[linux-block.git] / include / linux / huge_mm.h

/* SPDX-License-Identifier: GPL-2.0 */
#ifndef _LINUX_HUGE_MM_H
#define _LINUX_HUGE_MM_H

#include <linux/sched/coredump.h>
#include <linux/mm_types.h>

#include <linux/fs.h> /* only for vma_is_dax() */

vm_fault_t do_huge_pmd_anonymous_page(struct vm_fault *vmf);
int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm,
		  pmd_t *dst_pmd, pmd_t *src_pmd, unsigned long addr,
		  struct vm_area_struct *dst_vma, struct vm_area_struct *src_vma);
void huge_pmd_set_accessed(struct vm_fault *vmf);
int copy_huge_pud(struct mm_struct *dst_mm, struct mm_struct *src_mm,
		  pud_t *dst_pud, pud_t *src_pud, unsigned long addr,
		  struct vm_area_struct *vma);

#ifdef CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
void huge_pud_set_accessed(struct vm_fault *vmf, pud_t orig_pud);
#else
static inline void huge_pud_set_accessed(struct vm_fault *vmf, pud_t orig_pud)
{
}
#endif

vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf);
struct page *follow_trans_huge_pmd(struct vm_area_struct *vma,
				   unsigned long addr, pmd_t *pmd,
				   unsigned int flags);
bool madvise_free_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
			   pmd_t *pmd, unsigned long addr, unsigned long next);
int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma, pmd_t *pmd,
		 unsigned long addr);
int zap_huge_pud(struct mmu_gather *tlb, struct vm_area_struct *vma, pud_t *pud,
		 unsigned long addr);
bool move_huge_pmd(struct vm_area_struct *vma, unsigned long old_addr,
		   unsigned long new_addr, pmd_t *old_pmd, pmd_t *new_pmd);
int change_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
		    pmd_t *pmd, unsigned long addr, pgprot_t newprot,
		    unsigned long cp_flags);
vm_fault_t vmf_insert_pfn_pmd_prot(struct vm_fault *vmf, pfn_t pfn,
				   pgprot_t pgprot, bool write);

/**
 * vmf_insert_pfn_pmd - insert a pmd size pfn
 * @vmf: Structure describing the fault
 * @pfn: pfn to insert
 * @pgprot: page protection to use
 * @write: whether it's a write fault
 *
 * Insert a pmd size pfn. See vmf_insert_pfn() for additional info.
 *
 * Return: vm_fault_t value.
 */
static inline vm_fault_t vmf_insert_pfn_pmd(struct vm_fault *vmf, pfn_t pfn,
					    bool write)
{
	return vmf_insert_pfn_pmd_prot(vmf, pfn, vmf->vma->vm_page_prot, write);
}
vm_fault_t vmf_insert_pfn_pud_prot(struct vm_fault *vmf, pfn_t pfn,
				   pgprot_t pgprot, bool write);

/**
 * vmf_insert_pfn_pud - insert a pud size pfn
 * @vmf: Structure describing the fault
 * @pfn: pfn to insert
 * @pgprot: page protection to use
 * @write: whether it's a write fault
 *
 * Insert a pud size pfn. See vmf_insert_pfn() for additional info.
 *
 * Return: vm_fault_t value.
 */
static inline vm_fault_t vmf_insert_pfn_pud(struct vm_fault *vmf, pfn_t pfn,
					    bool write)
{
	return vmf_insert_pfn_pud_prot(vmf, pfn, vmf->vma->vm_page_prot, write);
}

enum transparent_hugepage_flag {
	TRANSPARENT_HUGEPAGE_NEVER_DAX,
	TRANSPARENT_HUGEPAGE_FLAG,
	TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG,
	TRANSPARENT_HUGEPAGE_DEFRAG_DIRECT_FLAG,
	TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_FLAG,
	TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_OR_MADV_FLAG,
	TRANSPARENT_HUGEPAGE_DEFRAG_REQ_MADV_FLAG,
	TRANSPARENT_HUGEPAGE_DEFRAG_KHUGEPAGED_FLAG,
	TRANSPARENT_HUGEPAGE_USE_ZERO_PAGE_FLAG,
};

struct kobject;
struct kobj_attribute;

ssize_t single_hugepage_flag_store(struct kobject *kobj,
				   struct kobj_attribute *attr,
				   const char *buf, size_t count,
				   enum transparent_hugepage_flag flag);
ssize_t single_hugepage_flag_show(struct kobject *kobj,
				  struct kobj_attribute *attr, char *buf,
				  enum transparent_hugepage_flag flag);
extern struct kobj_attribute shmem_enabled_attr;

#define HPAGE_PMD_ORDER (HPAGE_PMD_SHIFT-PAGE_SHIFT)
#define HPAGE_PMD_NR (1<<HPAGE_PMD_ORDER)

#ifdef CONFIG_TRANSPARENT_HUGEPAGE
#define HPAGE_PMD_SHIFT PMD_SHIFT
#define HPAGE_PMD_SIZE	((1UL) << HPAGE_PMD_SHIFT)
#define HPAGE_PMD_MASK	(~(HPAGE_PMD_SIZE - 1))

