[people/ms/linux.git] / fs / xfs / xfs_buf.h

// SPDX-License-Identifier: GPL-2.0
/*
 * Copyright (c) 2000-2005 Silicon Graphics, Inc.
 * All Rights Reserved.
 */
#ifndef __XFS_BUF_H__
#define __XFS_BUF_H__

#include <linux/list.h>
#include <linux/types.h>
#include <linux/spinlock.h>
#include <linux/mm.h>
#include <linux/fs.h>
#include <linux/dax.h>
#include <linux/uio.h>
#include <linux/list_lru.h>

/*
 *	Base types
 */

#define XFS_BUF_DADDR_NULL	((xfs_daddr_t) (-1LL))

typedef enum {
	XBRW_READ = 1,			/* transfer into target memory */
	XBRW_WRITE = 2,			/* transfer from target memory */
	XBRW_ZERO = 3,			/* Zero target memory */
} xfs_buf_rw_t;

#define XBF_READ	 (1 << 0) /* buffer intended for reading from device */
#define XBF_WRITE	 (1 << 1) /* buffer intended for writing to device */
#define XBF_READ_AHEAD	 (1 << 2) /* asynchronous read-ahead */
#define XBF_NO_IOACCT	 (1 << 3) /* bypass I/O accounting (non-LRU bufs) */
#define XBF_ASYNC	 (1 << 4) /* initiator will not wait for completion */
#define XBF_DONE	 (1 << 5) /* all pages in the buffer uptodate */
#define XBF_STALE	 (1 << 6) /* buffer has been staled, do not find it */
#define XBF_WRITE_FAIL	 (1 << 24)/* async writes have failed on this buffer */

/* I/O hints for the BIO layer */
#define XBF_SYNCIO	 (1 << 10)/* treat this buffer as synchronous I/O */
#define XBF_FUA		 (1 << 11)/* force cache write through mode */
#define XBF_FLUSH	 (1 << 12)/* flush the disk cache before a write */

/* flags used only as arguments to access routines */
#define XBF_TRYLOCK	 (1 << 16)/* lock requested, but do not wait */
#define XBF_UNMAPPED	 (1 << 17)/* do not map the buffer */

/* flags used only internally */
#define _XBF_PAGES	 (1 << 20)/* backed by refcounted pages */
#define _XBF_KMEM	 (1 << 21)/* backed by heap memory */
#define _XBF_DELWRI_Q	 (1 << 22)/* buffer on a delwri queue */
#define _XBF_COMPOUND	 (1 << 23)/* compound buffer */

typedef unsigned int xfs_buf_flags_t;

#define XFS_BUF_FLAGS \
	{ XBF_READ,		"READ" }, \
	{ XBF_WRITE,		"WRITE" }, \
	{ XBF_READ_AHEAD,	"READ_AHEAD" }, \
	{ XBF_NO_IOACCT,	"NO_IOACCT" }, \
	{ XBF_ASYNC,		"ASYNC" }, \
	{ XBF_DONE,		"DONE" }, \
	{ XBF_STALE,		"STALE" }, \
	{ XBF_WRITE_FAIL,	"WRITE_FAIL" }, \
	{ XBF_SYNCIO,		"SYNCIO" }, \
	{ XBF_FUA,		"FUA" }, \
	{ XBF_FLUSH,		"FLUSH" }, \
	{ XBF_TRYLOCK,		"TRYLOCK" },	/* should never be set */\
	{ XBF_UNMAPPED,		"UNMAPPED" },	/* ditto */\
	{ _XBF_PAGES,		"PAGES" }, \
	{ _XBF_KMEM,		"KMEM" }, \
	{ _XBF_DELWRI_Q,	"DELWRI_Q" }, \
	{ _XBF_COMPOUND,	"COMPOUND" }


/*
 * Internal state flags.
 */
#define XFS_BSTATE_DISPOSE	 (1 << 0)	/* buffer being discarded */
#define XFS_BSTATE_IN_FLIGHT	 (1 << 1)	/* I/O in flight */

/*
 * The xfs_buftarg contains 2 notions of "sector size" -
 *
 * 1) The metadata sector size, which is the minimum unit and
 *    alignment of IO which will be performed by metadata operations.
 * 2) The device logical sector size
 *
 * The first is specified at mkfs time, and is stored on-disk in the
 * superblock's sb_sectsize.
 *
 * The latter is derived from the underlying device, and controls direct IO
 * alignment constraints.
 */
typedef struct xfs_buftarg {
	dev_t			bt_dev;
	struct block_device	*bt_bdev;
	struct dax_device	*bt_daxdev;
	struct xfs_mount	*bt_mount;
	unsigned int		bt_meta_sectorsize;
	size_t			bt_meta_sectormask;
	size_t			bt_logical_sectorsize;
	size_t			bt_logical_sectormask;

