[deliverable/linux.git] / include / linux / pagemap.h

#ifndef _LINUX_PAGEMAP_H
#define _LINUX_PAGEMAP_H

/*
 * Copyright 1995 Linus Torvalds
 */
#include <linux/mm.h>
#include <linux/fs.h>
#include <linux/list.h>
#include <linux/highmem.h>
#include <linux/compiler.h>
#include <asm/uaccess.h>
#include <linux/gfp.h>
#include <linux/bitops.h>
#include <linux/hardirq.h> /* for in_interrupt() */

/*
 * Bits in mapping->flags.  The lower __GFP_BITS_SHIFT bits are the page
 * allocation mode flags.
 */
#define	AS_EIO		(__GFP_BITS_SHIFT + 0)	/* IO error on async write */
#define AS_ENOSPC	(__GFP_BITS_SHIFT + 1)	/* ENOSPC on async write */
#define AS_MM_ALL_LOCKS	(__GFP_BITS_SHIFT + 2)	/* under mm_take_all_locks() */

static inline void mapping_set_error(struct address_space *mapping, int error)
{
	if (unlikely(error)) {
		if (error == -ENOSPC)
			set_bit(AS_ENOSPC, &mapping->flags);
		else
			set_bit(AS_EIO, &mapping->flags);
	}
}

static inline gfp_t mapping_gfp_mask(struct address_space * mapping)
{
	return (__force gfp_t)mapping->flags & __GFP_BITS_MASK;
}

/*
 * This is non-atomic.  Only to be used before the mapping is activated.
 * Probably needs a barrier...
 */
static inline void mapping_set_gfp_mask(struct address_space *m, gfp_t mask)
{
	m->flags = (m->flags & ~(__force unsigned long)__GFP_BITS_MASK) |
				(__force unsigned long)mask;
}

/*
 * The page cache can done in larger chunks than
 * one page, because it allows for more efficient
 * throughput (it can then be mapped into user
 * space in smaller chunks for same flexibility).
 *
 * Or rather, it _will_ be done in larger chunks.
 */
#define PAGE_CACHE_SHIFT	PAGE_SHIFT
#define PAGE_CACHE_SIZE		PAGE_SIZE
#define PAGE_CACHE_MASK		PAGE_MASK
#define PAGE_CACHE_ALIGN(addr)	(((addr)+PAGE_CACHE_SIZE-1)&PAGE_CACHE_MASK)

#define page_cache_get(page)		get_page(page)
#define page_cache_release(page)	put_page(page)
void release_pages(struct page **pages, int nr, int cold);

/*
 * speculatively take a reference to a page.
 * If the page is free (_count == 0), then _count is untouched, and 0
 * is returned. Otherwise, _count is incremented by 1 and 1 is returned.
 *
 * This function must be called inside the same rcu_read_lock() section as has
 * been used to lookup the page in the pagecache radix-tree (or page table):
 * this allows allocators to use a synchronize_rcu() to stabilize _count.
 *
 * Unless an RCU grace period has passed, the count of all pages coming out
 * of the allocator must be considered unstable. page_count may return higher
 * than expected, and put_page must be able to do the right thing when the
 * page has been finished with, no matter what it is subsequently allocated
 * for (because put_page is what is used here to drop an invalid speculative
 * reference).
 *
 * This is the interesting part of the lockless pagecache (and lockless
 * get_user_pages) locking protocol, where the lookup-side (eg. find_get_page)
 * has the following pattern:
 * 1. find page in radix tree
 * 2. conditionally increment refcount
 * 3. check the page is still in pagecache (if no, goto 1)
 *
 * Remove-side that cares about stability of _count (eg. reclaim) has the
 * following (with tree_lock held for write):
 * A. atomically check refcount is correct and set it to 0 (atomic_cmpxchg)
 * B. remove page from pagecache
 * C. free the page
 *
 * There are 2 critical interleavings that matter:
 * - 2 runs before A: in this case, A sees elevated refcount and bails out
 * - A runs before 2: in this case, 2 sees zero refcount and retries;
 *   subsequently, B will complete and 1 will find no page, causing the
 *   lookup to return NULL.
 *
 * It is possible that between 1 and 2, the page is removed then the exact same
 * page is inserted into the same position in pagecache. That's OK: the
 * old find_get_page using tree_lock could equally have run before or after
 * such a re-insertion, depending on order that locks are granted.
 *
 * Lookups racing against pagecache insertion isn't a big problem: either 1
 * will find the page or it will not. Likewise, the old find_get_page could run
 * either before the insertion or afterwards, depending on timing.
 */
static inline int page_cache_get_speculative(struct page *page)
{
	VM_BUG_ON(in_interrupt());

