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

/*
 * Copyright (c) 2000-2005 Silicon Graphics, Inc.
 * All Rights Reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it would be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#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/buffer_head.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 */
#define _XBF_IN_FLIGHT	 (1 << 25) /* I/O in flight, for accounting purposes */

typedef unsigned int xfs_buf_flags_t;

#define XFS_BUF_FLAGS \
	{ XBF_READ,		"READ" }, \
	{ XBF_WRITE,		"WRITE" }, \
	{ XBF_READ_AHEAD,	"READ_AHEAD" }, \
	{ 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" }, \
	{ _XBF_IN_FLIGHT,	"IN_FLIGHT" }


/*
 * Internal state flags.
 */
#define XFS_BSTATE_DISPOSE	 (1 << 0)	/* buffer being discarded */

/*
 * 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 backing_dev_info	*bt_bdi;
	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;
	void (*verify_read)(struct xfs_buf *);
	void (*verify_write)(struct xfs_buf *);
};

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 rb_node		b_rbnode;	/* rbtree 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_fspriv;
	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;

#ifdef XFS_BUF_LOCK_TRACKING
	int			b_last_holder;
#endif
} xfs_buf_t;

/* Finding and Reading Buffers */
struct xfs_buf *_xfs_buf_find(struct xfs_buftarg *target,
			      struct xfs_buf_map *map, int nmaps,
			      xfs_buf_flags_t flags, struct xfs_buf *new_bp);

static inline struct xfs_buf *
xfs_incore(
	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_find(target, &map, 1, flags, NULL);
}

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);
}

struct xfs_buf *xfs_buf_get_empty(struct xfs_buftarg *target, size_t numblks);
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(xfs_buf_t *, int);
extern void xfs_buf_ioerror_alert(struct xfs_buf *, const char *func);
extern void xfs_buf_submit(struct xfs_buf *bp);
extern int xfs_buf_submit_wait(struct xfs_buf *bp);
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 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 *);

/* 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))

static inline void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref)
{
	atomic_set(&bp->b_lru_ref, lru_ref);
}

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 *);
extern void xfs_free_buftarg(struct xfs_mount *, 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)

