[people/ms/linux.git] / fs / xfs / linux-2.6 / 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 <asm/system.h>
#include <linux/mm.h>
#include <linux/fs.h>
#include <linux/buffer_head.h>
#include <linux/uio.h>

/*
 *	Base types
 */

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

#define xfs_buf_ctob(pp)	((pp) * PAGE_CACHE_SIZE)
#define xfs_buf_btoc(dd)	(((dd) + PAGE_CACHE_SIZE-1) >> PAGE_CACHE_SHIFT)
#define xfs_buf_btoct(dd)	((dd) >> PAGE_CACHE_SHIFT)
#define xfs_buf_poff(aa)	((aa) & ~PAGE_CACHE_MASK)

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;

typedef enum {
	XBF_READ = (1 << 0),	/* buffer intended for reading from device */
	XBF_WRITE = (1 << 1),	/* buffer intended for writing to device   */
	XBF_MAPPED = (1 << 2),  /* buffer mapped (b_addr valid)            */
	XBF_ASYNC = (1 << 4),   /* initiator will not wait for completion  */
	XBF_DONE = (1 << 5),    /* all pages in the buffer uptodate	   */
	XBF_DELWRI = (1 << 6),  /* buffer has dirty pages                  */
	XBF_STALE = (1 << 7),	/* buffer has been staled, do not find it  */
	XBF_FS_MANAGED = (1 << 8),  /* filesystem controls freeing memory  */
 	XBF_ORDERED = (1 << 11),    /* use ordered writes		   */
	XBF_READ_AHEAD = (1 << 12), /* asynchronous read-ahead		   */

	/* flags used only as arguments to access routines */
	XBF_LOCK = (1 << 14),       /* lock requested			   */
	XBF_TRYLOCK = (1 << 15),    /* lock requested, but do not wait	   */
	XBF_DONT_BLOCK = (1 << 16), /* do not block in current thread	   */

	/* flags used only internally */
	_XBF_PAGE_CACHE = (1 << 17),/* backed by pagecache		   */
	_XBF_PAGES = (1 << 18),	    /* backed by refcounted pages	   */
	_XBF_RUN_QUEUES = (1 << 19),/* run block device task queue	   */
	_XBF_DELWRI_Q = (1 << 21),   /* buffer on delwri queue		   */

	/*
	 * Special flag for supporting metadata blocks smaller than a FSB.
	 *
	 * In this case we can have multiple xfs_buf_t on a single page and
	 * need to lock out concurrent xfs_buf_t readers as they only
	 * serialise access to the buffer.
	 *
	 * If the FSB size >= PAGE_CACHE_SIZE case, we have no serialisation
	 * between reads of the page. Hence we can have one thread read the
	 * page and modify it, but then race with another thread that thinks
	 * the page is not up-to-date and hence reads it again.
	 *
	 * The result is that the first modifcation to the page is lost.
	 * This sort of AGF/AGI reading race can happen when unlinking inodes
	 * that require truncation and results in the AGI unlinked list
	 * modifications being lost.
	 */
	_XBF_PAGE_LOCKED = (1 << 22),

	/*
	 * If we try a barrier write, but it fails we have to communicate
	 * this to the upper layers.  Unfortunately b_error gets overwritten
	 * when the buffer is re-issued so we have to add another flag to
	 * keep this information.
	 */
	_XFS_BARRIER_FAILED = (1 << 23),
} xfs_buf_flags_t;

typedef enum {
	XBT_FORCE_SLEEP = 0,
	XBT_FORCE_FLUSH = 1,
} xfs_buftarg_flags_t;

typedef struct xfs_bufhash {
	struct list_head	bh_list;
	spinlock_t		bh_lock;
} xfs_bufhash_t;

typedef struct xfs_buftarg {
	dev_t			bt_dev;
	struct block_device	*bt_bdev;
	struct address_space	*bt_mapping;
	unsigned int		bt_bsize;
	unsigned int		bt_sshift;
	size_t			bt_smask;

	/* per device buffer hash table */
	uint			bt_hashmask;
	uint			bt_hashshift;
	xfs_bufhash_t		*bt_hash;

	/* per device delwri queue */
	struct task_struct	*bt_task;
	struct list_head	bt_list;
	struct list_head	bt_delwrite_queue;
	spinlock_t		bt_delwrite_lock;
	unsigned long		bt_flags;
} xfs_buftarg_t;

/*
 *	xfs_buf_t:  Buffer structure for pagecache-based buffers
 *
 * This buffer structure is used by the pagecache buffer management routines
 * to refer to an assembly of pages forming a logical buffer.
 *
 * The buffer structure is used on a temporary basis only, and discarded when
 * released.  The real data storage is recorded in the pagecache. Buffers are
 * hashed to the block device on which the file system resides.
 */

struct xfs_buf;
typedef void (*xfs_buf_iodone_t)(struct xfs_buf *);
typedef void (*xfs_buf_relse_t)(struct xfs_buf *);
typedef int (*xfs_buf_bdstrat_t)(struct xfs_buf *);

