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

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

#include <linux/list.h>
#include <linux/types.h>
#include <linux/spinlock.h>
#include <linux/mm.h>
#include <linux/fs.h>
#include <linux/dax.h>
#include <linux/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 list_head	b_li_list;	/* Log items list head */
	struct xfs_trans	*b_transp;
	struct page		**b_pages;	/* array of page pointers */
	struct page		*b_page_array[XB_PAGES]; /* inline pages */
	struct xfs_buf_map	*b_maps;	/* compound buffer map */
	struct xfs_buf_map	__b_map;	/* inline compound buffer map */
	int			b_map_count;
	int			b_io_length;	/* IO size in BBs */
	atomic_t		b_pin_count;	/* pin count */
	atomic_t		b_io_remaining;	/* #outstanding I/O requests */
	unsigned int		b_page_count;	/* size of page array */
	unsigned int		b_offset;	/* page offset in first page */
	int			b_error;	/* error code on I/O */

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

	const struct xfs_buf_ops	*b_ops;

#ifdef XFS_BUF_LOCK_TRACKING
	int			b_last_holder;
#endif
} xfs_buf_t;

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

#endif	/* __XFS_BUF_H__ */
Commit	Line	Data
0b61f8a4	1	// SPDX-License-Identifier: GPL-2.0
1da177e4	2	/*
7b718769 NS	3	* Copyright (c) 2000-2005 Silicon Graphics, Inc.
7b718769 NS	4	* All Rights Reserved.
1da177e4	5	*/
1da177e4 LT	6	#ifndef __XFS_BUF_H__
	7	#define __XFS_BUF_H__
	8
1da177e4 LT	9	#include <linux/list.h>
	10	#include <linux/types.h>
	11	#include <linux/spinlock.h>
1da177e4 LT	12	#include <linux/mm.h>
1da177e4 LT	13	#include <linux/fs.h>
c94c2acf	14	#include <linux/dax.h>
1da177e4 LT	15	#include <linux/buffer_head.h>
1da177e4 LT	16	#include <linux/uio.h>
e80dfa19	17	#include <linux/list_lru.h>
1da177e4 LT	18
	19	/*
	20	* Base types
	21	*/
	22
ce8e922c NS	23	#define XFS_BUF_DADDR_NULL ((xfs_daddr_t) (-1LL))
ce8e922c NS	24
ce8e922c NS	25	typedef enum {
	26	XBRW_READ = 1, /* transfer into target memory */
	27	XBRW_WRITE = 2, /* transfer from target memory */
	28	XBRW_ZERO = 3, /* Zero target memory */
	29	} xfs_buf_rw_t;
	30
6fb8a90a CM	31	#define XBF_READ (1 << 0) /* buffer intended for reading from device */
	32	#define XBF_WRITE (1 << 1) /* buffer intended for writing to device */
	33	#define XBF_READ_AHEAD (1 << 2) /* asynchronous read-ahead */
c891c30a	34	#define XBF_NO_IOACCT (1 << 3) /* bypass I/O accounting (non-LRU bufs) */
6fb8a90a CM	35	#define XBF_ASYNC (1 << 4) /* initiator will not wait for completion */
	36	#define XBF_DONE (1 << 5) /* all pages in the buffer uptodate */
	37	#define XBF_STALE (1 << 6) /* buffer has been staled, do not find it */
ac8809f9	38	#define XBF_WRITE_FAIL (1 << 24)/* async writes have failed on this buffer */
1d5ae5df CH	39
1d5ae5df CH	40	/* I/O hints for the BIO layer */
6fb8a90a CM	41	#define XBF_SYNCIO (1 << 10)/* treat this buffer as synchronous I/O */
	42	#define XBF_FUA (1 << 11)/* force cache write through mode */
	43	#define XBF_FLUSH (1 << 12)/* flush the disk cache before a write */
1da177e4	44
807cbbdb	45	/* flags used only as arguments to access routines */
6fb8a90a CM	46	#define XBF_TRYLOCK (1 << 16)/* lock requested, but do not wait */
6fb8a90a CM	47	#define XBF_UNMAPPED (1 << 17)/* do not map the buffer */
1da177e4	48
807cbbdb	49	/* flags used only internally */
6fb8a90a CM	50	#define _XBF_PAGES (1 << 20)/* backed by refcounted pages */