#define HPAGE_PUD_SHIFT PUD_SHIFT
#define HPAGE_PUD_SIZE	((1UL) << HPAGE_PUD_SHIFT)
#define HPAGE_PUD_MASK	(~(HPAGE_PUD_SIZE - 1))

extern unsigned long transparent_hugepage_flags;

#define hugepage_flags_enabled()					       \
	(transparent_hugepage_flags &				       \
	 ((1<<TRANSPARENT_HUGEPAGE_FLAG) |		       \
	  (1<<TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG)))
#define hugepage_flags_always()				\
	(transparent_hugepage_flags &			\
	 (1<<TRANSPARENT_HUGEPAGE_FLAG))

/*
 * Do the below checks:
 *   - For file vma, check if the linear page offset of vma is
 *     HPAGE_PMD_NR aligned within the file.  The hugepage is
 *     guaranteed to be hugepage-aligned within the file, but we must
 *     check that the PMD-aligned addresses in the VMA map to
 *     PMD-aligned offsets within the file, else the hugepage will
 *     not be PMD-mappable.
 *   - For all vmas, check if the haddr is in an aligned HPAGE_PMD_SIZE
 *     area.
 */
static inline bool transhuge_vma_suitable(struct vm_area_struct *vma,
		unsigned long addr)
{
	unsigned long haddr;

	/* Don't have to check pgoff for anonymous vma */
	if (!vma_is_anonymous(vma)) {
		if (!IS_ALIGNED((vma->vm_start >> PAGE_SHIFT) - vma->vm_pgoff,
				HPAGE_PMD_NR))
			return false;
	}

	haddr = addr & HPAGE_PMD_MASK;

	if (haddr < vma->vm_start || haddr + HPAGE_PMD_SIZE > vma->vm_end)
		return false;
	return true;
}

static inline bool file_thp_enabled(struct vm_area_struct *vma)
{
	struct inode *inode;

	if (!vma->vm_file)
		return false;

	inode = vma->vm_file->f_inode;

	return (IS_ENABLED(CONFIG_READ_ONLY_THP_FOR_FS)) &&
	       (vma->vm_flags & VM_EXEC) &&
	       !inode_is_open_for_write(inode) && S_ISREG(inode->i_mode);
}

bool hugepage_vma_check(struct vm_area_struct *vma, unsigned long vm_flags,
			bool smaps, bool in_pf, bool enforce_sysfs);

#define transparent_hugepage_use_zero_page()				\
	(transparent_hugepage_flags &					\
	 (1<<TRANSPARENT_HUGEPAGE_USE_ZERO_PAGE_FLAG))

unsigned long thp_get_unmapped_area(struct file *filp, unsigned long addr,
		unsigned long len, unsigned long pgoff, unsigned long flags);

void prep_transhuge_page(struct page *page);
void free_transhuge_page(struct page *page);

bool can_split_folio(struct folio *folio, int *pextra_pins);
int split_huge_page_to_list(struct page *page, struct list_head *list);
static inline int split_huge_page(struct page *page)
{
	return split_huge_page_to_list(page, NULL);
}
void deferred_split_huge_page(struct page *page);

void __split_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd,
		unsigned long address, bool freeze, struct folio *folio);

#define split_huge_pmd(__vma, __pmd, __address)				\
	do {								\
		pmd_t *____pmd = (__pmd);				\
		if (is_swap_pmd(*____pmd) || pmd_trans_huge(*____pmd)	\
					|| pmd_devmap(*____pmd))	\
			__split_huge_pmd(__vma, __pmd, __address,	\
						false, NULL);		\
	}  while (0)


void split_huge_pmd_address(struct vm_area_struct *vma, unsigned long address,
		bool freeze, struct folio *folio);

void __split_huge_pud(struct vm_area_struct *vma, pud_t *pud,
		unsigned long address);

#define split_huge_pud(__vma, __pud, __address)				\
	do {								\
		pud_t *____pud = (__pud);				\
		if (pud_trans_huge(*____pud)				\
					|| pud_devmap(*____pud))	\
			__split_huge_pud(__vma, __pud, __address);	\
	}  while (0)

int hugepage_madvise(struct vm_area_struct *vma, unsigned long *vm_flags,
		     int advice);
int madvise_collapse(struct vm_area_struct *vma,
		     struct vm_area_struct **prev,
		     unsigned long start, unsigned long end);
void vma_adjust_trans_huge(struct vm_area_struct *vma, unsigned long start,
			   unsigned long end, long adjust_next);
spinlock_t *__pmd_trans_huge_lock(pmd_t *pmd, struct vm_area_struct *vma);
spinlock_t *__pud_trans_huge_lock(pud_t *pud, struct vm_area_struct *vma);

static inline int is_swap_pmd(pmd_t pmd)
{
	return !pmd_none(pmd) && !pmd_present(pmd);
}