#define XB_PAGES	2

typedef struct xfs_buf {
	struct semaphore	b_sema;		/* semaphore for lockables */
	unsigned long		b_queuetime;	/* time buffer was queued */
	atomic_t		b_pin_count;	/* pin count */
	wait_queue_head_t	b_waiters;	/* unpin waiters */
	struct list_head	b_list;
	xfs_buf_flags_t		b_flags;	/* status flags */
	struct list_head	b_hash_list;	/* hash table list */
	xfs_bufhash_t		*b_hash;	/* hash table list start */
	xfs_buftarg_t		*b_target;	/* buffer target (device) */
	atomic_t		b_hold;		/* reference count */
	xfs_daddr_t		b_bn;		/* block number for I/O */
	xfs_off_t		b_file_offset;	/* offset in file */
	size_t			b_buffer_length;/* size of buffer in bytes */
	size_t			b_count_desired;/* desired transfer size */
	void			*b_addr;	/* virtual address of buffer */
	struct work_struct	b_iodone_work;
	atomic_t		b_io_remaining;	/* #outstanding I/O requests */
	xfs_buf_iodone_t	b_iodone;	/* I/O completion function */
	xfs_buf_relse_t		b_relse;	/* releasing function */
	xfs_buf_bdstrat_t	b_strat;	/* pre-write function */
	struct completion	b_iowait;	/* queue for I/O waiters */
	void			*b_fspriv;
	void			*b_fspriv2;
	struct xfs_mount	*b_mount;
	unsigned short		b_error;	/* error code on I/O */
	unsigned int		b_page_count;	/* size of page array */
	unsigned int		b_offset;	/* page offset in first page */
	struct page		**b_pages;	/* array of page pointers */
	struct page		*b_page_array[XB_PAGES]; /* inline pages */
#ifdef XFS_BUF_LOCK_TRACKING
	int			b_last_holder;
#endif
} xfs_buf_t;


/* Finding and Reading Buffers */
extern xfs_buf_t *_xfs_buf_find(xfs_buftarg_t *, xfs_off_t, size_t,
				xfs_buf_flags_t, xfs_buf_t *);
#define xfs_incore(buftarg,blkno,len,lockit) \
	_xfs_buf_find(buftarg, blkno ,len, lockit, NULL)

extern xfs_buf_t *xfs_buf_get(xfs_buftarg_t *, xfs_off_t, size_t,
				xfs_buf_flags_t);
extern xfs_buf_t *xfs_buf_read(xfs_buftarg_t *, xfs_off_t, size_t,
				xfs_buf_flags_t);

extern xfs_buf_t *xfs_buf_get_empty(size_t, xfs_buftarg_t *);
extern xfs_buf_t *xfs_buf_get_noaddr(size_t, xfs_buftarg_t *);
extern int xfs_buf_associate_memory(xfs_buf_t *, void *, size_t);
extern void xfs_buf_hold(xfs_buf_t *);
extern void xfs_buf_readahead(xfs_buftarg_t *, xfs_off_t, size_t,
				xfs_buf_flags_t);

/* 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_cond_lock(xfs_buf_t *);
extern int xfs_buf_lock_value(xfs_buf_t *);
extern void xfs_buf_lock(xfs_buf_t *);
extern void xfs_buf_unlock(xfs_buf_t *);

/* Buffer Read and Write Routines */
extern int xfs_bawrite(void *mp, xfs_buf_t *bp);
extern void xfs_bdwrite(void *mp, xfs_buf_t *bp);
extern void xfs_buf_ioend(xfs_buf_t *,	int);
extern void xfs_buf_ioerror(xfs_buf_t *, int);
extern int xfs_buf_iorequest(xfs_buf_t *);
extern int xfs_buf_iowait(xfs_buf_t *);
extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, xfs_caddr_t,
				xfs_buf_rw_t);

static inline int xfs_buf_iostrategy(xfs_buf_t *bp)
{
	return bp->b_strat ? bp->b_strat(bp) : xfs_buf_iorequest(bp);
}

static inline int xfs_buf_geterror(xfs_buf_t *bp)
{
	return bp ? bp->b_error : ENOMEM;
}

/* Buffer Utility Routines */
extern xfs_caddr_t xfs_buf_offset(xfs_buf_t *, size_t);

/* Pinning Buffer Storage in Memory */
extern void xfs_buf_pin(xfs_buf_t *);
extern void xfs_buf_unpin(xfs_buf_t *);
extern int xfs_buf_ispin(xfs_buf_t *);

/* Delayed Write Buffer Routines */
extern void xfs_buf_delwri_dequeue(xfs_buf_t *);

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

#ifdef XFS_BUF_TRACE
extern ktrace_t *xfs_buf_trace_buf;
extern void xfs_buf_trace(xfs_buf_t *, char *, void *, void *);
#else
#define xfs_buf_trace(bp,id,ptr,ra)	do { } while (0)
#endif

#define xfs_buf_target_name(target)	\
	({ char __b[BDEVNAME_SIZE]; bdevname((target)->bt_bdev, __b); __b; })


#define XFS_B_ASYNC		XBF_ASYNC
#define XFS_B_DELWRI		XBF_DELWRI
#define XFS_B_READ		XBF_READ
#define XFS_B_WRITE		XBF_WRITE
#define XFS_B_STALE		XBF_STALE

#define XFS_BUF_TRYLOCK		XBF_TRYLOCK
#define XFS_INCORE_TRYLOCK	XBF_TRYLOCK
#define XFS_BUF_LOCK		XBF_LOCK
#define XFS_BUF_MAPPED		XBF_MAPPED

#define BUF_BUSY		XBF_DONT_BLOCK

#define XFS_BUF_BFLAGS(bp)	((bp)->b_flags)
#define XFS_BUF_ZEROFLAGS(bp)	((bp)->b_flags &= \
		~(XBF_READ|XBF_WRITE|XBF_ASYNC|XBF_DELWRI|XBF_ORDERED))

#define XFS_BUF_STALE(bp)	((bp)->b_flags |= XFS_B_STALE)
#define XFS_BUF_UNSTALE(bp)	((bp)->b_flags &= ~XFS_B_STALE)
#define XFS_BUF_ISSTALE(bp)	((bp)->b_flags & XFS_B_STALE)
#define XFS_BUF_SUPER_STALE(bp)	do {				\
					XFS_BUF_STALE(bp);	\
					xfs_buf_delwri_dequeue(bp);	\
					XFS_BUF_DONE(bp);	\
				} while (0)

