blkdev.h 46.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3
#ifndef _LINUX_BLKDEV_H
#define _LINUX_BLKDEV_H

4 5
#include <linux/sched.h>

J
Jens Axboe 已提交
6 7
#ifdef CONFIG_BLOCK

L
Linus Torvalds 已提交
8 9 10
#include <linux/major.h>
#include <linux/genhd.h>
#include <linux/list.h>
11
#include <linux/llist.h>
L
Linus Torvalds 已提交
12 13 14
#include <linux/timer.h>
#include <linux/workqueue.h>
#include <linux/pagemap.h>
15
#include <linux/backing-dev-defs.h>
L
Linus Torvalds 已提交
16 17 18 19
#include <linux/wait.h>
#include <linux/mempool.h>
#include <linux/bio.h>
#include <linux/stringify.h>
20
#include <linux/gfp.h>
21
#include <linux/bsg.h>
22
#include <linux/smp.h>
T
Tejun Heo 已提交
23
#include <linux/rcupdate.h>
24
#include <linux/percpu-refcount.h>
C
Christoph Hellwig 已提交
25
#include <linux/scatterlist.h>
L
Linus Torvalds 已提交
26

27
struct module;
28 29
struct scsi_ioctl_command;

L
Linus Torvalds 已提交
30 31
struct request_queue;
struct elevator_queue;
32
struct blk_trace;
33 34
struct request;
struct sg_io_hdr;
M
Mike Christie 已提交
35
struct bsg_job;
T
Tejun Heo 已提交
36
struct blkcg_gq;
37
struct blk_flush_queue;
L
Linus Torvalds 已提交
38 39 40 41

#define BLKDEV_MIN_RQ	4
#define BLKDEV_MAX_RQ	128	/* Default maximum */

T
Tejun Heo 已提交
42 43 44 45 46 47
/*
 * Maximum number of blkcg policies allowed to be registered concurrently.
 * Defined here to simplify include dependency.
 */
#define BLKCG_MAX_POLS		2

L
Linus Torvalds 已提交
48
struct request;
49
typedef void (rq_end_io_fn)(struct request *, int);
L
Linus Torvalds 已提交
50

51 52 53
#define BLK_RL_SYNCFULL		(1U << 0)
#define BLK_RL_ASYNCFULL	(1U << 1)

L
Linus Torvalds 已提交
54
struct request_list {
55
	struct request_queue	*q;	/* the queue this rl belongs to */
56 57 58
#ifdef CONFIG_BLK_CGROUP
	struct blkcg_gq		*blkg;	/* blkg this request pool belongs to */
#endif
59 60 61 62
	/*
	 * count[], starved[], and wait[] are indexed by
	 * BLK_RW_SYNC/BLK_RW_ASYNC
	 */
63 64 65 66
	int			count[2];
	int			starved[2];
	mempool_t		*rq_pool;
	wait_queue_head_t	wait[2];
67
	unsigned int		flags;
L
Linus Torvalds 已提交
68 69
};

70 71 72 73 74 75
/*
 * request command types
 */
enum rq_cmd_type_bits {
	REQ_TYPE_FS		= 1,	/* fs request */
	REQ_TYPE_BLOCK_PC,		/* scsi command */
76
	REQ_TYPE_DRV_PRIV,		/* driver defined types from here */
77 78
};

L
Linus Torvalds 已提交
79 80 81
#define BLK_MAX_CDB	16

/*
82 83 84 85
 * Try to put the fields that are referenced together in the same cacheline.
 *
 * If you modify this structure, make sure to update blk_rq_init() and
 * especially blk_mq_rq_ctx_init() to take care of the added fields.
L
Linus Torvalds 已提交
86 87
 */
struct request {
88
	struct list_head queuelist;
89 90
	union {
		struct call_single_data csd;
91
		unsigned long fifo_time;
92
	};
93

94
	struct request_queue *q;
95
	struct blk_mq_ctx *mq_ctx;
96

J
Jens Axboe 已提交
97
	u64 cmd_flags;
98
	unsigned cmd_type;
J
Jens Axboe 已提交
99
	unsigned long atomic_flags;
L
Linus Torvalds 已提交
100

101 102
	int cpu;

103 104
	/* the following two fields are internal, NEVER access directly */
	unsigned int __data_len;	/* total data len */
105
	sector_t __sector;		/* sector cursor */
L
Linus Torvalds 已提交
106 107 108 109

	struct bio *bio;
	struct bio *biotail;

110 111 112 113 114 115 116 117 118 119 120 121
	/*
	 * The hash is used inside the scheduler, and killed once the
	 * request reaches the dispatch list. The ipi_list is only used
	 * to queue the request for softirq completion, which is long
	 * after the request has been unhashed (and even removed from
	 * the dispatch list).
	 */
	union {
		struct hlist_node hash;	/* merge hash */
		struct list_head ipi_list;
	};

122 123 124
	/*
	 * The rb_node is only used inside the io scheduler, requests
	 * are pruned when moved to the dispatch queue. So let the
125
	 * completion_data share space with the rb_node.
126 127 128
	 */
	union {
		struct rb_node rb_node;	/* sort/lookup */
129
		void *completion_data;
130
	};
131

132
	/*
133
	 * Three pointers are available for the IO schedulers, if they need
134 135
	 * more they have to dynamically allocate it.  Flush requests are
	 * never put on the IO scheduler. So let the flush fields share
136
	 * space with the elevator data.
137
	 */
138
	union {
139 140 141 142 143
		struct {
			struct io_cq		*icq;
			void			*priv[2];
		} elv;

144 145 146
		struct {
			unsigned int		seq;
			struct list_head	list;
147
			rq_end_io_fn		*saved_end_io;
148 149
		} flush;
	};
150

151
	struct gendisk *rq_disk;
152
	struct hd_struct *part;
L
Linus Torvalds 已提交
153
	unsigned long start_time;
154
#ifdef CONFIG_BLK_CGROUP
155
	struct request_list *rl;		/* rl this rq is alloced from */
156 157 158
	unsigned long long start_time_ns;
	unsigned long long io_start_time_ns;    /* when passed to hardware */
#endif
L
Linus Torvalds 已提交
159 160 161 162
	/* Number of scatter-gather DMA addr+len pairs after
	 * physical address coalescing is performed.
	 */
	unsigned short nr_phys_segments;
163 164 165
#if defined(CONFIG_BLK_DEV_INTEGRITY)
	unsigned short nr_integrity_segments;
#endif
L
Linus Torvalds 已提交
166

167 168
	unsigned short ioprio;

T
Tejun Heo 已提交
169
	void *special;		/* opaque pointer available for LLD use */
L
Linus Torvalds 已提交
170

171 172 173
	int tag;
	int errors;

L
Linus Torvalds 已提交
174 175 176
	/*
	 * when request is used as a packet command carrier
	 */
177 178
	unsigned char __cmd[BLK_MAX_CDB];
	unsigned char *cmd;
179
	unsigned short cmd_len;
L
Linus Torvalds 已提交
180

181
	unsigned int extra_len;	/* length of alignment and padding */
L
Linus Torvalds 已提交
182
	unsigned int sense_len;
T
Tejun Heo 已提交
183
	unsigned int resid_len;	/* residual count */
L
Linus Torvalds 已提交
184 185
	void *sense;

J
Jens Axboe 已提交
186 187
	unsigned long deadline;
	struct list_head timeout_list;
L
Linus Torvalds 已提交
188
	unsigned int timeout;
189
	int retries;
L
Linus Torvalds 已提交
190 191

	/*
192
	 * completion callback.
L
Linus Torvalds 已提交
193 194 195
	 */
	rq_end_io_fn *end_io;
	void *end_io_data;
196 197 198

	/* for bidi */
	struct request *next_rq;
L
Linus Torvalds 已提交
199 200
};

201 202 203 204 205
static inline unsigned short req_get_ioprio(struct request *req)
{
	return req->ioprio;
}

L
Linus Torvalds 已提交
206 207
#include <linux/elevator.h>

208 209
struct blk_queue_ctx;

210
typedef void (request_fn_proc) (struct request_queue *q);
211
typedef void (make_request_fn) (struct request_queue *q, struct bio *bio);
212
typedef int (prep_rq_fn) (struct request_queue *, struct request *);
213
typedef void (unprep_rq_fn) (struct request_queue *, struct request *);
L
Linus Torvalds 已提交
214 215

struct bio_vec;
216
typedef void (softirq_done_fn)(struct request *);
217
typedef int (dma_drain_needed_fn)(struct request *);
218
typedef int (lld_busy_fn) (struct request_queue *q);
M
Mike Christie 已提交
219
typedef int (bsg_job_fn) (struct bsg_job *);
L
Linus Torvalds 已提交
220