/* mmap_lock must be held on entry */
static inline spinlock_t *pmd_trans_huge_lock(pmd_t *pmd,
		struct vm_area_struct *vma)
{
	if (is_swap_pmd(*pmd) || pmd_trans_huge(*pmd) || pmd_devmap(*pmd))
		return __pmd_trans_huge_lock(pmd, vma);
	else
		return NULL;
}
static inline spinlock_t *pud_trans_huge_lock(pud_t *pud,
		struct vm_area_struct *vma)
{
	if (pud_trans_huge(*pud) || pud_devmap(*pud))
		return __pud_trans_huge_lock(pud, vma);
	else
		return NULL;
}

/**
 * folio_test_pmd_mappable - Can we map this folio with a PMD?
 * @folio: The folio to test
 */
static inline bool folio_test_pmd_mappable(struct folio *folio)
{
	return folio_order(folio) >= HPAGE_PMD_ORDER;
}

struct page *follow_devmap_pmd(struct vm_area_struct *vma, unsigned long addr,
		pmd_t *pmd, int flags, struct dev_pagemap **pgmap);
struct page *follow_devmap_pud(struct vm_area_struct *vma, unsigned long addr,
		pud_t *pud, int flags, struct dev_pagemap **pgmap);

vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf);

extern struct page *huge_zero_page;
extern unsigned long huge_zero_pfn;

static inline bool is_huge_zero_page(struct page *page)
{
	return READ_ONCE(huge_zero_page) == page;
}

static inline bool is_huge_zero_pmd(pmd_t pmd)
{
	return pmd_present(pmd) && READ_ONCE(huge_zero_pfn) == pmd_pfn(pmd);
}

static inline bool is_huge_zero_pud(pud_t pud)
{
	return false;
}

struct page *mm_get_huge_zero_page(struct mm_struct *mm);
void mm_put_huge_zero_page(struct mm_struct *mm);

#define mk_huge_pmd(page, prot) pmd_mkhuge(mk_pmd(page, prot))

static inline bool thp_migration_supported(void)
{
	return IS_ENABLED(CONFIG_ARCH_ENABLE_THP_MIGRATION);
}

static inline struct list_head *page_deferred_list(struct page *page)
{
	/*
	 * See organization of tail pages of compound page in
	 * "struct page" definition.
	 */
	return &page[2].deferred_list;
}

#else /* CONFIG_TRANSPARENT_HUGEPAGE */
#define HPAGE_PMD_SHIFT ({ BUILD_BUG(); 0; })
#define HPAGE_PMD_MASK ({ BUILD_BUG(); 0; })
#define HPAGE_PMD_SIZE ({ BUILD_BUG(); 0; })

#define HPAGE_PUD_SHIFT ({ BUILD_BUG(); 0; })
#define HPAGE_PUD_MASK ({ BUILD_BUG(); 0; })
#define HPAGE_PUD_SIZE ({ BUILD_BUG(); 0; })

static inline bool folio_test_pmd_mappable(struct folio *folio)
{
	return false;
}

static inline bool transhuge_vma_suitable(struct vm_area_struct *vma,
		unsigned long addr)
{
	return false;
}

static inline bool hugepage_vma_check(struct vm_area_struct *vma,
				      unsigned long vm_flags, bool smaps,
				      bool in_pf, bool enforce_sysfs)
{
	return false;
}

static inline void prep_transhuge_page(struct page *page) {}

#define transparent_hugepage_flags 0UL

#define thp_get_unmapped_area	NULL

static inline bool
can_split_folio(struct folio *folio, int *pextra_pins)
{
	return false;
}
static inline int
split_huge_page_to_list(struct page *page, struct list_head *list)
{
	return 0;
}
static inline int split_huge_page(struct page *page)
{
	return 0;
}
static inline void deferred_split_huge_page(struct page *page) {}
#define split_huge_pmd(__vma, __pmd, __address)	\
	do { } while (0)

static inline void __split_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd,
		unsigned long address, bool freeze, struct folio *folio) {}
static inline void split_huge_pmd_address(struct vm_area_struct *vma,
		unsigned long address, bool freeze, struct folio *folio) {}

#define split_huge_pud(__vma, __pmd, __address)	\
	do { } while (0)

static inline int hugepage_madvise(struct vm_area_struct *vma,
				   unsigned long *vm_flags, int advice)
{
	return -EINVAL;
}

static inline int madvise_collapse(struct vm_area_struct *vma,
				   struct vm_area_struct **prev,
				   unsigned long start, unsigned long end)
{
	return -EINVAL;
}

static inline void vma_adjust_trans_huge(struct vm_area_struct *vma,
					 unsigned long start,
					 unsigned long end,
					 long adjust_next)
{
}
static inline int is_swap_pmd(pmd_t pmd)
{
	return 0;
}
static inline spinlock_t *pmd_trans_huge_lock(pmd_t *pmd,
		struct vm_area_struct *vma)
{
	return NULL;
}
static inline spinlock_t *pud_trans_huge_lock(pud_t *pud,
		struct vm_area_struct *vma)
{
	return NULL;
}