#define XFS_BUF_MANAGE		XBF_FS_MANAGED
#define XFS_BUF_UNMANAGE(bp)	((bp)->b_flags &= ~XBF_FS_MANAGED)

#define XFS_BUF_DELAYWRITE(bp)		((bp)->b_flags |= XBF_DELWRI)
#define XFS_BUF_UNDELAYWRITE(bp)	xfs_buf_delwri_dequeue(bp)
#define XFS_BUF_ISDELAYWRITE(bp)	((bp)->b_flags & XBF_DELWRI)

#define XFS_BUF_ERROR(bp,no)	xfs_buf_ioerror(bp,no)
#define XFS_BUF_GETERROR(bp)	xfs_buf_geterror(bp)
#define XFS_BUF_ISERROR(bp)	(xfs_buf_geterror(bp) ? 1 : 0)

#define XFS_BUF_DONE(bp)	((bp)->b_flags |= XBF_DONE)
#define XFS_BUF_UNDONE(bp)	((bp)->b_flags &= ~XBF_DONE)
#define XFS_BUF_ISDONE(bp)	((bp)->b_flags & XBF_DONE)

#define XFS_BUF_BUSY(bp)	do { } while (0)
#define XFS_BUF_UNBUSY(bp)	do { } while (0)
#define XFS_BUF_ISBUSY(bp)	(1)

#define XFS_BUF_ASYNC(bp)	((bp)->b_flags |= XBF_ASYNC)
#define XFS_BUF_UNASYNC(bp)	((bp)->b_flags &= ~XBF_ASYNC)
#define XFS_BUF_ISASYNC(bp)	((bp)->b_flags & XBF_ASYNC)

#define XFS_BUF_ORDERED(bp)	((bp)->b_flags |= XBF_ORDERED)
#define XFS_BUF_UNORDERED(bp)	((bp)->b_flags &= ~XBF_ORDERED)
#define XFS_BUF_ISORDERED(bp)	((bp)->b_flags & XBF_ORDERED)

#define XFS_BUF_HOLD(bp)	xfs_buf_hold(bp)
#define XFS_BUF_READ(bp)	((bp)->b_flags |= XBF_READ)
#define XFS_BUF_UNREAD(bp)	((bp)->b_flags &= ~XBF_READ)
#define XFS_BUF_ISREAD(bp)	((bp)->b_flags & XBF_READ)

#define XFS_BUF_WRITE(bp)	((bp)->b_flags |= XBF_WRITE)
#define XFS_BUF_UNWRITE(bp)	((bp)->b_flags &= ~XBF_WRITE)
#define XFS_BUF_ISWRITE(bp)	((bp)->b_flags & XBF_WRITE)

#define XFS_BUF_IODONE_FUNC(bp)			((bp)->b_iodone)
#define XFS_BUF_SET_IODONE_FUNC(bp, func)	((bp)->b_iodone = (func))
#define XFS_BUF_CLR_IODONE_FUNC(bp)		((bp)->b_iodone = NULL)
#define XFS_BUF_SET_BDSTRAT_FUNC(bp, func)	((bp)->b_strat = (func))
#define XFS_BUF_CLR_BDSTRAT_FUNC(bp)		((bp)->b_strat = NULL)

#define XFS_BUF_FSPRIVATE(bp, type)		((type)(bp)->b_fspriv)
#define XFS_BUF_SET_FSPRIVATE(bp, val)		((bp)->b_fspriv = (void*)(val))
#define XFS_BUF_FSPRIVATE2(bp, type)		((type)(bp)->b_fspriv2)
#define XFS_BUF_SET_FSPRIVATE2(bp, val)		((bp)->b_fspriv2 = (void*)(val))
#define XFS_BUF_SET_START(bp)			do { } while (0)
#define XFS_BUF_SET_BRELSE_FUNC(bp, func)	((bp)->b_relse = (func))

#define XFS_BUF_PTR(bp)			(xfs_caddr_t)((bp)->b_addr)
#define XFS_BUF_SET_PTR(bp, val, cnt)	xfs_buf_associate_memory(bp, val, cnt)
#define XFS_BUF_ADDR(bp)		((bp)->b_bn)
#define XFS_BUF_SET_ADDR(bp, bno)	((bp)->b_bn = (xfs_daddr_t)(bno))
#define XFS_BUF_OFFSET(bp)		((bp)->b_file_offset)
#define XFS_BUF_SET_OFFSET(bp, off)	((bp)->b_file_offset = (off))
#define XFS_BUF_COUNT(bp)		((bp)->b_count_desired)
#define XFS_BUF_SET_COUNT(bp, cnt)	((bp)->b_count_desired = (cnt))
#define XFS_BUF_SIZE(bp)		((bp)->b_buffer_length)
#define XFS_BUF_SET_SIZE(bp, cnt)	((bp)->b_buffer_length = (cnt))

#define XFS_BUF_SET_VTYPE_REF(bp, type, ref)	do { } while (0)
#define XFS_BUF_SET_VTYPE(bp, type)		do { } while (0)
#define XFS_BUF_SET_REF(bp, ref)		do { } while (0)

#define XFS_BUF_ISPINNED(bp)	xfs_buf_ispin(bp)