J
Jens Axboe 已提交
221 222 223 224 225 226 227 228
enum blk_eh_timer_return {
	BLK_EH_NOT_HANDLED,
	BLK_EH_HANDLED,
	BLK_EH_RESET_TIMER,
};

typedef enum blk_eh_timer_return (rq_timed_out_fn)(struct request *);

L
Linus Torvalds 已提交
229 230 231 232 233 234 235 236 237 238
enum blk_queue_state {
	Queue_down,
	Queue_up,
};

struct blk_queue_tag {
	struct request **tag_index;	/* map of busy tags */
	unsigned long *tag_map;		/* bit map of free/busy tags */
	int busy;			/* current depth */
	int max_depth;			/* what we will send to device */
239
	int real_max_depth;		/* what the array can hold */
L
Linus Torvalds 已提交
240
	atomic_t refcnt;		/* map can be shared */
241 242
	int alloc_policy;		/* tag allocation policy */
	int next_tag;			/* next tag */
L
Linus Torvalds 已提交
243
};
244 245
#define BLK_TAG_ALLOC_FIFO 0 /* allocate starting from 0 */
#define BLK_TAG_ALLOC_RR 1 /* allocate starting from last allocated tag */
L
Linus Torvalds 已提交
246

247 248 249
#define BLK_SCSI_MAX_CMDS	(256)
#define BLK_SCSI_CMD_PER_LONG	(BLK_SCSI_MAX_CMDS / (sizeof(long) * 8))

250 251 252
struct queue_limits {
	unsigned long		bounce_pfn;
	unsigned long		seg_boundary_mask;
253
	unsigned long		virt_boundary_mask;
254 255

	unsigned int		max_hw_sectors;
256
	unsigned int		chunk_sectors;
257 258
	unsigned int		max_sectors;
	unsigned int		max_segment_size;
259 260 261 262
	unsigned int		physical_block_size;
	unsigned int		alignment_offset;
	unsigned int		io_min;
	unsigned int		io_opt;
263
	unsigned int		max_discard_sectors;
264
	unsigned int		max_hw_discard_sectors;
265
	unsigned int		max_write_same_sectors;
266 267
	unsigned int		discard_granularity;
	unsigned int		discard_alignment;
268 269

	unsigned short		logical_block_size;
270
	unsigned short		max_segments;
271
	unsigned short		max_integrity_segments;
272

273
	unsigned char		misaligned;
274
	unsigned char		discard_misaligned;
275
	unsigned char		cluster;
276
	unsigned char		discard_zeroes_data;
277
	unsigned char		raid_partial_stripes_expensive;
278 279
};

280
struct request_queue {
L
Linus Torvalds 已提交
281 282 283 284 285
	/*
	 * Together with queue_head for cacheline sharing
	 */
	struct list_head	queue_head;
	struct request		*last_merge;
J
Jens Axboe 已提交
286
	struct elevator_queue	*elevator;
287 288
	int			nr_rqs[2];	/* # allocated [a]sync rqs */
	int			nr_rqs_elvpriv;	/* # allocated rqs w/ elvpriv */
L
Linus Torvalds 已提交
289 290

	/*
291 292 293 294
	 * If blkcg is not used, @q->root_rl serves all requests.  If blkcg
	 * is used, root blkg allocates from @q->root_rl and all other
	 * blkgs from their own blkg->rl.  Which one to use should be
	 * determined using bio_request_list().
L
Linus Torvalds 已提交
295
	 */
296
	struct request_list	root_rl;
L
Linus Torvalds 已提交
297 298 299 300

	request_fn_proc		*request_fn;
	make_request_fn		*make_request_fn;
	prep_rq_fn		*prep_rq_fn;
301
	unprep_rq_fn		*unprep_rq_fn;
302
	softirq_done_fn		*softirq_done_fn;
J
Jens Axboe 已提交
303
	rq_timed_out_fn		*rq_timed_out_fn;
304
	dma_drain_needed_fn	*dma_drain_needed;
305
	lld_busy_fn		*lld_busy_fn;
L
Linus Torvalds 已提交
306

307 308 309 310 311
	struct blk_mq_ops	*mq_ops;

	unsigned int		*mq_map;

	/* sw queues */
312
	struct blk_mq_ctx __percpu	*queue_ctx;
313 314 315 316 317 318
	unsigned int		nr_queues;

	/* hw dispatch queues */
	struct blk_mq_hw_ctx	**queue_hw_ctx;
	unsigned int		nr_hw_queues;

319 320 321
	/*
	 * Dispatch queue sorting
	 */
J
Jens Axboe 已提交
322
	sector_t		end_sector;
323 324
	struct request		*boundary_rq;

L
Linus Torvalds 已提交
325
	/*
326
	 * Delayed queue handling
L
Linus Torvalds 已提交
327
	 */
328
	struct delayed_work	delay_work;
L
Linus Torvalds 已提交
329 330 331 332 333 334 335 336 337 338

	struct backing_dev_info	backing_dev_info;

	/*
	 * The queue owner gets to use this for whatever they like.
	 * ll_rw_blk doesn't touch it.
	 */
	void			*queuedata;

	/*
339
	 * various queue flags, see QUEUE_* below
L
Linus Torvalds 已提交
340
	 */
341
	unsigned long		queue_flags;
L
Linus Torvalds 已提交
342

343 344 345 346 347 348
	/*
	 * ida allocated id for this queue.  Used to index queues from
	 * ioctx.
	 */
	int			id;

L
Linus Torvalds 已提交
349
	/*
350
	 * queue needs bounce pages for pages above this limit
L
Linus Torvalds 已提交
351
	 */
352
	gfp_t			bounce_gfp;
L
Linus Torvalds 已提交
353 354

	/*
已提交
355 356 357
	 * protects queue structures from reentrancy. ->__queue_lock should
	 * _never_ be used directly, it is queue private. always use
	 * ->queue_lock.
L
Linus Torvalds 已提交
358
	 */
已提交
359
	spinlock_t		__queue_lock;
L
Linus Torvalds 已提交
360 361 362 363 364 365 366
	spinlock_t		*queue_lock;

	/*
	 * queue kobject
	 */
	struct kobject kobj;

367 368 369 370 371
	/*
	 * mq queue kobject
	 */
	struct kobject mq_kobj;

372
#ifdef CONFIG_PM
L
Lin Ming 已提交
373 374 375 376 377
	struct device		*dev;
	int			rpm_status;
	unsigned int		nr_pending;
#endif

L
Linus Torvalds 已提交
378 379 380 381 382 383 384 385
	/*
	 * queue settings
	 */
	unsigned long		nr_requests;	/* Max # of requests */
	unsigned int		nr_congestion_on;
	unsigned int		nr_congestion_off;
	unsigned int		nr_batching;

J
James Bottomley 已提交
386
	unsigned int		dma_drain_size;
387
	void			*dma_drain_buffer;
388
	unsigned int		dma_pad_mask;
L
Linus Torvalds 已提交
389 390 391
	unsigned int		dma_alignment;

	struct blk_queue_tag	*queue_tags;
392
	struct list_head	tag_busy_list;
L
Linus Torvalds 已提交
393

394
	unsigned int		nr_sorted;
395
	unsigned int		in_flight[2];
396 397 398 399 400 401
	/*
	 * Number of active block driver functions for which blk_drain_queue()
	 * must wait. Must be incremented around functions that unlock the
	 * queue_lock internally, e.g. scsi_request_fn().
	 */
	unsigned int		request_fn_active;
L
Linus Torvalds 已提交
402

J
Jens Axboe 已提交
403 404 405 406
	unsigned int		rq_timeout;
	struct timer_list	timeout;
	struct list_head	timeout_list;

407
	struct list_head	icq_list;
408
#ifdef CONFIG_BLK_CGROUP
409
	DECLARE_BITMAP		(blkcg_pols, BLKCG_MAX_POLS);
T
Tejun Heo 已提交
410
	struct blkcg_gq		*root_blkg;
411
	struct list_head	blkg_list;
412
#endif
413

414 415
	struct queue_limits	limits;

L
Linus Torvalds 已提交
416 417 418 419 420
	/*
	 * sg stuff
	 */
	unsigned int		sg_timeout;
	unsigned int		sg_reserved_size;
421
	int			node;
422
#ifdef CONFIG_BLK_DEV_IO_TRACE
423
	struct blk_trace	*blk_trace;
424
#endif
L
Linus Torvalds 已提交
425
	/*
426
	 * for flush operations
L
Linus Torvalds 已提交
427
	 */
428
	unsigned int		flush_flags;
429
	unsigned int		flush_not_queueable:1;
430
	struct blk_flush_queue	*fq;
431

