blkdev.h 48.3 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
#include <linux/wait.h>
#include <linux/mempool.h>
D
Dan Williams 已提交
18
#include <linux/pfn.h>
L
Linus Torvalds 已提交
19 20
#include <linux/bio.h>
#include <linux/stringify.h>
21
#include <linux/gfp.h>
22
#include <linux/bsg.h>
23
#include <linux/smp.h>
T
Tejun Heo 已提交
24
#include <linux/rcupdate.h>
25
#include <linux/percpu-refcount.h>
C
Christoph Hellwig 已提交
26
#include <linux/scatterlist.h>
L
Linus Torvalds 已提交
27

28
struct module;
29 30
struct scsi_ioctl_command;

L
Linus Torvalds 已提交
31 32
struct request_queue;
struct elevator_queue;
33
struct blk_trace;
34 35
struct request;
struct sg_io_hdr;
M
Mike Christie 已提交
36
struct bsg_job;
T
Tejun Heo 已提交
37
struct blkcg_gq;
38
struct blk_flush_queue;
39
struct pr_ops;
L
Linus Torvalds 已提交
40 41 42 43

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

T
Tejun Heo 已提交
44 45 46 47 48 49
/*
 * 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
50
struct request;
51
typedef void (rq_end_io_fn)(struct request *, int);
L
Linus Torvalds 已提交
52

53 54 55
#define BLK_RL_SYNCFULL		(1U << 0)
#define BLK_RL_ASYNCFULL	(1U << 1)

L
Linus Torvalds 已提交
56
struct request_list {
57
	struct request_queue	*q;	/* the queue this rl belongs to */
58 59 60
#ifdef CONFIG_BLK_CGROUP
	struct blkcg_gq		*blkg;	/* blkg this request pool belongs to */
#endif
61 62 63 64
	/*
	 * count[], starved[], and wait[] are indexed by
	 * BLK_RW_SYNC/BLK_RW_ASYNC
	 */
65 66 67 68
	int			count[2];
	int			starved[2];
	mempool_t		*rq_pool;
	wait_queue_head_t	wait[2];
69
	unsigned int		flags;
L
Linus Torvalds 已提交
70 71
};

72 73 74 75 76 77
/*
 * request command types
 */
enum rq_cmd_type_bits {
	REQ_TYPE_FS		= 1,	/* fs request */
	REQ_TYPE_BLOCK_PC,		/* scsi command */
78
	REQ_TYPE_DRV_PRIV,		/* driver defined types from here */
79 80
};

L
Linus Torvalds 已提交
81 82 83
#define BLK_MAX_CDB	16

/*
84 85 86 87
 * 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
88 89
 */
struct request {
90
	struct list_head queuelist;
91 92
	union {
		struct call_single_data csd;
93
		unsigned long fifo_time;
94
	};
95

96
	struct request_queue *q;
97
	struct blk_mq_ctx *mq_ctx;
98

J
Jens Axboe 已提交
99
	u64 cmd_flags;
100
	unsigned cmd_type;
J
Jens Axboe 已提交
101
	unsigned long atomic_flags;
L
Linus Torvalds 已提交
102

103 104
	int cpu;

105 106
	/* the following two fields are internal, NEVER access directly */
	unsigned int __data_len;	/* total data len */
107
	sector_t __sector;		/* sector cursor */
L
Linus Torvalds 已提交
108 109 110 111

	struct bio *bio;
	struct bio *biotail;

112 113 114 115 116 117 118 119 120 121 122 123
	/*
	 * 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;
	};

124 125 126
	/*
	 * The rb_node is only used inside the io scheduler, requests
	 * are pruned when moved to the dispatch queue. So let the
127
	 * completion_data share space with the rb_node.
128 129 130
	 */
	union {
		struct rb_node rb_node;	/* sort/lookup */
131
		void *completion_data;
132
	};
133

134
	/*
135
	 * Three pointers are available for the IO schedulers, if they need
136 137
	 * more they have to dynamically allocate it.  Flush requests are
	 * never put on the IO scheduler. So let the flush fields share
138
	 * space with the elevator data.
139
	 */
140
	union {
141 142 143 144 145
		struct {
			struct io_cq		*icq;
			void			*priv[2];
		} elv;

146 147 148
		struct {
			unsigned int		seq;
			struct list_head	list;
149
			rq_end_io_fn		*saved_end_io;
150 151
		} flush;
	};
152

153
	struct gendisk *rq_disk;
154
	struct hd_struct *part;
L
Linus Torvalds 已提交
155
	unsigned long start_time;
156
#ifdef CONFIG_BLK_CGROUP
157
	struct request_list *rl;		/* rl this rq is alloced from */
158 159 160
	unsigned long long start_time_ns;
	unsigned long long io_start_time_ns;    /* when passed to hardware */
#endif
L
Linus Torvalds 已提交
161 162 163 164
	/* Number of scatter-gather DMA addr+len pairs after
	 * physical address coalescing is performed.
	 */
	unsigned short nr_phys_segments;
165 166 167
#if defined(CONFIG_BLK_DEV_INTEGRITY)
	unsigned short nr_integrity_segments;
#endif
L
Linus Torvalds 已提交
168

169 170
	unsigned short ioprio;

T
Tejun Heo 已提交
171
	void *special;		/* opaque pointer available for LLD use */
L
Linus Torvalds 已提交
172