#define XFS_BUF_VALUSEMA(bp)	xfs_buf_lock_value(bp)
#define XFS_BUF_CPSEMA(bp)	(xfs_buf_cond_lock(bp) == 0)
#define XFS_BUF_VSEMA(bp)	xfs_buf_unlock(bp)
#define XFS_BUF_PSEMA(bp,x)	xfs_buf_lock(bp)
#define XFS_BUF_FINISH_IOWAIT(bp)	complete(&bp->b_iowait);

#define XFS_BUF_SET_TARGET(bp, target)	((bp)->b_target = (target))
#define XFS_BUF_TARGET(bp)		((bp)->b_target)
#define XFS_BUFTARG_NAME(target)	xfs_buf_target_name(target)

static inline void xfs_buf_relse(xfs_buf_t *bp)
{
	if (!bp->b_relse)
		xfs_buf_unlock(bp);
	xfs_buf_rele(bp);
}

#define xfs_bpin(bp)		xfs_buf_pin(bp)
#define xfs_bunpin(bp)		xfs_buf_unpin(bp)

#define xfs_buftrace(id, bp)	\
	    xfs_buf_trace(bp, id, NULL, (void *)__builtin_return_address(0))

#define xfs_biodone(bp)		xfs_buf_ioend(bp, 0)

#define xfs_biomove(bp, off, len, data, rw) \
	    xfs_buf_iomove((bp), (off), (len), (data), \
		((rw) == XFS_B_WRITE) ? XBRW_WRITE : XBRW_READ)

#define xfs_biozero(bp, off, len) \
	    xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)


static inline int XFS_bwrite(xfs_buf_t *bp)
{
	int	iowait = (bp->b_flags & XBF_ASYNC) == 0;
	int	error = 0;

	if (!iowait)
		bp->b_flags |= _XBF_RUN_QUEUES;

	xfs_buf_delwri_dequeue(bp);
	xfs_buf_iostrategy(bp);
	if (iowait) {
		error = xfs_buf_iowait(bp);
		xfs_buf_relse(bp);
	}
	return error;
}

#define XFS_bdstrat(bp) xfs_buf_iorequest(bp)

#define xfs_iowait(bp)	xfs_buf_iowait(bp)

#define xfs_baread(target, rablkno, ralen)  \
	xfs_buf_readahead((target), (rablkno), (ralen), XBF_DONT_BLOCK)


/*
 *	Handling of buftargs.
 */
extern xfs_buftarg_t *xfs_alloc_buftarg(struct block_device *, int);
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, unsigned int);
extern int xfs_flush_buftarg(xfs_buftarg_t *, int);
#ifdef CONFIG_KDB_MODULES
extern struct list_head *xfs_get_buftarg_list(void);
#endif

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

#define xfs_binval(buftarg)		xfs_flush_buftarg(buftarg, 1)
#define XFS_bflush(buftarg)		xfs_flush_buftarg(buftarg, 1)