432 433 434 435
	struct list_head	requeue_list;
	spinlock_t		requeue_lock;
	struct work_struct	requeue_work;

436
	struct mutex		sysfs_lock;
437

438
	int			bypass_depth;
439
	atomic_t		mq_freeze_depth;
440

441
#if defined(CONFIG_BLK_DEV_BSG)
M
Mike Christie 已提交
442 443
	bsg_job_fn		*bsg_job_fn;
	int			bsg_job_size;
444 445
	struct bsg_class_device bsg_dev;
#endif
446 447 448 449 450

#ifdef CONFIG_BLK_DEV_THROTTLING
	/* Throttle data */
	struct throtl_data *td;
#endif
T
Tejun Heo 已提交
451
	struct rcu_head		rcu_head;
452
	wait_queue_head_t	mq_freeze_wq;
453
	struct percpu_ref	mq_usage_counter;
454
	struct list_head	all_q_node;
455 456 457

	struct blk_mq_tag_set	*tag_set;
	struct list_head	tag_set_list;
458
	struct bio_set		*bio_split;
L
Linus Torvalds 已提交
459 460 461 462
};

#define QUEUE_FLAG_QUEUED	1	/* uses generic tag queueing */
#define QUEUE_FLAG_STOPPED	2	/* queue is stopped */
463 464
#define	QUEUE_FLAG_SYNCFULL	3	/* read queue has been filled */
#define QUEUE_FLAG_ASYNCFULL	4	/* write queue has been filled */
B
Bart Van Assche 已提交
465
#define QUEUE_FLAG_DYING	5	/* queue being torn down */
466
#define QUEUE_FLAG_BYPASS	6	/* act as dumb FIFO queue */
467 468
#define QUEUE_FLAG_BIDI		7	/* queue supports bidi requests */
#define QUEUE_FLAG_NOMERGES     8	/* disable merge attempts */
D
Dan Williams 已提交
469
#define QUEUE_FLAG_SAME_COMP	9	/* complete on same CPU-group */
470 471 472
#define QUEUE_FLAG_FAIL_IO     10	/* fake timeout */
#define QUEUE_FLAG_STACKABLE   11	/* supports request stacking */
#define QUEUE_FLAG_NONROT      12	/* non-rotational device (SSD) */
473
#define QUEUE_FLAG_VIRT        QUEUE_FLAG_NONROT /* paravirt device */
474 475 476 477 478
#define QUEUE_FLAG_IO_STAT     13	/* do IO stats */
#define QUEUE_FLAG_DISCARD     14	/* supports DISCARD */
#define QUEUE_FLAG_NOXMERGES   15	/* No extended merges */
#define QUEUE_FLAG_ADD_RANDOM  16	/* Contributes to random pool */
#define QUEUE_FLAG_SECDISCARD  17	/* supports SECDISCARD */
D
Dan Williams 已提交
479
#define QUEUE_FLAG_SAME_FORCE  18	/* force complete on same CPU */
480
#define QUEUE_FLAG_DEAD        19	/* queue tear-down finished */
481
#define QUEUE_FLAG_INIT_DONE   20	/* queue is initialized */
482
#define QUEUE_FLAG_NO_SG_MERGE 21	/* don't attempt to merge SG segments*/
483 484

#define QUEUE_FLAG_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
485
				 (1 << QUEUE_FLAG_STACKABLE)	|	\
486 487
				 (1 << QUEUE_FLAG_SAME_COMP)	|	\
				 (1 << QUEUE_FLAG_ADD_RANDOM))
488

489
#define QUEUE_FLAG_MQ_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
490
				 (1 << QUEUE_FLAG_STACKABLE)	|	\
491 492
				 (1 << QUEUE_FLAG_SAME_COMP))

493
static inline void queue_lockdep_assert_held(struct request_queue *q)
494
{
495 496
	if (q->queue_lock)
		lockdep_assert_held(q->queue_lock);
497 498
}

N
Nick Piggin 已提交
499 500 501 502 503 504
static inline void queue_flag_set_unlocked(unsigned int flag,
					   struct request_queue *q)
{
	__set_bit(flag, &q->queue_flags);
}

J
Jens Axboe 已提交
505 506 507
static inline int queue_flag_test_and_clear(unsigned int flag,
					    struct request_queue *q)
{
508
	queue_lockdep_assert_held(q);
J
Jens Axboe 已提交
509 510 511 512 513 514 515 516 517 518 519 520

	if (test_bit(flag, &q->queue_flags)) {
		__clear_bit(flag, &q->queue_flags);
		return 1;
	}

	return 0;
}

static inline int queue_flag_test_and_set(unsigned int flag,
					  struct request_queue *q)
{
521
	queue_lockdep_assert_held(q);
J
Jens Axboe 已提交
522 523 524 525 526 527 528 529 530

	if (!test_bit(flag, &q->queue_flags)) {
		__set_bit(flag, &q->queue_flags);
		return 0;
	}

	return 1;
}

N
Nick Piggin 已提交
531 532
static inline void queue_flag_set(unsigned int flag, struct request_queue *q)
{
533
	queue_lockdep_assert_held(q);
N
Nick Piggin 已提交
534 535 536 537 538 539 540 541 542
	__set_bit(flag, &q->queue_flags);
}

static inline void queue_flag_clear_unlocked(unsigned int flag,
					     struct request_queue *q)
{
	__clear_bit(flag, &q->queue_flags);
}

543 544 545 546 547
static inline int queue_in_flight(struct request_queue *q)
{
	return q->in_flight[0] + q->in_flight[1];
}

N
Nick Piggin 已提交
548 549
static inline void queue_flag_clear(unsigned int flag, struct request_queue *q)
{
550
	queue_lockdep_assert_held(q);
N
Nick Piggin 已提交
551 552 553
	__clear_bit(flag, &q->queue_flags);
}

L
Linus Torvalds 已提交
554 555
#define blk_queue_tagged(q)	test_bit(QUEUE_FLAG_QUEUED, &(q)->queue_flags)
#define blk_queue_stopped(q)	test_bit(QUEUE_FLAG_STOPPED, &(q)->queue_flags)
B
Bart Van Assche 已提交
556
#define blk_queue_dying(q)	test_bit(QUEUE_FLAG_DYING, &(q)->queue_flags)
557
#define blk_queue_dead(q)	test_bit(QUEUE_FLAG_DEAD, &(q)->queue_flags)
558
#define blk_queue_bypass(q)	test_bit(QUEUE_FLAG_BYPASS, &(q)->queue_flags)
559
#define blk_queue_init_done(q)	test_bit(QUEUE_FLAG_INIT_DONE, &(q)->queue_flags)
560
#define blk_queue_nomerges(q)	test_bit(QUEUE_FLAG_NOMERGES, &(q)->queue_flags)
561 562
#define blk_queue_noxmerges(q)	\
	test_bit(QUEUE_FLAG_NOXMERGES, &(q)->queue_flags)
563
#define blk_queue_nonrot(q)	test_bit(QUEUE_FLAG_NONROT, &(q)->queue_flags)
564
#define blk_queue_io_stat(q)	test_bit(QUEUE_FLAG_IO_STAT, &(q)->queue_flags)
565
#define blk_queue_add_random(q)	test_bit(QUEUE_FLAG_ADD_RANDOM, &(q)->queue_flags)
566 567
#define blk_queue_stackable(q)	\
	test_bit(QUEUE_FLAG_STACKABLE, &(q)->queue_flags)
568
#define blk_queue_discard(q)	test_bit(QUEUE_FLAG_DISCARD, &(q)->queue_flags)
A
Adrian Hunter 已提交
569 570
#define blk_queue_secdiscard(q)	(blk_queue_discard(q) && \
	test_bit(QUEUE_FLAG_SECDISCARD, &(q)->queue_flags))
L
Linus Torvalds 已提交
571

572 573 574 575 576 577
#define blk_noretry_request(rq) \
	((rq)->cmd_flags & (REQ_FAILFAST_DEV|REQ_FAILFAST_TRANSPORT| \
			     REQ_FAILFAST_DRIVER))

#define blk_account_rq(rq) \
	(((rq)->cmd_flags & REQ_STARTED) && \
578
	 ((rq)->cmd_type == REQ_TYPE_FS))
579

580
#define blk_rq_cpu_valid(rq)	((rq)->cpu != -1)
581
#define blk_bidi_rq(rq)		((rq)->next_rq != NULL)
582 583
/* rq->queuelist of dequeued request must be list_empty() */
#define blk_queued_rq(rq)	(!list_empty(&(rq)->queuelist))
L
Linus Torvalds 已提交
584 585 586

#define list_entry_rq(ptr)	list_entry((ptr), struct request, queuelist)

