct kref refs;
36
	spinlock_t lock;
37
	struct list_head regions;
38 39 40
	long adds_in_progress;
	struct list_head region_cache;
	long region_cache_count;
41 42 43 44
};
extern struct resv_map *resv_map_alloc(void);
void resv_map_release(struct kref *ref);

45 46 47 48 49
extern spinlock_t hugetlb_lock;
extern int hugetlb_max_hstate __read_mostly;
#define for_each_hstate(h) \
	for ((h) = hstates; (h) < &hstates[hugetlb_max_hstate]; (h)++)

50 51
struct hugepage_subpool *hugepage_new_subpool(struct hstate *h, long max_hpages,
						long min_hpages);
52 53
void hugepage_put_subpool(struct hugepage_subpool *spool);

54
void reset_vma_resv_huge_pages(struct vm_area_struct *vma);
55 56 57
int hugetlb_sysctl_handler(struct ctl_table *, int, void __user *, size_t *, loff_t *);
int hugetlb_overcommit_handler(struct ctl_table *, int, void __user *, size_t *, loff_t *);
int hugetlb_treat_movable_handler(struct ctl_table *, int, void __user *, size_t *, loff_t *);
58 59 60 61 62 63

#ifdef CONFIG_NUMA
int hugetlb_mempolicy_sysctl_handler(struct ctl_table *, int,
					void __user *, size_t *, loff_t *);
#endif

Linus Torvalds's avatar
Linus Torvalds committed
64
int copy_hugetlb_page_range(struct mm_struct *, struct mm_struct *, struct vm_area_struct *);
65 66 67
long follow_hugetlb_page(struct mm_struct *, struct vm_area_struct *,
			 struct page **, struct vm_area_struct **,
			 unsigned long *, unsigned long *, long, unsigned int);
68
void unmap_hugepage_range(struct vm_area_struct *,
69
			  unsigned long, unsigned long, struct page *);
70 71 72 73
void __unmap_hugepage_range_final(struct mmu_gather *tlb,
			  struct vm_area_struct *vma,
			  unsigned long start, unsigned long end,
			  struct page *ref_page);
74 75 76
void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
				unsigned long start, unsigned long end,
				struct page *ref_page);
77
void hugetlb_report_meminfo(struct seq_file *);
Linus Torvalds's avatar
Linus Torvalds committed
78
int hugetlb_report_node_meminfo(int, char *);
79
void hugetlb_show_meminfo(void);
Linus Torvalds's avatar
Linus Torvalds committed
80
unsigned long hugetlb_total_pages(void);
81
int hugetlb_fault(struct mm_struct *mm, struct vm_area_struct *vma,
82
			unsigned long address, unsigned int flags);
83
int hugetlb_reserve_pages(struct inode *inode, long from, long to,
84
						struct vm_area_struct *vma,
85
						vm_flags_t vm_flags);
86
void hugetlb_unreserve_pages(struct inode *inode, long offset, long freed);
87
int dequeue_hwpoisoned_huge_page(struct page *page);
88 89
bool isolate_huge_page(struct page *page, struct list_head *list);
void putback_active_hugepage(struct page *page);
90
void free_huge_page(struct page *page);
Linus Torvalds's avatar
Linus Torvalds committed
91

92 93 94 95
#ifdef CONFIG_ARCH_WANT_HUGE_PMD_SHARE
pte_t *huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud);
#endif

96
extern int hugepages_treat_as_movable;
Linus Torvalds's avatar
Linus Torvalds committed
97
extern int sysctl_hugetlb_shm_group;
98
extern struct list_head huge_boot_pages;
Linus Torvalds's avatar
Linus Torvalds committed
99