#if !defined(CONFIG_SMP) && defined(CONFIG_CLASSIC_RCU)
# ifdef CONFIG_PREEMPT
	VM_BUG_ON(!in_atomic());
# endif
	/*
	 * Preempt must be disabled here - we rely on rcu_read_lock doing
	 * this for us.
	 *
	 * Pagecache won't be truncated from interrupt context, so if we have
	 * found a page in the radix tree here, we have pinned its refcount by
	 * disabling preempt, and hence no need for the "speculative get" that
	 * SMP requires.
	 */
	VM_BUG_ON(page_count(page) == 0);
	atomic_inc(&page->_count);

#else
	if (unlikely(!get_page_unless_zero(page))) {
		/*
		 * Either the page has been freed, or will be freed.
		 * In either case, retry here and the caller should
		 * do the right thing (see comments above).
		 */
		return 0;
	}
#endif
	VM_BUG_ON(PageTail(page));

	return 1;
}

/*
 * Same as above, but add instead of inc (could just be merged)
 */
static inline int page_cache_add_speculative(struct page *page, int count)
{
	VM_BUG_ON(in_interrupt());

#if !defined(CONFIG_SMP) && defined(CONFIG_CLASSIC_RCU)
# ifdef CONFIG_PREEMPT
	VM_BUG_ON(!in_atomic());
# endif
	VM_BUG_ON(page_count(page) == 0);
	atomic_add(count, &page->_count);

#else
	if (unlikely(!atomic_add_unless(&page->_count, count, 0)))
		return 0;
#endif
	VM_BUG_ON(PageCompound(page) && page != compound_head(page));

	return 1;
}

static inline int page_freeze_refs(struct page *page, int count)
{
	return likely(atomic_cmpxchg(&page->_count, count, 0) == count);
}

static inline void page_unfreeze_refs(struct page *page, int count)
{
	VM_BUG_ON(page_count(page) != 0);
	VM_BUG_ON(count == 0);

	atomic_set(&page->_count, count);
}

#ifdef CONFIG_NUMA
extern struct page *__page_cache_alloc(gfp_t gfp);
#else
static inline struct page *__page_cache_alloc(gfp_t gfp)
{
	return alloc_pages(gfp, 0);
}
#endif

static inline struct page *page_cache_alloc(struct address_space *x)
{
	return __page_cache_alloc(mapping_gfp_mask(x));
}

static inline struct page *page_cache_alloc_cold(struct address_space *x)
{
	return __page_cache_alloc(mapping_gfp_mask(x)|__GFP_COLD);
}

typedef int filler_t(void *, struct page *);

extern struct page * find_get_page(struct address_space *mapping,
				pgoff_t index);
extern struct page * find_lock_page(struct address_space *mapping,
				pgoff_t index);
extern struct page * find_or_create_page(struct address_space *mapping,
				pgoff_t index, gfp_t gfp_mask);
unsigned find_get_pages(struct address_space *mapping, pgoff_t start,
			unsigned int nr_pages, struct page **pages);
unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t start,
			       unsigned int nr_pages, struct page **pages);
unsigned find_get_pages_tag(struct address_space *mapping, pgoff_t *index,
			int tag, unsigned int nr_pages, struct page **pages);

struct page *__grab_cache_page(struct address_space *mapping, pgoff_t index);