	51	#define _XBF_KMEM (1 << 21)/* backed by heap memory */
	52	#define _XBF_DELWRI_Q (1 << 22)/* buffer on a delwri queue */
	53	#define _XBF_COMPOUND (1 << 23)/* compound buffer */
6ab455ee	54
807cbbdb	55	typedef unsigned int xfs_buf_flags_t;
1da177e4	56
0b1b213f CH	57	#define XFS_BUF_FLAGS \
	58	{ XBF_READ, "READ" }, \
	59	{ XBF_WRITE, "WRITE" }, \
1d5ae5df	60	{ XBF_READ_AHEAD, "READ_AHEAD" }, \
1247ec4c	61	{ XBF_NO_IOACCT, "NO_IOACCT" }, \
0b1b213f CH	62	{ XBF_ASYNC, "ASYNC" }, \
0b1b213f CH	63	{ XBF_DONE, "DONE" }, \
0b1b213f	64	{ XBF_STALE, "STALE" }, \
ac8809f9	65	{ XBF_WRITE_FAIL, "WRITE_FAIL" }, \
1d5ae5df CH	66	{ XBF_SYNCIO, "SYNCIO" }, \
	67	{ XBF_FUA, "FUA" }, \
	68	{ XBF_FLUSH, "FLUSH" }, \
6fb8a90a	69	{ XBF_TRYLOCK, "TRYLOCK" }, /* should never be set */\
611c9946	70	{ XBF_UNMAPPED, "UNMAPPED" }, /* ditto */\
0b1b213f	71	{ _XBF_PAGES, "PAGES" }, \
0e6e847f	72	{ _XBF_KMEM, "KMEM" }, \
cbb7baab	73	{ _XBF_DELWRI_Q, "DELWRI_Q" }, \
63db7c81	74	{ _XBF_COMPOUND, "COMPOUND" }
a4082357	75
ac8809f9	76
a4082357 DC	77	/*
	78	* Internal state flags.
	79	*/
	80	#define XFS_BSTATE_DISPOSE (1 << 0) /* buffer being discarded */
63db7c81	81	#define XFS_BSTATE_IN_FLIGHT (1 << 1) /* I/O in flight */
0b1b213f	82
7c71ee78 ES	83	/*
	84	* The xfs_buftarg contains 2 notions of "sector size" -
	85	*
	86	* 1) The metadata sector size, which is the minimum unit and
	87	* alignment of IO which will be performed by metadata operations.
	88	* 2) The device logical sector size
	89	*
	90	* The first is specified at mkfs time, and is stored on-disk in the
	91	* superblock's sb_sectsize.
	92	*
	93	* The latter is derived from the underlying device, and controls direct IO
	94	* alignment constraints.
	95	*/
1da177e4	96	typedef struct xfs_buftarg {
ce8e922c NS	97	dev_t bt_dev;
ce8e922c NS	98	struct block_device *bt_bdev;
486aff5e	99	struct dax_device *bt_daxdev;
ebad861b	100	struct xfs_mount *bt_mount;
6da54179 ES	101	unsigned int bt_meta_sectorsize;
6da54179 ES	102	size_t bt_meta_sectormask;
7c71ee78 ES	103	size_t bt_logical_sectorsize;
7c71ee78 ES	104	size_t bt_logical_sectormask;
ce8e922c	105
ff57ab21 DC	106	/* LRU control structures */
ff57ab21 DC	107	struct shrinker bt_shrinker;
e80dfa19	108	struct list_lru bt_lru;
9c7504aa BF	109
9c7504aa BF	110	struct percpu_counter bt_io_count;
1da177e4 LT	111	} xfs_buftarg_t;
1da177e4 LT	112
1da177e4	113	struct xfs_buf;
ce8e922c	114	typedef void (xfs_buf_iodone_t)(struct xfs_buf );
1da177e4	115
c3f8fc73	116
ce8e922c	117	#define XB_PAGES 2
1da177e4	118
cbb7baab DC	119	struct xfs_buf_map {
	120	xfs_daddr_t bm_bn; /* block number for I/O */
	121	int bm_len; /* size of I/O */
	122	};
	123
3e85c868 DC	124	#define DEFINE_SINGLE_BUF_MAP(map, blkno, numblk) \
	125	struct xfs_buf_map (map) = { .bm_bn = (blkno), .bm_len = (numblk) };
	126
1813dd64	127	struct xfs_buf_ops {
233135b7	128	char *name;
1813dd64 DC	129	void (verify_read)(struct xfs_buf );
1813dd64 DC	130	void (verify_write)(struct xfs_buf );
b5572597	131	xfs_failaddr_t (verify_struct)(struct xfs_buf bp);
1813dd64 DC	132	};
1813dd64 DC	133
1da177e4	134	typedef struct xfs_buf {
50f59e8e DC	135	/*
	136	* first cacheline holds all the fields needed for an uncontended cache
	137	* hit to be fully processed. The semaphore straddles the cacheline
	138	* boundary, but the counter and lock sits on the first cacheline,
	139	* which is the only bit that is touched if we hit the semaphore
	140	* fast-path on locking.