David Gibson's avatar
David Gibson committed
100 101
/* arch callbacks */

102 103
pte_t *huge_pte_alloc(struct mm_struct *mm,
			unsigned long addr, unsigned long sz);
David Gibson's avatar
David Gibson committed
104
pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr);
105
int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, pte_t *ptep);
David Gibson's avatar
David Gibson committed
106 107 108
struct page *follow_huge_addr(struct mm_struct *mm, unsigned long address,
			      int write);
struct page *follow_huge_pmd(struct mm_struct *mm, unsigned long address,
109
				pmd_t *pmd, int flags);
Andi Kleen's avatar
Andi Kleen committed
110
struct page *follow_huge_pud(struct mm_struct *mm, unsigned long address,
111
				pud_t *pud, int flags);
David Gibson's avatar
David Gibson committed
112
int pmd_huge(pmd_t pmd);
Andi Kleen's avatar
Andi Kleen committed
113
int pud_huge(pud_t pmd);
114
unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
115
		unsigned long address, unsigned long end, pgprot_t newprot);
David Gibson's avatar
David Gibson committed
116

Linus Torvalds's avatar
Linus Torvalds committed
117 118
#else /* !CONFIG_HUGETLB_PAGE */

119 120 121 122
static inline void reset_vma_resv_huge_pages(struct vm_area_struct *vma)
{
}

Linus Torvalds's avatar
Linus Torvalds committed
123 124 125 126 127
static inline unsigned long hugetlb_total_pages(void)
{
	return 0;
}

128
#define follow_hugetlb_page(m,v,p,vs,a,b,i,w)	({ BUG(); 0; })
Linus Torvalds's avatar
Linus Torvalds committed
129 130
#define follow_huge_addr(mm, addr, write)	ERR_PTR(-EINVAL)
#define copy_hugetlb_page_range(src, dst, vma)	({ BUG(); 0; })
131 132 133
static inline void hugetlb_report_meminfo(struct seq_file *m)
{
}
Linus Torvalds's avatar
Linus Torvalds committed
134
#define hugetlb_report_node_meminfo(n, buf)	0
135 136 137
static inline void hugetlb_show_meminfo(void)
{
}
138 139
#define follow_huge_pmd(mm, addr, pmd, flags)	NULL
#define follow_huge_pud(mm, addr, pud, flags)	NULL
140
#define prepare_hugepage_range(file, addr, len)	(-EINVAL)
Linus Torvalds's avatar
Linus Torvalds committed
141
#define pmd_huge(x)	0
Andi Kleen's avatar
Andi Kleen committed
142
#define pud_huge(x)	0
Linus Torvalds's avatar
Linus Torvalds committed
143
#define is_hugepage_only_range(mm, addr, len)	0
144
#define hugetlb_free_pgd_range(tlb, addr, end, floor, ceiling) ({BUG(); 0; })
145
#define hugetlb_fault(mm, vma, addr, flags)	({ BUG(); 0; })
146
#define huge_pte_offset(mm, address)	0
147 148 149 150 151
static inline int dequeue_hwpoisoned_huge_page(struct page *page)
{
	return 0;
}

152 153 154 155
static inline bool isolate_huge_page(struct page *page, struct list_head *list)
{
	return false;
}
156
#define putback_active_hugepage(p)	do {} while (0)
Linus Torvalds's avatar
Linus Torvalds committed
157

158 159 160 161 162
static inline unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
		unsigned long address, unsigned long end, pgprot_t newprot)
{
	return 0;
}
163

164 165 166 167 168 169 170
static inline void __unmap_hugepage_range_final(struct mmu_gather *tlb,
			struct vm_area_struct *vma, unsigned long start,
			unsigned long end, struct page *ref_page)
{
	BUG();
}

171 172 173 174 175 176 177
static inline void __unmap_hugepage_range(struct mmu_gather *tlb,
			struct vm_area_struct *vma, unsigned long start,
			unsigned long end, struct page *ref_page)
{
	BUG();
}

Linus Torvalds's avatar
Linus Torvalds committed
178
#endif /* !CONFIG_HUGETLB_PAGE */
179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224
/*
 * hugepages at page global directory. If arch support
 * hugepages at pgd level, they need to define this.
 */
#ifndef pgd_huge
#define pgd_huge(x)	0
#endif

