[people/ms/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 */

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;
	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 xfs_log_item	*b_li_list;
	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);
}

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 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 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);

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_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 */
6ab455ee	66
807cbbdb	67	typedef unsigned int xfs_buf_flags_t;
1da177e4	68
0b1b213f CH	69	#define XFS_BUF_FLAGS \
	70	{ XBF_READ, "READ" }, \
	71	{ XBF_WRITE, "WRITE" }, \
1d5ae5df	72	{ XBF_READ_AHEAD, "READ_AHEAD" }, \
1247ec4c	73	{ XBF_NO_IOACCT, "NO_IOACCT" }, \
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" }, \
63db7c81	86	{ _XBF_COMPOUND, "COMPOUND" }
a4082357	87
ac8809f9	88
a4082357 DC	89	/*
	90	* Internal state flags.
	91	*/
	92	#define XFS_BSTATE_DISPOSE (1 << 0) /* buffer being discarded */
63db7c81	93	#define XFS_BSTATE_IN_FLIGHT (1 << 1) /* I/O in flight */
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;
486aff5e	111	struct dax_device *bt_daxdev;
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 );
1813dd64 DC	142	void (verify_write)(struct xfs_buf );
b5572597	143	xfs_failaddr_t (verify_struct)(struct xfs_buf bp);
1813dd64 DC	144	};
1813dd64 DC	145
1da177e4	146	typedef struct xfs_buf {
50f59e8e DC	147	/*
	148	* first cacheline holds all the fields needed for an uncontended cache
	149	* hit to be fully processed. The semaphore straddles the cacheline
	150	* boundary, but the counter and lock sits on the first cacheline,
	151	* which is the only bit that is touched if we hit the semaphore
	152	* fast-path on locking.
	153	*/
6031e73a	154	struct rhash_head b_rhash_head; /* pag buffer hash node */
cbb7baab	155	xfs_daddr_t b_bn; /* block number of buffer */
4e94b71b	156	int b_length; /* size of buffer in BBs */
50f59e8e	157	atomic_t b_hold; /* reference count */
430cbeb8	158	atomic_t b_lru_ref; /* lru reclaim ref count */
50f59e8e	159	xfs_buf_flags_t b_flags; /* status flags */
ce8e922c	160	struct semaphore b_sema; /* semaphore for lockables */
50f59e8e	161
6fb8a90a CM	162	/*
	163	* concurrent access to b_lru and b_lru_flags are protected by
	164	* bt_lru_lock and not by b_sema
	165	*/
430cbeb8	166	struct list_head b_lru; /* lru list */
a4082357 DC	167	spinlock_t b_lock; /* internal state lock */
a4082357 DC	168	unsigned int b_state; /* internal state flags */
61be9c52	169	int b_io_error; /* internal IO error state */
ce8e922c NS	170	wait_queue_head_t b_waiters; /* unpin waiters */
ce8e922c NS	171	struct list_head b_list;
74f75a0c	172	struct xfs_perag b_pag; / contains rbtree root */
ce8e922c	173	xfs_buftarg_t b_target; / buffer target (device) */
ce8e922c	174	void b_addr; / virtual address of buffer */
b29c70f5 BF	175	struct work_struct b_ioend_work;
b29c70f5 BF	176	struct workqueue_struct b_ioend_wq; / I/O completion wq */
ce8e922c	177	xfs_buf_iodone_t b_iodone; /* I/O completion function */
b4dd330b	178	struct completion b_iowait; /* queue for I/O waiters */
fb1755a6 CM	179	void *b_log_item;
fb1755a6 CM	180	struct xfs_log_item *b_li_list;
bf9d9013	181	struct xfs_trans *b_transp;
ce8e922c NS	182	struct page *b_pages; / array of page pointers */
ce8e922c NS	183	struct page b_page_array[XB_PAGES]; / inline pages */
3e85c868	184	struct xfs_buf_map b_maps; / compound buffer map */
d44d9bc6	185	struct xfs_buf_map __b_map; /* inline compound buffer map */
3e85c868	186	int b_map_count;
aa0e8833	187	int b_io_length; /* IO size in BBs */
50f59e8e DC	188	atomic_t b_pin_count; /* pin count */
	189	atomic_t b_io_remaining; /* #outstanding I/O requests */
	190	unsigned int b_page_count; /* size of page array */
	191	unsigned int b_offset; /* page offset in first page */
2451337d	192	int b_error; /* error code on I/O */
a5ea70d2 CM	193
	194	/*
	195	* async write failure retry count. Initialised to zero on the first
	196	* failure, then when it exceeds the maximum configured without a
	197	* success the write is considered to be failed permanently and the
	198	* iodone handler will take appropriate action.
	199	*
	200	* For retry timeouts, we record the jiffie of the first failure. This
	201	* means that we can change the retry timeout for buffers already under
	202	* I/O and thus avoid getting stuck in a retry loop with a long timeout.