173 174 175
	int tag;
	int errors;

L
Linus Torvalds 已提交
176 177 178
	/*
	 * when request is used as a packet command carrier
	 */
179 180
	unsigned char __cmd[BLK_MAX_CDB];
	unsigned char *cmd;
181
	unsigned short cmd_len;
L
Linus Torvalds 已提交
182

183
	unsigned int extra_len;	/* length of alignment and padding */
L
Linus Torvalds 已提交
184
	unsigned int sense_len;
T
Tejun Heo 已提交
185
	unsigned int resid_len;	/* residual count */
L
Linus Torvalds 已提交
186 187
	void *sense;

J
Jens Axboe 已提交
188 189
	unsigned long deadline;
	struct list_head timeout_list;
L
Linus Torvalds 已提交
190
	unsigned int timeout;
191
	int retries;
L
Linus Torvalds 已提交
192 193

	/*
194
	 * completion callback.
L
Linus Torvalds 已提交
195 196 197
	 */
	rq_end_io_fn *end_io;
	void *end_io_data;
198 199 200

	/* for bidi */
	struct request *next_rq;
L
Linus Torvalds 已提交
201 202
};

203 204 205 206 207
static inline unsigned short req_get_ioprio(struct request *req)
{
	return req->ioprio;
}

L
Linus Torvalds 已提交
208 209
#include <linux/elevator.h>

210 211
struct blk_queue_ctx;

212
typedef void (request_fn_proc) (struct request_queue *q);
213
typedef blk_qc_t (make_request_fn) (struct request_queue *q, struct bio *bio);
214
typedef int (prep_rq_fn) (struct request_queue *, struct request *);
215
typedef void (unprep_rq_fn) (struct request_queue *, struct request *);
L
Linus Torvalds 已提交
216 217

struct bio_vec;
218
typedef void (softirq_done_fn)(struct request *);
219
typedef int (dma_drain_needed_fn)(struct request *);
220
typedef int (lld_busy_fn) (struct request_queue *q);
M
Mike Christie 已提交
221
typedef int (bsg_job_fn) (struct bsg_job *);
L
Linus Torvalds 已提交
222

J
Jens Axboe 已提交
223 224 225 226 227 228 229 230
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
231 232 233 234 235 236 237 238 239 240
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 */
241
	int real_max_depth;		/* what the array can hold */
L
Linus Torvalds 已提交
242
	atomic_t refcnt;		/* map can be shared */
243 244
	int alloc_policy;		/* tag allocation policy */
	int next_tag;			/* next tag */
L
Linus Torvalds 已提交
245
};
246 247
#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
248

249 250 251
#define BLK_SCSI_MAX_CMDS	(256)
#define BLK_SCSI_CMD_PER_LONG	(BLK_SCSI_MAX_CMDS / (sizeof(long) * 8))

252 253 254
struct queue_limits {
	unsigned long		bounce_pfn;
	unsigned long		seg_boundary_mask;
255
	unsigned long		virt_boundary_mask;
256 257

	unsigned int		max_hw_sectors;
258
	unsigned int		max_dev_sectors;
259
	unsigned int		chunk_sectors;
260 261
	unsigned int		max_sectors;
	unsigned int		max_segment_size;
262 263 264 265
	unsigned int		physical_block_size;
	unsigned int		alignment_offset;
	unsigned int		io_min;
	unsigned int		io_opt;
266
	unsigned int		max_discard_sectors;
267
	unsigned int		max_hw_discard_sectors;
268
	unsigned int		max_write_same_sectors;
269 270
	unsigned int		discard_granularity;
	unsigned int		discard_alignment;
271 272

	unsigned short		logical_block_size;
273
	unsigned short		max_segments;
274
	unsigned short		max_integrity_segments;
275

276
	unsigned char		misaligned;
277
	unsigned char		discard_misaligned;
278
	unsigned char		cluster;
279
	unsigned char		discard_zeroes_data;
280
	unsigned char		raid_partial_stripes_expensive;
281 282
};

283
struct request_queue {
L
Linus Torvalds 已提交
284 285 286 287 288
	/*
	 * Together with queue_head for cacheline sharing
	 */
	struct list_head	queue_head;
	struct request		*last_merge;
J
Jens Axboe 已提交
289
	struct elevator_queue	*elevator;
290 291
	int			nr_rqs[2];	/* # allocated [a]sync rqs */
	int			nr_rqs_elvpriv;	/* # allocated rqs w/ elvpriv */
L
Linus Torvalds 已提交
292 293

	/*
294 295 296 297
	 * 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
298
	 */
299
	struct request_list	root_rl;
L
Linus Torvalds 已提交
300 301 302 303

	request_fn_proc		*request_fn;
	make_request_fn		*make_request_fn;
	prep_rq_fn		*prep_rq_fn;
304
	unprep_rq_fn		*unprep_rq_fn;
305
	softirq_done_fn		*softirq_done_fn;
J
Jens Axboe 已提交
306
	rq_timed_out_fn		*rq_timed_out_fn;
307
	dma_drain_needed_fn	*dma_drain_needed;
308
	lld_busy_fn		*lld_busy_fn;
L
Linus Torvalds 已提交
309

310 311 312 313 314
	struct blk_mq_ops	*mq_ops;

	unsigned int		*mq_map;