	/* LRU control structures */
	struct shrinker		bt_shrinker;
	struct list_lru		bt_lru;

	struct percpu_counter	bt_io_count;
} xfs_buftarg_t;

struct xfs_buf;
typedef void (*xfs_buf_iodone_t)(struct xfs_buf *);


#define XB_PAGES	2

struct xfs_buf_map {
	xfs_daddr_t		bm_bn;	/* block number for I/O */
	int			bm_len;	/* size of I/O */
};

#define DEFINE_SINGLE_BUF_MAP(map, blkno, numblk) \
	struct xfs_buf_map (map) = { .bm_bn = (blkno), .bm_len = (numblk) };

struct xfs_buf_ops {
	char *name;
	union {
		__be32 magic[2];	/* v4 and v5 on disk magic values */
		__be16 magic16[2];	/* v4 and v5 on disk magic values */
	};
	void (*verify_read)(struct xfs_buf *);
	void (*verify_write)(struct xfs_buf *);
	xfs_failaddr_t (*verify_struct)(struct xfs_buf *bp);
};

typedef struct xfs_buf {
	/*
	 * first cacheline holds all the fields needed for an uncontended cache
	 * hit to be fully processed. The semaphore straddles the cacheline
	 * boundary, but the counter and lock sits on the first cacheline,
	 * which is the only bit that is touched if we hit the semaphore
	 * fast-path on locking.
	 */
	struct rhash_head	b_rhash_head;	/* pag buffer hash node */
	xfs_daddr_t		b_bn;		/* block number of buffer */
	int			b_length;	/* size of buffer in BBs */
	atomic_t		b_hold;		/* reference count */
	atomic_t		b_lru_ref;	/* lru reclaim ref count */
	xfs_buf_flags_t		b_flags;	/* status flags */
	struct semaphore	b_sema;		/* semaphore for lockables */

	/*
	 * concurrent access to b_lru and b_lru_flags are protected by
	 * bt_lru_lock and not by b_sema
	 */
	struct list_head	b_lru;		/* lru list */
	spinlock_t		b_lock;		/* internal state lock */
	unsigned int		b_state;	/* internal state flags */
	int			b_io_error;	/* internal IO error state */
	wait_queue_head_t	b_waiters;	/* unpin waiters */
	struct list_head	b_list;
	struct xfs_perag	*b_pag;		/* contains rbtree root */
	xfs_buftarg_t		*b_target;	/* buffer target (device) */
	void			*b_addr;	/* virtual address of buffer */
	struct work_struct	b_ioend_work;
	struct workqueue_struct	*b_ioend_wq;	/* I/O completion wq */
	xfs_buf_iodone_t	b_iodone;	/* I/O completion function */
	struct completion	b_iowait;	/* queue for I/O waiters */
	void			*b_log_item;
	struct list_head	b_li_list;	/* Log items list head */
	struct xfs_trans	*b_transp;
	struct page		**b_pages;	/* array of page pointers */
	struct page		*b_page_array[XB_PAGES]; /* inline pages */
	struct xfs_buf_map	*b_maps;	/* compound buffer map */
	struct xfs_buf_map	__b_map;	/* inline compound buffer map */
	int			b_map_count;
	int			b_io_length;	/* IO size in BBs */
	atomic_t		b_pin_count;	/* pin count */
	atomic_t		b_io_remaining;	/* #outstanding I/O requests */
	unsigned int		b_page_count;	/* size of page array */
	unsigned int		b_offset;	/* page offset in first page */
	int			b_error;	/* error code on I/O */

	/*
	 * async write failure retry count. Initialised to zero on the first
	 * failure, then when it exceeds the maximum configured without a
	 * success the write is considered to be failed permanently and the
	 * iodone handler will take appropriate action.
	 *
	 * For retry timeouts, we record the jiffie of the first failure. This
	 * means that we can change the retry timeout for buffers already under
	 * I/O and thus avoid getting stuck in a retry loop with a long timeout.
	 *
	 * last_error is used to ensure that we are getting repeated errors, not
	 * different errors. e.g. a block device might change ENOSPC to EIO when
	 * a failure timeout occurs, so we want to re-initialise the error
	 * retry behaviour appropriately when that happens.
	 */
	int			b_retries;
	unsigned long		b_first_retry_time; /* in jiffies */
	int			b_last_error;

	const struct xfs_buf_ops	*b_ops;
} xfs_buf_t;