/*
 * Returns locked page at given index in given cache, creating it if needed.
 */
static inline struct page *grab_cache_page(struct address_space *mapping,
								pgoff_t index)
{
	return find_or_create_page(mapping, index, mapping_gfp_mask(mapping));
}

extern struct page * grab_cache_page_nowait(struct address_space *mapping,
				pgoff_t index);
extern struct page * read_cache_page_async(struct address_space *mapping,
				pgoff_t index, filler_t *filler,
				void *data);
extern struct page * read_cache_page(struct address_space *mapping,
				pgoff_t index, filler_t *filler,
				void *data);
extern int read_cache_pages(struct address_space *mapping,
		struct list_head *pages, filler_t *filler, void *data);

static inline struct page *read_mapping_page_async(
						struct address_space *mapping,
						     pgoff_t index, void *data)
{
	filler_t *filler = (filler_t *)mapping->a_ops->readpage;
	return read_cache_page_async(mapping, index, filler, data);
}

static inline struct page *read_mapping_page(struct address_space *mapping,
					     pgoff_t index, void *data)
{
	filler_t *filler = (filler_t *)mapping->a_ops->readpage;
	return read_cache_page(mapping, index, filler, data);
}

/*
 * Return byte-offset into filesystem object for page.
 */
static inline loff_t page_offset(struct page *page)
{
	return ((loff_t)page->index) << PAGE_CACHE_SHIFT;
}

static inline pgoff_t linear_page_index(struct vm_area_struct *vma,
					unsigned long address)
{
	pgoff_t pgoff = (address - vma->vm_start) >> PAGE_SHIFT;
	pgoff += vma->vm_pgoff;
	return pgoff >> (PAGE_CACHE_SHIFT - PAGE_SHIFT);
}

extern void __lock_page(struct page *page);
extern int __lock_page_killable(struct page *page);
extern void __lock_page_nosync(struct page *page);
extern void unlock_page(struct page *page);

static inline void set_page_locked(struct page *page)
{
	set_bit(PG_locked, &page->flags);
}

static inline void clear_page_locked(struct page *page)
{
	clear_bit(PG_locked, &page->flags);
}

static inline int trylock_page(struct page *page)
{
	return !test_and_set_bit(PG_locked, &page->flags);
}

/*
 * lock_page may only be called if we have the page's inode pinned.
 */
static inline void lock_page(struct page *page)
{
	might_sleep();
	if (!trylock_page(page))
		__lock_page(page);
}

/*
 * lock_page_killable is like lock_page but can be interrupted by fatal
 * signals.  It returns 0 if it locked the page and -EINTR if it was
 * killed while waiting.
 */
static inline int lock_page_killable(struct page *page)
{
	might_sleep();
	if (!trylock_page(page))
		return __lock_page_killable(page);
	return 0;
}

/*
 * lock_page_nosync should only be used if we can't pin the page's inode.
 * Doesn't play quite so well with block device plugging.
 */
static inline void lock_page_nosync(struct page *page)
{
	might_sleep();
	if (!trylock_page(page))
		__lock_page_nosync(page);
}
	
/*
 * This is exported only for wait_on_page_locked/wait_on_page_writeback.
 * Never use this directly!
 */
extern void wait_on_page_bit(struct page *page, int bit_nr);

/* 
 * Wait for a page to be unlocked.
 *
 * This must be called with the caller "holding" the page,
 * ie with increased "page->count" so that the page won't
 * go away during the wait..
 */
static inline void wait_on_page_locked(struct page *page)
{
	if (PageLocked(page))
		wait_on_page_bit(page, PG_locked);
}

/* 
 * Wait for a page to complete writeback
 */
static inline void wait_on_page_writeback(struct page *page)
{
	if (PageWriteback(page))
		wait_on_page_bit(page, PG_writeback);
}

extern void end_page_writeback(struct page *page);

/*
 * Fault a userspace page into pagetables.  Return non-zero on a fault.
 *
 * This assumes that two userspace pages are always sufficient.  That's
 * not true if PAGE_CACHE_SIZE > PAGE_SIZE.
 */
static inline int fault_in_pages_writeable(char __user *uaddr, int size)
{
	int ret;

	if (unlikely(size == 0))
		return 0;

