blk_types.h 11 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
2
3
4
5
6
7
8
9
/*
 * Block data types and constants.  Directly include this file only to
 * break include dependency loop.
 */
#ifndef __LINUX_BLK_TYPES_H
#define __LINUX_BLK_TYPES_H

#include <linux/types.h>
10
#include <linux/bvec.h>
11
12
13
14
15
16

struct bio_set;
struct bio;
struct bio_integrity_payload;
struct page;
struct block_device;
17
18
struct io_context;
struct cgroup_subsys_state;
19
typedef void (bio_end_io_t) (struct bio *);
20

21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
/*
 * Block error status values.  See block/blk-core:blk_errors for the details.
 */
typedef u8 __bitwise blk_status_t;
#define	BLK_STS_OK 0
#define BLK_STS_NOTSUPP		((__force blk_status_t)1)
#define BLK_STS_TIMEOUT		((__force blk_status_t)2)
#define BLK_STS_NOSPC		((__force blk_status_t)3)
#define BLK_STS_TRANSPORT	((__force blk_status_t)4)
#define BLK_STS_TARGET		((__force blk_status_t)5)
#define BLK_STS_NEXUS		((__force blk_status_t)6)
#define BLK_STS_MEDIUM		((__force blk_status_t)7)
#define BLK_STS_PROTECTION	((__force blk_status_t)8)
#define BLK_STS_RESOURCE	((__force blk_status_t)9)
#define BLK_STS_IOERR		((__force blk_status_t)10)

37
38
39
/* hack for device mapper, don't use elsewhere: */
#define BLK_STS_DM_REQUEUE    ((__force blk_status_t)11)

40
41
#define BLK_STS_AGAIN		((__force blk_status_t)12)

42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
/*
 * BLK_STS_DEV_RESOURCE is returned from the driver to the block layer if
 * device related resources are unavailable, but the driver can guarantee
 * that the queue will be rerun in the future once resources become
 * available again. This is typically the case for device specific
 * resources that are consumed for IO. If the driver fails allocating these
 * resources, we know that inflight (or pending) IO will free these
 * resource upon completion.
 *
 * This is different from BLK_STS_RESOURCE in that it explicitly references
 * a device specific resource. For resources of wider scope, allocation
 * failure can happen without having pending IO. This means that we can't
 * rely on request completions freeing these resources, as IO may not be in
 * flight. Examples of that are kernel memory allocations, DMA mappings, or
 * any other system wide resources.
 */
#define BLK_STS_DEV_RESOURCE	((__force blk_status_t)13)

60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
/**
 * blk_path_error - returns true if error may be path related
 * @error: status the request was completed with
 *
 * Description:
 *     This classifies block error status into non-retryable errors and ones
 *     that may be successful if retried on a failover path.
 *
 * Return:
 *     %false - retrying failover path will not help
 *     %true  - may succeed if retried
 */
static inline bool blk_path_error(blk_status_t error)
{
	switch (error) {
	case BLK_STS_NOTSUPP:
	case BLK_STS_NOSPC:
	case BLK_STS_TARGET:
	case BLK_STS_NEXUS:
	case BLK_STS_MEDIUM:
	case BLK_STS_PROTECTION:
		return false;
	}

	/* Anything else could be a path failure, so should be retried */
	return true;
}

88
89
90
91
struct blk_issue_stat {
	u64 stat;
};

92
93
94
95
96
97
/*
 * main unit of I/O for the block layer and lower layers (ie drivers and
 * stacking drivers)
 */
struct bio {
	struct bio		*bi_next;	/* request queue link */
98
	struct gendisk		*bi_disk;
Jens Axboe's avatar
Jens Axboe committed
99
100
101
	unsigned int		bi_opf;		/* bottom bits req flags,
						 * top bits REQ_OP. Use
						 * accessors.
102
						 */
103
	unsigned short		bi_flags;	/* status, etc and bvec pool number */
104
	unsigned short		bi_ioprio;
105
	unsigned short		bi_write_hint;
106
107
	blk_status_t		bi_status;
	u8			bi_partno;
108
109
110
111
112
113
114
115
116
117
118
119
120

	/* Number of segments in this BIO after
	 * physical address coalescing is performed.
	 */
	unsigned int		bi_phys_segments;