	/* sw queues */
315
	struct blk_mq_ctx __percpu	*queue_ctx;
316 317 318 319 320 321
	unsigned int		nr_queues;

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

322 323 324
	/*
	 * Dispatch queue sorting
	 */
J
Jens Axboe 已提交
325
	sector_t		end_sector;
326 327
	struct request		*boundary_rq;

L
Linus Torvalds 已提交
328
	/*
329
	 * Delayed queue handling
L
Linus Torvalds 已提交
330
	 */
331
	struct delayed_work	delay_work;
L
Linus Torvalds 已提交
332 333 334 335 336 337 338 339 340 341

	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;

	/*
342
	 * various queue flags, see QUEUE_* below
L
Linus Torvalds 已提交
343
	 */
344
	unsigned long		queue_flags;
L
Linus Torvalds 已提交
345

346 347 348 349 350 351
	/*
	 * ida allocated id for this queue.  Used to index queues from
	 * ioctx.
	 */
	int			id;

L
Linus Torvalds 已提交
352
	/*
353
	 * queue needs bounce pages for pages above this limit
L
Linus Torvalds 已提交
354
	 */
355
	gfp_t			bounce_gfp;
L
Linus Torvalds 已提交
356 357

	/*
已提交
358 359 360
	 * 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
361
	 */
已提交
362
	spinlock_t		__queue_lock;
L
Linus Torvalds 已提交
363 364 365 366 367 368 369
	spinlock_t		*queue_lock;

	/*
	 * queue kobject
	 */
	struct kobject kobj;

370 371 372 373 374
	/*
	 * mq queue kobject
	 */
	struct kobject mq_kobj;

375 376 377 378
#ifdef  CONFIG_BLK_DEV_INTEGRITY
	struct blk_integrity integrity;
#endif	/* CONFIG_BLK_DEV_INTEGRITY */

379
#ifdef CONFIG_PM
L
Lin Ming 已提交
380 381 382 383 384
	struct device		*dev;
	int			rpm_status;
	unsigned int		nr_pending;
#endif

L
Linus Torvalds 已提交
385 386 387 388 389 390 391 392
	/*
	 * 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
393
	unsigned int		dma_drain_size;
394
	void			*dma_drain_buffer;
395
	unsigned int		dma_pad_mask;
L
Linus Torvalds 已提交
396 397 398
	unsigned int		dma_alignment;

	struct blk_queue_tag	*queue_tags;
399
	struct list_head	tag_busy_list;
L
Linus Torvalds 已提交
400

401
	unsigned int		nr_sorted;
402
	unsigned int		in_flight[2];
403 404 405 406 407 408
	/*
	 * 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
409

J
Jens Axboe 已提交
410 411 412 413
	unsigned int		rq_timeout;
	struct timer_list	timeout;
	struct list_head	timeout_list;

414
	struct list_head	icq_list;
415
#ifdef CONFIG_BLK_CGROUP
416
	DECLARE_BITMAP		(blkcg_pols, BLKCG_MAX_POLS);
T
Tejun Heo 已提交
417
	struct blkcg_gq		*root_blkg;
418
	struct list_head	blkg_list;
419
#endif
420

421 422
	struct queue_limits	limits;

L
Linus Torvalds 已提交
423 424 425 426 427
	/*
	 * sg stuff
	 */
	unsigned int		sg_timeout;
	unsigned int		sg_reserved_size;
428
	int			node;
429
#ifdef CONFIG_BLK_DEV_IO_TRACE
430
	struct blk_trace	*blk_trace;
431
#endif
L
Linus Torvalds 已提交
432
	/*
433
	 * for flush operations
L
Linus Torvalds 已提交
434
	 */
435
	unsigned int		flush_flags;
436
	unsigned int		flush_not_queueable:1;
437
	struct blk_flush_queue	*fq;
438

439 440 441 442
	struct list_head	requeue_list;
	spinlock_t		requeue_lock;
	struct work_struct	requeue_work;

443
	struct mutex		sysfs_lock;
444

445
	int			bypass_depth;
446
	atomic_t		mq_freeze_depth;
447

448
#if defined(CONFIG_BLK_DEV_BSG)
M
Mike Christie 已提交
449 450
	bsg_job_fn		*bsg_job_fn;
	int			bsg_job_size;
451 452
	struct bsg_class_device bsg_dev;
#endif
453 454 455 456 457

#ifdef CONFIG_BLK_DEV_THROTTLING
	/* Throttle data */
	struct throtl_data *td;
#endif
T
Tejun Heo 已提交
458
	struct rcu_head		rcu_head;
459
	wait_queue_head_t	mq_freeze_wq;
460
	struct percpu_ref	q_usage_counter;
461
	struct list_head	all_q_node;
462 463 464

	struct blk_mq_tag_set	*tag_set;
	struct list_head	tag_set_list;
465
	struct bio_set		*bio_split;
466 467

	bool			mq_sysfs_init_done;
L
Linus Torvalds 已提交
468 469 470 471
};