#endif	/* __XFS_BUF_H__ */
Commit	Line	Data
1da177e4	1	/*
7b718769 NS	2	* Copyright (c) 2000-2005 Silicon Graphics, Inc.
7b718769 NS	3	* All Rights Reserved.
1da177e4	4	*
7b718769 NS	5	* This program is free software; you can redistribute it and/or
7b718769 NS	6	* modify it under the terms of the GNU General Public License as
1da177e4 LT	7	* published by the Free Software Foundation.
1da177e4 LT	8	*
7b718769 NS	9	* This program is distributed in the hope that it would be useful,
	10	* but WITHOUT ANY WARRANTY; without even the implied warranty of
	11	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
	12	* GNU General Public License for more details.
1da177e4	13	*
7b718769 NS	14	* You should have received a copy of the GNU General Public License
	15	* along with this program; if not, write the Free Software Foundation,
	16	* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
1da177e4	17	*/
1da177e4 LT	18	#ifndef __XFS_BUF_H__
	19	#define __XFS_BUF_H__
	20
1da177e4 LT	21	#include <linux/list.h>
	22	#include <linux/types.h>
	23	#include <linux/spinlock.h>
1da177e4 LT	24	#include <linux/mm.h>
1da177e4 LT	25	#include <linux/fs.h>
c94c2acf	26	#include <linux/dax.h>
1da177e4 LT	27	#include <linux/buffer_head.h>
1da177e4 LT	28	#include <linux/uio.h>
e80dfa19	29	#include <linux/list_lru.h>
1da177e4 LT	30
	31	/*
	32	* Base types
	33	*/
	34
ce8e922c NS	35	#define XFS_BUF_DADDR_NULL ((xfs_daddr_t) (-1LL))
ce8e922c NS	36
ce8e922c NS	37	typedef enum {
	38	XBRW_READ = 1, /* transfer into target memory */
	39	XBRW_WRITE = 2, /* transfer from target memory */
	40	XBRW_ZERO = 3, /* Zero target memory */
	41	} xfs_buf_rw_t;
	42
6fb8a90a CM	43	#define XBF_READ (1 << 0) /* buffer intended for reading from device */
	44	#define XBF_WRITE (1 << 1) /* buffer intended for writing to device */
	45	#define XBF_READ_AHEAD (1 << 2) /* asynchronous read-ahead */
c891c30a	46	#define XBF_NO_IOACCT (1 << 3) /* bypass I/O accounting (non-LRU bufs) */
6fb8a90a CM	47	#define XBF_ASYNC (1 << 4) /* initiator will not wait for completion */
	48	#define XBF_DONE (1 << 5) /* all pages in the buffer uptodate */
	49	#define XBF_STALE (1 << 6) /* buffer has been staled, do not find it */
ac8809f9	50	#define XBF_WRITE_FAIL (1 << 24)/* async writes have failed on this buffer */
1d5ae5df CH	51
1d5ae5df CH	52	/* I/O hints for the BIO layer */
6fb8a90a CM	53	#define XBF_SYNCIO (1 << 10)/* treat this buffer as synchronous I/O */
	54	#define XBF_FUA (1 << 11)/* force cache write through mode */
	55	#define XBF_FLUSH (1 << 12)/* flush the disk cache before a write */
1da177e4	56
807cbbdb	57	/* flags used only as arguments to access routines */
6fb8a90a CM	58	#define XBF_TRYLOCK (1 << 16)/* lock requested, but do not wait */
6fb8a90a CM	59	#define XBF_UNMAPPED (1 << 17)/* do not map the buffer */
1da177e4	60
807cbbdb	61	/* flags used only internally */
6fb8a90a CM	62	#define _XBF_PAGES (1 << 20)/* backed by refcounted pages */
	63	#define _XBF_KMEM (1 << 21)/* backed by heap memory */
	64	#define _XBF_DELWRI_Q (1 << 22)/* buffer on a delwri queue */
	65	#define _XBF_COMPOUND (1 << 23)/* compound buffer */
9c7504aa	66	#define _XBF_IN_FLIGHT (1 << 25) /* I/O in flight, for accounting purposes */
6ab455ee	67
807cbbdb	68	typedef unsigned int xfs_buf_flags_t;
1da177e4	69