/* Finding and Reading Buffers */
struct xfs_buf *xfs_buf_incore(struct xfs_buftarg *target,
			   xfs_daddr_t blkno, size_t numblks,
			   xfs_buf_flags_t flags);

struct xfs_buf *_xfs_buf_alloc(struct xfs_buftarg *target,
			       struct xfs_buf_map *map, int nmaps,
			       xfs_buf_flags_t flags);

static inline struct xfs_buf *
xfs_buf_alloc(
	struct xfs_buftarg	*target,
	xfs_daddr_t		blkno,
	size_t			numblks,
	xfs_buf_flags_t		flags)
{
	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	return _xfs_buf_alloc(target, &map, 1, flags);
}

struct xfs_buf *xfs_buf_get_map(struct xfs_buftarg *target,
			       struct xfs_buf_map *map, int nmaps,
			       xfs_buf_flags_t flags);
struct xfs_buf *xfs_buf_read_map(struct xfs_buftarg *target,
			       struct xfs_buf_map *map, int nmaps,
			       xfs_buf_flags_t flags,
			       const struct xfs_buf_ops *ops);
void xfs_buf_readahead_map(struct xfs_buftarg *target,
			       struct xfs_buf_map *map, int nmaps,
			       const struct xfs_buf_ops *ops);

static inline struct xfs_buf *
xfs_buf_get(
	struct xfs_buftarg	*target,
	xfs_daddr_t		blkno,
	size_t			numblks,
	xfs_buf_flags_t		flags)
{
	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	return xfs_buf_get_map(target, &map, 1, flags);
}

static inline struct xfs_buf *
xfs_buf_read(
	struct xfs_buftarg	*target,
	xfs_daddr_t		blkno,
	size_t			numblks,
	xfs_buf_flags_t		flags,
	const struct xfs_buf_ops *ops)
{
	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	return xfs_buf_read_map(target, &map, 1, flags, ops);
}

static inline void
xfs_buf_readahead(
	struct xfs_buftarg	*target,
	xfs_daddr_t		blkno,
	size_t			numblks,
	const struct xfs_buf_ops *ops)
{
	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	return xfs_buf_readahead_map(target, &map, 1, ops);
}

void xfs_buf_set_empty(struct xfs_buf *bp, size_t numblks);
int xfs_buf_associate_memory(struct xfs_buf *bp, void *mem, size_t length);

struct xfs_buf *xfs_buf_get_uncached(struct xfs_buftarg *target, size_t numblks,
				int flags);
int xfs_buf_read_uncached(struct xfs_buftarg *target, xfs_daddr_t daddr,
			  size_t numblks, int flags, struct xfs_buf **bpp,
			  const struct xfs_buf_ops *ops);
void xfs_buf_hold(struct xfs_buf *bp);

/* Releasing Buffers */
extern void xfs_buf_free(xfs_buf_t *);
extern void xfs_buf_rele(xfs_buf_t *);

/* Locking and Unlocking Buffers */
extern int xfs_buf_trylock(xfs_buf_t *);
extern void xfs_buf_lock(xfs_buf_t *);
extern void xfs_buf_unlock(xfs_buf_t *);
#define xfs_buf_islocked(bp) \
	((bp)->b_sema.count <= 0)

/* Buffer Read and Write Routines */
extern int xfs_bwrite(struct xfs_buf *bp);
extern void xfs_buf_ioend(struct xfs_buf *bp);
extern void __xfs_buf_ioerror(struct xfs_buf *bp, int error,
		xfs_failaddr_t failaddr);
#define xfs_buf_ioerror(bp, err) __xfs_buf_ioerror((bp), (err), __this_address)
extern void xfs_buf_ioerror_alert(struct xfs_buf *, const char *func);

extern int __xfs_buf_submit(struct xfs_buf *bp, bool);
static inline int xfs_buf_submit(struct xfs_buf *bp)
{
	bool wait = bp->b_flags & XBF_ASYNC ? false : true;
	return __xfs_buf_submit(bp, wait);
}

extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, void *,
				xfs_buf_rw_t);
#define xfs_buf_zero(bp, off, len) \
	    xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)

/* Buffer Utility Routines */
extern void *xfs_buf_offset(struct xfs_buf *, size_t);
extern void xfs_buf_stale(struct xfs_buf *bp);

/* Delayed Write Buffer Routines */
extern void xfs_buf_delwri_cancel(struct list_head *);
extern bool xfs_buf_delwri_queue(struct xfs_buf *, struct list_head *);
extern int xfs_buf_delwri_submit(struct list_head *);
extern int xfs_buf_delwri_submit_nowait(struct list_head *);
extern int xfs_buf_delwri_pushbuf(struct xfs_buf *, struct list_head *);

/* Buffer Daemon Setup Routines */
extern int xfs_buf_init(void);
extern void xfs_buf_terminate(void);

/*
 * These macros use the IO block map rather than b_bn. b_bn is now really
 * just for the buffer cache index for cached buffers. As IO does not use b_bn
 * anymore, uncached buffers do not use b_bn at all and hence must modify the IO
 * map directly. Uncached buffers are not allowed to be discontiguous, so this
 * is safe to do.
 *
 * In future, uncached buffers will pass the block number directly to the io
 * request function and hence these macros will go away at that point.
 */
#define XFS_BUF_ADDR(bp)		((bp)->b_maps[0].bm_bn)
#define XFS_BUF_SET_ADDR(bp, bno)	((bp)->b_maps[0].bm_bn = (xfs_daddr_t)(bno))

void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref);