J
Jens Axboe 已提交
587
#define rq_data_dir(rq)		(((rq)->cmd_flags & 1) != 0)
L
Linus Torvalds 已提交
588

589 590 591 592 593 594 595 596 597
/*
 * Driver can handle struct request, if it either has an old style
 * request_fn defined, or is blk-mq based.
 */
static inline bool queue_is_rq_based(struct request_queue *q)
{
	return q->request_fn || q->mq_ops;
}

598 599 600 601 602
static inline unsigned int blk_queue_cluster(struct request_queue *q)
{
	return q->limits.cluster;
}

603
/*
604
 * We regard a request as sync, if either a read or a sync write
605
 */
606 607
static inline bool rw_is_sync(unsigned int rw_flags)
{
608
	return !(rw_flags & REQ_WRITE) || (rw_flags & REQ_SYNC);
609 610 611 612 613 614 615
}

static inline bool rq_is_sync(struct request *rq)
{
	return rw_is_sync(rq->cmd_flags);
}

616
static inline bool blk_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
617
{
618 619 620
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	return rl->flags & flag;
L
Linus Torvalds 已提交
621 622
}

623
static inline void blk_set_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
624
{
625 626 627
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	rl->flags |= flag;
L
Linus Torvalds 已提交
628 629
}

630
static inline void blk_clear_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
631
{
632 633 634
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	rl->flags &= ~flag;
L
Linus Torvalds 已提交
635 636
}

637 638 639 640
static inline bool rq_mergeable(struct request *rq)
{
	if (rq->cmd_type != REQ_TYPE_FS)
		return false;
L
Linus Torvalds 已提交
641

642 643 644 645 646
	if (rq->cmd_flags & REQ_NOMERGE_FLAGS)
		return false;

	return true;
}
L
Linus Torvalds 已提交
647

648 649 650 651 652 653 654 655 656
static inline bool blk_check_merge_flags(unsigned int flags1,
					 unsigned int flags2)
{
	if ((flags1 & REQ_DISCARD) != (flags2 & REQ_DISCARD))
		return false;

	if ((flags1 & REQ_SECURE) != (flags2 & REQ_SECURE))
		return false;

657 658 659
	if ((flags1 & REQ_WRITE_SAME) != (flags2 & REQ_WRITE_SAME))
		return false;

660 661 662
	return true;
}

663 664 665 666 667 668 669 670
static inline bool blk_write_same_mergeable(struct bio *a, struct bio *b)
{
	if (bio_data(a) == bio_data(b))
		return true;

	return false;
}

L
Linus Torvalds 已提交
671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686
/*
 * q->prep_rq_fn return values
 */
#define BLKPREP_OK		0	/* serve it */
#define BLKPREP_KILL		1	/* fatal error, kill */
#define BLKPREP_DEFER		2	/* leave on queue */

extern unsigned long blk_max_low_pfn, blk_max_pfn;

/*
 * standard bounce addresses:
 *
 * BLK_BOUNCE_HIGH	: bounce all highmem pages
 * BLK_BOUNCE_ANY	: don't bounce anything
 * BLK_BOUNCE_ISA	: bounce pages above ISA DMA boundary
 */
687 688

#if BITS_PER_LONG == 32
L
Linus Torvalds 已提交
689
#define BLK_BOUNCE_HIGH		((u64)blk_max_low_pfn << PAGE_SHIFT)
690 691 692 693
#else
#define BLK_BOUNCE_HIGH		-1ULL
#endif
#define BLK_BOUNCE_ANY		(-1ULL)
694
#define BLK_BOUNCE_ISA		(DMA_BIT_MASK(24))
L
Linus Torvalds 已提交
695

696 697 698 699
/*
 * default timeout for SG_IO if none specified
 */
#define BLK_DEFAULT_SG_TIMEOUT	(60 * HZ)
L
Linus Torvalds 已提交
700
#define BLK_MIN_SG_TIMEOUT	(7 * HZ)
701

702
#ifdef CONFIG_BOUNCE
L
Linus Torvalds 已提交
703
extern int init_emergency_isa_pool(void);
704
extern void blk_queue_bounce(struct request_queue *q, struct bio **bio);
L
Linus Torvalds 已提交
705 706 707 708 709
#else
static inline int init_emergency_isa_pool(void)
{
	return 0;
}
710
static inline void blk_queue_bounce(struct request_queue *q, struct bio **bio)
L
Linus Torvalds 已提交
711 712 713 714
{
}
#endif /* CONFIG_MMU */

715 716 717 718
struct rq_map_data {
	struct page **pages;
	int page_order;
	int nr_entries;
719
	unsigned long offset;
720
	int null_mapped;
721
	int from_user;
722 723
};

724
struct req_iterator {
725
	struct bvec_iter iter;
726 727 728 729
	struct bio *bio;
};

/* This should not be used directly - use rq_for_each_segment */
730 731
#define for_each_bio(_bio)		\
	for (; _bio; _bio = _bio->bi_next)
732
#define __rq_for_each_bio(_bio, rq)	\
L
Linus Torvalds 已提交
733 734 735
	if ((rq->bio))			\
		for (_bio = (rq)->bio; _bio; _bio = _bio->bi_next)

736 737
#define rq_for_each_segment(bvl, _rq, _iter)			\
	__rq_for_each_bio(_iter.bio, _rq)			\
738
		bio_for_each_segment(bvl, _iter.bio, _iter.iter)
739

K
Kent Overstreet 已提交
740
#define rq_iter_last(bvec, _iter)				\
741
		(_iter.bio->bi_next == NULL &&			\
K
Kent Overstreet 已提交
742
		 bio_iter_last(bvec, _iter.iter))
743

744 745 746 747 748 749 750 751 752 753 754
#ifndef ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE
# error	"You should define ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE for your platform"
#endif
#if ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE
extern void rq_flush_dcache_pages(struct request *rq);
#else
static inline void rq_flush_dcache_pages(struct request *rq)
{
}
#endif

L
Linus Torvalds 已提交
755 756 757
extern int blk_register_queue(struct gendisk *disk);
extern void blk_unregister_queue(struct gendisk *disk);
extern void generic_make_request(struct bio *bio);
758
extern void blk_rq_init(struct request_queue *q, struct request *rq);
L
Linus Torvalds 已提交
759
extern void blk_put_request(struct request *);
760 761
extern void __blk_put_request(struct request_queue *, struct request *);
extern struct request *blk_get_request(struct request_queue *, int, gfp_t);
762 763
extern struct request *blk_make_request(struct request_queue *, struct bio *,
					gfp_t);
J
Jens Axboe 已提交
764
extern void blk_rq_set_block_pc(struct request *);
765
extern void blk_requeue_request(struct request_queue *, struct request *);
766 767
extern void blk_add_request_payload(struct request *rq, struct page *page,
		unsigned int len);
768
extern int blk_rq_check_limits(struct request_queue *q, struct request *rq);
769
extern int blk_lld_busy(struct request_queue *q);
770 771 772 773 774
extern int blk_rq_prep_clone(struct request *rq, struct request *rq_src,
			     struct bio_set *bs, gfp_t gfp_mask,
			     int (*bio_ctr)(struct bio *, struct bio *, void *),
			     void *data);
extern void blk_rq_unprep_clone(struct request *rq);
775 776
extern int blk_insert_cloned_request(struct request_queue *q,
				     struct request *rq);
777
extern void blk_delay_queue(struct request_queue *, unsigned long);
778 779
extern void blk_queue_split(struct request_queue *, struct bio **,
			    struct bio_set *);
780
extern void blk_recount_segments(struct request_queue *, struct bio *);
781
extern int scsi_verify_blk_ioctl(struct block_device *, unsigned int);
782 783
extern int scsi_cmd_blk_ioctl(struct block_device *, fmode_t,
			      unsigned int, void __user *);
784 785
extern int scsi_cmd_ioctl(struct request_queue *, struct gendisk *, fmode_t,
			  unsigned int, void __user *);
786 787
extern int sg_scsi_ioctl(struct request_queue *, struct gendisk *, fmode_t,
			 struct scsi_ioctl_command __user *);
788

