xref: /kernel/linux/linux-6.6/fs/xfs/xfs_buf.h (revision 62306a36)
1// SPDX-License-Identifier: GPL-2.0
2/*
3 * Copyright (c) 2000-2005 Silicon Graphics, Inc.
4 * All Rights Reserved.
5 */
6#ifndef __XFS_BUF_H__
7#define __XFS_BUF_H__
8
9#include <linux/list.h>
10#include <linux/types.h>
11#include <linux/spinlock.h>
12#include <linux/mm.h>
13#include <linux/fs.h>
14#include <linux/dax.h>
15#include <linux/uio.h>
16#include <linux/list_lru.h>
17
18extern struct kmem_cache *xfs_buf_cache;
19
20/*
21 *	Base types
22 */
23struct xfs_buf;
24
25#define XFS_BUF_DADDR_NULL	((xfs_daddr_t) (-1LL))
26
27#define XBF_READ	 (1u << 0) /* buffer intended for reading from device */
28#define XBF_WRITE	 (1u << 1) /* buffer intended for writing to device */
29#define XBF_READ_AHEAD	 (1u << 2) /* asynchronous read-ahead */
30#define XBF_NO_IOACCT	 (1u << 3) /* bypass I/O accounting (non-LRU bufs) */
31#define XBF_ASYNC	 (1u << 4) /* initiator will not wait for completion */
32#define XBF_DONE	 (1u << 5) /* all pages in the buffer uptodate */
33#define XBF_STALE	 (1u << 6) /* buffer has been staled, do not find it */
34#define XBF_WRITE_FAIL	 (1u << 7) /* async writes have failed on this buffer */
35
36/* buffer type flags for write callbacks */
37#define _XBF_INODES	 (1u << 16)/* inode buffer */
38#define _XBF_DQUOTS	 (1u << 17)/* dquot buffer */
39#define _XBF_LOGRECOVERY (1u << 18)/* log recovery buffer */
40
41/* flags used only internally */
42#define _XBF_PAGES	 (1u << 20)/* backed by refcounted pages */
43#define _XBF_KMEM	 (1u << 21)/* backed by heap memory */
44#define _XBF_DELWRI_Q	 (1u << 22)/* buffer on a delwri queue */
45
46/* flags used only as arguments to access routines */
47/*
48 * Online fsck is scanning the buffer cache for live buffers.  Do not warn
49 * about length mismatches during lookups and do not return stale buffers.
50 */
51#define XBF_LIVESCAN	 (1u << 28)
52#define XBF_INCORE	 (1u << 29)/* lookup only, return if found in cache */
53#define XBF_TRYLOCK	 (1u << 30)/* lock requested, but do not wait */
54#define XBF_UNMAPPED	 (1u << 31)/* do not map the buffer */
55
56
57typedef unsigned int xfs_buf_flags_t;
58
59#define XFS_BUF_FLAGS \
60	{ XBF_READ,		"READ" }, \
61	{ XBF_WRITE,		"WRITE" }, \
62	{ XBF_READ_AHEAD,	"READ_AHEAD" }, \
63	{ XBF_NO_IOACCT,	"NO_IOACCT" }, \
64	{ XBF_ASYNC,		"ASYNC" }, \
65	{ XBF_DONE,		"DONE" }, \
66	{ XBF_STALE,		"STALE" }, \
67	{ XBF_WRITE_FAIL,	"WRITE_FAIL" }, \
68	{ _XBF_INODES,		"INODES" }, \
69	{ _XBF_DQUOTS,		"DQUOTS" }, \
70	{ _XBF_LOGRECOVERY,	"LOG_RECOVERY" }, \
71	{ _XBF_PAGES,		"PAGES" }, \
72	{ _XBF_KMEM,		"KMEM" }, \
73	{ _XBF_DELWRI_Q,	"DELWRI_Q" }, \
74	/* The following interface flags should never be set */ \
75	{ XBF_LIVESCAN,		"LIVESCAN" }, \
76	{ XBF_INCORE,		"INCORE" }, \
77	{ XBF_TRYLOCK,		"TRYLOCK" }, \
78	{ XBF_UNMAPPED,		"UNMAPPED" }
79
80/*
81 * Internal state flags.