#ifndef pgd_write
static inline int pgd_write(pgd_t pgd)
{
	BUG();
	return 0;
}
#endif

#ifndef pud_write
static inline int pud_write(pud_t pud)
{
	BUG();
	return 0;
}
#endif

#ifndef is_hugepd
/*
 * Some architectures requires a hugepage directory format that is
 * required to support multiple hugepage sizes. For example
 * a4fe3ce76 "powerpc/mm: Allow more flexible layouts for hugepage pagetables"
 * introduced the same on powerpc. This allows for a more flexible hugepage
 * pagetable layout.
 */
typedef struct { unsigned long pd; } hugepd_t;
#define is_hugepd(hugepd) (0)
#define __hugepd(x) ((hugepd_t) { (x) })
static inline int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
			      unsigned pdshift, unsigned long end,
			      int write, struct page **pages, int *nr)
{
	return 0;
}
#else
extern int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
		       unsigned pdshift, unsigned long end,
		       int write, struct page **pages, int *nr);
#endif
Linus Torvalds's avatar
Linus Torvalds committed
225

226 227
#define HUGETLB_ANON_FILE "anon_hugepage"

228 229 230 231 232 233
enum {
	/*
	 * The file will be used as an shm file so shmfs accounting rules
	 * apply
	 */
	HUGETLB_SHMFS_INODE     = 1,
234 235 236 237 238
	/*
	 * The file is being created on the internal vfs mount and shmfs
	 * accounting rules do not apply
	 */
	HUGETLB_ANONHUGE_INODE  = 2,
239 240
};

Linus Torvalds's avatar
Linus Torvalds committed
241 242 243 244 245
#ifdef CONFIG_HUGETLBFS
struct hugetlbfs_sb_info {
	long	max_inodes;   /* inodes allowed */
	long	free_inodes;  /* inodes free */
	spinlock_t	stat_lock;
246
	struct hstate *hstate;
247
	struct hugepage_subpool *spool;
Linus Torvalds's avatar
Linus Torvalds committed
248 249 250 251 252 253 254
};

static inline struct hugetlbfs_sb_info *HUGETLBFS_SB(struct super_block *sb)
{
	return sb->s_fs_info;
}

255
extern const struct file_operations hugetlbfs_file_operations;
256
extern const struct vm_operations_struct hugetlb_vm_ops;
257
struct file *hugetlb_file_setup(const char *name, size_t size, vm_flags_t acct,
258 259
				struct user_struct **user, int creat_flags,
				int page_size_log);
Linus Torvalds's avatar
Linus Torvalds committed
260 261 262

static inline int is_file_hugepages(struct file *file)
{
263 264 265 266 267 268
	if (file->f_op == &hugetlbfs_file_operations)
		return 1;
	if (is_file_shm_hugepages(file))
		return 1;

	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
269 270
}

271

Linus Torvalds's avatar
Linus Torvalds committed
272 273
#else /* !CONFIG_HUGETLBFS */

274
#define is_file_hugepages(file)			0
275
static inline struct file *
276 277
hugetlb_file_setup(const char *name, size_t size, vm_flags_t acctflag,
		struct user_struct **user, int creat_flags,
278
		int page_size_log)
279 280 281
{
	return ERR_PTR(-ENOSYS);
}
Linus Torvalds's avatar
Linus Torvalds committed
282 283 284

#endif /* !CONFIG_HUGETLBFS */

285 286 287 288 289 290
#ifdef HAVE_ARCH_HUGETLB_UNMAPPED_AREA
unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr,
					unsigned long len, unsigned long pgoff,
					unsigned long flags);
#endif /* HAVE_ARCH_HUGETLB_UNMAPPED_AREA */