789 790
extern void blk_start_queue(struct request_queue *q);
extern void blk_stop_queue(struct request_queue *q);
L
Linus Torvalds 已提交
791
extern void blk_sync_queue(struct request_queue *q);
792
extern void __blk_stop_queue(struct request_queue *q);
793
extern void __blk_run_queue(struct request_queue *q);
794
extern void __blk_run_queue_uncond(struct request_queue *q);
795
extern void blk_run_queue(struct request_queue *);
796
extern void blk_run_queue_async(struct request_queue *q);
797
extern int blk_rq_map_user(struct request_queue *, struct request *,
798 799
			   struct rq_map_data *, void __user *, unsigned long,
			   gfp_t);
J
Jens Axboe 已提交
800
extern int blk_rq_unmap_user(struct bio *);
801 802
extern int blk_rq_map_kern(struct request_queue *, struct request *, void *, unsigned int, gfp_t);
extern int blk_rq_map_user_iov(struct request_queue *, struct request *,
803 804
			       struct rq_map_data *, const struct iov_iter *,
			       gfp_t);
805
extern int blk_execute_rq(struct request_queue *, struct gendisk *,
806
			  struct request *, int);
807
extern void blk_execute_rq_nowait(struct request_queue *, struct gendisk *,
808
				  struct request *, int, rq_end_io_fn *);
809

810
static inline struct request_queue *bdev_get_queue(struct block_device *bdev)
L
Linus Torvalds 已提交
811
{
812
	return bdev->bd_disk->queue;	/* this is never NULL */
L
Linus Torvalds 已提交
813 814
}

815
/*
816 817 818 819 820 821
 * blk_rq_pos()			: the current sector
 * blk_rq_bytes()		: bytes left in the entire request
 * blk_rq_cur_bytes()		: bytes left in the current segment
 * blk_rq_err_bytes()		: bytes left till the next error boundary
 * blk_rq_sectors()		: sectors left in the entire request
 * blk_rq_cur_sectors()		: sectors left in the current segment
822
 */
823 824
static inline sector_t blk_rq_pos(const struct request *rq)
{
825
	return rq->__sector;
826 827 828 829
}

static inline unsigned int blk_rq_bytes(const struct request *rq)
{
830
	return rq->__data_len;
831 832
}

833 834 835 836
static inline int blk_rq_cur_bytes(const struct request *rq)
{
	return rq->bio ? bio_cur_bytes(rq->bio) : 0;
}
837

838 839
extern unsigned int blk_rq_err_bytes(const struct request *rq);

840 841
static inline unsigned int blk_rq_sectors(const struct request *rq)
{
842
	return blk_rq_bytes(rq) >> 9;
843 844 845 846
}

static inline unsigned int blk_rq_cur_sectors(const struct request *rq)
{
847
	return blk_rq_cur_bytes(rq) >> 9;
848 849
}

850 851 852 853
static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
						     unsigned int cmd_flags)
{
	if (unlikely(cmd_flags & REQ_DISCARD))
854
		return min(q->limits.max_discard_sectors, UINT_MAX >> 9);
855

856 857 858
	if (unlikely(cmd_flags & REQ_WRITE_SAME))
		return q->limits.max_write_same_sectors;

859 860 861
	return q->limits.max_sectors;
}

862 863 864 865 866 867 868 869
/*
 * Return maximum size of a request at given offset. Only valid for
 * file system requests.
 */
static inline unsigned int blk_max_size_offset(struct request_queue *q,
					       sector_t offset)
{
	if (!q->limits.chunk_sectors)
870
		return q->limits.max_sectors;
871 872 873 874 875

	return q->limits.chunk_sectors -
			(offset & (q->limits.chunk_sectors - 1));
}

876 877 878 879 880 881 882
static inline unsigned int blk_rq_get_max_sectors(struct request *rq)
{
	struct request_queue *q = rq->q;

	if (unlikely(rq->cmd_type == REQ_TYPE_BLOCK_PC))
		return q->limits.max_hw_sectors;

883
	if (!q->limits.chunk_sectors || (rq->cmd_flags & REQ_DISCARD))
884 885 886 887
		return blk_queue_get_max_sectors(q, rq->cmd_flags);

	return min(blk_max_size_offset(q, blk_rq_pos(rq)),
			blk_queue_get_max_sectors(q, rq->cmd_flags));
888 889
}

890 891 892 893 894 895 896 897 898 899 900
static inline unsigned int blk_rq_count_bios(struct request *rq)
{
	unsigned int nr_bios = 0;
	struct bio *bio;

	__rq_for_each_bio(bio, rq)
		nr_bios++;

	return nr_bios;
}

901 902 903 904 905 906 907
/*
 * Request issue related functions.
 */
extern struct request *blk_peek_request(struct request_queue *q);
extern void blk_start_request(struct request *rq);
extern struct request *blk_fetch_request(struct request_queue *q);

L
Linus Torvalds 已提交
908
/*
909 910 911 912 913
 * Request completion related functions.
 *
 * blk_update_request() completes given number of bytes and updates
 * the request without completing it.
 *
914 915
 * blk_end_request() and friends.  __blk_end_request() must be called
 * with the request queue spinlock acquired.
L
Linus Torvalds 已提交
916 917
 *
 * Several drivers define their own end_request and call
918 919
 * blk_end_request() for parts of the original function.
 * This prevents code duplication in drivers.
L
Linus Torvalds 已提交
920
 */
921 922
extern bool blk_update_request(struct request *rq, int error,
			       unsigned int nr_bytes);
923
extern void blk_finish_request(struct request *rq, int error);
924 925 926 927
extern bool blk_end_request(struct request *rq, int error,
			    unsigned int nr_bytes);
extern void blk_end_request_all(struct request *rq, int error);
extern bool blk_end_request_cur(struct request *rq, int error);
928
extern bool blk_end_request_err(struct request *rq, int error);
929 930 931 932
extern bool __blk_end_request(struct request *rq, int error,
			      unsigned int nr_bytes);
extern void __blk_end_request_all(struct request *rq, int error);
extern bool __blk_end_request_cur(struct request *rq, int error);
933
extern bool __blk_end_request_err(struct request *rq, int error);
934

935
extern void blk_complete_request(struct request *);
J
Jens Axboe 已提交
936 937
extern void __blk_complete_request(struct request *);
extern void blk_abort_request(struct request *);
938
extern void blk_unprep_request(struct request *);
939

L
Linus Torvalds 已提交
940 941 942
/*
 * Access functions for manipulating queue properties
 */
943
extern struct request_queue *blk_init_queue_node(request_fn_proc *rfn,
944
					spinlock_t *lock, int node_id);
945
extern struct request_queue *blk_init_queue(request_fn_proc *, spinlock_t *);
946 947
extern struct request_queue *blk_init_allocated_queue(struct request_queue *,
						      request_fn_proc *, spinlock_t *);
948 949 950
extern void blk_cleanup_queue(struct request_queue *);
extern void blk_queue_make_request(struct request_queue *, make_request_fn *);
extern void blk_queue_bounce_limit(struct request_queue *, u64);
951
extern void blk_limits_max_hw_sectors(struct queue_limits *, unsigned int);
952
extern void blk_queue_max_hw_sectors(struct request_queue *, unsigned int);
953
extern void blk_queue_chunk_sectors(struct request_queue *, unsigned int);
954
extern void blk_queue_max_segments(struct request_queue *, unsigned short);
955
extern void blk_queue_max_segment_size(struct request_queue *, unsigned int);
956 957
extern void blk_queue_max_discard_sectors(struct request_queue *q,
		unsigned int max_discard_sectors);
958 959
extern void blk_queue_max_write_same_sectors(struct request_queue *q,
		unsigned int max_write_same_sectors);
960
extern void blk_queue_logical_block_size(struct request_queue *, unsigned short);
961
extern void blk_queue_physical_block_size(struct request_queue *, unsigned int);
962 963
extern void blk_queue_alignment_offset(struct request_queue *q,
				       unsigned int alignment);
964
extern void blk_limits_io_min(struct queue_limits *limits, unsigned int min);
965
extern void blk_queue_io_min(struct request_queue *q, unsigned int min);
966
extern void blk_limits_io_opt(struct queue_limits *limits, unsigned int opt);
967
extern void blk_queue_io_opt(struct request_queue *q, unsigned int opt);
968
extern void blk_set_default_limits(struct queue_limits *lim);
969
extern void blk_set_stacking_limits(struct queue_limits *lim);
970 971
extern int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
			    sector_t offset);
972 973
extern int bdev_stack_limits(struct queue_limits *t, struct block_device *bdev,
			    sector_t offset);
974 975
extern void disk_stack_limits(struct gendisk *disk, struct block_device *bdev,
			      sector_t offset);
976
extern void blk_queue_stack_limits(struct request_queue *t, struct request_queue *b);
977
extern void blk_queue_dma_pad(struct request_queue *, unsigned int);
978
extern void blk_queue_update_dma_pad(struct request_queue *, unsigned int);
979 980 981
extern int blk_queue_dma_drain(struct request_queue *q,
			       dma_drain_needed_fn *dma_drain_needed,
			       void *buf, unsigned int size);
982
extern void blk_queue_lld_busy(struct request_queue *q, lld_busy_fn *fn);
983
extern void blk_queue_segment_boundary(struct request_queue *, unsigned long);
984
extern void blk_queue_virt_boundary(struct request_queue *, unsigned long);
985
extern void blk_queue_prep_rq(struct request_queue *, prep_rq_fn *pfn);
986
extern void blk_queue_unprep_rq(struct request_queue *, unprep_rq_fn *ufn);
987
extern void blk_queue_dma_alignment(struct request_queue *, int);
988
extern void blk_queue_update_dma_alignment(struct request_queue *, int);
989
extern void blk_queue_softirq_done(struct request_queue *, softirq_done_fn *);
J
Jens Axboe 已提交
990 991
extern void blk_queue_rq_timed_out(struct request_queue *, rq_timed_out_fn *);
extern void blk_queue_rq_timeout(struct request_queue *, unsigned int);
992
extern void blk_queue_flush(struct request_queue *q, unsigned int flush);
993
extern void blk_queue_flush_queueable(struct request_queue *q, bool queueable);
L
Linus Torvalds 已提交
994 995
extern struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev);