	/*
	 * To keep track of the max segment size, we account for the
	 * sizes of the first and last mergeable segments in this bio.
	 */
	unsigned int		bi_seg_front_size;
	unsigned int		bi_seg_back_size;

121
	struct bvec_iter	bi_iter;
Kent Overstreet's avatar
Kent Overstreet committed
122

123
	atomic_t		__bi_remaining;
124
125
126
	bio_end_io_t		*bi_end_io;

	void			*bi_private;
127
128
129
130
131
132
133
#ifdef CONFIG_BLK_CGROUP
	/*
	 * Optional ioc and css associated with this bio.  Put on bio
	 * release.  Read comment on top of bio_associate_current().
	 */
	struct io_context	*bi_ioc;
	struct cgroup_subsys_state *bi_css;
134
135
#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
	void			*bi_cg_private;
136
	struct blk_issue_stat	bi_issue_stat;
137
#endif
138
#endif
139
	union {
140
#if defined(CONFIG_BLK_DEV_INTEGRITY)
141
		struct bio_integrity_payload *bi_integrity; /* data integrity */
142
#endif
143
	};
144

145
146
	unsigned short		bi_vcnt;	/* how many bio_vec's */

Kent Overstreet's avatar
Kent Overstreet committed
147
148
149
150
	/*
	 * Everything starting with bi_max_vecs will be preserved by bio_reset()
	 */

151
	unsigned short		bi_max_vecs;	/* max bvl_vecs we can hold */
Kent Overstreet's avatar
Kent Overstreet committed
152

153
	atomic_t		__bi_cnt;	/* pin count */
Kent Overstreet's avatar
Kent Overstreet committed
154
155
156

	struct bio_vec		*bi_io_vec;	/* the actual vec list */

157
158
	struct bio_set		*bi_pool;

159
160
161
162
163
164
165
166
	/*
	 * We can inline a number of vecs at the end of the bio, to avoid
	 * double allocations for a small number of bio_vecs. This member
	 * MUST obviously be kept at the very end of the bio.
	 */
	struct bio_vec		bi_inline_vecs[0];
};

Kent Overstreet's avatar
Kent Overstreet committed
167
168
#define BIO_RESET_BYTES		offsetof(struct bio, bi_max_vecs)

169
170
171
/*
 * bio flags
 */
Jens Axboe's avatar
Jens Axboe committed
172
173
174
175
176
177
#define BIO_SEG_VALID	1	/* bi_phys_segments valid */
#define BIO_CLONED	2	/* doesn't own data */
#define BIO_BOUNCED	3	/* bio is a bounce bio */
#define BIO_USER_MAPPED 4	/* contains user pages */
#define BIO_NULL_MAPPED 5	/* contains invalid user pages */
#define BIO_QUIET	6	/* Make BIO Quiet */
178
179
#define BIO_CHAIN	7	/* chained bio, ->bi_remaining in effect */
#define BIO_REFFED	8	/* bio has elevated ->bi_cnt */
180
181
#define BIO_THROTTLED	9	/* This bio has already been subjected to
				 * throttling rules. Don't do it again. */
182
183
#define BIO_TRACE_COMPLETION 10	/* bio_endio() should trace the final completion
				 * of this bio. */
184
/* See BVEC_POOL_OFFSET below before adding new flags */
Kent Overstreet's avatar
Kent Overstreet committed
185

186
/*
187
188
 * We support 6 different bvec pools, the last one is magic in that it
 * is backed by a mempool.
189
 */
190
191
192
193
#define BVEC_POOL_NR		6
#define BVEC_POOL_MAX		(BVEC_POOL_NR - 1)

/*
194
 * Top 3 bits of bio flags indicate the pool the bvecs came from.  We add
195
196
197
 * 1 to the actual index so that 0 indicates that there are no bvecs to be
 * freed.
 */
198
#define BVEC_POOL_BITS		(3)
199
#define BVEC_POOL_OFFSET	(16 - BVEC_POOL_BITS)
200
#define BVEC_POOL_IDX(bio)	((bio)->bi_flags >> BVEC_POOL_OFFSET)
201
202
203
204
205
206
207
208
209
#if (1<< BVEC_POOL_BITS) < (BVEC_POOL_NR+1)
# error "BVEC_POOL_BITS is too small"
#endif