/*
 * If the buffer is already on the LRU, do nothing. Otherwise set the buffer
 * up with a reference count of 0 so it will be tossed from the cache when
 * released.
 */
static inline void xfs_buf_oneshot(struct xfs_buf *bp)
{
	if (!list_empty(&bp->b_lru) || atomic_read(&bp->b_lru_ref) > 1)
		return;
	atomic_set(&bp->b_lru_ref, 0);
}

static inline int xfs_buf_ispinned(struct xfs_buf *bp)
{
	return atomic_read(&bp->b_pin_count);
}

static inline void xfs_buf_relse(xfs_buf_t *bp)
{
	xfs_buf_unlock(bp);
	xfs_buf_rele(bp);
}

static inline int
xfs_buf_verify_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
{
	return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
				cksum_offset);
}

static inline void
xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
{
	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
			 cksum_offset);
}

/*
 *	Handling of buftargs.
 */
extern xfs_buftarg_t *xfs_alloc_buftarg(struct xfs_mount *,
			struct block_device *, struct dax_device *);
extern void xfs_free_buftarg(struct xfs_buftarg *);
extern void xfs_wait_buftarg(xfs_buftarg_t *);
extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int);

#define xfs_getsize_buftarg(buftarg)	block_size((buftarg)->bt_bdev)
#define xfs_readonly_buftarg(buftarg)	bdev_read_only((buftarg)->bt_bdev)

int xfs_buf_reverify(struct xfs_buf *bp, const struct xfs_buf_ops *ops);
bool xfs_verify_magic(struct xfs_buf *bp, __be32 dmagic);
bool xfs_verify_magic16(struct xfs_buf *bp, __be16 dmagic);