	/*
	 * Writing zeroes into userspace here is OK, because we know that if
	 * the zero gets there, we'll be overwriting it.
	 */
	ret = __put_user(0, uaddr);
	if (ret == 0) {
		char __user *end = uaddr + size - 1;

		/*
		 * If the page was already mapped, this will get a cache miss
		 * for sure, so try to avoid doing it.
		 */
		if (((unsigned long)uaddr & PAGE_MASK) !=
				((unsigned long)end & PAGE_MASK))
		 	ret = __put_user(0, end);
	}
	return ret;
}

static inline int fault_in_pages_readable(const char __user *uaddr, int size)
{
	volatile char c;
	int ret;

	if (unlikely(size == 0))
		return 0;

	ret = __get_user(c, uaddr);
	if (ret == 0) {
		const char __user *end = uaddr + size - 1;

		if (((unsigned long)uaddr & PAGE_MASK) !=
				((unsigned long)end & PAGE_MASK))
		 	ret = __get_user(c, end);
	}
	return ret;
}

int add_to_page_cache_locked(struct page *page, struct address_space *mapping,
				pgoff_t index, gfp_t gfp_mask);
int add_to_page_cache_lru(struct page *page, struct address_space *mapping,
				pgoff_t index, gfp_t gfp_mask);
extern void remove_from_page_cache(struct page *page);
extern void __remove_from_page_cache(struct page *page);

/*
 * Like add_to_page_cache_locked, but used to add newly allocated pages:
 * the page is new, so we can just run set_page_locked() against it.
 */
static inline int add_to_page_cache(struct page *page,
		struct address_space *mapping, pgoff_t offset, gfp_t gfp_mask)
{
	int error;

	set_page_locked(page);
	error = add_to_page_cache_locked(page, mapping, offset, gfp_mask);
	if (unlikely(error))
		clear_page_locked(page);
	return error;
}