0b1b213f CH	70	#define XFS_BUF_FLAGS \
	71	{ XBF_READ, "READ" }, \
	72	{ XBF_WRITE, "WRITE" }, \
1d5ae5df	73	{ XBF_READ_AHEAD, "READ_AHEAD" }, \
0b1b213f CH	74	{ XBF_ASYNC, "ASYNC" }, \
0b1b213f CH	75	{ XBF_DONE, "DONE" }, \
0b1b213f	76	{ XBF_STALE, "STALE" }, \
ac8809f9	77	{ XBF_WRITE_FAIL, "WRITE_FAIL" }, \
1d5ae5df CH	78	{ XBF_SYNCIO, "SYNCIO" }, \
	79	{ XBF_FUA, "FUA" }, \
	80	{ XBF_FLUSH, "FLUSH" }, \
6fb8a90a	81	{ XBF_TRYLOCK, "TRYLOCK" }, /* should never be set */\
611c9946	82	{ XBF_UNMAPPED, "UNMAPPED" }, /* ditto */\
0b1b213f	83	{ _XBF_PAGES, "PAGES" }, \
0e6e847f	84	{ _XBF_KMEM, "KMEM" }, \
cbb7baab	85	{ _XBF_DELWRI_Q, "DELWRI_Q" }, \
9c7504aa BF	86	{ _XBF_COMPOUND, "COMPOUND" }, \
9c7504aa BF	87	{ _XBF_IN_FLIGHT, "IN_FLIGHT" }
a4082357	88
ac8809f9	89
a4082357 DC	90	/*
	91	* Internal state flags.
	92	*/
	93	#define XFS_BSTATE_DISPOSE (1 << 0) /* buffer being discarded */
0b1b213f	94
7c71ee78 ES	95	/*
	96	* The xfs_buftarg contains 2 notions of "sector size" -
	97	*
	98	* 1) The metadata sector size, which is the minimum unit and
	99	* alignment of IO which will be performed by metadata operations.
	100	* 2) The device logical sector size
	101	*
	102	* The first is specified at mkfs time, and is stored on-disk in the
	103	* superblock's sb_sectsize.
	104	*
	105	* The latter is derived from the underlying device, and controls direct IO
	106	* alignment constraints.
	107	*/
1da177e4	108	typedef struct xfs_buftarg {
ce8e922c NS	109	dev_t bt_dev;
ce8e922c NS	110	struct block_device *bt_bdev;
0e6e847f	111	struct backing_dev_info *bt_bdi;
ebad861b	112	struct xfs_mount *bt_mount;
6da54179 ES	113	unsigned int bt_meta_sectorsize;
6da54179 ES	114	size_t bt_meta_sectormask;
7c71ee78 ES	115	size_t bt_logical_sectorsize;
7c71ee78 ES	116	size_t bt_logical_sectormask;
ce8e922c	117
ff57ab21 DC	118	/* LRU control structures */
ff57ab21 DC	119	struct shrinker bt_shrinker;
e80dfa19	120	struct list_lru bt_lru;
9c7504aa BF	121
9c7504aa BF	122	struct percpu_counter bt_io_count;
1da177e4 LT	123	} xfs_buftarg_t;
1da177e4 LT	124
1da177e4	125	struct xfs_buf;
ce8e922c	126	typedef void (xfs_buf_iodone_t)(struct xfs_buf );
1da177e4	127
c3f8fc73	128
ce8e922c	129	#define XB_PAGES 2
1da177e4	130
cbb7baab DC	131	struct xfs_buf_map {
	132	xfs_daddr_t bm_bn; /* block number for I/O */
	133	int bm_len; /* size of I/O */
	134	};
	135
3e85c868 DC	136	#define DEFINE_SINGLE_BUF_MAP(map, blkno, numblk) \
	137	struct xfs_buf_map (map) = { .bm_bn = (blkno), .bm_len = (numblk) };
	138
1813dd64	139	struct xfs_buf_ops {
233135b7	140	char *name;
1813dd64 DC	141	void (verify_read)(struct xfs_buf );
	142	void (verify_write)(struct xfs_buf );
	143	};
	144
1da177e4	145	typedef struct xfs_buf {
50f59e8e DC	146	/*
	147	* first cacheline holds all the fields needed for an uncontended cache
	148	* hit to be fully processed. The semaphore straddles the cacheline
	149	* boundary, but the counter and lock sits on the first cacheline,
	150	* which is the only bit that is touched if we hit the semaphore
	151	* fast-path on locking.
	152	*/
	153	struct rb_node b_rbnode; /* rbtree node */
cbb7baab	154	xfs_daddr_t b_bn; /* block number of buffer */
4e94b71b	155	int b_length; /* size of buffer in BBs */
50f59e8e	156	atomic_t b_hold; /* reference count */
430cbeb8	157	atomic_t b_lru_ref; /* lru reclaim ref count */