#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>
	24	#include <asm/system.h>
	25	#include <linux/mm.h>
	26	#include <linux/fs.h>
	27	#include <linux/buffer_head.h>
	28	#include <linux/uio.h>
	29
	30	/*
	31	* Base types
	32	*/
	33
ce8e922c NS	34	#define XFS_BUF_DADDR_NULL ((xfs_daddr_t) (-1LL))
	35
	36	#define xfs_buf_ctob(pp) ((pp) * PAGE_CACHE_SIZE)
	37	#define xfs_buf_btoc(dd) (((dd) + PAGE_CACHE_SIZE-1) >> PAGE_CACHE_SHIFT)
	38	#define xfs_buf_btoct(dd) ((dd) >> PAGE_CACHE_SHIFT)
	39	#define xfs_buf_poff(aa) ((aa) & ~PAGE_CACHE_MASK)
	40
	41	typedef enum {
	42	XBRW_READ = 1, /* transfer into target memory */
	43	XBRW_WRITE = 2, /* transfer from target memory */
	44	XBRW_ZERO = 3, /* Zero target memory */
	45	} xfs_buf_rw_t;
	46
	47	typedef enum {
	48	XBF_READ = (1 << 0), /* buffer intended for reading from device */
	49	XBF_WRITE = (1 << 1), /* buffer intended for writing to device */
	50	XBF_MAPPED = (1 << 2), /* buffer mapped (b_addr valid) */
	51	XBF_ASYNC = (1 << 4), /* initiator will not wait for completion */
	52	XBF_DONE = (1 << 5), /* all pages in the buffer uptodate */
	53	XBF_DELWRI = (1 << 6), /* buffer has dirty pages */
	54	XBF_STALE = (1 << 7), /* buffer has been staled, do not find it */
	55	XBF_FS_MANAGED = (1 << 8), /* filesystem controls freeing memory */
	56	XBF_ORDERED = (1 << 11), /* use ordered writes */
	57	XBF_READ_AHEAD = (1 << 12), /* asynchronous read-ahead */
1da177e4 LT	58
1da177e4 LT	59	/* flags used only as arguments to access routines */
ce8e922c NS	60	XBF_LOCK = (1 << 14), /* lock requested */
	61	XBF_TRYLOCK = (1 << 15), /* lock requested, but do not wait */
	62	XBF_DONT_BLOCK = (1 << 16), /* do not block in current thread */
1da177e4 LT	63
1da177e4 LT	64	/* flags used only internally */
ce8e922c	65	_XBF_PAGE_CACHE = (1 << 17),/* backed by pagecache */
1fa40b01	66	_XBF_PAGES = (1 << 18), /* backed by refcounted pages */
ce8e922c NS	67	_XBF_RUN_QUEUES = (1 << 19),/* run block device task queue */
ce8e922c NS	68	_XBF_DELWRI_Q = (1 << 21), /* buffer on delwri queue */
6ab455ee CH	69
	70	/*
	71	* Special flag for supporting metadata blocks smaller than a FSB.
	72	*
	73	* In this case we can have multiple xfs_buf_t on a single page and
	74	* need to lock out concurrent xfs_buf_t readers as they only
	75	* serialise access to the buffer.
	76	*
	77	* If the FSB size >= PAGE_CACHE_SIZE case, we have no serialisation
	78	* between reads of the page. Hence we can have one thread read the
	79	* page and modify it, but then race with another thread that thinks
	80	* the page is not up-to-date and hence reads it again.
	81	*
	82	* The result is that the first modifcation to the page is lost.
	83	* This sort of AGF/AGI reading race can happen when unlinking inodes
	84	* that require truncation and results in the AGI unlinked list
	85	* modifications being lost.
	86	*/
	87	_XBF_PAGE_LOCKED = (1 << 22),
73f6aa4d CH	88
	89	/*
	90	* If we try a barrier write, but it fails we have to communicate
	91	* this to the upper layers. Unfortunately b_error gets overwritten
	92	* when the buffer is re-issued so we have to add another flag to
	93	* keep this information.
	94	*/
	95	_XFS_BARRIER_FAILED = (1 << 23),
ce8e922c	96	} xfs_buf_flags_t;
1da177e4	97
ce8e922c	98	typedef enum {
5e6a07df DC	99	XBT_FORCE_SLEEP = 0,
5e6a07df DC	100	XBT_FORCE_FLUSH = 1,
ce8e922c	101	} xfs_buftarg_flags_t;
1da177e4 LT	102
	103	typedef struct xfs_bufhash {
	104	struct list_head bh_list;
	105	spinlock_t bh_lock;
	106	} xfs_bufhash_t;
	107
	108	typedef struct xfs_buftarg {
ce8e922c NS	109	dev_t bt_dev;
	110	struct block_device *bt_bdev;
	111	struct address_space *bt_mapping;
	112	unsigned int bt_bsize;
	113	unsigned int bt_sshift;
	114	size_t bt_smask;
	115
	116	/* per device buffer hash table */
1da177e4 LT	117	uint bt_hashmask;
	118	uint bt_hashshift;
	119	xfs_bufhash_t *bt_hash;
a6867a68 DC	120
	121	/* per device delwri queue */
	122	struct task_struct *bt_task;
	123	struct list_head bt_list;
	124	struct list_head bt_delwrite_queue;
	125	spinlock_t bt_delwrite_lock;
ce8e922c	126	unsigned long bt_flags;
1da177e4 LT	127	} xfs_buftarg_t;
	128
	129	/*
ce8e922c	130	* xfs_buf_t: Buffer structure for pagecache-based buffers
1da177e4	131	*
ce8e922c NS	132	* This buffer structure is used by the pagecache buffer management routines