#endif	/* __XFS_BUF_H__ */
Commit	Line	Data
0b61f8a4	1	// SPDX-License-Identifier: GPL-2.0
1da177e4	2	/*
7b718769 NS	3	* Copyright (c) 2000-2005 Silicon Graphics, Inc.
7b718769 NS	4	* All Rights Reserved.
1da177e4	5	*/
1da177e4 LT	6	#ifndef __XFS_BUF_H__
	7	#define __XFS_BUF_H__
	8
1da177e4 LT	9	#include <linux/list.h>
	10	#include <linux/types.h>
	11	#include <linux/spinlock.h>
1da177e4 LT	12	#include <linux/mm.h>
1da177e4 LT	13	#include <linux/fs.h>
c94c2acf	14	#include <linux/dax.h>
1da177e4	15	#include <linux/uio.h>
e80dfa19	16	#include <linux/list_lru.h>
1da177e4 LT	17
	18	/*
	19	* Base types
	20	*/
	21
ce8e922c NS	22	#define XFS_BUF_DADDR_NULL ((xfs_daddr_t) (-1LL))
ce8e922c NS	23
ce8e922c NS	24	typedef enum {
	25	XBRW_READ = 1, /* transfer into target memory */
	26	XBRW_WRITE = 2, /* transfer from target memory */
	27	XBRW_ZERO = 3, /* Zero target memory */
	28	} xfs_buf_rw_t;
	29
6fb8a90a CM	30	#define XBF_READ (1 << 0) /* buffer intended for reading from device */
	31	#define XBF_WRITE (1 << 1) /* buffer intended for writing to device */
	32	#define XBF_READ_AHEAD (1 << 2) /* asynchronous read-ahead */
c891c30a	33	#define XBF_NO_IOACCT (1 << 3) /* bypass I/O accounting (non-LRU bufs) */
6fb8a90a CM	34	#define XBF_ASYNC (1 << 4) /* initiator will not wait for completion */
	35	#define XBF_DONE (1 << 5) /* all pages in the buffer uptodate */
	36	#define XBF_STALE (1 << 6) /* buffer has been staled, do not find it */
ac8809f9	37	#define XBF_WRITE_FAIL (1 << 24)/* async writes have failed on this buffer */
1d5ae5df CH	38
1d5ae5df CH	39	/* I/O hints for the BIO layer */
6fb8a90a CM	40	#define XBF_SYNCIO (1 << 10)/* treat this buffer as synchronous I/O */
	41	#define XBF_FUA (1 << 11)/* force cache write through mode */
	42	#define XBF_FLUSH (1 << 12)/* flush the disk cache before a write */
1da177e4	43
807cbbdb	44	/* flags used only as arguments to access routines */
6fb8a90a CM	45	#define XBF_TRYLOCK (1 << 16)/* lock requested, but do not wait */
6fb8a90a CM	46	#define XBF_UNMAPPED (1 << 17)/* do not map the buffer */
1da177e4	47
807cbbdb	48	/* flags used only internally */
6fb8a90a CM	49	#define _XBF_PAGES (1 << 20)/* backed by refcounted pages */
	50	#define _XBF_KMEM (1 << 21)/* backed by heap memory */
	51	#define _XBF_DELWRI_Q (1 << 22)/* buffer on a delwri queue */
	52	#define _XBF_COMPOUND (1 << 23)/* compound buffer */
6ab455ee	53
807cbbdb	54	typedef unsigned int xfs_buf_flags_t;
1da177e4	55
0b1b213f CH	56	#define XFS_BUF_FLAGS \
	57	{ XBF_READ, "READ" }, \
	58	{ XBF_WRITE, "WRITE" }, \
1d5ae5df	59	{ XBF_READ_AHEAD, "READ_AHEAD" }, \
1247ec4c	60	{ XBF_NO_IOACCT, "NO_IOACCT" }, \
0b1b213f CH	61	{ XBF_ASYNC, "ASYNC" }, \
0b1b213f CH	62	{ XBF_DONE, "DONE" }, \
0b1b213f	63	{ XBF_STALE, "STALE" }, \
ac8809f9	64	{ XBF_WRITE_FAIL, "WRITE_FAIL" }, \
1d5ae5df CH	65	{ XBF_SYNCIO, "SYNCIO" }, \
	66	{ XBF_FUA, "FUA" }, \
	67	{ XBF_FLUSH, "FLUSH" }, \
6fb8a90a	68	{ XBF_TRYLOCK, "TRYLOCK" }, /* should never be set */\
611c9946	69	{ XBF_UNMAPPED, "UNMAPPED" }, /* ditto */\
0b1b213f	70	{ _XBF_PAGES, "PAGES" }, \
0e6e847f	71	{ _XBF_KMEM, "KMEM" }, \
cbb7baab	72	{ _XBF_DELWRI_Q, "DELWRI_Q" }, \
63db7c81	73	{ _XBF_COMPOUND, "COMPOUND" }
a4082357	74
ac8809f9	75
a4082357 DC	76	/*
	77	* Internal state flags.
	78	*/
	79	#define XFS_BSTATE_DISPOSE (1 << 0) /* buffer being discarded */
63db7c81	80	#define XFS_BSTATE_IN_FLIGHT (1 << 1) /* I/O in flight */
0b1b213f	81
7c71ee78 ES	82	/*
	83	* The xfs_buftarg contains 2 notions of "sector size" -
	84	*
	85	* 1) The metadata sector size, which is the minimum unit and
	86	* alignment of IO which will be performed by metadata operations.
	87	* 2) The device logical sector size
	88	*
	89	* The first is specified at mkfs time, and is stored on-disk in the
	90	* superblock's sb_sectsize.
	91	*
	92	* The latter is derived from the underlying device, and controls direct IO
	93	* alignment constraints.
	94	*/
1da177e4	95	typedef struct xfs_buftarg {
ce8e922c NS	96	dev_t bt_dev;
ce8e922c NS	97	struct block_device *bt_bdev;
486aff5e	98	struct dax_device *bt_daxdev;
ebad861b	99	struct xfs_mount *bt_mount;
6da54179 ES	100	unsigned int bt_meta_sectorsize;