50f59e8e	158	xfs_buf_flags_t b_flags; /* status flags */
ce8e922c	159	struct semaphore b_sema; /* semaphore for lockables */
50f59e8e	160
6fb8a90a CM	161	/*
	162	* concurrent access to b_lru and b_lru_flags are protected by
	163	* bt_lru_lock and not by b_sema
	164	*/
430cbeb8	165	struct list_head b_lru; /* lru list */
a4082357 DC	166	spinlock_t b_lock; /* internal state lock */
a4082357 DC	167	unsigned int b_state; /* internal state flags */
61be9c52	168	int b_io_error; /* internal IO error state */
ce8e922c NS	169	wait_queue_head_t b_waiters; /* unpin waiters */
ce8e922c NS	170	struct list_head b_list;
74f75a0c	171	struct xfs_perag b_pag; / contains rbtree root */
ce8e922c	172	xfs_buftarg_t b_target; / buffer target (device) */
ce8e922c	173	void b_addr; / virtual address of buffer */
b29c70f5 BF	174	struct work_struct b_ioend_work;
b29c70f5 BF	175	struct workqueue_struct b_ioend_wq; / I/O completion wq */
ce8e922c	176	xfs_buf_iodone_t b_iodone; /* I/O completion function */
b4dd330b	177	struct completion b_iowait; /* queue for I/O waiters */
ce8e922c	178	void *b_fspriv;
bf9d9013	179	struct xfs_trans *b_transp;
ce8e922c NS	180	struct page *b_pages; / array of page pointers */
ce8e922c NS	181	struct page b_page_array[XB_PAGES]; / inline pages */
3e85c868	182	struct xfs_buf_map b_maps; / compound buffer map */
d44d9bc6	183	struct xfs_buf_map __b_map; /* inline compound buffer map */
3e85c868	184	int b_map_count;
aa0e8833	185	int b_io_length; /* IO size in BBs */
50f59e8e DC	186	atomic_t b_pin_count; /* pin count */
	187	atomic_t b_io_remaining; /* #outstanding I/O requests */
	188	unsigned int b_page_count; /* size of page array */
	189	unsigned int b_offset; /* page offset in first page */
2451337d	190	int b_error; /* error code on I/O */
a5ea70d2 CM	191
	192	/*
	193	* async write failure retry count. Initialised to zero on the first
	194	* failure, then when it exceeds the maximum configured without a
	195	* success the write is considered to be failed permanently and the
	196	* iodone handler will take appropriate action.
	197	*
	198	* For retry timeouts, we record the jiffie of the first failure. This
	199	* means that we can change the retry timeout for buffers already under
	200	* I/O and thus avoid getting stuck in a retry loop with a long timeout.
	201	*
	202	* last_error is used to ensure that we are getting repeated errors, not
	203	* different errors. e.g. a block device might change ENOSPC to EIO when
	204	* a failure timeout occurs, so we want to re-initialise the error
	205	* retry behaviour appropriately when that happens.
	206	*/
	207	int b_retries;
	208	unsigned long b_first_retry_time; /* in jiffies */
	209	int b_last_error;
	210
1813dd64	211	const struct xfs_buf_ops *b_ops;
cfb02852	212
ce8e922c NS	213	#ifdef XFS_BUF_LOCK_TRACKING
ce8e922c NS	214	int b_last_holder;
1da177e4 LT	215	#endif
	216	} xfs_buf_t;
	217
1da177e4	218	/* Finding and Reading Buffers */
3e85c868 DC	219	struct xfs_buf _xfs_buf_find(struct xfs_buftarg target,
	220	struct xfs_buf_map *map, int nmaps,
	221	xfs_buf_flags_t flags, struct xfs_buf *new_bp);
	222
	223	static inline struct xfs_buf *
	224	xfs_incore(
	225	struct xfs_buftarg *target,
	226	xfs_daddr_t blkno,
	227	size_t numblks,
	228	xfs_buf_flags_t flags)