996
extern int blk_rq_map_sg(struct request_queue *, struct request *, struct scatterlist *);
L
Linus Torvalds 已提交
997 998 999
extern void blk_dump_rq_flags(struct request *, char *);
extern long nr_blockdev_pages(void);

T
Tejun Heo 已提交
1000
bool __must_check blk_get_queue(struct request_queue *);
1001 1002 1003
struct request_queue *blk_alloc_queue(gfp_t);
struct request_queue *blk_alloc_queue_node(gfp_t, int);
extern void blk_put_queue(struct request_queue *);
1004
extern void blk_set_queue_dying(struct request_queue *);
L
Linus Torvalds 已提交
1005

L
Lin Ming 已提交
1006 1007 1008
/*
 * block layer runtime pm functions
 */
1009
#ifdef CONFIG_PM
L
Lin Ming 已提交
1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026
extern void blk_pm_runtime_init(struct request_queue *q, struct device *dev);
extern int blk_pre_runtime_suspend(struct request_queue *q);
extern void blk_post_runtime_suspend(struct request_queue *q, int err);
extern void blk_pre_runtime_resume(struct request_queue *q);
extern void blk_post_runtime_resume(struct request_queue *q, int err);
#else
static inline void blk_pm_runtime_init(struct request_queue *q,
	struct device *dev) {}
static inline int blk_pre_runtime_suspend(struct request_queue *q)
{
	return -ENOSYS;
}
static inline void blk_post_runtime_suspend(struct request_queue *q, int err) {}
static inline void blk_pre_runtime_resume(struct request_queue *q) {}
static inline void blk_post_runtime_resume(struct request_queue *q, int err) {}
#endif

1027
/*
S
Suresh Jayaraman 已提交
1028 1029 1030 1031 1032 1033 1034 1035 1036 1037
 * blk_plug permits building a queue of related requests by holding the I/O
 * fragments for a short period. This allows merging of sequential requests
 * into single larger request. As the requests are moved from a per-task list to
 * the device's request_queue in a batch, this results in improved scalability
 * as the lock contention for request_queue lock is reduced.
 *
 * It is ok not to disable preemption when adding the request to the plug list
 * or when attempting a merge, because blk_schedule_flush_list() will only flush
 * the plug list when the task sleeps by itself. For details, please see
 * schedule() where blk_schedule_flush_plug() is called.
1038
 */
1039
struct blk_plug {
S
Suresh Jayaraman 已提交
1040
	struct list_head list; /* requests */
1041
	struct list_head mq_list; /* blk-mq requests */
S
Suresh Jayaraman 已提交
1042
	struct list_head cb_list; /* md requires an unplug callback */
1043
};
1044 1045
#define BLK_MAX_REQUEST_COUNT 16

1046
struct blk_plug_cb;
1047
typedef void (*blk_plug_cb_fn)(struct blk_plug_cb *, bool);
1048 1049
struct blk_plug_cb {
	struct list_head list;
1050 1051
	blk_plug_cb_fn callback;
	void *data;
1052
};
1053 1054
extern struct blk_plug_cb *blk_check_plugged(blk_plug_cb_fn unplug,
					     void *data, int size);
1055 1056
extern void blk_start_plug(struct blk_plug *);
extern void blk_finish_plug(struct blk_plug *);
1057
extern void blk_flush_plug_list(struct blk_plug *, bool);
1058 1059 1060 1061 1062

static inline void blk_flush_plug(struct task_struct *tsk)
{
	struct blk_plug *plug = tsk->plug;

1063 1064 1065 1066 1067 1068 1069 1070
	if (plug)
		blk_flush_plug_list(plug, false);
}

static inline void blk_schedule_flush_plug(struct task_struct *tsk)
{
	struct blk_plug *plug = tsk->plug;

1071
	if (plug)
1072
		blk_flush_plug_list(plug, true);
1073 1074 1075 1076 1077 1078
}

static inline bool blk_needs_flush_plug(struct task_struct *tsk)
{
	struct blk_plug *plug = tsk->plug;

1079 1080 1081 1082
	return plug &&
		(!list_empty(&plug->list) ||
		 !list_empty(&plug->mq_list) ||
		 !list_empty(&plug->cb_list));
1083 1084
}

L
Linus Torvalds 已提交
1085 1086 1087
/*
 * tag stuff
 */
1088 1089 1090
extern int blk_queue_start_tag(struct request_queue *, struct request *);
extern struct request *blk_queue_find_tag(struct request_queue *, int);
extern void blk_queue_end_tag(struct request_queue *, struct request *);
1091
extern int blk_queue_init_tags(struct request_queue *, int, struct blk_queue_tag *, int);
1092 1093 1094
extern void blk_queue_free_tags(struct request_queue *);
extern int blk_queue_resize_tags(struct request_queue *, int);
extern void blk_queue_invalidate_tags(struct request_queue *);
1095
extern struct blk_queue_tag *blk_init_tags(int, int);
1096
extern void blk_free_tags(struct blk_queue_tag *);
L
Linus Torvalds 已提交
1097

1098 1099 1100 1101 1102 1103 1104
static inline struct request *blk_map_queue_find_tag(struct blk_queue_tag *bqt,
						int tag)
{
	if (unlikely(bqt == NULL || tag >= bqt->real_max_depth))
		return NULL;
	return bqt->tag_index[tag];
}
1105 1106 1107 1108

#define BLKDEV_DISCARD_SECURE  0x01    /* secure discard */

extern int blkdev_issue_flush(struct block_device *, gfp_t, sector_t *);
1109 1110
extern int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, unsigned long flags);
1111 1112
extern int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, struct page *page);
1113
extern int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
1114
		sector_t nr_sects, gfp_t gfp_mask, bool discard);
1115 1116
static inline int sb_issue_discard(struct super_block *sb, sector_t block,
		sector_t nr_blocks, gfp_t gfp_mask, unsigned long flags)
D
David Woodhouse 已提交
1117
{
1118 1119 1120
	return blkdev_issue_discard(sb->s_bdev, block << (sb->s_blocksize_bits - 9),
				    nr_blocks << (sb->s_blocksize_bits - 9),
				    gfp_mask, flags);
D
David Woodhouse 已提交
1121
}
1122
static inline int sb_issue_zeroout(struct super_block *sb, sector_t block,
1123
		sector_t nr_blocks, gfp_t gfp_mask)
1124 1125 1126 1127
{
	return blkdev_issue_zeroout(sb->s_bdev,
				    block << (sb->s_blocksize_bits - 9),
				    nr_blocks << (sb->s_blocksize_bits - 9),
1128
				    gfp_mask, true);
1129
}
L
Linus Torvalds 已提交
1130

1131
extern int blk_verify_command(unsigned char *cmd, fmode_t has_write_perm);
1132

1133 1134 1135
enum blk_default_limits {
	BLK_MAX_SEGMENTS	= 128,
	BLK_SAFE_MAX_SECTORS	= 255,
1136
	BLK_DEF_MAX_SECTORS	= 2560,
1137 1138 1139
	BLK_MAX_SEGMENT_SIZE	= 65536,
	BLK_SEG_BOUNDARY_MASK	= 0xFFFFFFFFUL,
};
1140

L
Linus Torvalds 已提交
1141 1142
#define blkdev_entry_to_request(entry) list_entry((entry), struct request, queuelist)

1143 1144
static inline unsigned long queue_bounce_pfn(struct request_queue *q)
{
1145
	return q->limits.bounce_pfn;
1146 1147 1148 1149
}

static inline unsigned long queue_segment_boundary(struct request_queue *q)
{
1150
	return q->limits.seg_boundary_mask;
1151 1152
}

1153 1154 1155 1156 1157
static inline unsigned long queue_virt_boundary(struct request_queue *q)
{
	return q->limits.virt_boundary_mask;
}