#endif /* _LINUX_PAGEMAP_H */
Commit	Line	Data
1da177e4 LT	1	#ifndef _LINUX_PAGEMAP_H
	2	#define _LINUX_PAGEMAP_H
	3
	4	/*
	5	* Copyright 1995 Linus Torvalds
	6	*/
	7	#include <linux/mm.h>
	8	#include <linux/fs.h>
	9	#include <linux/list.h>
	10	#include <linux/highmem.h>
	11	#include <linux/compiler.h>
	12	#include <asm/uaccess.h>
	13	#include <linux/gfp.h>
3e9f45bd	14	#include <linux/bitops.h>
e286781d	15	#include <linux/hardirq.h> /* for in_interrupt() */
1da177e4 LT	16
	17	/*
	18	* Bits in mapping->flags. The lower __GFP_BITS_SHIFT bits are the page
	19	* allocation mode flags.
	20	*/
	21	#define AS_EIO (__GFP_BITS_SHIFT + 0) /* IO error on async write */
	22	#define AS_ENOSPC (__GFP_BITS_SHIFT + 1) /* ENOSPC on async write */
7906d00c	23	#define AS_MM_ALL_LOCKS (__GFP_BITS_SHIFT + 2) /* under mm_take_all_locks() */
1da177e4	24
3e9f45bd GC	25	static inline void mapping_set_error(struct address_space *mapping, int error)
3e9f45bd GC	26	{
2185e69f	27	if (unlikely(error)) {
3e9f45bd GC	28	if (error == -ENOSPC)
	29	set_bit(AS_ENOSPC, &mapping->flags);
	30	else
	31	set_bit(AS_EIO, &mapping->flags);
	32	}
	33	}
	34
dd0fc66f	35	static inline gfp_t mapping_gfp_mask(struct address_space * mapping)
1da177e4	36	{
260b2367	37	return (__force gfp_t)mapping->flags & __GFP_BITS_MASK;
1da177e4 LT	38	}
	39
	40	/*
	41	* This is non-atomic. Only to be used before the mapping is activated.
	42	* Probably needs a barrier...
	43	*/
260b2367	44	static inline void mapping_set_gfp_mask(struct address_space *m, gfp_t mask)
1da177e4	45	{
260b2367 AV	46	m->flags = (m->flags & ~(__force unsigned long)__GFP_BITS_MASK) \|
260b2367 AV	47	(__force unsigned long)mask;
1da177e4 LT	48	}
	49
	50	/*
	51	* The page cache can done in larger chunks than
	52	* one page, because it allows for more efficient
	53	* throughput (it can then be mapped into user
	54	* space in smaller chunks for same flexibility).
	55	*
	56	* Or rather, it _will_ be done in larger chunks.
	57	*/
	58	#define PAGE_CACHE_SHIFT PAGE_SHIFT
	59	#define PAGE_CACHE_SIZE PAGE_SIZE
	60	#define PAGE_CACHE_MASK PAGE_MASK
	61	#define PAGE_CACHE_ALIGN(addr) (((addr)+PAGE_CACHE_SIZE-1)&PAGE_CACHE_MASK)
	62
	63	#define page_cache_get(page) get_page(page)
	64	#define page_cache_release(page) put_page(page)
	65	void release_pages(struct page **pages, int nr, int cold);
	66
e286781d NP	67	/*
	68	* speculatively take a reference to a page.
	69	* If the page is free (_count == 0), then _count is untouched, and 0
	70	* is returned. Otherwise, _count is incremented by 1 and 1 is returned.
	71	*
	72	* This function must be called inside the same rcu_read_lock() section as has
	73	* been used to lookup the page in the pagecache radix-tree (or page table):
	74	* this allows allocators to use a synchronize_rcu() to stabilize _count.
	75	*
	76	* Unless an RCU grace period has passed, the count of all pages coming out
	77	* of the allocator must be considered unstable. page_count may return higher
	78	* than expected, and put_page must be able to do the right thing when the
	79	* page has been finished with, no matter what it is subsequently allocated
	80	* for (because put_page is what is used here to drop an invalid speculative
	81	* reference).
	82	*
	83	* This is the interesting part of the lockless pagecache (and lockless
	84	* get_user_pages) locking protocol, where the lookup-side (eg. find_get_page)
	85	* has the following pattern:
	86	* 1. find page in radix tree
	87	* 2. conditionally increment refcount
	88	* 3. check the page is still in pagecache (if no, goto 1)
	89	*
	90	* Remove-side that cares about stability of _count (eg. reclaim) has the
	91	* following (with tree_lock held for write):
	92	* A. atomically check refcount is correct and set it to 0 (atomic_cmpxchg)
	93	* B. remove page from pagecache