	133	* to refer to an assembly of pages forming a logical buffer.
	134	*
	135	* The buffer structure is used on a temporary basis only, and discarded when
	136	* released. The real data storage is recorded in the pagecache. Buffers are
1da177e4 LT	137	* hashed to the block device on which the file system resides.
	138	*/
	139
	140	struct xfs_buf;
ce8e922c NS	141	typedef void (xfs_buf_iodone_t)(struct xfs_buf );
	142	typedef void (xfs_buf_relse_t)(struct xfs_buf );
	143	typedef int (xfs_buf_bdstrat_t)(struct xfs_buf );
1da177e4	144
ce8e922c	145	#define XB_PAGES 2
1da177e4 LT	146
1da177e4 LT	147	typedef struct xfs_buf {
ce8e922c NS	148	struct semaphore b_sema; /* semaphore for lockables */
	149	unsigned long b_queuetime; /* time buffer was queued */
	150	atomic_t b_pin_count; /* pin count */
	151	wait_queue_head_t b_waiters; /* unpin waiters */
	152	struct list_head b_list;
	153	xfs_buf_flags_t b_flags; /* status flags */
	154	struct list_head b_hash_list; /* hash table list */
	155	xfs_bufhash_t b_hash; / hash table list start */
	156	xfs_buftarg_t b_target; / buffer target (device) */
	157	atomic_t b_hold; /* reference count */
	158	xfs_daddr_t b_bn; /* block number for I/O */
	159	xfs_off_t b_file_offset; /* offset in file */
	160	size_t b_buffer_length;/* size of buffer in bytes */
	161	size_t b_count_desired;/* desired transfer size */
	162	void b_addr; / virtual address of buffer */
	163	struct work_struct b_iodone_work;
	164	atomic_t b_io_remaining; /* #outstanding I/O requests */
	165	xfs_buf_iodone_t b_iodone; /* I/O completion function */
	166	xfs_buf_relse_t b_relse; /* releasing function */
	167	xfs_buf_bdstrat_t b_strat; /* pre-write function */
b4dd330b	168	struct completion b_iowait; /* queue for I/O waiters */
ce8e922c NS	169	void *b_fspriv;
ce8e922c NS	170	void *b_fspriv2;
15ac08a8	171	struct xfs_mount *b_mount;
ce8e922c	172	unsigned short b_error; /* error code on I/O */
ce8e922c NS	173	unsigned int b_page_count; /* size of page array */
	174	unsigned int b_offset; /* page offset in first page */
	175	struct page *b_pages; / array of page pointers */
	176	struct page b_page_array[XB_PAGES]; / inline pages */
	177	#ifdef XFS_BUF_LOCK_TRACKING
	178	int b_last_holder;
1da177e4 LT	179	#endif
	180	} xfs_buf_t;
	181
	182
	183	/* Finding and Reading Buffers */
ce8e922c NS	184	extern xfs_buf_t _xfs_buf_find(xfs_buftarg_t , xfs_off_t, size_t,
ce8e922c NS	185	xfs_buf_flags_t, xfs_buf_t *);
1da177e4	186	#define xfs_incore(buftarg,blkno,len,lockit) \
ce8e922c	187	_xfs_buf_find(buftarg, blkno ,len, lockit, NULL)
1da177e4	188
6ad112bf	189	extern xfs_buf_t xfs_buf_get(xfs_buftarg_t , xfs_off_t, size_t,
ce8e922c	190	xfs_buf_flags_t);
6ad112bf	191	extern xfs_buf_t xfs_buf_read(xfs_buftarg_t , xfs_off_t, size_t,
ce8e922c	192	xfs_buf_flags_t);
1da177e4	193
ce8e922c NS	194	extern xfs_buf_t xfs_buf_get_empty(size_t, xfs_buftarg_t );
	195	extern xfs_buf_t xfs_buf_get_noaddr(size_t, xfs_buftarg_t );
	196	extern int xfs_buf_associate_memory(xfs_buf_t , void , size_t);
	197	extern void xfs_buf_hold(xfs_buf_t *);
	198	extern void xfs_buf_readahead(xfs_buftarg_t *, xfs_off_t, size_t,
	199	xfs_buf_flags_t);
1da177e4 LT	200
1da177e4 LT	201	/* Releasing Buffers */
ce8e922c NS	202	extern void xfs_buf_free(xfs_buf_t *);
ce8e922c NS	203	extern void xfs_buf_rele(xfs_buf_t *);
1da177e4 LT	204
1da177e4 LT	205	/* Locking and Unlocking Buffers */
ce8e922c NS	206	extern int xfs_buf_cond_lock(xfs_buf_t *);
	207	extern int xfs_buf_lock_value(xfs_buf_t *);
	208	extern void xfs_buf_lock(xfs_buf_t *);
	209	extern void xfs_buf_unlock(xfs_buf_t *);
1da177e4 LT	210
1da177e4 LT	211	/* Buffer Read and Write Routines */
5d765b97 CH	212	extern int xfs_bawrite(void mp, xfs_buf_t bp);
5d765b97 CH	213	extern void xfs_bdwrite(void mp, xfs_buf_t bp);
ce8e922c NS	214	extern void xfs_buf_ioend(xfs_buf_t *, int);
ce8e922c NS	215	extern void xfs_buf_ioerror(xfs_buf_t *, int);
ce8e922c NS	216	extern int xfs_buf_iorequest(xfs_buf_t *);
	217	extern int xfs_buf_iowait(xfs_buf_t *);
	218	extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, xfs_caddr_t,
	219	xfs_buf_rw_t);
	220
	221	static inline int xfs_buf_iostrategy(xfs_buf_t *bp)
1da177e4	222	{
ce8e922c	223	return bp->b_strat ? bp->b_strat(bp) : xfs_buf_iorequest(bp);
1da177e4 LT	224	}