	141	*/
6031e73a	142	struct rhash_head b_rhash_head; /* pag buffer hash node */
cbb7baab	143	xfs_daddr_t b_bn; /* block number of buffer */
4e94b71b	144	int b_length; /* size of buffer in BBs */
50f59e8e	145	atomic_t b_hold; /* reference count */
430cbeb8	146	atomic_t b_lru_ref; /* lru reclaim ref count */
50f59e8e	147	xfs_buf_flags_t b_flags; /* status flags */
ce8e922c	148	struct semaphore b_sema; /* semaphore for lockables */
50f59e8e	149
6fb8a90a CM	150	/*
	151	* concurrent access to b_lru and b_lru_flags are protected by
	152	* bt_lru_lock and not by b_sema
	153	*/
430cbeb8	154	struct list_head b_lru; /* lru list */
a4082357 DC	155	spinlock_t b_lock; /* internal state lock */
a4082357 DC	156	unsigned int b_state; /* internal state flags */
61be9c52	157	int b_io_error; /* internal IO error state */
ce8e922c NS	158	wait_queue_head_t b_waiters; /* unpin waiters */
ce8e922c NS	159	struct list_head b_list;
74f75a0c	160	struct xfs_perag b_pag; / contains rbtree root */
ce8e922c	161	xfs_buftarg_t b_target; / buffer target (device) */
ce8e922c	162	void b_addr; / virtual address of buffer */
b29c70f5 BF	163	struct work_struct b_ioend_work;
b29c70f5 BF	164	struct workqueue_struct b_ioend_wq; / I/O completion wq */
ce8e922c	165	xfs_buf_iodone_t b_iodone; /* I/O completion function */
b4dd330b	166	struct completion b_iowait; /* queue for I/O waiters */
fb1755a6	167	void *b_log_item;
643c8c05	168	struct list_head b_li_list; /* Log items list head */
bf9d9013	169	struct xfs_trans *b_transp;
ce8e922c NS	170	struct page *b_pages; / array of page pointers */
ce8e922c NS	171	struct page b_page_array[XB_PAGES]; / inline pages */
3e85c868	172	struct xfs_buf_map b_maps; / compound buffer map */
d44d9bc6	173	struct xfs_buf_map __b_map; /* inline compound buffer map */
3e85c868	174	int b_map_count;
aa0e8833	175	int b_io_length; /* IO size in BBs */
50f59e8e DC	176	atomic_t b_pin_count; /* pin count */
	177	atomic_t b_io_remaining; /* #outstanding I/O requests */
	178	unsigned int b_page_count; /* size of page array */
	179	unsigned int b_offset; /* page offset in first page */
2451337d	180	int b_error; /* error code on I/O */
a5ea70d2 CM	181
	182	/*
	183	* async write failure retry count. Initialised to zero on the first
	184	* failure, then when it exceeds the maximum configured without a
	185	* success the write is considered to be failed permanently and the
	186	* iodone handler will take appropriate action.
	187	*
	188	* For retry timeouts, we record the jiffie of the first failure. This
	189	* means that we can change the retry timeout for buffers already under
	190	* I/O and thus avoid getting stuck in a retry loop with a long timeout.
	191	*
	192	* last_error is used to ensure that we are getting repeated errors, not
	193	* different errors. e.g. a block device might change ENOSPC to EIO when
	194	* a failure timeout occurs, so we want to re-initialise the error
	195	* retry behaviour appropriately when that happens.