#define QUEUE_FLAG_QUEUED	1	/* uses generic tag queueing */
#define QUEUE_FLAG_STOPPED	2	/* queue is stopped */
472 473
#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
474
#define QUEUE_FLAG_DYING	5	/* queue being torn down */
475
#define QUEUE_FLAG_BYPASS	6	/* act as dumb FIFO queue */
476 477
#define QUEUE_FLAG_BIDI		7	/* queue supports bidi requests */
#define QUEUE_FLAG_NOMERGES     8	/* disable merge attempts */
D
Dan Williams 已提交
478
#define QUEUE_FLAG_SAME_COMP	9	/* complete on same CPU-group */
479 480 481
#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) */
482
#define QUEUE_FLAG_VIRT        QUEUE_FLAG_NONROT /* paravirt device */
483 484 485 486 487
#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
488
#define QUEUE_FLAG_SAME_FORCE  18	/* force complete on same CPU */
489
#define QUEUE_FLAG_DEAD        19	/* queue tear-down finished */
490
#define QUEUE_FLAG_INIT_DONE   20	/* queue is initialized */
491
#define QUEUE_FLAG_NO_SG_MERGE 21	/* don't attempt to merge SG segments*/
J
Jens Axboe 已提交
492
#define QUEUE_FLAG_POLL	       22	/* IO polling enabled if set */
493 494

#define QUEUE_FLAG_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
495
				 (1 << QUEUE_FLAG_STACKABLE)	|	\
496 497
				 (1 << QUEUE_FLAG_SAME_COMP)	|	\
				 (1 << QUEUE_FLAG_ADD_RANDOM))
498

499
#define QUEUE_FLAG_MQ_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
500
				 (1 << QUEUE_FLAG_STACKABLE)	|	\
501 502
				 (1 << QUEUE_FLAG_SAME_COMP))

503
static inline void queue_lockdep_assert_held(struct request_queue *q)
504
{
505 506
	if (q->queue_lock)
		lockdep_assert_held(q->queue_lock);
507 508
}

N
Nick Piggin 已提交
509 510 511 512 513 514
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
515 516 517
static inline int queue_flag_test_and_clear(unsigned int flag,
					    struct request_queue *q)
{
518
	queue_lockdep_assert_held(q);
J
Jens Axboe 已提交
519 520 521 522 523 524 525 526 527 528 529 530

	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)
{
531
	queue_lockdep_assert_held(q);
J
Jens Axboe 已提交
532 533 534 535 536 537 538 539 540

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

	return 1;
}

N
Nick Piggin 已提交
541 542
static inline void queue_flag_set(unsigned int flag, struct request_queue *q)
{
543
	queue_lockdep_assert_held(q);
N
Nick Piggin 已提交
544 545 546 547 548 549 550 551 552
	__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);
}

553 554 555 556 557
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
558 559
static inline void queue_flag_clear(unsigned int flag, struct request_queue *q)
{
560
	queue_lockdep_assert_held(q);
N
Nick Piggin 已提交
561 562 563
	__clear_bit(flag, &q->queue_flags);
}

L
Linus Torvalds 已提交
564 565
#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
566
#define blk_queue_dying(q)	test_bit(QUEUE_FLAG_DYING, &(q)->queue_flags)
567
#define blk_queue_dead(q)	test_bit(QUEUE_FLAG_DEAD, &(q)->queue_flags)
568
#define blk_queue_bypass(q)	test_bit(QUEUE_FLAG_BYPASS, &(q)->queue_flags)
569
#define blk_queue_init_done(q)	test_bit(QUEUE_FLAG_INIT_DONE, &(q)->queue_flags)
570
#define blk_queue_nomerges(q)	test_bit(QUEUE_FLAG_NOMERGES, &(q)->queue_flags)
571 572
#define blk_queue_noxmerges(q)	\
	test_bit(QUEUE_FLAG_NOXMERGES, &(q)->queue_flags)
573
#define blk_queue_nonrot(q)	test_bit(QUEUE_FLAG_NONROT, &(q)->queue_flags)
574
#define blk_queue_io_stat(q)	test_bit(QUEUE_FLAG_IO_STAT, &(q)->queue_flags)
575
#define blk_queue_add_random(q)	test_bit(QUEUE_FLAG_ADD_RANDOM, &(q)->queue_flags)
576 577
#define blk_queue_stackable(q)	\
	test_bit(QUEUE_FLAG_STACKABLE, &(q)->queue_flags)
578
#define blk_queue_discard(q)	test_bit(QUEUE_FLAG_DISCARD, &(q)->queue_flags)
A
Adrian Hunter 已提交
579 580
#define blk_queue_secdiscard(q)	(blk_queue_discard(q) && \
	test_bit(QUEUE_FLAG_SECDISCARD, &(q)->queue_flags))
L
Linus Torvalds 已提交
581

582 583 584 585 586 587
#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) && \
588
	 ((rq)->cmd_type == REQ_TYPE_FS))
589

590
#define blk_rq_cpu_valid(rq)	((rq)->cpu != -1)
591
#define blk_bidi_rq(rq)		((rq)->next_rq != NULL)
592 593
/* rq->queuelist of dequeued request must be list_empty() */
#define blk_queued_rq(rq)	(!list_empty(&(rq)->queuelist))
L
Linus Torvalds 已提交
594 595 596

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

597
#define rq_data_dir(rq)		((int)((rq)->cmd_flags & 1))
L
Linus Torvalds 已提交
598

599 600 601 602 603 604 605 606 607
/*
 * 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;
}

608 609 610 611 612
static inline unsigned int blk_queue_cluster(struct request_queue *q)
{
	return q->limits.cluster;
}