	94	* C. free the page
	95	*
	96	* There are 2 critical interleavings that matter:
	97	* - 2 runs before A: in this case, A sees elevated refcount and bails out
	98	* - A runs before 2: in this case, 2 sees zero refcount and retries;
	99	* subsequently, B will complete and 1 will find no page, causing the
	100	* lookup to return NULL.
	101	*
	102	* It is possible that between 1 and 2, the page is removed then the exact same
	103	* page is inserted into the same position in pagecache. That's OK: the
	104	* old find_get_page using tree_lock could equally have run before or after
	105	* such a re-insertion, depending on order that locks are granted.
	106	*
	107	* Lookups racing against pagecache insertion isn't a big problem: either 1
	108	* will find the page or it will not. Likewise, the old find_get_page could run
	109	* either before the insertion or afterwards, depending on timing.
	110	*/
	111	static inline int page_cache_get_speculative(struct page *page)
	112	{
	113	VM_BUG_ON(in_interrupt());
	114
	115	#if !defined(CONFIG_SMP) && defined(CONFIG_CLASSIC_RCU)
	116	# ifdef CONFIG_PREEMPT
	117	VM_BUG_ON(!in_atomic());
	118	# endif
	119	/*
	120	* Preempt must be disabled here - we rely on rcu_read_lock doing
	121	* this for us.
	122	*
	123	* Pagecache won't be truncated from interrupt context, so if we have
	124	* found a page in the radix tree here, we have pinned its refcount by
	125	* disabling preempt, and hence no need for the "speculative get" that
	126	* SMP requires.
	127	*/
	128	VM_BUG_ON(page_count(page) == 0);
	129	atomic_inc(&page->_count);
	130
131	#else
132	if (unlikely(!get_page_unless_zero(page))) {
133	/*
134	* Either the page has been freed, or will be freed.
135	* In either case, retry here and the caller should
136	* do the right thing (see comments above).
137	*/
138	return 0;
139	}
140	#endif
141	VM_BUG_ON(PageTail(page));
142
143	return 1;
144	}
145
ce0ad7f0 NP	146	/*
	147	* Same as above, but add instead of inc (could just be merged)
	148	*/
	149	static inline int page_cache_add_speculative(struct page *page, int count)
	150	{
	151	VM_BUG_ON(in_interrupt());
	152
	153	#if !defined(CONFIG_SMP) && defined(CONFIG_CLASSIC_RCU)
	154	# ifdef CONFIG_PREEMPT
	155	VM_BUG_ON(!in_atomic());
	156	# endif
	157	VM_BUG_ON(page_count(page) == 0);
	158	atomic_add(count, &page->_count);
	159
	160	#else
	161	if (unlikely(!atomic_add_unless(&page->_count, count, 0)))
	162	return 0;
	163	#endif
	164	VM_BUG_ON(PageCompound(page) && page != compound_head(page));
	165
	166	return 1;
	167	}
	168
e286781d NP	169	static inline int page_freeze_refs(struct page *page, int count)
	170	{
	171	return likely(atomic_cmpxchg(&page->_count, count, 0) == count);
	172	}
	173
	174	static inline void page_unfreeze_refs(struct page *page, int count)
	175	{
	176	VM_BUG_ON(page_count(page) != 0);
	177	VM_BUG_ON(count == 0);
	178
	179	atomic_set(&page->_count, count);
	180	}
	181
44110fe3	182	#ifdef CONFIG_NUMA
2ae88149	183	extern struct page *__page_cache_alloc(gfp_t gfp);
44110fe3	184	#else
2ae88149 NP	185	static inline struct page *__page_cache_alloc(gfp_t gfp)
	186	{
	187	return alloc_pages(gfp, 0);
	188	}
	189	#endif
	190
1da177e4 LT	191	static inline struct page page_cache_alloc(struct address_space x)
1da177e4 LT	192	{
2ae88149	193	return __page_cache_alloc(mapping_gfp_mask(x));
1da177e4 LT	194	}
	195
	196	static inline struct page page_cache_alloc_cold(struct address_space x)
	197	{
2ae88149	198	return __page_cache_alloc(mapping_gfp_mask(x)\|__GFP_COLD);
1da177e4 LT	199	}
	200
	201	typedef int filler_t(void , struct page );
	202
	203	extern struct page * find_get_page(struct address_space *mapping,
57f6b96c	204	pgoff_t index);
1da177e4	205	extern struct page * find_lock_page(struct address_space *mapping,