	196	*/
	197	int b_retries;
	198	unsigned long b_first_retry_time; /* in jiffies */
	199	int b_last_error;
	200
1813dd64	201	const struct xfs_buf_ops *b_ops;
cfb02852	202
ce8e922c NS	203	#ifdef XFS_BUF_LOCK_TRACKING
ce8e922c NS	204	int b_last_holder;
1da177e4 LT	205	#endif
	206	} xfs_buf_t;
	207
1da177e4	208	/* Finding and Reading Buffers */
8925a3dc DC	209	struct xfs_buf xfs_buf_incore(struct xfs_buftarg target,
	210	xfs_daddr_t blkno, size_t numblks,
	211	xfs_buf_flags_t flags);
3e85c868 DC	212
	213	struct xfs_buf _xfs_buf_alloc(struct xfs_buftarg target,
	214	struct xfs_buf_map *map, int nmaps,
	215	xfs_buf_flags_t flags);
	216
	217	static inline struct xfs_buf *
	218	xfs_buf_alloc(
	219	struct xfs_buftarg *target,
	220	xfs_daddr_t blkno,
	221	size_t numblks,
	222	xfs_buf_flags_t flags)
	223	{
	224	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	225	return _xfs_buf_alloc(target, &map, 1, flags);
	226	}
1da177e4	227
6dde2707 DC	228	struct xfs_buf xfs_buf_get_map(struct xfs_buftarg target,
	229	struct xfs_buf_map *map, int nmaps,
	230	xfs_buf_flags_t flags);
	231	struct xfs_buf xfs_buf_read_map(struct xfs_buftarg target,
	232	struct xfs_buf_map *map, int nmaps,
1813dd64 DC	233	xfs_buf_flags_t flags,
1813dd64 DC	234	const struct xfs_buf_ops *ops);
6dde2707	235	void xfs_buf_readahead_map(struct xfs_buftarg *target,
c3f8fc73	236	struct xfs_buf_map *map, int nmaps,
1813dd64	237	const struct xfs_buf_ops *ops);
6dde2707 DC	238
	239	static inline struct xfs_buf *
	240	xfs_buf_get(
	241	struct xfs_buftarg *target,
	242	xfs_daddr_t blkno,
	243	size_t numblks,
	244	xfs_buf_flags_t flags)
	245	{
	246	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
	247	return xfs_buf_get_map(target, &map, 1, flags);
	248	}
	249
	250	static inline struct xfs_buf *
	251	xfs_buf_read(
	252	struct xfs_buftarg *target,
	253	xfs_daddr_t blkno,
	254	size_t numblks,
c3f8fc73	255	xfs_buf_flags_t flags,
1813dd64	256	const struct xfs_buf_ops *ops)
6dde2707 DC	257	{
6dde2707 DC	258	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	259	return xfs_buf_read_map(target, &map, 1, flags, ops);
6dde2707 DC	260	}
	261
	262	static inline void
	263	xfs_buf_readahead(
	264	struct xfs_buftarg *target,
	265	xfs_daddr_t blkno,
c3f8fc73	266	size_t numblks,
1813dd64	267	const struct xfs_buf_ops *ops)
6dde2707 DC	268	{
6dde2707 DC	269	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
1813dd64	270	return xfs_buf_readahead_map(target, &map, 1, ops);
6dde2707	271	}
e70b73f8	272
e70b73f8 DC	273	void xfs_buf_set_empty(struct xfs_buf *bp, size_t numblks);
	274	int xfs_buf_associate_memory(struct xfs_buf bp, void mem, size_t length);
	275
	276	struct xfs_buf xfs_buf_get_uncached(struct xfs_buftarg target, size_t numblks,
	277	int flags);
ba372674 DC	278	int xfs_buf_read_uncached(struct xfs_buftarg *target, xfs_daddr_t daddr,
	279	size_t numblks, int flags, struct xfs_buf **bpp,
	280	const struct xfs_buf_ops *ops);
e70b73f8	281	void xfs_buf_hold(struct xfs_buf *bp);
1da177e4 LT	282
1da177e4 LT	283	/* Releasing Buffers */
ce8e922c NS	284	extern void xfs_buf_free(xfs_buf_t *);
ce8e922c NS	285	extern void xfs_buf_rele(xfs_buf_t *);
1da177e4 LT	286
1da177e4 LT	287	/* Locking and Unlocking Buffers */
0c842ad4	288	extern int xfs_buf_trylock(xfs_buf_t *);
ce8e922c NS	289	extern void xfs_buf_lock(xfs_buf_t *);
ce8e922c NS	290	extern void xfs_buf_unlock(xfs_buf_t *);
0c842ad4 CH	291	#define xfs_buf_islocked(bp) \
0c842ad4 CH	292	((bp)->b_sema.count <= 0)
1da177e4 LT	293
1da177e4 LT	294	/* Buffer Read and Write Routines */
c2b006c1	295	extern int xfs_bwrite(struct xfs_buf *bp);
e8aaba9a	296	extern void xfs_buf_ioend(struct xfs_buf *bp);
31ca03c9 DW	297	extern void __xfs_buf_ioerror(struct xfs_buf *bp, int error,
	298	xfs_failaddr_t failaddr);