	203	*
	204	* last_error is used to ensure that we are getting repeated errors, not
	205	* different errors. e.g. a block device might change ENOSPC to EIO when
	206	* a failure timeout occurs, so we want to re-initialise the error
	207	* retry behaviour appropriately when that happens.
	208	*/
	209	int b_retries;
	210	unsigned long b_first_retry_time; /* in jiffies */
	211	int b_last_error;
	212
1813dd64	213	const struct xfs_buf_ops *b_ops;
cfb02852	214
ce8e922c NS	215	#ifdef XFS_BUF_LOCK_TRACKING
ce8e922c NS	216	int b_last_holder;
1da177e4 LT	217	#endif
	218	} xfs_buf_t;
	219
1da177e4	220	/* Finding and Reading Buffers */
3e85c868 DC	221	struct xfs_buf _xfs_buf_find(struct xfs_buftarg target,
	222	struct xfs_buf_map *map, int nmaps,
	223	xfs_buf_flags_t flags, struct xfs_buf *new_bp);
	224
	225	static inline struct xfs_buf *
	226	xfs_incore(
	227	struct xfs_buftarg *target,
	228	xfs_daddr_t blkno,
	229	size_t numblks,
	230	xfs_buf_flags_t flags)
	231	{
	232	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	233	return _xfs_buf_find(target, &map, 1, flags, NULL);
	234	}
	235
	236	struct xfs_buf _xfs_buf_alloc(struct xfs_buftarg target,
	237	struct xfs_buf_map *map, int nmaps,
	238	xfs_buf_flags_t flags);
	239
	240	static inline struct xfs_buf *
	241	xfs_buf_alloc(
	242	struct xfs_buftarg *target,
	243	xfs_daddr_t blkno,
	244	size_t numblks,
	245	xfs_buf_flags_t flags)
	246	{
	247	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	248	return _xfs_buf_alloc(target, &map, 1, flags);
	249	}
1da177e4	250
6dde2707 DC	251	struct xfs_buf xfs_buf_get_map(struct xfs_buftarg target,
	252	struct xfs_buf_map *map, int nmaps,
	253	xfs_buf_flags_t flags);
	254	struct xfs_buf xfs_buf_read_map(struct xfs_buftarg target,
	255	struct xfs_buf_map *map, int nmaps,
1813dd64 DC	256	xfs_buf_flags_t flags,
1813dd64 DC	257	const struct xfs_buf_ops *ops);
6dde2707	258	void xfs_buf_readahead_map(struct xfs_buftarg *target,
c3f8fc73	259	struct xfs_buf_map *map, int nmaps,
1813dd64	260	const struct xfs_buf_ops *ops);
6dde2707 DC	261
	262	static inline struct xfs_buf *
	263	xfs_buf_get(
	264	struct xfs_buftarg *target,
	265	xfs_daddr_t blkno,
	266	size_t numblks,
	267	xfs_buf_flags_t flags)
	268	{
	269	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	270	return xfs_buf_get_map(target, &map, 1, flags);
	271	}
	272
	273	static inline struct xfs_buf *
	274	xfs_buf_read(
	275	struct xfs_buftarg *target,
	276	xfs_daddr_t blkno,
	277	size_t numblks,
c3f8fc73	278	xfs_buf_flags_t flags,
1813dd64	279	const struct xfs_buf_ops *ops)
6dde2707 DC	280	{
6dde2707 DC	281	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	282	return xfs_buf_read_map(target, &map, 1, flags, ops);
6dde2707 DC	283	}
	284
	285	static inline void
	286	xfs_buf_readahead(
	287	struct xfs_buftarg *target,
	288	xfs_daddr_t blkno,
c3f8fc73	289	size_t numblks,
1813dd64	290	const struct xfs_buf_ops *ops)
6dde2707 DC	291	{
6dde2707 DC	292	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	293	return xfs_buf_readahead_map(target, &map, 1, ops);
6dde2707	294	}
e70b73f8	295
e70b73f8 DC	296	void xfs_buf_set_empty(struct xfs_buf *bp, size_t numblks);
	297	int xfs_buf_associate_memory(struct xfs_buf bp, void mem, size_t length);
	298
	299	struct xfs_buf xfs_buf_get_uncached(struct xfs_buftarg target, size_t numblks,
	300	int flags);
ba372674 DC	301	int xfs_buf_read_uncached(struct xfs_buftarg *target, xfs_daddr_t daddr,
	302	size_t numblks, int flags, struct xfs_buf **bpp,
	303	const struct xfs_buf_ops *ops);
e70b73f8	304	void xfs_buf_hold(struct xfs_buf *bp);
1da177e4 LT	305
1da177e4 LT	306	/* Releasing Buffers */
ce8e922c NS	307	extern void xfs_buf_free(xfs_buf_t *);
ce8e922c NS	308	extern void xfs_buf_rele(xfs_buf_t *);
1da177e4 LT	309
1da177e4 LT	310	/* Locking and Unlocking Buffers */
0c842ad4	311	extern int xfs_buf_trylock(xfs_buf_t *);
ce8e922c NS	312	extern void xfs_buf_lock(xfs_buf_t *);