613
/*
614
 * We regard a request as sync, if either a read or a sync write
615
 */
616 617
static inline bool rw_is_sync(unsigned int rw_flags)
{
618
	return !(rw_flags & REQ_WRITE) || (rw_flags & REQ_SYNC);
619 620 621 622 623 624 625
}

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

626
static inline bool blk_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
627
{
628 629 630
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	return rl->flags & flag;
L
Linus Torvalds 已提交
631 632
}

633
static inline void blk_set_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
634
{
635 636 637
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	rl->flags |= flag;
L
Linus Torvalds 已提交
638 639
}

640
static inline void blk_clear_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
641
{
642 643 644
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	rl->flags &= ~flag;
L
Linus Torvalds 已提交
645 646
}

647 648 649 650
static inline bool rq_mergeable(struct request *rq)
{
	if (rq->cmd_type != REQ_TYPE_FS)
		return false;
L
Linus Torvalds 已提交
651

652 653 654 655 656
	if (rq->cmd_flags & REQ_NOMERGE_FLAGS)
		return false;

	return true;
}
L
Linus Torvalds 已提交
657

658 659 660 661 662 663 664 665 666
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;

667 668 669
	if ((flags1 & REQ_WRITE_SAME) != (flags2 & REQ_WRITE_SAME))
		return false;

670 671 672
	return true;
}

673 674 675 676 677 678 679 680
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
681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696
/*
 * 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
 */
697 698

#if BITS_PER_LONG == 32
L
Linus Torvalds 已提交
699
#define BLK_BOUNCE_HIGH		((u64)blk_max_low_pfn << PAGE_SHIFT)
700 701 702 703
#else
#define BLK_BOUNCE_HIGH		-1ULL
#endif
#define BLK_BOUNCE_ANY		(-1ULL)
704
#define BLK_BOUNCE_ISA		(DMA_BIT_MASK(24))
L
Linus Torvalds 已提交
705

706 707 708 709
/*
 * default timeout for SG_IO if none specified
 */
#define BLK_DEFAULT_SG_TIMEOUT	(60 * HZ)
L
Linus Torvalds 已提交
710
#define BLK_MIN_SG_TIMEOUT	(7 * HZ)
711

712
#ifdef CONFIG_BOUNCE
L
Linus Torvalds 已提交
713
extern int init_emergency_isa_pool(void);
714
extern void blk_queue_bounce(struct request_queue *q, struct bio **bio);
L
Linus Torvalds 已提交
715 716 717 718 719
#else
static inline int init_emergency_isa_pool(void)
{
	return 0;
}
720
static inline void blk_queue_bounce(struct request_queue *q, struct bio **bio)
L
Linus Torvalds 已提交
721 722 723 724
{
}
#endif /* CONFIG_MMU */

725 726 727 728
struct rq_map_data {
	struct page **pages;
	int page_order;
	int nr_entries;
729
	unsigned long offset;
730
	int null_mapped;
731
	int from_user;
732 733
};

734
struct req_iterator {
735
	struct bvec_iter iter;
736 737 738 739
	struct bio *bio;
};

/* This should not be used directly - use rq_for_each_segment */
740 741
#define for_each_bio(_bio)		\
	for (; _bio; _bio = _bio->bi_next)
742
#define __rq_for_each_bio(_bio, rq)	\
L
Linus Torvalds 已提交
743 744 745
	if ((rq->bio))			\
		for (_bio = (rq)->bio; _bio; _bio = _bio->bi_next)

746 747
#define rq_for_each_segment(bvl, _rq, _iter)			\
	__rq_for_each_bio(_iter.bio, _rq)			\
748
		bio_for_each_segment(bvl, _iter.bio, _iter.iter)
749

K
Kent Overstreet 已提交
750
#define rq_iter_last(bvec, _iter)				\
751
		(_iter.bio->bi_next == NULL &&			\
K
Kent Overstreet 已提交
752
		 bio_iter_last(bvec, _iter.iter))
753

754 755 756 757 758 759 760 761 762 763 764
#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
765 766
extern int blk_register_queue(struct gendisk *disk);
extern void blk_unregister_queue(struct gendisk *disk);
767
extern blk_qc_t generic_make_request(struct bio *bio);
768
extern void blk_rq_init(struct request_queue *q, struct request *rq);
L
Linus Torvalds 已提交
769
extern void blk_put_request(struct request *);
770 771
extern void __blk_put_request(struct request_queue *, struct request *);
extern struct request *blk_get_request(struct request_queue *, int, gfp_t);
772 773
extern struct request *blk_make_request(struct request_queue *, struct bio *,
					gfp_t);
J
Jens Axboe 已提交
774
extern void blk_rq_set_block_pc(struct request *);
775
extern void blk_requeue_request(struct request_queue *, struct request *);
776 777
extern void blk_add_request_payload(struct request *rq, struct page *page,
		unsigned int len);
778
extern int blk_lld_busy(struct request_queue *q);
779 780 781 782 783
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);
784 785
extern int blk_insert_cloned_request(struct request_queue *q,
				     struct request *rq);
786
extern void blk_delay_queue(struct request_queue *, unsigned long);
787 788
extern void blk_queue_split(struct request_queue *, struct bio **,
			    struct bio_set *);
789
extern void blk_recount_segments(struct request_queue *, struct bio *);
790
extern int scsi_verify_blk_ioctl(struct block_device *, unsigned int);
791 792
extern int scsi_cmd_blk_ioctl(struct block_device *, fmode_t,
			      unsigned int, void __user *);
793 794
extern int scsi_cmd_ioctl(struct request_queue *, struct gendisk *, fmode_t,
			  unsigned int, void __user *);
795 796
extern int sg_scsi_ioctl(struct request_queue *, struct gendisk *, fmode_t,
			 struct scsi_ioctl_command __user *);
797