291 292
#ifdef CONFIG_HUGETLB_PAGE

293
#define HSTATE_NAME_LEN 32
294 295
/* Defines one hugetlb page size */
struct hstate {
296 297
	int next_nid_to_alloc;
	int next_nid_to_free;
298 299 300 301 302 303 304 305
	unsigned int order;
	unsigned long mask;
	unsigned long max_huge_pages;
	unsigned long nr_huge_pages;
	unsigned long free_huge_pages;
	unsigned long resv_huge_pages;
	unsigned long surplus_huge_pages;
	unsigned long nr_overcommit_huge_pages;
306
	struct list_head hugepage_activelist;
307 308 309 310
	struct list_head hugepage_freelists[MAX_NUMNODES];
	unsigned int nr_huge_pages_node[MAX_NUMNODES];
	unsigned int free_huge_pages_node[MAX_NUMNODES];
	unsigned int surplus_huge_pages_node[MAX_NUMNODES];
311 312 313 314
#ifdef CONFIG_CGROUP_HUGETLB
	/* cgroup control files */
	struct cftype cgroup_files[5];
#endif
315
	char name[HSTATE_NAME_LEN];
316 317
};

318 319 320
struct huge_bootmem_page {
	struct list_head list;
	struct hstate *hstate;
321 322 323
#ifdef CONFIG_HIGHMEM
	phys_addr_t phys;
#endif
324 325
};

326
struct page *alloc_huge_page_node(struct hstate *h, int nid);
327 328
struct page *alloc_huge_page_noerr(struct vm_area_struct *vma,
				unsigned long addr, int avoid_reserve);
329

330 331 332
/* arch callback */
int __init alloc_bootmem_huge_page(struct hstate *h);

333 334 335 336 337 338 339 340 341 342 343
void __init hugetlb_add_hstate(unsigned order);
struct hstate *size_to_hstate(unsigned long size);

#ifndef HUGE_MAX_HSTATE
#define HUGE_MAX_HSTATE 1
#endif

extern struct hstate hstates[HUGE_MAX_HSTATE];
extern unsigned int default_hstate_idx;

#define default_hstate (hstates[default_hstate_idx])
344

345
static inline struct hstate *hstate_inode(struct inode *i)
346
{
347 348 349
	struct hugetlbfs_sb_info *hsb;
	hsb = HUGETLBFS_SB(i->i_sb);
	return hsb->hstate;
350 351 352 353
}

static inline struct hstate *hstate_file(struct file *f)
{
Al Viro's avatar
Al Viro committed
354
	return hstate_inode(file_inode(f));
355 356
}

357 358 359 360
static inline struct hstate *hstate_sizelog(int page_size_log)
{
	if (!page_size_log)
		return &default_hstate;
361 362

	return size_to_hstate(1UL << page_size_log);
363 364
}

365
static inline struct hstate *hstate_vma(struct vm_area_struct *vma)
366
{
367
	return hstate_file(vma->vm_file);
368 369 370 371 372 373 374
}

static inline unsigned long huge_page_size(struct hstate *h)
{
	return (unsigned long)PAGE_SIZE << h->order;
}

375 376
extern unsigned long vma_kernel_pagesize(struct vm_area_struct *vma);

377 378
extern unsigned long vma_mmu_pagesize(struct vm_area_struct *vma);

379 380 381 382 383 384 385 386 387 388 389 390 391 392 393
static inline unsigned long huge_page_mask(struct hstate *h)
{
	return h->mask;
}

static inline unsigned int huge_page_order(struct hstate *h)
{
	return h->order;
}

static inline unsigned huge_page_shift(struct hstate *h)
{
	return h->order + PAGE_SHIFT;
}

394 395 396 397 398
static inline bool hstate_is_gigantic(struct hstate *h)
{
	return huge_page_order(h) >= MAX_ORDER;
}

399 400 401 402 403 404 405 406 407 408 409 410
static inline unsigned int pages_per_huge_page(struct hstate *h)
{
	return 1 << h->order;
}

static inline unsigned int blocks_per_huge_page(struct hstate *h)
{
	return huge_page_size(h) / 512;
}

#include <asm/hugetlb.h>

411 412 413 414 415 416 417 418
#ifndef arch_make_huge_pte
static inline pte_t arch_make_huge_pte(pte_t entry, struct vm_area_struct *vma,
				       struct page *page, int writable)
{
	return entry;
}
#endif