static inline vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf)
{
	return 0;
}

static inline bool is_huge_zero_page(struct page *page)
{
	return false;
}

static inline bool is_huge_zero_pmd(pmd_t pmd)
{
	return false;
}

static inline bool is_huge_zero_pud(pud_t pud)
{
	return false;
}

static inline void mm_put_huge_zero_page(struct mm_struct *mm)
{
	return;
}

static inline struct page *follow_devmap_pmd(struct vm_area_struct *vma,
	unsigned long addr, pmd_t *pmd, int flags, struct dev_pagemap **pgmap)
{
	return NULL;
}

static inline struct page *follow_devmap_pud(struct vm_area_struct *vma,
	unsigned long addr, pud_t *pud, int flags, struct dev_pagemap **pgmap)
{
	return NULL;
}

static inline bool thp_migration_supported(void)
{
	return false;
}
#endif /* CONFIG_TRANSPARENT_HUGEPAGE */

static inline int split_folio_to_list(struct folio *folio,
		struct list_head *list)
{
	return split_huge_page_to_list(&folio->page, list);
}

static inline int split_folio(struct folio *folio)
{
	return split_folio_to_list(folio, NULL);
}

/*
 * archs that select ARCH_WANTS_THP_SWAP but don't support THP_SWP due to
 * limitations in the implementation like arm64 MTE can override this to
 * false
 */
#ifndef arch_thp_swp_supported
static inline bool arch_thp_swp_supported(void)
{
	return true;
}
#endif