57f6b96c	206	pgoff_t index);
1da177e4	207	extern struct page * find_or_create_page(struct address_space *mapping,
57f6b96c	208	pgoff_t index, gfp_t gfp_mask);
1da177e4 LT	209	unsigned find_get_pages(struct address_space *mapping, pgoff_t start,
1da177e4 LT	210	unsigned int nr_pages, struct page **pages);
ebf43500 JA	211	unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t start,
ebf43500 JA	212	unsigned int nr_pages, struct page **pages);
1da177e4 LT	213	unsigned find_get_pages_tag(struct address_space mapping, pgoff_t index,
	214	int tag, unsigned int nr_pages, struct page **pages);
	215
afddba49 NP	216	struct page __grab_cache_page(struct address_space mapping, pgoff_t index);
afddba49 NP	217
1da177e4 LT	218	/*
	219	* Returns locked page at given index in given cache, creating it if needed.
	220	*/
57f6b96c FW	221	static inline struct page grab_cache_page(struct address_space mapping,
57f6b96c FW	222	pgoff_t index)
1da177e4 LT	223	{
	224	return find_or_create_page(mapping, index, mapping_gfp_mask(mapping));
	225	}
	226
	227	extern struct page * grab_cache_page_nowait(struct address_space *mapping,
57f6b96c	228	pgoff_t index);
6fe6900e	229	extern struct page * read_cache_page_async(struct address_space *mapping,
57f6b96c	230	pgoff_t index, filler_t *filler,
6fe6900e	231	void *data);
1da177e4	232	extern struct page * read_cache_page(struct address_space *mapping,
57f6b96c	233	pgoff_t index, filler_t *filler,
1da177e4 LT	234	void *data);
	235	extern int read_cache_pages(struct address_space *mapping,
	236	struct list_head pages, filler_t filler, void *data);
	237
6fe6900e NP	238	static inline struct page *read_mapping_page_async(
6fe6900e NP	239	struct address_space *mapping,
57f6b96c	240	pgoff_t index, void *data)
6fe6900e NP	241	{
	242	filler_t filler = (filler_t )mapping->a_ops->readpage;
	243	return read_cache_page_async(mapping, index, filler, data);
	244	}
	245
090d2b18	246	static inline struct page read_mapping_page(struct address_space mapping,
57f6b96c	247	pgoff_t index, void *data)
090d2b18 PE	248	{
	249	filler_t filler = (filler_t )mapping->a_ops->readpage;
	250	return read_cache_page(mapping, index, filler, data);
	251	}
	252
1da177e4 LT	253	/*
	254	* Return byte-offset into filesystem object for page.
	255	*/
	256	static inline loff_t page_offset(struct page *page)
	257	{
	258	return ((loff_t)page->index) << PAGE_CACHE_SHIFT;
	259	}
	260
	261	static inline pgoff_t linear_page_index(struct vm_area_struct *vma,
	262	unsigned long address)
	263	{
	264	pgoff_t pgoff = (address - vma->vm_start) >> PAGE_SHIFT;
	265	pgoff += vma->vm_pgoff;
	266	return pgoff >> (PAGE_CACHE_SHIFT - PAGE_SHIFT);
	267	}
	268
b3c97528 HH	269	extern void __lock_page(struct page *page);
	270	extern int __lock_page_killable(struct page *page);
	271	extern void __lock_page_nosync(struct page *page);
	272	extern void unlock_page(struct page *page);
1da177e4	273
529ae9aa NP	274	static inline void set_page_locked(struct page *page)
	275	{
	276	set_bit(PG_locked, &page->flags);
	277	}
	278
	279	static inline void clear_page_locked(struct page *page)
	280	{
	281	clear_bit(PG_locked, &page->flags);
	282	}
	283
	284	static inline int trylock_page(struct page *page)
	285	{
	286	return !test_and_set_bit(PG_locked, &page->flags);
	287	}
	288
db37648c NP	289	/*
	290	* lock_page may only be called if we have the page's inode pinned.
	291	*/
1da177e4 LT	292	static inline void lock_page(struct page *page)
	293	{
	294	might_sleep();
529ae9aa	295	if (!trylock_page(page))
1da177e4 LT	296	__lock_page(page);
1da177e4 LT	297	}
db37648c	298
2687a356 MW	299	/*
	300	* lock_page_killable is like lock_page but can be interrupted by fatal
	301	* signals. It returns 0 if it locked the page and -EINTR if it was
	302	* killed while waiting.
	303	*/