1da177e4 LT	225
ce8e922c	226	static inline int xfs_buf_geterror(xfs_buf_t *bp)
1da177e4	227	{
ce8e922c	228	return bp ? bp->b_error : ENOMEM;
1da177e4 LT	229	}
	230
	231	/* Buffer Utility Routines */
ce8e922c	232	extern xfs_caddr_t xfs_buf_offset(xfs_buf_t *, size_t);
1da177e4 LT	233
1da177e4 LT	234	/* Pinning Buffer Storage in Memory */
ce8e922c NS	235	extern void xfs_buf_pin(xfs_buf_t *);
	236	extern void xfs_buf_unpin(xfs_buf_t *);
	237	extern int xfs_buf_ispin(xfs_buf_t *);
1da177e4 LT	238
1da177e4 LT	239	/* Delayed Write Buffer Routines */
ce8e922c	240	extern void xfs_buf_delwri_dequeue(xfs_buf_t *);
1da177e4 LT	241
1da177e4 LT	242	/* Buffer Daemon Setup Routines */
ce8e922c NS	243	extern int xfs_buf_init(void);
ce8e922c NS	244	extern void xfs_buf_terminate(void);
1da177e4	245
ce8e922c NS	246	#ifdef XFS_BUF_TRACE
	247	extern ktrace_t *xfs_buf_trace_buf;
	248	extern void xfs_buf_trace(xfs_buf_t , char , void , void );
1da177e4	249	#else
ce8e922c	250	#define xfs_buf_trace(bp,id,ptr,ra) do { } while (0)
1da177e4 LT	251	#endif
1da177e4 LT	252
ce8e922c NS	253	#define xfs_buf_target_name(target) \
ce8e922c NS	254	({ char __b[BDEVNAME_SIZE]; bdevname((target)->bt_bdev, __b); __b; })
1da177e4 LT	255
1da177e4 LT	256
ce8e922c NS	257	#define XFS_B_ASYNC XBF_ASYNC
	258	#define XFS_B_DELWRI XBF_DELWRI
	259	#define XFS_B_READ XBF_READ
	260	#define XFS_B_WRITE XBF_WRITE
	261	#define XFS_B_STALE XBF_STALE
1da177e4	262
ce8e922c NS	263	#define XFS_BUF_TRYLOCK XBF_TRYLOCK
	264	#define XFS_INCORE_TRYLOCK XBF_TRYLOCK
	265	#define XFS_BUF_LOCK XBF_LOCK
	266	#define XFS_BUF_MAPPED XBF_MAPPED
1da177e4	267
ce8e922c NS	268	#define BUF_BUSY XBF_DONT_BLOCK
	269
	270	#define XFS_BUF_BFLAGS(bp) ((bp)->b_flags)
f5faad79 NS	271	#define XFS_BUF_ZEROFLAGS(bp) ((bp)->b_flags &= \
f5faad79 NS	272	~(XBF_READ\|XBF_WRITE\|XBF_ASYNC\|XBF_DELWRI\|XBF_ORDERED))
ce8e922c NS	273
	274	#define XFS_BUF_STALE(bp) ((bp)->b_flags \|= XFS_B_STALE)
	275	#define XFS_BUF_UNSTALE(bp) ((bp)->b_flags &= ~XFS_B_STALE)
	276	#define XFS_BUF_ISSTALE(bp) ((bp)->b_flags & XFS_B_STALE)
	277	#define XFS_BUF_SUPER_STALE(bp) do { \
	278	XFS_BUF_STALE(bp); \
	279	xfs_buf_delwri_dequeue(bp); \
	280	XFS_BUF_DONE(bp); \
	281	} while (0)
1da177e4	282
ce8e922c NS	283	#define XFS_BUF_MANAGE XBF_FS_MANAGED
	284	#define XFS_BUF_UNMANAGE(bp) ((bp)->b_flags &= ~XBF_FS_MANAGED)
	285
	286	#define XFS_BUF_DELAYWRITE(bp) ((bp)->b_flags \|= XBF_DELWRI)
	287	#define XFS_BUF_UNDELAYWRITE(bp) xfs_buf_delwri_dequeue(bp)
	288	#define XFS_BUF_ISDELAYWRITE(bp) ((bp)->b_flags & XBF_DELWRI)
	289
	290	#define XFS_BUF_ERROR(bp,no) xfs_buf_ioerror(bp,no)
	291	#define XFS_BUF_GETERROR(bp) xfs_buf_geterror(bp)
	292	#define XFS_BUF_ISERROR(bp) (xfs_buf_geterror(bp) ? 1 : 0)
	293
	294	#define XFS_BUF_DONE(bp) ((bp)->b_flags \|= XBF_DONE)
	295	#define XFS_BUF_UNDONE(bp) ((bp)->b_flags &= ~XBF_DONE)
	296	#define XFS_BUF_ISDONE(bp) ((bp)->b_flags & XBF_DONE)
	297
	298	#define XFS_BUF_BUSY(bp) do { } while (0)
	299	#define XFS_BUF_UNBUSY(bp) do { } while (0)
	300	#define XFS_BUF_ISBUSY(bp) (1)
	301
	302	#define XFS_BUF_ASYNC(bp) ((bp)->b_flags \|= XBF_ASYNC)
	303	#define XFS_BUF_UNASYNC(bp) ((bp)->b_flags &= ~XBF_ASYNC)
	304	#define XFS_BUF_ISASYNC(bp) ((bp)->b_flags & XBF_ASYNC)
	305
	306	#define XFS_BUF_ORDERED(bp) ((bp)->b_flags \|= XBF_ORDERED)
	307	#define XFS_BUF_UNORDERED(bp) ((bp)->b_flags &= ~XBF_ORDERED)
	308	#define XFS_BUF_ISORDERED(bp) ((bp)->b_flags & XBF_ORDERED)
ce8e922c NS	309
	310	#define XFS_BUF_HOLD(bp) xfs_buf_hold(bp)
	311	#define XFS_BUF_READ(bp) ((bp)->b_flags \|= XBF_READ)
	312	#define XFS_BUF_UNREAD(bp) ((bp)->b_flags &= ~XBF_READ)
	313	#define XFS_BUF_ISREAD(bp) ((bp)->b_flags & XBF_READ)
	314
	315	#define XFS_BUF_WRITE(bp) ((bp)->b_flags \|= XBF_WRITE)
	316	#define XFS_BUF_UNWRITE(bp) ((bp)->b_flags &= ~XBF_WRITE)
	317	#define XFS_BUF_ISWRITE(bp) ((bp)->b_flags & XBF_WRITE)
	318
ce8e922c NS	319	#define XFS_BUF_IODONE_FUNC(bp) ((bp)->b_iodone)
	320	#define XFS_BUF_SET_IODONE_FUNC(bp, func) ((bp)->b_iodone = (func))
	321	#define XFS_BUF_CLR_IODONE_FUNC(bp) ((bp)->b_iodone = NULL)
	322	#define XFS_BUF_SET_BDSTRAT_FUNC(bp, func) ((bp)->b_strat = (func))
	323	#define XFS_BUF_CLR_BDSTRAT_FUNC(bp) ((bp)->b_strat = NULL)
	324
	325	#define XFS_BUF_FSPRIVATE(bp, type) ((type)(bp)->b_fspriv)