	299	#define xfs_buf_ioerror(bp, err) __xfs_buf_ioerror((bp), (err), __this_address)
901796af	300	extern void xfs_buf_ioerror_alert(struct xfs_buf , const char func);
595bff75 DC	301	extern void xfs_buf_submit(struct xfs_buf *bp);
595bff75 DC	302	extern int xfs_buf_submit_wait(struct xfs_buf *bp);
b9c48649	303	extern void xfs_buf_iomove(xfs_buf_t , size_t, size_t, void ,
ce8e922c	304	xfs_buf_rw_t);
1a1a3e97 CH	305	#define xfs_buf_zero(bp, off, len) \
1a1a3e97 CH	306	xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)
ce8e922c	307
1da177e4	308	/* Buffer Utility Routines */
88ee2df7	309	extern void xfs_buf_offset(struct xfs_buf , size_t);
5cfd28b6	310	extern void xfs_buf_stale(struct xfs_buf *bp);
1da177e4	311
1da177e4	312	/* Delayed Write Buffer Routines */
20e8a063	313	extern void xfs_buf_delwri_cancel(struct list_head *);
43ff2122 CH	314	extern bool xfs_buf_delwri_queue(struct xfs_buf , struct list_head );
	315	extern int xfs_buf_delwri_submit(struct list_head *);
	316	extern int xfs_buf_delwri_submit_nowait(struct list_head *);
7912e7fe	317	extern int xfs_buf_delwri_pushbuf(struct xfs_buf , struct list_head );
1da177e4 LT	318
1da177e4 LT	319	/* Buffer Daemon Setup Routines */
ce8e922c NS	320	extern int xfs_buf_init(void);
ce8e922c NS	321	extern void xfs_buf_terminate(void);
1da177e4	322
cbb7baab DC	323	/*
	324	* These macros use the IO block map rather than b_bn. b_bn is now really
	325	* just for the buffer cache index for cached buffers. As IO does not use b_bn
	326	* anymore, uncached buffers do not use b_bn at all and hence must modify the IO
	327	* map directly. Uncached buffers are not allowed to be discontiguous, so this
	328	* is safe to do.
	329	*
	330	* In future, uncached buffers will pass the block number directly to the io
	331	* request function and hence these macros will go away at that point.
	332	*/
d44d9bc6 MT	333	#define XFS_BUF_ADDR(bp) ((bp)->b_maps[0].bm_bn)
d44d9bc6 MT	334	#define XFS_BUF_SET_ADDR(bp, bno) ((bp)->b_maps[0].bm_bn = (xfs_daddr_t)(bno))
ce8e922c	335
7561d27e	336	void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref);
ce8e922c	337
879de98e DC	338	/*
	339	* If the buffer is already on the LRU, do nothing. Otherwise set the buffer
	340	* up with a reference count of 0 so it will be tossed from the cache when
	341	* released.
	342	*/
	343	static inline void xfs_buf_oneshot(struct xfs_buf *bp)
	344	{
	345	if (!list_empty(&bp->b_lru) \|\| atomic_read(&bp->b_lru_ref) > 1)
	346	return;
	347	atomic_set(&bp->b_lru_ref, 0);
	348	}
	349
811e64c7 CS	350	static inline int xfs_buf_ispinned(struct xfs_buf *bp)
	351	{
	352	return atomic_read(&bp->b_pin_count);
	353	}
ce8e922c	354
ce8e922c	355	static inline void xfs_buf_relse(xfs_buf_t *bp)
1da177e4	356	{
bfc60177	357	xfs_buf_unlock(bp);
ce8e922c	358	xfs_buf_rele(bp);
1da177e4 LT	359	}
1da177e4 LT	360
51582170 ES	361	static inline int
	362	xfs_buf_verify_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	363	{
	364	return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
	365	cksum_offset);
	366	}
	367
f1dbcd7e ES	368	static inline void
	369	xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
	370	{
	371	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
	372	cksum_offset);
	373	}
	374
1da177e4 LT	375	/*
	376	* Handling of buftargs.
	377	*/
ebad861b	378	extern xfs_buftarg_t xfs_alloc_buftarg(struct xfs_mount ,
486aff5e	379	struct block_device , struct dax_device );
a1f69417	380	extern void xfs_free_buftarg(struct xfs_buftarg *);
1da177e4	381	extern void xfs_wait_buftarg(xfs_buftarg_t *);
a96c4151	382	extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int);
d808f617	383
ce8e922c NS	384	#define xfs_getsize_buftarg(buftarg) block_size((buftarg)->bt_bdev)
	385	#define xfs_readonly_buftarg(buftarg) bdev_read_only((buftarg)->bt_bdev)
	386
1da177e4	387	#endif /* __XFS_BUF_H__ */