1158 1159
static inline unsigned int queue_max_sectors(struct request_queue *q)
{
1160
	return q->limits.max_sectors;
1161 1162 1163 1164
}

static inline unsigned int queue_max_hw_sectors(struct request_queue *q)
{
1165
	return q->limits.max_hw_sectors;
1166 1167
}

1168
static inline unsigned short queue_max_segments(struct request_queue *q)
1169
{
1170
	return q->limits.max_segments;
1171 1172 1173 1174
}

static inline unsigned int queue_max_segment_size(struct request_queue *q)
{
1175
	return q->limits.max_segment_size;
1176 1177
}

1178
static inline unsigned short queue_logical_block_size(struct request_queue *q)
L
Linus Torvalds 已提交
1179 1180 1181
{
	int retval = 512;

1182 1183
	if (q && q->limits.logical_block_size)
		retval = q->limits.logical_block_size;
L
Linus Torvalds 已提交
1184 1185 1186 1187

	return retval;
}

1188
static inline unsigned short bdev_logical_block_size(struct block_device *bdev)
L
Linus Torvalds 已提交
1189
{
1190
	return queue_logical_block_size(bdev_get_queue(bdev));
L
Linus Torvalds 已提交
1191 1192
}

1193 1194 1195 1196 1197
static inline unsigned int queue_physical_block_size(struct request_queue *q)
{
	return q->limits.physical_block_size;
}

1198
static inline unsigned int bdev_physical_block_size(struct block_device *bdev)
M
Martin K. Petersen 已提交
1199 1200 1201 1202
{
	return queue_physical_block_size(bdev_get_queue(bdev));
}

1203 1204 1205 1206 1207
static inline unsigned int queue_io_min(struct request_queue *q)
{
	return q->limits.io_min;
}

M
Martin K. Petersen 已提交
1208 1209 1210 1211 1212
static inline int bdev_io_min(struct block_device *bdev)
{
	return queue_io_min(bdev_get_queue(bdev));
}

1213 1214 1215 1216 1217
static inline unsigned int queue_io_opt(struct request_queue *q)
{
	return q->limits.io_opt;
}

M
Martin K. Petersen 已提交
1218 1219 1220 1221 1222
static inline int bdev_io_opt(struct block_device *bdev)
{
	return queue_io_opt(bdev_get_queue(bdev));
}

1223 1224
static inline int queue_alignment_offset(struct request_queue *q)
{
M
Martin K. Petersen 已提交
1225
	if (q->limits.misaligned)
1226 1227
		return -1;

M
Martin K. Petersen 已提交
1228
	return q->limits.alignment_offset;
1229 1230
}

1231
static inline int queue_limit_alignment_offset(struct queue_limits *lim, sector_t sector)
1232 1233
{
	unsigned int granularity = max(lim->physical_block_size, lim->io_min);
1234
	unsigned int alignment = sector_div(sector, granularity >> 9) << 9;
1235

1236
	return (granularity + lim->alignment_offset - alignment) % granularity;
1237 1238
}

M
Martin K. Petersen 已提交
1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
static inline int bdev_alignment_offset(struct block_device *bdev)
{
	struct request_queue *q = bdev_get_queue(bdev);

	if (q->limits.misaligned)
		return -1;

	if (bdev != bdev->bd_contains)
		return bdev->bd_part->alignment_offset;

	return q->limits.alignment_offset;
}

1252 1253 1254 1255 1256 1257 1258 1259
static inline int queue_discard_alignment(struct request_queue *q)
{
	if (q->limits.discard_misaligned)
		return -1;

	return q->limits.discard_alignment;
}

1260
static inline int queue_limit_discard_alignment(struct queue_limits *lim, sector_t sector)
1261
{
1262
	unsigned int alignment, granularity, offset;
1263

1264 1265 1266
	if (!lim->max_discard_sectors)
		return 0;

1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280
	/* Why are these in bytes, not sectors? */
	alignment = lim->discard_alignment >> 9;
	granularity = lim->discard_granularity >> 9;
	if (!granularity)
		return 0;

	/* Offset of the partition start in 'granularity' sectors */
	offset = sector_div(sector, granularity);

	/* And why do we do this modulus *again* in blkdev_issue_discard()? */
	offset = (granularity + alignment - offset) % granularity;

	/* Turn it back into bytes, gaah */
	return offset << 9;
1281 1282
}

1283 1284 1285 1286 1287 1288 1289 1290 1291 1292
static inline int bdev_discard_alignment(struct block_device *bdev)
{
	struct request_queue *q = bdev_get_queue(bdev);

	if (bdev != bdev->bd_contains)
		return bdev->bd_part->discard_alignment;

	return q->limits.discard_alignment;
}

1293 1294
static inline unsigned int queue_discard_zeroes_data(struct request_queue *q)
{
1295
	if (q->limits.max_discard_sectors && q->limits.discard_zeroes_data == 1)
1296 1297 1298 1299 1300 1301 1302 1303 1304 1305
		return 1;

	return 0;
}

static inline unsigned int bdev_discard_zeroes_data(struct block_device *bdev)
{
	return queue_discard_zeroes_data(bdev_get_queue(bdev));
}

1306 1307 1308 1309 1310 1311 1312 1313 1314 1315
static inline unsigned int bdev_write_same(struct block_device *bdev)
{
	struct request_queue *q = bdev_get_queue(bdev);

	if (q)
		return q->limits.max_write_same_sectors;

	return 0;
}

1316
static inline int queue_dma_alignment(struct request_queue *q)
L
Linus Torvalds 已提交
1317
{
1318
	return q ? q->dma_alignment : 511;
L
Linus Torvalds 已提交
1319 1320
}

1321
static inline int blk_rq_aligned(struct request_queue *q, unsigned long addr,
1322 1323 1324
				 unsigned int len)
{
	unsigned int alignment = queue_dma_alignment(q) | q->dma_pad_mask;
1325
	return !(addr & alignment) && !(len & alignment);
1326 1327
}

L
Linus Torvalds 已提交
1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
/* assumes size > 256 */
static inline unsigned int blksize_bits(unsigned int size)
{
	unsigned int bits = 8;
	do {
		bits++;
		size >>= 1;
	} while (size > 256);
	return bits;
}

1339
static inline unsigned int block_size(struct block_device *bdev)
L
Linus Torvalds 已提交
1340 1341 1342 1343
{
	return bdev->bd_block_size;
}

1344 1345 1346 1347 1348
static inline bool queue_flush_queueable(struct request_queue *q)
{
	return !q->flush_not_queueable;
}

L
Linus Torvalds 已提交
1349 1350 1351 1352 1353 1354 1355 1356 1357
typedef struct {struct page *v;} Sector;

unsigned char *read_dev_sector(struct block_device *, sector_t, Sector *);

static inline void put_dev_sector(Sector p)
{
	page_cache_release(p.v);
}

1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370
/*
 * Check if adding a bio_vec after bprv with offset would create a gap in
 * the SG list. Most drivers don't care about this, but some do.
 */
static inline bool bvec_gap_to_prev(struct request_queue *q,
				struct bio_vec *bprv, unsigned int offset)
{
	if (!queue_virt_boundary(q))
		return false;
	return offset ||
		((bprv->bv_offset + bprv->bv_len) & queue_virt_boundary(q));
}

1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390
static inline bool bio_will_gap(struct request_queue *q, struct bio *prev,
			 struct bio *next)
{
	if (!bio_has_data(prev))
		return false;

	return bvec_gap_to_prev(q, &prev->bi_io_vec[prev->bi_vcnt - 1],
				next->bi_io_vec[0].bv_offset);
}

static inline bool req_gap_back_merge(struct request *req, struct bio *bio)
{
	return bio_will_gap(req->q, req->biotail, bio);
}

static inline bool req_gap_front_merge(struct request *req, struct bio *bio)
{
	return bio_will_gap(req->q, bio, req->bio);
}

L
Linus Torvalds 已提交
1391
struct work_struct;
1392 1393
int kblockd_schedule_work(struct work_struct *work);
int kblockd_schedule_delayed_work(struct delayed_work *dwork, unsigned long delay);
1394
int kblockd_schedule_delayed_work_on(int cpu, struct delayed_work *dwork, unsigned long delay);
L
Linus Torvalds 已提交
1395