	229	{
	230	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	231	return _xfs_buf_find(target, &map, 1, flags, NULL);
	232	}
	233
	234	struct xfs_buf _xfs_buf_alloc(struct xfs_buftarg target,
	235	struct xfs_buf_map *map, int nmaps,
	236	xfs_buf_flags_t flags);
	237
	238	static inline struct xfs_buf *
	239	xfs_buf_alloc(
	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_alloc(target, &map, 1, flags);
	247	}
1da177e4	248
6dde2707 DC	249	struct xfs_buf xfs_buf_get_map(struct xfs_buftarg target,
	250	struct xfs_buf_map *map, int nmaps,
	251	xfs_buf_flags_t flags);
	252	struct xfs_buf xfs_buf_read_map(struct xfs_buftarg target,
	253	struct xfs_buf_map *map, int nmaps,
1813dd64 DC	254	xfs_buf_flags_t flags,
1813dd64 DC	255	const struct xfs_buf_ops *ops);
6dde2707	256	void xfs_buf_readahead_map(struct xfs_buftarg *target,
c3f8fc73	257	struct xfs_buf_map *map, int nmaps,
1813dd64	258	const struct xfs_buf_ops *ops);
6dde2707 DC	259
	260	static inline struct xfs_buf *
	261	xfs_buf_get(
	262	struct xfs_buftarg *target,
	263	xfs_daddr_t blkno,
	264	size_t numblks,
	265	xfs_buf_flags_t flags)
	266	{
	267	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	268	return xfs_buf_get_map(target, &map, 1, flags);
	269	}
	270
	271	static inline struct xfs_buf *
	272	xfs_buf_read(
	273	struct xfs_buftarg *target,
	274	xfs_daddr_t blkno,
	275	size_t numblks,
c3f8fc73	276	xfs_buf_flags_t flags,
1813dd64	277	const struct xfs_buf_ops *ops)
6dde2707 DC	278	{
6dde2707 DC	279	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	280	return xfs_buf_read_map(target, &map, 1, flags, ops);
6dde2707 DC	281	}
	282
	283	static inline void
	284	xfs_buf_readahead(
	285	struct xfs_buftarg *target,
	286	xfs_daddr_t blkno,
c3f8fc73	287	size_t numblks,
1813dd64	288	const struct xfs_buf_ops *ops)
6dde2707 DC	289	{
6dde2707 DC	290	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	291	return xfs_buf_readahead_map(target, &map, 1, ops);
6dde2707	292	}
e70b73f8 DC	293
e70b73f8 DC	294	struct xfs_buf xfs_buf_get_empty(struct xfs_buftarg target, size_t numblks);
e70b73f8 DC	295	void xfs_buf_set_empty(struct xfs_buf *bp, size_t numblks);
	296	int xfs_buf_associate_memory(struct xfs_buf bp, void mem, size_t length);
	297
	298	struct xfs_buf xfs_buf_get_uncached(struct xfs_buftarg target, size_t numblks,
	299	int flags);
ba372674 DC	300	int xfs_buf_read_uncached(struct xfs_buftarg *target, xfs_daddr_t daddr,
	301	size_t numblks, int flags, struct xfs_buf **bpp,
	302	const struct xfs_buf_ops *ops);
e70b73f8	303	void xfs_buf_hold(struct xfs_buf *bp);
1da177e4 LT	304
1da177e4 LT	305	/* Releasing Buffers */
ce8e922c NS	306	extern void xfs_buf_free(xfs_buf_t *);
ce8e922c NS	307	extern void xfs_buf_rele(xfs_buf_t *);
1da177e4 LT	308
1da177e4 LT	309	/* Locking and Unlocking Buffers */
0c842ad4	310	extern int xfs_buf_trylock(xfs_buf_t *);
ce8e922c NS	311	extern void xfs_buf_lock(xfs_buf_t *);
ce8e922c NS	312	extern void xfs_buf_unlock(xfs_buf_t *);
0c842ad4 CH	313	#define xfs_buf_islocked(bp) \
0c842ad4 CH	314	((bp)->b_sema.count <= 0)
1da177e4 LT	315
1da177e4 LT	316	/* Buffer Read and Write Routines */
c2b006c1	317	extern int xfs_bwrite(struct xfs_buf *bp);
e8aaba9a	318	extern void xfs_buf_ioend(struct xfs_buf *bp);
ce8e922c	319	extern void xfs_buf_ioerror(xfs_buf_t *, int);