ce8e922c NS	313	extern void xfs_buf_unlock(xfs_buf_t *);
0c842ad4 CH	314	#define xfs_buf_islocked(bp) \
0c842ad4 CH	315	((bp)->b_sema.count <= 0)
1da177e4 LT	316
1da177e4 LT	317	/* Buffer Read and Write Routines */
c2b006c1	318	extern int xfs_bwrite(struct xfs_buf *bp);
e8aaba9a	319	extern void xfs_buf_ioend(struct xfs_buf *bp);
31ca03c9 DW	320	extern void __xfs_buf_ioerror(struct xfs_buf *bp, int error,
	321	xfs_failaddr_t failaddr);
	322	#define xfs_buf_ioerror(bp, err) __xfs_buf_ioerror((bp), (err), __this_address)
901796af	323	extern void xfs_buf_ioerror_alert(struct xfs_buf , const char func);
595bff75 DC	324	extern void xfs_buf_submit(struct xfs_buf *bp);
595bff75 DC	325	extern int xfs_buf_submit_wait(struct xfs_buf *bp);
b9c48649	326	extern void xfs_buf_iomove(xfs_buf_t , size_t, size_t, void ,
ce8e922c	327	xfs_buf_rw_t);
1a1a3e97 CH	328	#define xfs_buf_zero(bp, off, len) \
1a1a3e97 CH	329	xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)
ce8e922c	330
1da177e4	331	/* Buffer Utility Routines */
88ee2df7	332	extern void xfs_buf_offset(struct xfs_buf , size_t);
5cfd28b6	333	extern void xfs_buf_stale(struct xfs_buf *bp);
1da177e4	334
1da177e4	335	/* Delayed Write Buffer Routines */
20e8a063	336	extern void xfs_buf_delwri_cancel(struct list_head *);
43ff2122 CH	337	extern bool xfs_buf_delwri_queue(struct xfs_buf , struct list_head );
	338	extern int xfs_buf_delwri_submit(struct list_head *);
	339	extern int xfs_buf_delwri_submit_nowait(struct list_head *);
7912e7fe	340	extern int xfs_buf_delwri_pushbuf(struct xfs_buf , struct list_head );
1da177e4 LT	341
1da177e4 LT	342	/* Buffer Daemon Setup Routines */
ce8e922c NS	343	extern int xfs_buf_init(void);
ce8e922c NS	344	extern void xfs_buf_terminate(void);
1da177e4	345
cbb7baab DC	346	/*
	347	* These macros use the IO block map rather than b_bn. b_bn is now really
	348	* just for the buffer cache index for cached buffers. As IO does not use b_bn
	349	* anymore, uncached buffers do not use b_bn at all and hence must modify the IO
	350	* map directly. Uncached buffers are not allowed to be discontiguous, so this
	351	* is safe to do.
	352	*
	353	* In future, uncached buffers will pass the block number directly to the io
	354	* request function and hence these macros will go away at that point.
	355	*/
d44d9bc6 MT	356	#define XFS_BUF_ADDR(bp) ((bp)->b_maps[0].bm_bn)
d44d9bc6 MT	357	#define XFS_BUF_SET_ADDR(bp, bno) ((bp)->b_maps[0].bm_bn = (xfs_daddr_t)(bno))
ce8e922c	358
7561d27e	359	void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref);
ce8e922c	360
811e64c7 CS	361	static inline int xfs_buf_ispinned(struct xfs_buf *bp)
	362	{
	363	return atomic_read(&bp->b_pin_count);
	364	}
ce8e922c	365
ce8e922c	366	static inline void xfs_buf_relse(xfs_buf_t *bp)
1da177e4	367	{
bfc60177	368	xfs_buf_unlock(bp);
ce8e922c	369	xfs_buf_rele(bp);
1da177e4 LT	370	}
1da177e4 LT	371
51582170 ES	372	static inline int
	373	xfs_buf_verify_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	374	{
	375	return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
	376	cksum_offset);
	377	}
	378
f1dbcd7e ES	379	static inline void
	380	xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	381	{
	382	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
	383	cksum_offset);
	384	}
	385
1da177e4 LT	386	/*
	387	* Handling of buftargs.
	388	*/
ebad861b	389	extern xfs_buftarg_t xfs_alloc_buftarg(struct xfs_mount ,
486aff5e	390	struct block_device , struct dax_device );
c141b292	391	extern void xfs_free_buftarg(struct xfs_mount , struct xfs_buftarg );
1da177e4	392	extern void xfs_wait_buftarg(xfs_buftarg_t *);
a96c4151	393	extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int);
d808f617	394
ce8e922c NS	395	#define xfs_getsize_buftarg(buftarg) block_size((buftarg)->bt_bdev)
	396	#define xfs_readonly_buftarg(buftarg) bdev_read_only((buftarg)->bt_bdev)
	397
1da177e4	398	#endif /* __XFS_BUF_H__ */