82 */
83#define XFS_BSTATE_DISPOSE	 (1 << 0)	/* buffer being discarded */
84#define XFS_BSTATE_IN_FLIGHT	 (1 << 1)	/* I/O in flight */
85
86/*
87 * The xfs_buftarg contains 2 notions of "sector size" -
88 *
89 * 1) The metadata sector size, which is the minimum unit and
90 *    alignment of IO which will be performed by metadata operations.
91 * 2) The device logical sector size
92 *
93 * The first is specified at mkfs time, and is stored on-disk in the
94 * superblock's sb_sectsize.
95 *
96 * The latter is derived from the underlying device, and controls direct IO
97 * alignment constraints.
98 */
99typedef struct xfs_buftarg {
100	dev_t			bt_dev;
101	struct block_device	*bt_bdev;
102	struct dax_device	*bt_daxdev;
103	u64			bt_dax_part_off;
104	struct xfs_mount	*bt_mount;
105	unsigned int		bt_meta_sectorsize;
106	size_t			bt_meta_sectormask;
107	size_t			bt_logical_sectorsize;
108	size_t			bt_logical_sectormask;
109
110	/* LRU control structures */
111	struct shrinker		bt_shrinker;
112	struct list_lru		bt_lru;
113
114	struct percpu_counter	bt_io_count;
115	struct ratelimit_state	bt_ioerror_rl;
116} xfs_buftarg_t;
117
118#define XB_PAGES	2
119
120struct xfs_buf_map {
121	xfs_daddr_t		bm_bn;	/* block number for I/O */
122	int			bm_len;	/* size of I/O */
123	unsigned int		bm_flags;
124};
125
126/*
127 * Online fsck is scanning the buffer cache for live buffers.  Do not warn
128 * about length mismatches during lookups and do not return stale buffers.
129 */
130#define XBM_LIVESCAN		(1U << 0)
131
132#define DEFINE_SINGLE_BUF_MAP(map, blkno, numblk) \
133	struct xfs_buf_map (map) = { .bm_bn = (blkno), .bm_len = (numblk) };
134
135struct xfs_buf_ops {
136	char *name;
137	union {
138		__be32 magic[2];	/* v4 and v5 on disk magic values */
139		__be16 magic16[2];	/* v4 and v5 on disk magic values */
140	};
141	void (*verify_read)(struct xfs_buf *);
142	void (*verify_write)(struct xfs_buf *);
143	xfs_failaddr_t (*verify_struct)(struct xfs_buf *bp);
144};
145
146struct xfs_buf {
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	 */
154	struct rhash_head	b_rhash_head;	/* pag buffer hash node */
155
156	xfs_daddr_t		b_rhash_key;	/* buffer cache index */
157	int			b_length;	/* size of buffer in BBs */
158	atomic_t		b_hold;		/* reference count */
159	atomic_t		b_lru_ref;	/* lru reclaim ref count */
160	xfs_buf_flags_t		b_flags;	/* status flags */
161	struct semaphore	b_sema;		/* semaphore for lockables */
162
163	/*
164	 * concurrent access to b_lru and b_lru_flags are protected by
165	 * bt_lru_lock and not by b_sema
166	 */
167	struct list_head	b_lru;		/* lru list */
168	spinlock_t		b_lock;		/* internal state lock */
169	unsigned int		b_state;	/* internal state flags */
170	int			b_io_error;	/* internal IO error state */
171	wait_queue_head_t	b_waiters;	/* unpin waiters */
172	struct list_head	b_list;
173	struct xfs_perag	*b_pag;		/* contains rbtree root */
174	struct xfs_mount	*b_mount;
175	struct xfs_buftarg	*b_target;	/* buffer target (device) */
176	void			*b_addr;	/* virtual address of buffer */
177	struct work_struct	b_ioend_work;
178	struct completion	b_iowait;	/* queue for I/O waiters */
179	struct xfs_buf_log_item	*b_log_item;