419 420
static inline struct hstate *page_hstate(struct page *page)
{
421
	VM_BUG_ON_PAGE(!PageHuge(page), page);
422 423 424
	return size_to_hstate(PAGE_SIZE << compound_order(page));
}

425 426 427 428 429
static inline unsigned hstate_index_to_shift(unsigned index)
{
	return hstates[index].order + PAGE_SHIFT;
}

430 431 432 433 434
static inline int hstate_index(struct hstate *h)
{
	return h - hstates;
}

435 436 437 438 439 440 441 442 443 444 445
pgoff_t __basepage_index(struct page *page);

/* Return page->index in PAGE_SIZE units */
static inline pgoff_t basepage_index(struct page *page)
{
	if (!PageCompound(page))
		return page->index;

	return __basepage_index(page);
}

446 447
extern void dissolve_free_huge_pages(unsigned long start_pfn,
				     unsigned long end_pfn);
448
static inline int hugepage_migration_supported(struct hstate *h)
449
{
450 451 452 453 454
#ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION
	return huge_page_shift(h) == PMD_SHIFT;
#else
	return 0;
#endif
455
}
456

457 458 459 460 461 462 463 464 465
static inline spinlock_t *huge_pte_lockptr(struct hstate *h,
					   struct mm_struct *mm, pte_t *pte)
{
	if (huge_page_size(h) == PMD_SIZE)
		return pmd_lockptr(mm, (pmd_t *) pte);
	VM_BUG_ON(huge_page_size(h) == PAGE_SIZE);
	return &mm->page_table_lock;
}

466 467 468 469 470 471 472 473
#ifndef hugepages_supported
/*
 * Some platform decide whether they support huge pages at boot
 * time. Some of them, such as powerpc, set HPAGE_SHIFT to 0
 * when there is no such support
 */
#define hugepages_supported() (HPAGE_SHIFT != 0)
#endif
474

475
#else	/* CONFIG_HUGETLB_PAGE */
476
struct hstate {};
477
#define alloc_huge_page_node(h, nid) NULL
478
#define alloc_huge_page_noerr(v, a, r) NULL
479
#define alloc_bootmem_huge_page(h) NULL
480
#define hstate_file(f) NULL
481
#define hstate_sizelog(s) NULL
482 483
#define hstate_vma(v) NULL
#define hstate_inode(i) NULL
484
#define page_hstate(page) NULL
485 486
#define huge_page_size(h) PAGE_SIZE
#define huge_page_mask(h) PAGE_MASK
487
#define vma_kernel_pagesize(v) PAGE_SIZE
488
#define vma_mmu_pagesize(v) PAGE_SIZE
489 490
#define huge_page_order(h) 0
#define huge_page_shift(h) PAGE_SHIFT
491 492 493 494
static inline unsigned int pages_per_huge_page(struct hstate *h)
{
	return 1;
}
495
#define hstate_index_to_shift(index) 0
496
#define hstate_index(h) 0
497 498 499 500 501

static inline pgoff_t basepage_index(struct page *page)
{
	return page->index;
}
502
#define dissolve_free_huge_pages(s, e)	do {} while (0)
503
#define hugepage_migration_supported(h)	0
504 505 506 507 508 509

static inline spinlock_t *huge_pte_lockptr(struct hstate *h,
					   struct mm_struct *mm, pte_t *pte)
{
	return &mm->page_table_lock;
}
510
#endif	/* CONFIG_HUGETLB_PAGE */
511

512 513 514 515 516 517 518 519 520 521
static inline spinlock_t *huge_pte_lock(struct hstate *h,
					struct mm_struct *mm, pte_t *pte)
{
	spinlock_t *ptl;

	ptl = huge_pte_lockptr(h, mm, pte);
	spin_lock(ptl);
	return ptl;
}

Linus Torvalds's avatar
Linus Torvalds committed
522
#endif /* _LINUX_HUGETLB_H */