/*
 * Flags starting here get preserved by bio_reset() - this includes
 * only BVEC_POOL_IDX()
 */
#define BIO_RESET_BITS	BVEC_POOL_OFFSET
210

211
212
typedef __u32 __bitwise blk_mq_req_flags_t;

213
/*
214
215
 * Operations and flags common to the bio and request structures.
 * We use 8 bits for encoding the operation, and the remaining 24 for flags.
216
217
218
219
220
221
222
223
224
 *
 * The least significant bit of the operation number indicates the data
 * transfer direction:
 *
 *   - if the least significant bit is set transfers are TO the device
 *   - if the least significant bit is not set transfers are FROM the device
 *
 * If a operation does not transfer data the least significant bit has no
 * meaning.
225
 */
226
227
228
229
230
#define REQ_OP_BITS	8
#define REQ_OP_MASK	((1 << REQ_OP_BITS) - 1)
#define REQ_FLAG_BITS	24

enum req_opf {
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
	/* read sectors from the device */
	REQ_OP_READ		= 0,
	/* write sectors to the device */
	REQ_OP_WRITE		= 1,
	/* flush the volatile write cache */
	REQ_OP_FLUSH		= 2,
	/* discard sectors */
	REQ_OP_DISCARD		= 3,
	/* get zone information */
	REQ_OP_ZONE_REPORT	= 4,
	/* securely erase sectors */
	REQ_OP_SECURE_ERASE	= 5,
	/* seset a zone write pointer */
	REQ_OP_ZONE_RESET	= 6,
	/* write the same sector many times */
	REQ_OP_WRITE_SAME	= 7,
247
	/* write the zero filled sector many times */
248
	REQ_OP_WRITE_ZEROES	= 9,
249

250
251
252
253
254
255
256
	/* SCSI passthrough using struct scsi_request */
	REQ_OP_SCSI_IN		= 32,
	REQ_OP_SCSI_OUT		= 33,
	/* Driver private requests */
	REQ_OP_DRV_IN		= 34,
	REQ_OP_DRV_OUT		= 35,

257
258
259
260
261
262
	REQ_OP_LAST,
};

enum req_flag_bits {
	__REQ_FAILFAST_DEV =	/* no driver retries of device errors */
		REQ_OP_BITS,
263
264
265
266
	__REQ_FAILFAST_TRANSPORT, /* no driver retries of transport errors */
	__REQ_FAILFAST_DRIVER,	/* no driver retries of driver errors */
	__REQ_SYNC,		/* request is sync (sync write or read) */
	__REQ_META,		/* metadata io request */
267
	__REQ_PRIO,		/* boost priority in cfq */
268
	__REQ_NOMERGE,		/* don't touch this for merging */
269
	__REQ_IDLE,		/* anticipate more IO after this one */
270
	__REQ_INTEGRITY,	/* I/O includes block integrity payload */
271
	__REQ_FUA,		/* forced unit access */
272
	__REQ_PREFLUSH,		/* request for cache flush */
273
	__REQ_RAHEAD,		/* read ahead, can fail anytime */
Jens Axboe's avatar
Jens Axboe committed
274
	__REQ_BACKGROUND,	/* background IO */
Christoph Hellwig's avatar
Christoph Hellwig committed
275
	__REQ_NOWAIT,           /* Don't wait if request will block */
276
277
278
279

	/* command specific flags for REQ_OP_WRITE_ZEROES: */
	__REQ_NOUNMAP,		/* do not free blocks when zeroing */

Christoph Hellwig's avatar
Christoph Hellwig committed
280
281
282
	/* for driver use */
	__REQ_DRV,

283
284
285
	__REQ_NR_BITS,		/* stops here */
};