#endif /* _LINUX_HUGE_MM_H */
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
71e3aac0 AA	2	#ifndef _LINUX_HUGE_MM_H
	3	#define _LINUX_HUGE_MM_H
	4
16981d76	5	#include <linux/sched/coredump.h>
226ab561	6	#include <linux/mm_types.h>
16981d76	7
baabda26 DW	8	#include <linux/fs.h> /* only for vma_is_dax() */
baabda26 DW	9
ebfe1b8f RC	10	vm_fault_t do_huge_pmd_anonymous_page(struct vm_fault *vmf);
	11	int copy_huge_pmd(struct mm_struct dst_mm, struct mm_struct src_mm,
	12	pmd_t dst_pmd, pmd_t src_pmd, unsigned long addr,
8f34f1ea	13	struct vm_area_struct dst_vma, struct vm_area_struct src_vma);
5db4f15c	14	void huge_pmd_set_accessed(struct vm_fault *vmf);
ebfe1b8f RC	15	int copy_huge_pud(struct mm_struct dst_mm, struct mm_struct src_mm,
	16	pud_t dst_pud, pud_t src_pud, unsigned long addr,
	17	struct vm_area_struct *vma);
a00cc7d9 MW	18
a00cc7d9 MW	19	#ifdef CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
ebfe1b8f	20	void huge_pud_set_accessed(struct vm_fault *vmf, pud_t orig_pud);
a00cc7d9 MW	21	#else
	22	static inline void huge_pud_set_accessed(struct vm_fault *vmf, pud_t orig_pud)
	23	{
	24	}
	25	#endif
	26
5db4f15c	27	vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf);
ebfe1b8f RC	28	struct page follow_trans_huge_pmd(struct vm_area_struct vma,
	29	unsigned long addr, pmd_t *pmd,
	30	unsigned int flags);
	31	bool madvise_free_huge_pmd(struct mmu_gather tlb, struct vm_area_struct vma,
	32	pmd_t *pmd, unsigned long addr, unsigned long next);
	33	int zap_huge_pmd(struct mmu_gather tlb, struct vm_area_struct vma, pmd_t *pmd,
	34	unsigned long addr);
	35	int zap_huge_pud(struct mmu_gather tlb, struct vm_area_struct vma, pud_t *pud,
	36	unsigned long addr);
	37	bool move_huge_pmd(struct vm_area_struct *vma, unsigned long old_addr,
	38	unsigned long new_addr, pmd_t old_pmd, pmd_t new_pmd);
4a18419f NA	39	int change_huge_pmd(struct mmu_gather tlb, struct vm_area_struct vma,
	40	pmd_t *pmd, unsigned long addr, pgprot_t newprot,
	41	unsigned long cp_flags);
9a9731b1 THV	42	vm_fault_t vmf_insert_pfn_pmd_prot(struct vm_fault *vmf, pfn_t pfn,
	43	pgprot_t pgprot, bool write);
	44
	45	/**
	46	* vmf_insert_pfn_pmd - insert a pmd size pfn
	47	* @vmf: Structure describing the fault
	48	* @pfn: pfn to insert
	49	* @pgprot: page protection to use
	50	* @write: whether it's a write fault
	51	*
	52	* Insert a pmd size pfn. See vmf_insert_pfn() for additional info.
	53	*
	54	* Return: vm_fault_t value.
	55	*/
	56	static inline vm_fault_t vmf_insert_pfn_pmd(struct vm_fault *vmf, pfn_t pfn,
	57	bool write)
	58	{
	59	return vmf_insert_pfn_pmd_prot(vmf, pfn, vmf->vma->vm_page_prot, write);
	60	}
	61	vm_fault_t vmf_insert_pfn_pud_prot(struct vm_fault *vmf, pfn_t pfn,
	62	pgprot_t pgprot, bool write);
	63
	64	/**
	65	* vmf_insert_pfn_pud - insert a pud size pfn
	66	* @vmf: Structure describing the fault
	67	* @pfn: pfn to insert
	68	* @pgprot: page protection to use
	69	* @write: whether it's a write fault
	70	*
	71	* Insert a pud size pfn. See vmf_insert_pfn() for additional info.
	72	*
	73	* Return: vm_fault_t value.
	74	*/
	75	static inline vm_fault_t vmf_insert_pfn_pud(struct vm_fault *vmf, pfn_t pfn,
	76	bool write)
	77	{
	78	return vmf_insert_pfn_pud_prot(vmf, pfn, vmf->vma->vm_page_prot, write);
	79	}
	80
71e3aac0	81	enum transparent_hugepage_flag {
bae84953	82	TRANSPARENT_HUGEPAGE_NEVER_DAX,
71e3aac0 AA	83	TRANSPARENT_HUGEPAGE_FLAG,
71e3aac0 AA	84	TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG,
444eb2a4 MG	85	TRANSPARENT_HUGEPAGE_DEFRAG_DIRECT_FLAG,
444eb2a4 MG	86	TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_FLAG,
21440d7e	87	TRANSPARENT_HUGEPAGE_DEFRAG_KSWAPD_OR_MADV_FLAG,
71e3aac0	88	TRANSPARENT_HUGEPAGE_DEFRAG_REQ_MADV_FLAG,
ba76149f	89	TRANSPARENT_HUGEPAGE_DEFRAG_KHUGEPAGED_FLAG,
79da5407	90	TRANSPARENT_HUGEPAGE_USE_ZERO_PAGE_FLAG,
71e3aac0 AA	91	};
71e3aac0 AA	92
b46e756f KS	93	struct kobject;
	94	struct kobj_attribute;
	95
ebfe1b8f RC	96	ssize_t single_hugepage_flag_store(struct kobject *kobj,
	97	struct kobj_attribute *attr,
	98	const char *buf, size_t count,
	99	enum transparent_hugepage_flag flag);
	100	ssize_t single_hugepage_flag_show(struct kobject *kobj,
	101	struct kobj_attribute attr, char buf,
	102	enum transparent_hugepage_flag flag);