798
extern int blk_queue_enter(struct request_queue *q, bool nowait);
799
extern void blk_queue_exit(struct request_queue *q);
800
extern void blk_start_queue(struct request_queue *q);
J
Jens Axboe 已提交
801
extern void blk_start_queue_async(struct request_queue *q);
802
extern void blk_stop_queue(struct request_queue *q);
L
Linus Torvalds 已提交
803
extern void blk_sync_queue(struct request_queue *q);
804
extern void __blk_stop_queue(struct request_queue *q);
805
extern void __blk_run_queue(struct request_queue *q);
806
extern void __blk_run_queue_uncond(struct request_queue *q);
807
extern void blk_run_queue(struct request_queue *);
808
extern void blk_run_queue_async(struct request_queue *q);
809
extern int blk_rq_map_user(struct request_queue *, struct request *,
810 811
			   struct rq_map_data *, void __user *, unsigned long,
			   gfp_t);
J
Jens Axboe 已提交
812
extern int blk_rq_unmap_user(struct bio *);
813 814
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 *,
815 816
			       struct rq_map_data *, const struct iov_iter *,
			       gfp_t);
817
extern int blk_execute_rq(struct request_queue *, struct gendisk *,
818
			  struct request *, int);
819
extern void blk_execute_rq_nowait(struct request_queue *, struct gendisk *,
820
				  struct request *, int, rq_end_io_fn *);
821

J
Jens Axboe 已提交
822 823
bool blk_poll(struct request_queue *q, blk_qc_t cookie);

824
static inline struct request_queue *bdev_get_queue(struct block_device *bdev)
L
Linus Torvalds 已提交
825
{
826
	return bdev->bd_disk->queue;	/* this is never NULL */
L
Linus Torvalds 已提交
827 828
}

829
/*
830 831 832 833 834 835
 * 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
836
 */
837 838
static inline sector_t blk_rq_pos(const struct request *rq)
{
839
	return rq->__sector;
840 841 842 843
}

static inline unsigned int blk_rq_bytes(const struct request *rq)
{
844
	return rq->__data_len;
845 846
}

847 848 849 850
static inline int blk_rq_cur_bytes(const struct request *rq)
{
	return rq->bio ? bio_cur_bytes(rq->bio) : 0;
}
851

852 853
extern unsigned int blk_rq_err_bytes(const struct request *rq);

854 855
static inline unsigned int blk_rq_sectors(const struct request *rq)
{
856
	return blk_rq_bytes(rq) >> 9;
857 858 859 860
}

static inline unsigned int blk_rq_cur_sectors(const struct request *rq)
{
861
	return blk_rq_cur_bytes(rq) >> 9;
862 863
}

864 865 866 867
static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
						     unsigned int cmd_flags)
{
	if (unlikely(cmd_flags & REQ_DISCARD))
868
		return min(q->limits.max_discard_sectors, UINT_MAX >> 9);
869

870 871 872
	if (unlikely(cmd_flags & REQ_WRITE_SAME))
		return q->limits.max_write_same_sectors;

873 874 875
	return q->limits.max_sectors;
}

876 877 878 879 880 881 882 883
/*
 * 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)
884
		return q->limits.max_sectors;
885 886 887 888 889

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

890 891 892 893 894 895 896
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;

897
	if (!q->limits.chunk_sectors || (rq->cmd_flags & REQ_DISCARD))
898 899 900 901
		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));
902 903
}

904 905 906 907 908 909 910 911 912 913 914
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;
}

915 916 917 918 919 920 921
/*
 * 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
922
/*
923 924 925 926 927
 * Request completion related functions.
 *
 * blk_update_request() completes given number of bytes and updates
 * the request without completing it.
 *
928 929
 * blk_end_request() and friends.  __blk_end_request() must be called
 * with the request queue spinlock acquired.
L
Linus Torvalds 已提交
930 931
 *
 * Several drivers define their own end_request and call
932 933
 * blk_end_request() for parts of the original function.
 * This prevents code duplication in drivers.
L
Linus Torvalds 已提交
934
 */
935 936
extern bool blk_update_request(struct request *rq, int error,
			       unsigned int nr_bytes);
937
extern void blk_finish_request(struct request *rq, int error);
938 939 940 941
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);
942
extern bool blk_end_request_err(struct request *rq, int error);
943 944 945 946
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);
947
extern bool __blk_end_request_err(struct request *rq, int error);
948

949
extern void blk_complete_request(struct request *);
J
Jens Axboe 已提交
950 951
extern void __blk_complete_request(struct request *);
extern void blk_abort_request(struct request *);
952
extern void blk_unprep_request(struct request *);
953