180	struct list_head	b_li_list;	/* Log items list head */
181	struct xfs_trans	*b_transp;
182	struct page		**b_pages;	/* array of page pointers */
183	struct page		*b_page_array[XB_PAGES]; /* inline pages */
184	struct xfs_buf_map	*b_maps;	/* compound buffer map */
185	struct xfs_buf_map	__b_map;	/* inline compound buffer map */
186	int			b_map_count;
187	atomic_t		b_pin_count;	/* pin count */
188	atomic_t		b_io_remaining;	/* #outstanding I/O requests */
189	unsigned int		b_page_count;	/* size of page array */
190	unsigned int		b_offset;	/* page offset of b_addr,
191						   only for _XBF_KMEM buffers */
192	int			b_error;	/* error code on I/O */
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
213	const struct xfs_buf_ops	*b_ops;
214	struct rcu_head		b_rcu;
215};
216
217/* Finding and Reading Buffers */
218int xfs_buf_get_map(struct xfs_buftarg *target, struct xfs_buf_map *map,
219		int nmaps, xfs_buf_flags_t flags, struct xfs_buf **bpp);
220int xfs_buf_read_map(struct xfs_buftarg *target, struct xfs_buf_map *map,
221		int nmaps, xfs_buf_flags_t flags, struct xfs_buf **bpp,
222		const struct xfs_buf_ops *ops, xfs_failaddr_t fa);
223void xfs_buf_readahead_map(struct xfs_buftarg *target,
224			       struct xfs_buf_map *map, int nmaps,
225			       const struct xfs_buf_ops *ops);
226
227static inline int
228xfs_buf_incore(
229	struct xfs_buftarg	*target,
230	xfs_daddr_t		blkno,
231	size_t			numblks,
232	xfs_buf_flags_t		flags,
233	struct xfs_buf		**bpp)
234{
235	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
236
237	return xfs_buf_get_map(target, &map, 1, XBF_INCORE | flags, bpp);
238}
239
240static inline int
241xfs_buf_get(
242	struct xfs_buftarg	*target,
243	xfs_daddr_t		blkno,
244	size_t			numblks,
245	struct xfs_buf		**bpp)
246{
247	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
248
249	return xfs_buf_get_map(target, &map, 1, 0, bpp);
250}
251
252static inline int
253xfs_buf_read(
254	struct xfs_buftarg	*target,
255	xfs_daddr_t		blkno,
256	size_t			numblks,
257	xfs_buf_flags_t		flags,
258	struct xfs_buf		**bpp,
259	const struct xfs_buf_ops *ops)
260{
261	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
262
263	return xfs_buf_read_map(target, &map, 1, flags, bpp, ops,
264			__builtin_return_address(0));
265}
266
267static inline void
268xfs_buf_readahead(
269	struct xfs_buftarg	*target,
270	xfs_daddr_t		blkno,
271	size_t			numblks,
272	const struct xfs_buf_ops *ops)
273{
274	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
275	return xfs_buf_readahead_map(target, &map, 1, ops);
276}
277
278int xfs_buf_get_uncached(struct xfs_buftarg *target, size_t numblks,
279		xfs_buf_flags_t flags, struct xfs_buf **bpp);
280int xfs_buf_read_uncached(struct xfs_buftarg *target, xfs_daddr_t daddr,
281		size_t numblks, xfs_buf_flags_t flags, struct xfs_buf **bpp,
282		const struct xfs_buf_ops *ops);
283int _xfs_buf_read(struct xfs_buf *bp, xfs_buf_flags_t flags);
284void xfs_buf_hold(struct xfs_buf *bp);
285
286/* Releasing Buffers */
287extern void xfs_buf_rele(struct xfs_buf *);
288
289/* Locking and Unlocking Buffers */
290extern int xfs_buf_trylock(struct xfs_buf *);
291extern void xfs_buf_lock(struct xfs_buf *);