5a6e75f8 KS	103	extern struct kobj_attribute shmem_enabled_attr;
5a6e75f8 KS	104
d8c37c48 NH	105	#define HPAGE_PMD_ORDER (HPAGE_PMD_SHIFT-PAGE_SHIFT)
	106	#define HPAGE_PMD_NR (1<<HPAGE_PMD_ORDER)
	107
71e3aac0	108	#ifdef CONFIG_TRANSPARENT_HUGEPAGE
fde52796 AK	109	#define HPAGE_PMD_SHIFT PMD_SHIFT
	110	#define HPAGE_PMD_SIZE ((1UL) << HPAGE_PMD_SHIFT)
	111	#define HPAGE_PMD_MASK (~(HPAGE_PMD_SIZE - 1))
71e3aac0	112
a00cc7d9 MW	113	#define HPAGE_PUD_SHIFT PUD_SHIFT
	114	#define HPAGE_PUD_SIZE ((1UL) << HPAGE_PUD_SHIFT)
	115	#define HPAGE_PUD_MASK (~(HPAGE_PUD_SIZE - 1))
	116
16981d76 DW	117	extern unsigned long transparent_hugepage_flags;
16981d76 DW	118
1064026b YS	119	#define hugepage_flags_enabled() \
	120	(transparent_hugepage_flags & \
	121	((1<<TRANSPARENT_HUGEPAGE_FLAG) \| \
	122	(1<<TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG)))
	123	#define hugepage_flags_always() \
	124	(transparent_hugepage_flags & \
	125	(1<<TRANSPARENT_HUGEPAGE_FLAG))
	126
4fa6893f YS	127	/*
	128	* Do the below checks:
	129	* - For file vma, check if the linear page offset of vma is
	130	* HPAGE_PMD_NR aligned within the file. The hugepage is
	131	* guaranteed to be hugepage-aligned within the file, but we must
	132	* check that the PMD-aligned addresses in the VMA map to
	133	* PMD-aligned offsets within the file, else the hugepage will
	134	* not be PMD-mappable.
	135	* - For all vmas, check if the haddr is in an aligned HPAGE_PMD_SIZE
	136	* area.
	137	*/
e6be37b2	138	static inline bool transhuge_vma_suitable(struct vm_area_struct *vma,
c453d8c7	139	unsigned long addr)
e6be37b2	140	{
c453d8c7 YS	141	unsigned long haddr;
c453d8c7 YS	142
e6be37b2 ML	143	/* Don't have to check pgoff for anonymous vma */
	144	if (!vma_is_anonymous(vma)) {
	145	if (!IS_ALIGNED((vma->vm_start >> PAGE_SHIFT) - vma->vm_pgoff,
	146	HPAGE_PMD_NR))
	147	return false;
	148	}
	149
c453d8c7	150	haddr = addr & HPAGE_PMD_MASK;
e6be37b2	151
e6be37b2	152	if (haddr < vma->vm_start \|\| haddr + HPAGE_PMD_SIZE > vma->vm_end)
e6be37b2 ML	153	return false;
	154	return true;
	155	}
	156
78d12c19 YS	157	static inline bool file_thp_enabled(struct vm_area_struct *vma)
	158	{
	159	struct inode *inode;
	160
	161	if (!vma->vm_file)
	162	return false;
	163
	164	inode = vma->vm_file->f_inode;
	165
	166	return (IS_ENABLED(CONFIG_READ_ONLY_THP_FOR_FS)) &&
	167	(vma->vm_flags & VM_EXEC) &&
	168	!inode_is_open_for_write(inode) && S_ISREG(inode->i_mode);
	169	}
	170
a7f4e6e4 ZK	171	bool hugepage_vma_check(struct vm_area_struct *vma, unsigned long vm_flags,
a7f4e6e4 ZK	172	bool smaps, bool in_pf, bool enforce_sysfs);
43675e6f	173
79da5407 KS	174	#define transparent_hugepage_use_zero_page() \
	175	(transparent_hugepage_flags & \
	176	(1<<TRANSPARENT_HUGEPAGE_USE_ZERO_PAGE_FLAG))
71e3aac0	177
ebfe1b8f RC	178	unsigned long thp_get_unmapped_area(struct file *filp, unsigned long addr,
ebfe1b8f RC	179	unsigned long len, unsigned long pgoff, unsigned long flags);
74d2fad1	180
ebfe1b8f RC	181	void prep_transhuge_page(struct page *page);
ebfe1b8f RC	182	void free_transhuge_page(struct page *page);
9a982250	183
d4b4084a	184	bool can_split_folio(struct folio folio, int pextra_pins);
e9b61f19 KS	185	int split_huge_page_to_list(struct page page, struct list_head list);
	186	static inline int split_huge_page(struct page *page)
	187	{
	188	return split_huge_page_to_list(page, NULL);
	189	}
9a982250	190	void deferred_split_huge_page(struct page *page);
eef1b3ba KS	191
eef1b3ba KS	192	void __split_huge_pmd(struct vm_area_struct vma, pmd_t pmd,
af28a988	193	unsigned long address, bool freeze, struct folio *folio);
eef1b3ba KS	194
	195	#define split_huge_pmd(__vma, __pmd, __address) \
	196	do { \
	197	pmd_t *____pmd = (__pmd); \
84c3fc4e	198	if (is_swap_pmd(____pmd) \|\| pmd_trans_huge(____pmd) \
5c7fb56e	199	\|\| pmd_devmap(*____pmd)) \
fec89c10	200	__split_huge_pmd(__vma, __pmd, __address, \
33f4751e	201	false, NULL); \
eef1b3ba	202	} while (0)
ad0bed24	203
2a52bcbc	204
fec89c10	205	void split_huge_pmd_address(struct vm_area_struct *vma, unsigned long address,
af28a988	206	bool freeze, struct folio *folio);
2a52bcbc	207
a00cc7d9 MW	208	void __split_huge_pud(struct vm_area_struct vma, pud_t pud,
	209	unsigned long address);
	210