286
287
288
289
290
291
#define REQ_FAILFAST_DEV	(1ULL << __REQ_FAILFAST_DEV)
#define REQ_FAILFAST_TRANSPORT	(1ULL << __REQ_FAILFAST_TRANSPORT)
#define REQ_FAILFAST_DRIVER	(1ULL << __REQ_FAILFAST_DRIVER)
#define REQ_SYNC		(1ULL << __REQ_SYNC)
#define REQ_META		(1ULL << __REQ_META)
#define REQ_PRIO		(1ULL << __REQ_PRIO)
292
#define REQ_NOMERGE		(1ULL << __REQ_NOMERGE)
293
#define REQ_IDLE		(1ULL << __REQ_IDLE)
294
#define REQ_INTEGRITY		(1ULL << __REQ_INTEGRITY)
295
296
297
#define REQ_FUA			(1ULL << __REQ_FUA)
#define REQ_PREFLUSH		(1ULL << __REQ_PREFLUSH)
#define REQ_RAHEAD		(1ULL << __REQ_RAHEAD)
Jens Axboe's avatar
Jens Axboe committed
298
#define REQ_BACKGROUND		(1ULL << __REQ_BACKGROUND)
Christoph Hellwig's avatar
Christoph Hellwig committed
299
#define REQ_NOWAIT		(1ULL << __REQ_NOWAIT)
300

301
302
#define REQ_NOUNMAP		(1ULL << __REQ_NOUNMAP)

Christoph Hellwig's avatar
Christoph Hellwig committed
303
#define REQ_DRV			(1ULL << __REQ_DRV)
304

305
306
307
#define REQ_FAILFAST_MASK \
	(REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT | REQ_FAILFAST_DRIVER)

308
#define REQ_NOMERGE_FLAGS \
309
	(REQ_NOMERGE | REQ_PREFLUSH | REQ_FUA)
310

311
312
313
314
#define bio_op(bio) \
	((bio)->bi_opf & REQ_OP_MASK)
#define req_op(req) \
	((req)->cmd_flags & REQ_OP_MASK)
315

316
/* obsolete, don't use in new code */
317
318
319
320
321
static inline void bio_set_op_attrs(struct bio *bio, unsigned op,
		unsigned op_flags)
{
	bio->bi_opf = op | op_flags;
}
322

323
324
325
326
327
static inline bool op_is_write(unsigned int op)
{
	return (op & 1);
}

328
329
330
331
332
333
334
335
336
/*
 * Check if the bio or request is one that needs special treatment in the
 * flush state machine.
 */
static inline bool op_is_flush(unsigned int op)
{
	return op & (REQ_FUA | REQ_PREFLUSH);
}

337
338
339
340
341
/*
 * Reads are always treated as synchronous, as are requests with the FUA or
 * PREFLUSH flag.  Other operations may be marked as synchronous using the
 * REQ_SYNC flag.
 */
342
343
static inline bool op_is_sync(unsigned int op)
{
344
345
	return (op & REQ_OP_MASK) == REQ_OP_READ ||
		(op & (REQ_SYNC | REQ_FUA | REQ_PREFLUSH));
346
}
347

348
typedef unsigned int blk_qc_t;
349
350
351
#define BLK_QC_T_NONE		-1U
#define BLK_QC_T_SHIFT		16
#define BLK_QC_T_INTERNAL	(1U << 31)
352
353
354
355
356
357

static inline bool blk_qc_t_valid(blk_qc_t cookie)
{
	return cookie != BLK_QC_T_NONE;
}

358
359
static inline blk_qc_t blk_tag_to_qc_t(unsigned int tag, unsigned int queue_num,
				       bool internal)
360
{
361
362
363
364
365
366
	blk_qc_t ret = tag | (queue_num << BLK_QC_T_SHIFT);

	if (internal)
		ret |= BLK_QC_T_INTERNAL;

	return ret;
367
368
369
370
}

static inline unsigned int blk_qc_t_to_queue_num(blk_qc_t cookie)
{
371
	return (cookie & ~BLK_QC_T_INTERNAL) >> BLK_QC_T_SHIFT;
372
373
374
375
}

static inline unsigned int blk_qc_t_to_tag(blk_qc_t cookie)
{
376
	return cookie & ((1u << BLK_QC_T_SHIFT) - 1);
377
378
}

379
380
381
382
383
static inline bool blk_qc_t_is_internal(blk_qc_t cookie)
{
	return (cookie & BLK_QC_T_INTERNAL) != 0;
}

384
struct blk_rq_stat {
Shaohua Li's avatar
Shaohua Li committed
385
	u64 mean;
386
387
	u64 min;
	u64 max;
Shaohua Li's avatar
Shaohua Li committed
388
	u32 nr_samples;
389
390
391
	u64 batch;
};

392
#endif /* __LINUX_BLK_TYPES_H */