6da54179 ES	101	size_t bt_meta_sectormask;
7c71ee78 ES	102	size_t bt_logical_sectorsize;
7c71ee78 ES	103	size_t bt_logical_sectormask;
ce8e922c	104
ff57ab21 DC	105	/* LRU control structures */
ff57ab21 DC	106	struct shrinker bt_shrinker;
e80dfa19	107	struct list_lru bt_lru;
9c7504aa BF	108
9c7504aa BF	109	struct percpu_counter bt_io_count;
1da177e4 LT	110	} xfs_buftarg_t;
1da177e4 LT	111
1da177e4	112	struct xfs_buf;
ce8e922c	113	typedef void (xfs_buf_iodone_t)(struct xfs_buf );
1da177e4	114
c3f8fc73	115
ce8e922c	116	#define XB_PAGES 2
1da177e4	117
cbb7baab DC	118	struct xfs_buf_map {
	119	xfs_daddr_t bm_bn; /* block number for I/O */
	120	int bm_len; /* size of I/O */
	121	};
	122
3e85c868 DC	123	#define DEFINE_SINGLE_BUF_MAP(map, blkno, numblk) \
	124	struct xfs_buf_map (map) = { .bm_bn = (blkno), .bm_len = (numblk) };
	125
1813dd64	126	struct xfs_buf_ops {
233135b7	127	char *name;
15baadf7 DW	128	union {
	129	__be32 magic[2]; /* v4 and v5 on disk magic values */
	130	__be16 magic16[2]; /* v4 and v5 on disk magic values */
	131	};
1813dd64 DC	132	void (verify_read)(struct xfs_buf );
1813dd64 DC	133	void (verify_write)(struct xfs_buf );
b5572597	134	xfs_failaddr_t (verify_struct)(struct xfs_buf bp);
1813dd64 DC	135	};
1813dd64 DC	136
1da177e4	137	typedef struct xfs_buf {
50f59e8e DC	138	/*
	139	* first cacheline holds all the fields needed for an uncontended cache
	140	* hit to be fully processed. The semaphore straddles the cacheline
	141	* boundary, but the counter and lock sits on the first cacheline,
	142	* which is the only bit that is touched if we hit the semaphore
	143	* fast-path on locking.
	144	*/
6031e73a	145	struct rhash_head b_rhash_head; /* pag buffer hash node */
cbb7baab	146	xfs_daddr_t b_bn; /* block number of buffer */
4e94b71b	147	int b_length; /* size of buffer in BBs */
50f59e8e	148	atomic_t b_hold; /* reference count */
430cbeb8	149	atomic_t b_lru_ref; /* lru reclaim ref count */
50f59e8e	150	xfs_buf_flags_t b_flags; /* status flags */
ce8e922c	151	struct semaphore b_sema; /* semaphore for lockables */
50f59e8e	152
6fb8a90a CM	153	/*
	154	* concurrent access to b_lru and b_lru_flags are protected by
	155	* bt_lru_lock and not by b_sema
	156	*/
430cbeb8	157	struct list_head b_lru; /* lru list */
a4082357 DC	158	spinlock_t b_lock; /* internal state lock */
a4082357 DC	159	unsigned int b_state; /* internal state flags */
61be9c52	160	int b_io_error; /* internal IO error state */
ce8e922c NS	161	wait_queue_head_t b_waiters; /* unpin waiters */
ce8e922c NS	162	struct list_head b_list;
74f75a0c	163	struct xfs_perag b_pag; / contains rbtree root */
ce8e922c	164	xfs_buftarg_t b_target; / buffer target (device) */
ce8e922c	165	void b_addr; / virtual address of buffer */
b29c70f5 BF	166	struct work_struct b_ioend_work;
b29c70f5 BF	167	struct workqueue_struct b_ioend_wq; / I/O completion wq */
ce8e922c	168	xfs_buf_iodone_t b_iodone; /* I/O completion function */
b4dd330b	169	struct completion b_iowait; /* queue for I/O waiters */
fb1755a6	170	void *b_log_item;
643c8c05	171	struct list_head b_li_list; /* Log items list head */
bf9d9013	172	struct xfs_trans *b_transp;
ce8e922c NS	173	struct page *b_pages; / array of page pointers */
ce8e922c NS	174	struct page b_page_array[XB_PAGES]; / inline pages */
3e85c868	175	struct xfs_buf_map b_maps; / compound buffer map */
d44d9bc6	176	struct xfs_buf_map __b_map; /* inline compound buffer map */
3e85c868	177	int b_map_count;
aa0e8833	178	int b_io_length; /* IO size in BBs */
50f59e8e DC	179	atomic_t b_pin_count; /* pin count */
	180	atomic_t b_io_remaining; /* #outstanding I/O requests */
	181	unsigned int b_page_count; /* size of page array */
	182	unsigned int b_offset; /* page offset in first page */
2451337d	183	int b_error; /* error code on I/O */
a5ea70d2 CM	184
	185	/*
	186	* async write failure retry count. Initialised to zero on the first
	187	* failure, then when it exceeds the maximum configured without a
	188	* success the write is considered to be failed permanently and the
	189	* iodone handler will take appropriate action.
	190	*
	191	* For retry timeouts, we record the jiffie of the first failure. This