L
Linus Torvalds 已提交
954 955 956
/*
 * Access functions for manipulating queue properties
 */
957
extern struct request_queue *blk_init_queue_node(request_fn_proc *rfn,
958
					spinlock_t *lock, int node_id);
959
extern struct request_queue *blk_init_queue(request_fn_proc *, spinlock_t *);
960 961
extern struct request_queue *blk_init_allocated_queue(struct request_queue *,
						      request_fn_proc *, spinlock_t *);
962 963 964
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);
965
extern void blk_queue_max_hw_sectors(struct request_queue *, unsigned int);
966
extern void blk_queue_chunk_sectors(struct request_queue *, unsigned int);
967
extern void blk_queue_max_segments(struct request_queue *, unsigned short);
968
extern void blk_queue_max_segment_size(struct request_queue *, unsigned int);
969 970
extern void blk_queue_max_discard_sectors(struct request_queue *q,
		unsigned int max_discard_sectors);
971 972
extern void blk_queue_max_write_same_sectors(struct request_queue *q,
		unsigned int max_write_same_sectors);
973
extern void blk_queue_logical_block_size(struct request_queue *, unsigned short);
974
extern void blk_queue_physical_block_size(struct request_queue *, unsigned int);
975 976
extern void blk_queue_alignment_offset(struct request_queue *q,
				       unsigned int alignment);
977
extern void blk_limits_io_min(struct queue_limits *limits, unsigned int min);
978
extern void blk_queue_io_min(struct request_queue *q, unsigned int min);
979
extern void blk_limits_io_opt(struct queue_limits *limits, unsigned int opt);
980
extern void blk_queue_io_opt(struct request_queue *q, unsigned int opt);
981
extern void blk_set_default_limits(struct queue_limits *lim);
982
extern void blk_set_stacking_limits(struct queue_limits *lim);
983 984
extern int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
			    sector_t offset);
985 986
extern int bdev_stack_limits(struct queue_limits *t, struct block_device *bdev,
			    sector_t offset);
987 988
extern void disk_stack_limits(struct gendisk *disk, struct block_device *bdev,
			      sector_t offset);
989
extern void blk_queue_stack_limits(struct request_queue *t, struct request_queue *b);
990
extern void blk_queue_dma_pad(struct request_queue *, unsigned int);
991
extern void blk_queue_update_dma_pad(struct request_queue *, unsigned int);
992 993 994
extern int blk_queue_dma_drain(struct request_queue *q,
			       dma_drain_needed_fn *dma_drain_needed,
			       void *buf, unsigned int size);
995
extern void blk_queue_lld_busy(struct request_queue *q, lld_busy_fn *fn);
996
extern void blk_queue_segment_boundary(struct request_queue *, unsigned long);
997
extern void blk_queue_virt_boundary(struct request_queue *, unsigned long);
998
extern void blk_queue_prep_rq(struct request_queue *, prep_rq_fn *pfn);
999
extern void blk_queue_unprep_rq(struct request_queue *, unprep_rq_fn *ufn);
1000
extern void blk_queue_dma_alignment(struct request_queue *, int);
1001
extern void blk_queue_update_dma_alignment(struct request_queue *, int);
1002
extern void blk_queue_softirq_done(struct request_queue *, softirq_done_fn *);
J
Jens Axboe 已提交
1003 1004
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);
1005
extern void blk_queue_flush(struct request_queue *q, unsigned int flush);
1006
extern void blk_queue_flush_queueable(struct request_queue *q, bool queueable);
L
Linus Torvalds 已提交
1007 1008
extern struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev);

1009
extern int blk_rq_map_sg(struct request_queue *, struct request *, struct scatterlist *);
L
Linus Torvalds 已提交
1010 1011 1012
extern void blk_dump_rq_flags(struct request *, char *);
extern long nr_blockdev_pages(void);

T
Tejun Heo 已提交
1013
bool __must_check blk_get_queue(struct request_queue *);
1014 1015 1016
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 *);
1017
extern void blk_set_queue_dying(struct request_queue *);
L
Linus Torvalds 已提交
1018

L
Lin Ming 已提交
1019 1020 1021
/*
 * block layer runtime pm functions
 */
1022
#ifdef CONFIG_PM
L
Lin Ming 已提交
1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039
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

1040
/*
S
Suresh Jayaraman 已提交
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050
 * 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.
1051
 */
1052
struct blk_plug {
S
Suresh Jayaraman 已提交
1053
	struct list_head list; /* requests */
1054
	struct list_head mq_list; /* blk-mq requests */
S
Suresh Jayaraman 已提交
1055
	struct list_head cb_list; /* md requires an unplug callback */
1056
};
1057 1058
#define BLK_MAX_REQUEST_COUNT 16

1059
struct blk_plug_cb;
1060
typedef void (*blk_plug_cb_fn)(struct blk_plug_cb *, bool);
1061 1062
struct blk_plug_cb {
	struct list_head list;
1063 1064
	blk_plug_cb_fn callback;
	void *data;
1065
};
1066 1067
extern struct blk_plug_cb *blk_check_plugged(blk_plug_cb_fn unplug,
					     void *data, int size);
1068 1069
extern void blk_start_plug(struct blk_plug *);
extern void blk_finish_plug(struct blk_plug *);
1070
extern void blk_flush_plug_list(struct blk_plug *, bool);
1071 1072 1073 1074 1075

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

1076 1077 1078 1079 1080 1081 1082 1083
	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;