901796af	320	extern void xfs_buf_ioerror_alert(struct xfs_buf , const char func);
595bff75 DC	321	extern void xfs_buf_submit(struct xfs_buf *bp);
595bff75 DC	322	extern int xfs_buf_submit_wait(struct xfs_buf *bp);
b9c48649	323	extern void xfs_buf_iomove(xfs_buf_t , size_t, size_t, void ,
ce8e922c	324	xfs_buf_rw_t);
1a1a3e97 CH	325	#define xfs_buf_zero(bp, off, len) \
1a1a3e97 CH	326	xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)
ce8e922c	327
1da177e4	328	/* Buffer Utility Routines */
88ee2df7	329	extern void xfs_buf_offset(struct xfs_buf , size_t);
5cfd28b6	330	extern void xfs_buf_stale(struct xfs_buf *bp);
1da177e4	331
1da177e4	332	/* Delayed Write Buffer Routines */
43ff2122 CH	333	extern bool xfs_buf_delwri_queue(struct xfs_buf , struct list_head );
	334	extern int xfs_buf_delwri_submit(struct list_head *);
	335	extern int xfs_buf_delwri_submit_nowait(struct list_head *);
1da177e4 LT	336
1da177e4 LT	337	/* Buffer Daemon Setup Routines */
ce8e922c NS	338	extern int xfs_buf_init(void);
ce8e922c NS	339	extern void xfs_buf_terminate(void);
1da177e4	340
cbb7baab DC	341	/*
	342	* These macros use the IO block map rather than b_bn. b_bn is now really
	343	* just for the buffer cache index for cached buffers. As IO does not use b_bn
	344	* anymore, uncached buffers do not use b_bn at all and hence must modify the IO
	345	* map directly. Uncached buffers are not allowed to be discontiguous, so this
	346	* is safe to do.
	347	*
	348	* In future, uncached buffers will pass the block number directly to the io
	349	* request function and hence these macros will go away at that point.
	350	*/
d44d9bc6 MT	351	#define XFS_BUF_ADDR(bp) ((bp)->b_maps[0].bm_bn)
d44d9bc6 MT	352	#define XFS_BUF_SET_ADDR(bp, bno) ((bp)->b_maps[0].bm_bn = (xfs_daddr_t)(bno))
ce8e922c	353
38f23232	354	static inline void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref)
821eb21d DC	355	{
	356	atomic_set(&bp->b_lru_ref, lru_ref);
	357	}
ce8e922c	358
811e64c7 CS	359	static inline int xfs_buf_ispinned(struct xfs_buf *bp)
	360	{
	361	return atomic_read(&bp->b_pin_count);
	362	}
ce8e922c	363
ce8e922c	364	static inline void xfs_buf_relse(xfs_buf_t *bp)
1da177e4	365	{
bfc60177	366	xfs_buf_unlock(bp);
ce8e922c	367	xfs_buf_rele(bp);
1da177e4 LT	368	}
1da177e4 LT	369
51582170 ES	370	static inline int
	371	xfs_buf_verify_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	372	{
	373	return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
	374	cksum_offset);
	375	}
	376
f1dbcd7e ES	377	static inline void
	378	xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	379	{
	380	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
	381	cksum_offset);
	382	}
	383
1da177e4 LT	384	/*
	385	* Handling of buftargs.
	386	*/
ebad861b	387	extern xfs_buftarg_t xfs_alloc_buftarg(struct xfs_mount ,
34dcefd7	388	struct block_device *);
c141b292	389	extern void xfs_free_buftarg(struct xfs_mount , struct xfs_buftarg );
1da177e4	390	extern void xfs_wait_buftarg(xfs_buftarg_t *);
a96c4151	391	extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int);
d808f617	392
ce8e922c NS	393	#define xfs_getsize_buftarg(buftarg) block_size((buftarg)->bt_bdev)
	394	#define xfs_readonly_buftarg(buftarg) bdev_read_only((buftarg)->bt_bdev)
	395
1da177e4	396	#endif /* __XFS_BUF_H__ */