	192	* means that we can change the retry timeout for buffers already under
	193	* I/O and thus avoid getting stuck in a retry loop with a long timeout.
	194	*
	195	* last_error is used to ensure that we are getting repeated errors, not
	196	* different errors. e.g. a block device might change ENOSPC to EIO when
	197	* a failure timeout occurs, so we want to re-initialise the error
	198	* retry behaviour appropriately when that happens.
	199	*/
	200	int b_retries;
	201	unsigned long b_first_retry_time; /* in jiffies */
	202	int b_last_error;
	203
1813dd64	204	const struct xfs_buf_ops *b_ops;
1da177e4 LT	205	} xfs_buf_t;
1da177e4 LT	206
1da177e4	207	/* Finding and Reading Buffers */
8925a3dc DC	208	struct xfs_buf xfs_buf_incore(struct xfs_buftarg target,
	209	xfs_daddr_t blkno, size_t numblks,
	210	xfs_buf_flags_t flags);
3e85c868 DC	211
	212	struct xfs_buf _xfs_buf_alloc(struct xfs_buftarg target,
	213	struct xfs_buf_map *map, int nmaps,
	214	xfs_buf_flags_t flags);
	215
	216	static inline struct xfs_buf *
	217	xfs_buf_alloc(
	218	struct xfs_buftarg *target,
	219	xfs_daddr_t blkno,
	220	size_t numblks,
	221	xfs_buf_flags_t flags)
	222	{
	223	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	224	return _xfs_buf_alloc(target, &map, 1, flags);
	225	}
1da177e4	226
6dde2707 DC	227	struct xfs_buf xfs_buf_get_map(struct xfs_buftarg target,
	228	struct xfs_buf_map *map, int nmaps,
	229	xfs_buf_flags_t flags);
	230	struct xfs_buf xfs_buf_read_map(struct xfs_buftarg target,
	231	struct xfs_buf_map *map, int nmaps,
1813dd64 DC	232	xfs_buf_flags_t flags,
1813dd64 DC	233	const struct xfs_buf_ops *ops);
6dde2707	234	void xfs_buf_readahead_map(struct xfs_buftarg *target,
c3f8fc73	235	struct xfs_buf_map *map, int nmaps,
1813dd64	236	const struct xfs_buf_ops *ops);
6dde2707 DC	237
	238	static inline struct xfs_buf *
	239	xfs_buf_get(
	240	struct xfs_buftarg *target,
	241	xfs_daddr_t blkno,
	242	size_t numblks,
	243	xfs_buf_flags_t flags)
	244	{
	245	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	246	return xfs_buf_get_map(target, &map, 1, flags);
	247	}
	248
	249	static inline struct xfs_buf *
	250	xfs_buf_read(
	251	struct xfs_buftarg *target,
	252	xfs_daddr_t blkno,
	253	size_t numblks,
c3f8fc73	254	xfs_buf_flags_t flags,
1813dd64	255	const struct xfs_buf_ops *ops)
6dde2707 DC	256	{
6dde2707 DC	257	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	258	return xfs_buf_read_map(target, &map, 1, flags, ops);
6dde2707 DC	259	}
	260
	261	static inline void
	262	xfs_buf_readahead(
	263	struct xfs_buftarg *target,
	264	xfs_daddr_t blkno,
c3f8fc73	265	size_t numblks,
1813dd64	266	const struct xfs_buf_ops *ops)
6dde2707 DC	267	{
6dde2707 DC	268	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	269	return xfs_buf_readahead_map(target, &map, 1, ops);
6dde2707	270	}
e70b73f8	271
e70b73f8 DC	272	void xfs_buf_set_empty(struct xfs_buf *bp, size_t numblks);
	273	int xfs_buf_associate_memory(struct xfs_buf bp, void mem, size_t length);
	274
	275	struct xfs_buf xfs_buf_get_uncached(struct xfs_buftarg target, size_t numblks,
	276	int flags);
ba372674 DC	277	int xfs_buf_read_uncached(struct xfs_buftarg *target, xfs_daddr_t daddr,
	278	size_t numblks, int flags, struct xfs_buf **bpp,
	279	const struct xfs_buf_ops *ops);
e70b73f8	280	void xfs_buf_hold(struct xfs_buf *bp);
1da177e4 LT	281
1da177e4 LT	282	/* Releasing Buffers */
ce8e922c NS	283	extern void xfs_buf_free(xfs_buf_t *);
ce8e922c NS	284	extern void xfs_buf_rele(xfs_buf_t *);
1da177e4 LT	285
1da177e4 LT	286	/* Locking and Unlocking Buffers */
0c842ad4	287	extern int xfs_buf_trylock(xfs_buf_t *);
ce8e922c NS	288	extern void xfs_buf_lock(xfs_buf_t *);
ce8e922c NS	289	extern void xfs_buf_unlock(xfs_buf_t *);
0c842ad4 CH	290	#define xfs_buf_islocked(bp) \
0c842ad4 CH	291	((bp)->b_sema.count <= 0)
1da177e4 LT	292
1da177e4 LT	293	/* Buffer Read and Write Routines */
c2b006c1	294	extern int xfs_bwrite(struct xfs_buf *bp);
e8aaba9a	295	extern void xfs_buf_ioend(struct xfs_buf *bp);
31ca03c9 DW	296	extern void __xfs_buf_ioerror(struct xfs_buf *bp, int error,
	297	xfs_failaddr_t failaddr);
	298	#define xfs_buf_ioerror(bp, err) __xfs_buf_ioerror((bp), (err), __this_address)