	304	static inline int lock_page_killable(struct page *page)
	305	{
	306	might_sleep();
529ae9aa	307	if (!trylock_page(page))
2687a356 MW	308	return __lock_page_killable(page);
	309	return 0;
	310	}
	311
db37648c NP	312	/*
	313	* lock_page_nosync should only be used if we can't pin the page's inode.
	314	* Doesn't play quite so well with block device plugging.
	315	*/
	316	static inline void lock_page_nosync(struct page *page)
	317	{
	318	might_sleep();
529ae9aa	319	if (!trylock_page(page))
db37648c NP	320	__lock_page_nosync(page);
db37648c NP	321	}
1da177e4 LT	322
	323	/*
	324	* This is exported only for wait_on_page_locked/wait_on_page_writeback.
	325	* Never use this directly!
	326	*/
b3c97528	327	extern void wait_on_page_bit(struct page *page, int bit_nr);
1da177e4 LT	328
	329	/*
	330	* Wait for a page to be unlocked.
	331	*
	332	* This must be called with the caller "holding" the page,
	333	* ie with increased "page->count" so that the page won't
	334	* go away during the wait..
	335	*/
	336	static inline void wait_on_page_locked(struct page *page)
	337	{
	338	if (PageLocked(page))
	339	wait_on_page_bit(page, PG_locked);
	340	}
	341
	342	/*
	343	* Wait for a page to complete writeback
	344	*/
	345	static inline void wait_on_page_writeback(struct page *page)
	346	{
	347	if (PageWriteback(page))
	348	wait_on_page_bit(page, PG_writeback);
	349	}
	350
	351	extern void end_page_writeback(struct page *page);
	352
	353	/*
	354	* Fault a userspace page into pagetables. Return non-zero on a fault.
	355	*
	356	* This assumes that two userspace pages are always sufficient. That's
	357	* not true if PAGE_CACHE_SIZE > PAGE_SIZE.
	358	*/
	359	static inline int fault_in_pages_writeable(char __user *uaddr, int size)
	360	{
	361	int ret;
	362
08291429 NP	363	if (unlikely(size == 0))
	364	return 0;
	365
1da177e4 LT	366	/*
	367	* Writing zeroes into userspace here is OK, because we know that if
	368	* the zero gets there, we'll be overwriting it.
	369	*/
	370	ret = __put_user(0, uaddr);
	371	if (ret == 0) {
	372	char __user *end = uaddr + size - 1;
	373
	374	/*
	375	* If the page was already mapped, this will get a cache miss
	376	* for sure, so try to avoid doing it.
	377	*/
	378	if (((unsigned long)uaddr & PAGE_MASK) !=
	379	((unsigned long)end & PAGE_MASK))
	380	ret = __put_user(0, end);
	381	}
	382	return ret;
	383	}
	384
08291429	385	static inline int fault_in_pages_readable(const char __user *uaddr, int size)
1da177e4 LT	386	{
	387	volatile char c;
	388	int ret;
	389
08291429 NP	390	if (unlikely(size == 0))
	391	return 0;
	392
1da177e4 LT	393	ret = __get_user(c, uaddr);
	394	if (ret == 0) {
	395	const char __user *end = uaddr + size - 1;
	396
	397	if (((unsigned long)uaddr & PAGE_MASK) !=
	398	((unsigned long)end & PAGE_MASK))
08291429	399	ret = __get_user(c, end);
1da177e4	400	}
08291429	401	return ret;
1da177e4 LT	402	}
1da177e4 LT	403
529ae9aa NP	404	int add_to_page_cache_locked(struct page page, struct address_space mapping,
	405	pgoff_t index, gfp_t gfp_mask);
	406	int add_to_page_cache_lru(struct page page, struct address_space mapping,
	407	pgoff_t index, gfp_t gfp_mask);
	408	extern void remove_from_page_cache(struct page *page);
	409	extern void __remove_from_page_cache(struct page *page);
	410
	411	/*
	412	* Like add_to_page_cache_locked, but used to add newly allocated pages:
	413	* the page is new, so we can just run set_page_locked() against it.
	414	*/
	415	static inline int add_to_page_cache(struct page *page,
	416	struct address_space *mapping, pgoff_t offset, gfp_t gfp_mask)
	417	{
	418	int error;
	419
	420	set_page_locked(page);
	421	error = add_to_page_cache_locked(page, mapping, offset, gfp_mask);
	422	if (unlikely(error))
	423	clear_page_locked(page);
	424	return error;
	425	}
	426
1da177e4	427	#endif /* _LINUX_PAGEMAP_H */