	211	#define split_huge_pud(__vma, __pud, __address) \
	212	do { \
	213	pud_t *____pud = (__pud); \
	214	if (pud_trans_huge(*____pud) \
	215	\|\| pud_devmap(*____pud)) \
	216	__split_huge_pud(__vma, __pud, __address); \
	217	} while (0)
	218
ebfe1b8f RC	219	int hugepage_madvise(struct vm_area_struct vma, unsigned long vm_flags,
ebfe1b8f RC	220	int advice);
7d8faaf1 ZK	221	int madvise_collapse(struct vm_area_struct *vma,
	222	struct vm_area_struct **prev,
	223	unsigned long start, unsigned long end);
ebfe1b8f RC	224	void vma_adjust_trans_huge(struct vm_area_struct *vma, unsigned long start,
	225	unsigned long end, long adjust_next);
	226	spinlock_t __pmd_trans_huge_lock(pmd_t pmd, struct vm_area_struct *vma);
	227	spinlock_t __pud_trans_huge_lock(pud_t pud, struct vm_area_struct *vma);
84c3fc4e ZY	228
	229	static inline int is_swap_pmd(pmd_t pmd)
	230	{
	231	return !pmd_none(pmd) && !pmd_present(pmd);
	232	}
	233
c1e8d7c6	234	/* mmap_lock must be held on entry */
b6ec57f4 KS	235	static inline spinlock_t pmd_trans_huge_lock(pmd_t pmd,
b6ec57f4 KS	236	struct vm_area_struct *vma)
025c5b24	237	{
84c3fc4e	238	if (is_swap_pmd(pmd) \|\| pmd_trans_huge(pmd) \|\| pmd_devmap(*pmd))
b6ec57f4	239	return __pmd_trans_huge_lock(pmd, vma);
025c5b24	240	else
969e8d7e	241	return NULL;
025c5b24	242	}
a00cc7d9 MW	243	static inline spinlock_t pud_trans_huge_lock(pud_t pud,
	244	struct vm_area_struct *vma)
	245	{
a00cc7d9 MW	246	if (pud_trans_huge(pud) \|\| pud_devmap(pud))
	247	return __pud_trans_huge_lock(pud, vma);
	248	else
	249	return NULL;
	250	}
6ffbb458	251
5bf34d7c MWO	252	/**
	253	* folio_test_pmd_mappable - Can we map this folio with a PMD?
	254	* @folio: The folio to test
	255	*/
	256	static inline bool folio_test_pmd_mappable(struct folio *folio)
	257	{
	258	return folio_order(folio) >= HPAGE_PMD_ORDER;
	259	}
	260
a00cc7d9	261	struct page follow_devmap_pmd(struct vm_area_struct vma, unsigned long addr,
df06b37f	262	pmd_t pmd, int flags, struct dev_pagemap *pgmap);
a00cc7d9	263	struct page follow_devmap_pud(struct vm_area_struct vma, unsigned long addr,
df06b37f	264	pud_t pud, int flags, struct dev_pagemap *pgmap);
a00cc7d9	265
5db4f15c	266	vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf);
d10e63f2	267
56873f43	268	extern struct page *huge_zero_page;
3b77e8c8	269	extern unsigned long huge_zero_pfn;
56873f43 WY	270
	271	static inline bool is_huge_zero_page(struct page *page)
	272	{
6aa7de05	273	return READ_ONCE(huge_zero_page) == page;
56873f43 WY	274	}
56873f43 WY	275
fc437044 MW	276	static inline bool is_huge_zero_pmd(pmd_t pmd)
fc437044 MW	277	{
3ce4fee4	278	return pmd_present(pmd) && READ_ONCE(huge_zero_pfn) == pmd_pfn(pmd);
fc437044 MW	279	}
fc437044 MW	280
a00cc7d9 MW	281	static inline bool is_huge_zero_pud(pud_t pud)
	282	{
	283	return false;
	284	}
	285
6fcb52a5 AL	286	struct page mm_get_huge_zero_page(struct mm_struct mm);
6fcb52a5 AL	287	void mm_put_huge_zero_page(struct mm_struct *mm);
fc437044	288
10102459 KS	289	#define mk_huge_pmd(page, prot) pmd_mkhuge(mk_pmd(page, prot))
10102459 KS	290
9c670ea3 NH	291	static inline bool thp_migration_supported(void)
	292	{
	293	return IS_ENABLED(CONFIG_ARCH_ENABLE_THP_MIGRATION);
	294	}
	295
87eaceb3 YS	296	static inline struct list_head page_deferred_list(struct page page)
	297	{
	298	/*
121c1781 ML	299	* See organization of tail pages of compound page in
121c1781 ML	300	* "struct page" definition.
87eaceb3 YS	301	*/
	302	return &page[2].deferred_list;
	303	}
	304
71e3aac0	305	#else /* CONFIG_TRANSPARENT_HUGEPAGE */
d8c37c48 NH	306	#define HPAGE_PMD_SHIFT ({ BUILD_BUG(); 0; })
	307	#define HPAGE_PMD_MASK ({ BUILD_BUG(); 0; })
	308	#define HPAGE_PMD_SIZE ({ BUILD_BUG(); 0; })
71e3aac0	309
a00cc7d9 MW	310	#define HPAGE_PUD_SHIFT ({ BUILD_BUG(); 0; })
	311	#define HPAGE_PUD_MASK ({ BUILD_BUG(); 0; })
	312	#define HPAGE_PUD_SIZE ({ BUILD_BUG(); 0; })
	313
5bf34d7c MWO	314	static inline bool folio_test_pmd_mappable(struct folio *folio)
	315	{
	316	return false;
	317	}
	318
43675e6f	319	static inline bool transhuge_vma_suitable(struct vm_area_struct *vma,
c453d8c7	320	unsigned long addr)
43675e6f YS	321	{
	322	return false;
	323	}
	324
9fec5168	325	static inline bool hugepage_vma_check(struct vm_area_struct *vma,