292extern void xfs_buf_unlock(struct xfs_buf *);
293#define xfs_buf_islocked(bp) \
294	((bp)->b_sema.count <= 0)
295
296static inline void xfs_buf_relse(struct xfs_buf *bp)
297{
298	xfs_buf_unlock(bp);
299	xfs_buf_rele(bp);
300}
301
302/* Buffer Read and Write Routines */
303extern int xfs_bwrite(struct xfs_buf *bp);
304
305extern void __xfs_buf_ioerror(struct xfs_buf *bp, int error,
306		xfs_failaddr_t failaddr);
307#define xfs_buf_ioerror(bp, err) __xfs_buf_ioerror((bp), (err), __this_address)
308extern void xfs_buf_ioerror_alert(struct xfs_buf *bp, xfs_failaddr_t fa);
309void xfs_buf_ioend_fail(struct xfs_buf *);
310void xfs_buf_zero(struct xfs_buf *bp, size_t boff, size_t bsize);
311void __xfs_buf_mark_corrupt(struct xfs_buf *bp, xfs_failaddr_t fa);
312#define xfs_buf_mark_corrupt(bp) __xfs_buf_mark_corrupt((bp), __this_address)
313
314/* Buffer Utility Routines */
315extern void *xfs_buf_offset(struct xfs_buf *, size_t);
316extern void xfs_buf_stale(struct xfs_buf *bp);
317
318/* Delayed Write Buffer Routines */
319extern void xfs_buf_delwri_cancel(struct list_head *);
320extern bool xfs_buf_delwri_queue(struct xfs_buf *, struct list_head *);
321extern int xfs_buf_delwri_submit(struct list_head *);
322extern int xfs_buf_delwri_submit_nowait(struct list_head *);
323extern int xfs_buf_delwri_pushbuf(struct xfs_buf *, struct list_head *);
324
325static inline xfs_daddr_t xfs_buf_daddr(struct xfs_buf *bp)
326{
327	return bp->b_maps[0].bm_bn;
328}
329
330void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref);
331
332/*
333 * If the buffer is already on the LRU, do nothing. Otherwise set the buffer
334 * up with a reference count of 0 so it will be tossed from the cache when
335 * released.
336 */
337static inline void xfs_buf_oneshot(struct xfs_buf *bp)
338{
339	if (!list_empty(&bp->b_lru) || atomic_read(&bp->b_lru_ref) > 1)
340		return;
341	atomic_set(&bp->b_lru_ref, 0);
342}
343
344static inline int xfs_buf_ispinned(struct xfs_buf *bp)
345{
346	return atomic_read(&bp->b_pin_count);
347}
348
349static inline int
350xfs_buf_verify_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
351{
352	return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
353				cksum_offset);
354}
355
356static inline void
357xfs_buf_update_cksum(struct xfs_buf *bp, unsigned long cksum_offset)
358{
359	xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
360			 cksum_offset);
361}
362
363/*
364 *	Handling of buftargs.
365 */
366struct xfs_buftarg *xfs_alloc_buftarg(struct xfs_mount *mp,
367		struct block_device *bdev);
368extern void xfs_free_buftarg(struct xfs_buftarg *);
369extern void xfs_buftarg_wait(struct xfs_buftarg *);
370extern void xfs_buftarg_drain(struct xfs_buftarg *);
371extern int xfs_setsize_buftarg(struct xfs_buftarg *, unsigned int);
372
373#define xfs_getsize_buftarg(buftarg)	block_size((buftarg)->bt_bdev)
374#define xfs_readonly_buftarg(buftarg)	bdev_read_only((buftarg)->bt_bdev)
375
376int xfs_buf_reverify(struct xfs_buf *bp, const struct xfs_buf_ops *ops);
377bool xfs_verify_magic(struct xfs_buf *bp, __be32 dmagic);
378bool xfs_verify_magic16(struct xfs_buf *bp, __be16 dmagic);
379
380#endif	/* __XFS_BUF_H__ */
381