901796af	299	extern void xfs_buf_ioerror_alert(struct xfs_buf , const char func);
6af88cda BF	300
	301	extern int __xfs_buf_submit(struct xfs_buf *bp, bool);
	302	static inline int xfs_buf_submit(struct xfs_buf *bp)
	303	{
	304	bool wait = bp->b_flags & XBF_ASYNC ? false : true;
	305	return __xfs_buf_submit(bp, wait);
	306	}
	307
b9c48649	308	extern void xfs_buf_iomove(xfs_buf_t , size_t, size_t, void ,
ce8e922c	309	xfs_buf_rw_t);
1a1a3e97 CH	310	#define xfs_buf_zero(bp, off, len) \
1a1a3e97 CH	311	xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)
ce8e922c	312
1da177e4	313	/* Buffer Utility Routines */
88ee2df7	314	extern void xfs_buf_offset(struct xfs_buf , size_t);
5cfd28b6	315	extern void xfs_buf_stale(struct xfs_buf *bp);
1da177e4	316
1da177e4	317	/* Delayed Write Buffer Routines */
20e8a063	318	extern void xfs_buf_delwri_cancel(struct list_head *);
43ff2122 CH	319	extern bool xfs_buf_delwri_queue(struct xfs_buf , struct list_head );
	320	extern int xfs_buf_delwri_submit(struct list_head *);
	321	extern int xfs_buf_delwri_submit_nowait(struct list_head *);
7912e7fe	322	extern int xfs_buf_delwri_pushbuf(struct xfs_buf , struct list_head );
1da177e4 LT	323
1da177e4 LT	324	/* Buffer Daemon Setup Routines */
ce8e922c NS	325	extern int xfs_buf_init(void);
ce8e922c NS	326	extern void xfs_buf_terminate(void);
1da177e4	327
cbb7baab DC	328	/*
	329	* These macros use the IO block map rather than b_bn. b_bn is now really
	330	* just for the buffer cache index for cached buffers. As IO does not use b_bn
	331	* anymore, uncached buffers do not use b_bn at all and hence must modify the IO
	332	* map directly. Uncached buffers are not allowed to be discontiguous, so this
	333	* is safe to do.
	334	*
	335	* In future, uncached buffers will pass the block number directly to the io
	336	* request function and hence these macros will go away at that point.
	337	*/
d44d9bc6 MT	338	#define XFS_BUF_ADDR(bp) ((bp)->b_maps[0].bm_bn)
d44d9bc6 MT	339	#define XFS_BUF_SET_ADDR(bp, bno) ((bp)->b_maps[0].bm_bn = (xfs_daddr_t)(bno))
ce8e922c	340
7561d27e	341	void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref);
ce8e922c	342
879de98e DC	343	/*
	344	* If the buffer is already on the LRU, do nothing. Otherwise set the buffer
	345	* up with a reference count of 0 so it will be tossed from the cache when
	346	* released.
	347	*/
	348	static inline void xfs_buf_oneshot(struct xfs_buf *bp)
	349	{
	350	if (!list_empty(&bp->b_lru) \|\| atomic_read(&bp->b_lru_ref) > 1)
	351	return;
	352	atomic_set(&bp->b_lru_ref, 0);
	353	}
	354
811e64c7 CS	355	static inline int xfs_buf_ispinned(struct xfs_buf *bp)
	356	{
	357	return atomic_read(&bp->b_pin_count);
	358	}
ce8e922c	359
ce8e922c	360	static inline void xfs_buf_relse(xfs_buf_t *bp)
1da177e4	361	{
bfc60177	362	xfs_buf_unlock(bp);
ce8e922c	363	xfs_buf_rele(bp);
1da177e4 LT	364	}
1da177e4 LT	365
51582170 ES	366	static inline int
	367	xfs_buf_verify_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	368	{
	369	return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
	370	cksum_offset);
	371	}
	372
f1dbcd7e ES	373	static inline void
	374	xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	375	{
	376	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
	377	cksum_offset);
	378	}
	379
1da177e4 LT	380	/*
	381	* Handling of buftargs.
	382	*/
ebad861b	383	extern xfs_buftarg_t xfs_alloc_buftarg(struct xfs_mount ,
486aff5e	384	struct block_device , struct dax_device );
a1f69417	385	extern void xfs_free_buftarg(struct xfs_buftarg *);
1da177e4	386	extern void xfs_wait_buftarg(xfs_buftarg_t *);
a96c4151	387	extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int);
d808f617	388
ce8e922c NS	389	#define xfs_getsize_buftarg(buftarg) block_size((buftarg)->bt_bdev)
	390	#define xfs_readonly_buftarg(buftarg) bdev_read_only((buftarg)->bt_bdev)
	391
75d02303	392	int xfs_buf_reverify(struct xfs_buf bp, const struct xfs_buf_ops ops);
15baadf7 DW	393	bool xfs_verify_magic(struct xfs_buf *bp, __be32 dmagic);
15baadf7 DW	394	bool xfs_verify_magic16(struct xfs_buf *bp, __be16 dmagic);
1aff5696	395
1da177e4	396	#endif /* __XFS_BUF_H__ */