a7f4e6e4 ZK	326	unsigned long vm_flags, bool smaps,
a7f4e6e4 ZK	327	bool in_pf, bool enforce_sysfs)
e6be37b2 ML	328	{
	329	return false;
	330	}
	331
800d8c63 KS	332	static inline void prep_transhuge_page(struct page *page) {}
800d8c63 KS	333
71e3aac0	334	#define transparent_hugepage_flags 0UL
74d2fad1 TK	335
	336	#define thp_get_unmapped_area NULL
	337
b8f593cd	338	static inline bool
d4b4084a	339	can_split_folio(struct folio folio, int pextra_pins)
b8f593cd	340	{
b8f593cd HY	341	return false;
b8f593cd HY	342	}
5bc7b8ac SL	343	static inline int
	344	split_huge_page_to_list(struct page page, struct list_head list)
	345	{
	346	return 0;
	347	}
71e3aac0 AA	348	static inline int split_huge_page(struct page *page)
	349	{
	350	return 0;
	351	}
9a982250	352	static inline void deferred_split_huge_page(struct page *page) {}
78ddc534	353	#define split_huge_pmd(__vma, __pmd, __address) \
e180377f	354	do { } while (0)
2a52bcbc	355
fd60775a	356	static inline void __split_huge_pmd(struct vm_area_struct vma, pmd_t pmd,
af28a988	357	unsigned long address, bool freeze, struct folio *folio) {}
2a52bcbc	358	static inline void split_huge_pmd_address(struct vm_area_struct *vma,
af28a988	359	unsigned long address, bool freeze, struct folio *folio) {}
2a52bcbc	360
a00cc7d9 MW	361	#define split_huge_pud(__vma, __pmd, __address) \
	362	do { } while (0)
	363
60ab3244 AA	364	static inline int hugepage_madvise(struct vm_area_struct *vma,
60ab3244 AA	365	unsigned long *vm_flags, int advice)
0af4e98b	366	{
7d8faaf1	367	return -EINVAL;
0af4e98b	368	}
7d8faaf1 ZK	369
	370	static inline int madvise_collapse(struct vm_area_struct *vma,
	371	struct vm_area_struct **prev,
	372	unsigned long start, unsigned long end)
	373	{
	374	return -EINVAL;
	375	}
	376
94fcc585 AA	377	static inline void vma_adjust_trans_huge(struct vm_area_struct *vma,
	378	unsigned long start,
	379	unsigned long end,
	380	long adjust_next)
	381	{
	382	}
84c3fc4e ZY	383	static inline int is_swap_pmd(pmd_t pmd)
	384	{
	385	return 0;
	386	}
b6ec57f4 KS	387	static inline spinlock_t pmd_trans_huge_lock(pmd_t pmd,
b6ec57f4 KS	388	struct vm_area_struct *vma)
025c5b24	389	{
b6ec57f4	390	return NULL;
025c5b24	391	}
a00cc7d9 MW	392	static inline spinlock_t pud_trans_huge_lock(pud_t pud,
	393	struct vm_area_struct *vma)
	394	{
	395	return NULL;
	396	}
d10e63f2	397
5db4f15c	398	static inline vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf)
d10e63f2	399	{
4daae3b4	400	return 0;
d10e63f2 MG	401	}
d10e63f2 MG	402
56873f43 WY	403	static inline bool is_huge_zero_page(struct page *page)
	404	{
	405	return false;
	406	}
	407
3b77e8c8 HD	408	static inline bool is_huge_zero_pmd(pmd_t pmd)
	409	{
	410	return false;
	411	}
	412
a00cc7d9 MW	413	static inline bool is_huge_zero_pud(pud_t pud)
	414	{
	415	return false;
	416	}
	417
6fcb52a5	418	static inline void mm_put_huge_zero_page(struct mm_struct *mm)
aa88b68c	419	{
6fcb52a5	420	return;
aa88b68c	421	}
3565fce3 DW	422
3565fce3 DW	423	static inline struct page follow_devmap_pmd(struct vm_area_struct vma,
df06b37f	424	unsigned long addr, pmd_t pmd, int flags, struct dev_pagemap *pgmap)
3565fce3 DW	425	{
	426	return NULL;
	427	}
a00cc7d9 MW	428
a00cc7d9 MW	429	static inline struct page follow_devmap_pud(struct vm_area_struct vma,
df06b37f	430	unsigned long addr, pud_t pud, int flags, struct dev_pagemap *pgmap)
a00cc7d9 MW	431	{
	432	return NULL;
	433	}
9c670ea3 NH	434
	435	static inline bool thp_migration_supported(void)
	436	{
	437	return false;
	438	}
71e3aac0 AA	439	#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
71e3aac0 AA	440
346cf613 MWO	441	static inline int split_folio_to_list(struct folio *folio,
	442	struct list_head *list)
	443	{
	444	return split_huge_page_to_list(&folio->page, list);
	445	}
	446
d788f5b3 MWO	447	static inline int split_folio(struct folio *folio)
	448	{
	449	return split_folio_to_list(folio, NULL);
	450	}
	451
d0637c50 BS	452	/*
	453	* archs that select ARCH_WANTS_THP_SWAP but don't support THP_SWP due to
	454	* limitations in the implementation like arm64 MTE can override this to
	455	* false
	456	*/
	457	#ifndef arch_thp_swp_supported
	458	static inline bool arch_thp_swp_supported(void)
	459	{
	460	return true;
	461	}
	462	#endif
	463
71e3aac0	464	#endif /* _LINUX_HUGE_MM_H */