1396
#ifdef CONFIG_BLK_CGROUP
1397 1398 1399 1400 1401
/*
 * This should not be using sched_clock(). A real patch is in progress
 * to fix this up, until that is in place we need to disable preemption
 * around sched_clock() in this function and set_io_start_time_ns().
 */
1402 1403
static inline void set_start_time_ns(struct request *req)
{
1404
	preempt_disable();
1405
	req->start_time_ns = sched_clock();
1406
	preempt_enable();
1407 1408 1409 1410
}

static inline void set_io_start_time_ns(struct request *req)
{
1411
	preempt_disable();
1412
	req->io_start_time_ns = sched_clock();
1413
	preempt_enable();
1414
}
1415 1416 1417 1418 1419 1420 1421 1422 1423 1424

static inline uint64_t rq_start_time_ns(struct request *req)
{
        return req->start_time_ns;
}

static inline uint64_t rq_io_start_time_ns(struct request *req)
{
        return req->io_start_time_ns;
}
1425 1426 1427
#else
static inline void set_start_time_ns(struct request *req) {}
static inline void set_io_start_time_ns(struct request *req) {}
1428 1429 1430 1431 1432 1433 1434 1435
static inline uint64_t rq_start_time_ns(struct request *req)
{
	return 0;
}
static inline uint64_t rq_io_start_time_ns(struct request *req)
{
	return 0;
}
1436 1437
#endif

L
Linus Torvalds 已提交
1438 1439 1440 1441 1442
#define MODULE_ALIAS_BLOCKDEV(major,minor) \
	MODULE_ALIAS("block-major-" __stringify(major) "-" __stringify(minor))
#define MODULE_ALIAS_BLOCKDEV_MAJOR(major) \
	MODULE_ALIAS("block-major-" __stringify(major) "-*")

1443 1444
#if defined(CONFIG_BLK_DEV_INTEGRITY)

1445 1446 1447
enum blk_integrity_flags {
	BLK_INTEGRITY_VERIFY		= 1 << 0,
	BLK_INTEGRITY_GENERATE		= 1 << 1,
1448
	BLK_INTEGRITY_DEVICE_CAPABLE	= 1 << 2,
1449
	BLK_INTEGRITY_IP_CHECKSUM	= 1 << 3,
1450
};
1451

1452
struct blk_integrity_iter {
1453 1454
	void			*prot_buf;
	void			*data_buf;
1455
	sector_t		seed;
1456
	unsigned int		data_size;
1457
	unsigned short		interval;
1458 1459 1460
	const char		*disk_name;
};

1461
typedef int (integrity_processing_fn) (struct blk_integrity_iter *);
1462 1463

struct blk_integrity {
1464 1465
	integrity_processing_fn	*generate_fn;
	integrity_processing_fn	*verify_fn;
1466 1467 1468

	unsigned short		flags;
	unsigned short		tuple_size;
1469
	unsigned short		interval;
1470 1471 1472 1473 1474 1475 1476
	unsigned short		tag_size;

	const char		*name;

	struct kobject		kobj;
};

1477
extern bool blk_integrity_is_initialized(struct gendisk *);
1478 1479
extern int blk_integrity_register(struct gendisk *, struct blk_integrity *);
extern void blk_integrity_unregister(struct gendisk *);
1480
extern int blk_integrity_compare(struct gendisk *, struct gendisk *);
1481 1482 1483
extern int blk_rq_map_integrity_sg(struct request_queue *, struct bio *,
				   struct scatterlist *);
extern int blk_rq_count_integrity_sg(struct request_queue *, struct bio *);
1484 1485 1486 1487
extern bool blk_integrity_merge_rq(struct request_queue *, struct request *,
				   struct request *);
extern bool blk_integrity_merge_bio(struct request_queue *, struct request *,
				    struct bio *);
1488

1489 1490 1491 1492 1493 1494
static inline
struct blk_integrity *bdev_get_integrity(struct block_device *bdev)
{
	return bdev->bd_disk->integrity;
}

1495 1496 1497 1498 1499
static inline struct blk_integrity *blk_get_integrity(struct gendisk *disk)
{
	return disk->integrity;
}

1500
static inline bool blk_integrity_rq(struct request *rq)
1501
{
1502
	return rq->cmd_flags & REQ_INTEGRITY;
1503 1504
}

1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
static inline void blk_queue_max_integrity_segments(struct request_queue *q,
						    unsigned int segs)
{
	q->limits.max_integrity_segments = segs;
}

static inline unsigned short
queue_max_integrity_segments(struct request_queue *q)
{
	return q->limits.max_integrity_segments;
}

1517 1518
#else /* CONFIG_BLK_DEV_INTEGRITY */

1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540
struct bio;
struct block_device;
struct gendisk;
struct blk_integrity;

static inline int blk_integrity_rq(struct request *rq)
{
	return 0;
}
static inline int blk_rq_count_integrity_sg(struct request_queue *q,
					    struct bio *b)
{
	return 0;
}
static inline int blk_rq_map_integrity_sg(struct request_queue *q,
					  struct bio *b,
					  struct scatterlist *s)
{
	return 0;
}
static inline struct blk_integrity *bdev_get_integrity(struct block_device *b)
{
1541
	return NULL;
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566
}
static inline struct blk_integrity *blk_get_integrity(struct gendisk *disk)
{
	return NULL;
}
static inline int blk_integrity_compare(struct gendisk *a, struct gendisk *b)
{
	return 0;
}
static inline int blk_integrity_register(struct gendisk *d,
					 struct blk_integrity *b)
{
	return 0;
}
static inline void blk_integrity_unregister(struct gendisk *d)
{
}
static inline void blk_queue_max_integrity_segments(struct request_queue *q,
						    unsigned int segs)
{
}
static inline unsigned short queue_max_integrity_segments(struct request_queue *q)
{
	return 0;
}
1567 1568 1569
static inline bool blk_integrity_merge_rq(struct request_queue *rq,
					  struct request *r1,
					  struct request *r2)
1570
{
1571
	return true;
1572
}
1573 1574 1575
static inline bool blk_integrity_merge_bio(struct request_queue *rq,
					   struct request *r,
					   struct bio *b)
1576
{
1577
	return true;
1578 1579 1580 1581 1582
}
static inline bool blk_integrity_is_initialized(struct gendisk *g)
{
	return 0;
}
1583 1584 1585

#endif /* CONFIG_BLK_DEV_INTEGRITY */

1586
struct block_device_operations {
A
Al Viro 已提交
1587
	int (*open) (struct block_device *, fmode_t);
1588
	void (*release) (struct gendisk *, fmode_t);
1589
	int (*rw_page)(struct block_device *, sector_t, struct page *, int rw);
A
Al Viro 已提交
1590 1591
	int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
	int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
1592 1593
	long (*direct_access)(struct block_device *, sector_t,
					void **, unsigned long *pfn, long size);
1594 1595 1596
	unsigned int (*check_events) (struct gendisk *disk,
				      unsigned int clearing);
	/* ->media_changed() is DEPRECATED, use ->check_events() instead */
1597
	int (*media_changed) (struct gendisk *);
1598
	void (*unlock_native_capacity) (struct gendisk *);
1599 1600
	int (*revalidate_disk) (struct gendisk *);
	int (*getgeo)(struct block_device *, struct hd_geometry *);
1601 1602
	/* this callback is with swap_lock and sometimes page table lock held */
	void (*swap_slot_free_notify) (struct block_device *, unsigned long);
1603 1604 1605
	struct module *owner;
};

1606 1607
extern int __blkdev_driver_ioctl(struct block_device *, fmode_t, unsigned int,
				 unsigned long);
1608 1609 1610
extern int bdev_read_page(struct block_device *, sector_t, struct page *);
extern int bdev_write_page(struct block_device *, sector_t, struct page *,
						struct writeback_control *);
1611 1612
extern long bdev_direct_access(struct block_device *, sector_t, void **addr,
						unsigned long *pfn, long size);
1613
#else /* CONFIG_BLOCK */
1614 1615 1616

struct block_device;

1617 1618 1619 1620 1621 1622 1623 1624 1625 1626
/*
 * stubs for when the block layer is configured out
 */
#define buffer_heads_over_limit 0

static inline long nr_blockdev_pages(void)
{
	return 0;
}

1627 1628 1629 1630
struct blk_plug {
};

static inline void blk_start_plug(struct blk_plug *plug)
1631 1632 1633
{
}

1634
static inline void blk_finish_plug(struct blk_plug *plug)
1635 1636 1637
{
}

1638
static inline void blk_flush_plug(struct task_struct *task)
1639 1640 1641
{
}

1642 1643 1644 1645 1646
static inline void blk_schedule_flush_plug(struct task_struct *task)
{
}


1647 1648 1649 1650 1651
static inline bool blk_needs_flush_plug(struct task_struct *tsk)
{
	return false;
}

1652 1653 1654 1655 1656 1657
static inline int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask,
				     sector_t *error_sector)
{
	return 0;
}

1658 1659
#endif /* CONFIG_BLOCK */

L
Linus Torvalds 已提交
1660
#endif