	326	#define XFS_BUF_SET_FSPRIVATE(bp, val) ((bp)->b_fspriv = (void*)(val))
	327	#define XFS_BUF_FSPRIVATE2(bp, type) ((type)(bp)->b_fspriv2)
	328	#define XFS_BUF_SET_FSPRIVATE2(bp, val) ((bp)->b_fspriv2 = (void*)(val))
ce8e922c NS	329	#define XFS_BUF_SET_START(bp) do { } while (0)
	330	#define XFS_BUF_SET_BRELSE_FUNC(bp, func) ((bp)->b_relse = (func))
	331
	332	#define XFS_BUF_PTR(bp) (xfs_caddr_t)((bp)->b_addr)
	333	#define XFS_BUF_SET_PTR(bp, val, cnt) xfs_buf_associate_memory(bp, val, cnt)
	334	#define XFS_BUF_ADDR(bp) ((bp)->b_bn)
	335	#define XFS_BUF_SET_ADDR(bp, bno) ((bp)->b_bn = (xfs_daddr_t)(bno))
	336	#define XFS_BUF_OFFSET(bp) ((bp)->b_file_offset)
	337	#define XFS_BUF_SET_OFFSET(bp, off) ((bp)->b_file_offset = (off))
	338	#define XFS_BUF_COUNT(bp) ((bp)->b_count_desired)
	339	#define XFS_BUF_SET_COUNT(bp, cnt) ((bp)->b_count_desired = (cnt))
	340	#define XFS_BUF_SIZE(bp) ((bp)->b_buffer_length)
	341	#define XFS_BUF_SET_SIZE(bp, cnt) ((bp)->b_buffer_length = (cnt))
	342
	343	#define XFS_BUF_SET_VTYPE_REF(bp, type, ref) do { } while (0)
	344	#define XFS_BUF_SET_VTYPE(bp, type) do { } while (0)
	345	#define XFS_BUF_SET_REF(bp, ref) do { } while (0)
	346
	347	#define XFS_BUF_ISPINNED(bp) xfs_buf_ispin(bp)
	348
	349	#define XFS_BUF_VALUSEMA(bp) xfs_buf_lock_value(bp)
	350	#define XFS_BUF_CPSEMA(bp) (xfs_buf_cond_lock(bp) == 0)
	351	#define XFS_BUF_VSEMA(bp) xfs_buf_unlock(bp)
	352	#define XFS_BUF_PSEMA(bp,x) xfs_buf_lock(bp)
b4dd330b	353	#define XFS_BUF_FINISH_IOWAIT(bp) complete(&bp->b_iowait);
ce8e922c NS	354
	355	#define XFS_BUF_SET_TARGET(bp, target) ((bp)->b_target = (target))
	356	#define XFS_BUF_TARGET(bp) ((bp)->b_target)
	357	#define XFS_BUFTARG_NAME(target) xfs_buf_target_name(target)
	358
ce8e922c	359	static inline void xfs_buf_relse(xfs_buf_t *bp)
1da177e4	360	{
ce8e922c NS	361	if (!bp->b_relse)
	362	xfs_buf_unlock(bp);
	363	xfs_buf_rele(bp);
1da177e4 LT	364	}
1da177e4 LT	365
ce8e922c NS	366	#define xfs_bpin(bp) xfs_buf_pin(bp)
ce8e922c NS	367	#define xfs_bunpin(bp) xfs_buf_unpin(bp)
1da177e4 LT	368
1da177e4 LT	369	#define xfs_buftrace(id, bp) \
ce8e922c	370	xfs_buf_trace(bp, id, NULL, (void *)__builtin_return_address(0))
1da177e4	371
ce8e922c	372	#define xfs_biodone(bp) xfs_buf_ioend(bp, 0)
1da177e4	373
ce8e922c NS	374	#define xfs_biomove(bp, off, len, data, rw) \
	375	xfs_buf_iomove((bp), (off), (len), (data), \
	376	((rw) == XFS_B_WRITE) ? XBRW_WRITE : XBRW_READ)
1da177e4	377
ce8e922c NS	378	#define xfs_biozero(bp, off, len) \
ce8e922c NS	379	xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)
1da177e4 LT	380
1da177e4 LT	381
ce8e922c	382	static inline int XFS_bwrite(xfs_buf_t *bp)
1da177e4	383	{
ce8e922c	384	int iowait = (bp->b_flags & XBF_ASYNC) == 0;
1da177e4 LT	385	int error = 0;
	386
	387	if (!iowait)
ce8e922c	388	bp->b_flags \|= _XBF_RUN_QUEUES;
1da177e4	389
ce8e922c NS	390	xfs_buf_delwri_dequeue(bp);
ce8e922c NS	391	xfs_buf_iostrategy(bp);
1da177e4	392	if (iowait) {
ce8e922c NS	393	error = xfs_buf_iowait(bp);
ce8e922c NS	394	xfs_buf_relse(bp);
1da177e4 LT	395	}
	396	return error;
	397	}
	398
ce8e922c	399	#define XFS_bdstrat(bp) xfs_buf_iorequest(bp)
1da177e4	400
ce8e922c	401	#define xfs_iowait(bp) xfs_buf_iowait(bp)
1da177e4 LT	402
1da177e4 LT	403	#define xfs_baread(target, rablkno, ralen) \
ce8e922c	404	xfs_buf_readahead((target), (rablkno), (ralen), XBF_DONT_BLOCK)
1da177e4 LT	405
	406
	407	/*
	408	* Handling of buftargs.
	409	*/
1da177e4	410	extern xfs_buftarg_t xfs_alloc_buftarg(struct block_device , int);
c141b292	411	extern void xfs_free_buftarg(struct xfs_mount , struct xfs_buftarg );
1da177e4 LT	412	extern void xfs_wait_buftarg(xfs_buftarg_t *);
1da177e4 LT	413	extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int, unsigned int);
1da177e4	414	extern int xfs_flush_buftarg(xfs_buftarg_t *, int);
e6a0e9cd TS	415	#ifdef CONFIG_KDB_MODULES
	416	extern struct list_head *xfs_get_buftarg_list(void);
	417	#endif
1da177e4	418
ce8e922c NS	419	#define xfs_getsize_buftarg(buftarg) block_size((buftarg)->bt_bdev)
	420	#define xfs_readonly_buftarg(buftarg) bdev_read_only((buftarg)->bt_bdev)
	421
	422	#define xfs_binval(buftarg) xfs_flush_buftarg(buftarg, 1)
	423	#define XFS_bflush(buftarg) xfs_flush_buftarg(buftarg, 1)
1da177e4 LT	424
1da177e4 LT	425	#endif /* __XFS_BUF_H__ */