1084
	if (plug)
1085
		blk_flush_plug_list(plug, true);
1086 1087 1088 1089 1090 1091
}

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

1092 1093 1094 1095
	return plug &&
		(!list_empty(&plug->list) ||
		 !list_empty(&plug->mq_list) ||
		 !list_empty(&plug->cb_list));
1096 1097
}

L
Linus Torvalds 已提交
1098 1099 1100
/*
 * tag stuff
 */
1101 1102 1103
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 *);
1104
extern int blk_queue_init_tags(struct request_queue *, int, struct blk_queue_tag *, int);
1105 1106 1107
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 *);
1108
extern struct blk_queue_tag *blk_init_tags(int, int);
1109
extern void blk_free_tags(struct blk_queue_tag *);
L
Linus Torvalds 已提交
1110

1111 1112 1113 1114 1115 1116 1117
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];
}
1118 1119 1120 1121

#define BLKDEV_DISCARD_SECURE  0x01    /* secure discard */

extern int blkdev_issue_flush(struct block_device *, gfp_t, sector_t *);
1122 1123
extern int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, unsigned long flags);
1124 1125
extern int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, struct page *page);
1126
extern int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
1127
		sector_t nr_sects, gfp_t gfp_mask, bool discard);
1128 1129
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
1130
{
1131 1132 1133
	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
1134
}
1135
static inline int sb_issue_zeroout(struct super_block *sb, sector_t block,
1136
		sector_t nr_blocks, gfp_t gfp_mask)
1137 1138 1139 1140
{
	return blkdev_issue_zeroout(sb->s_bdev,
				    block << (sb->s_blocksize_bits - 9),
				    nr_blocks << (sb->s_blocksize_bits - 9),
1141
				    gfp_mask, true);
1142
}
L
Linus Torvalds 已提交
1143

1144
extern int blk_verify_command(unsigned char *cmd, fmode_t has_write_perm);
1145

1146 1147 1148
enum blk_default_limits {
	BLK_MAX_SEGMENTS	= 128,
	BLK_SAFE_MAX_SECTORS	= 255,
1149
	BLK_DEF_MAX_SECTORS	= 2560,
1150 1151 1152
	BLK_MAX_SEGMENT_SIZE	= 65536,
	BLK_SEG_BOUNDARY_MASK	= 0xFFFFFFFFUL,
};
1153

L
Linus Torvalds 已提交
1154 1155
#define blkdev_entry_to_request(entry) list_entry((entry), struct request, queuelist)

1156 1157
static inline unsigned long queue_bounce_pfn(struct request_queue *q)
{
1158
	return q->limits.bounce_pfn;
1159 1160 1161 1162
}

static inline unsigned long queue_segment_boundary(struct request_queue *q)
{
1163
	return q->limits.seg_boundary_mask;
1164 1165
}

1166 1167 1168 1169 1170
static inline unsigned long queue_virt_boundary(struct request_queue *q)
{
	return q->limits.virt_boundary_mask;
}

1171 1172
static inline unsigned int queue_max_sectors(struct request_queue *q)
{
1173
	return q->limits.max_sectors;
1174 1175 1176 1177
}

static inline unsigned int queue_max_hw_sectors(struct request_queue *q)
{
1178
	return q->limits.max_hw_sectors;
1179 1180
}

1181
static inline unsigned short queue_max_segments(struct request_queue *q)
1182
{
1183
	return q->limits.max_segments;
1184 1185 1186 1187
}

static inline unsigned int queue_max_segment_size(struct request_queue *q)
{
1188
	return q->limits.max_segment_size;
1189 1190
}

1191
static inline unsigned short queue_logical_block_size(struct request_queue *q)
L
Linus Torvalds 已提交
1192 1193 1194
{
	int retval = 512;

1195 1196
	if (q && q->limits.logical_block_size)
		retval = q->limits.logical_block_size;
L
Linus Torvalds 已提交
1197 1198 1199 1200

	return retval;
}

1201
static inline unsigned short bdev_logical_block_size(struct block_device *bdev)
L
Linus Torvalds 已提交
1202
{
1203
	return queue_logical_block_size(bdev_get_queue(bdev));
L
Linus Torvalds 已提交
1204 1205
}

1206 1207 1208 1209 1210
static inline unsigned int queue_physical_block_size(struct request_queue *q)
{
	return q->limits.physical_block_size;
}

1211
static inline unsigned int bdev_physical_block_size(struct block_device *bdev)
M
Martin K. Petersen 已提交
1212 1213 1214 1215
{
	return queue_physical_block_size(bdev_get_queue(bdev));
}

1216 1217 1218 1219 1220
static inline unsigned int queue_io_min(struct request_queue *q)
{
	return q->limits.io_min;
}

M
Martin K. Petersen 已提交
1221 1222 1223 1224 1225
static inline int bdev_io_min(struct block_device *bdev)
{
	return queue_io_min(bdev_get_queue(bdev));
}

1226 1227 1228 1229 1230
static inline unsigned int queue_io_opt(struct request_queue *q)
{
	return q->limits.io_opt;
}

M
Martin K. Petersen 已提交
1231 1232 1233 1234 1235
static inline int bdev_io_opt(struct block_device *bdev)
{
	return queue_io_opt(bdev_get_queue(bdev));
}

1236 1237
static inline int queue_alignment_offset(struct request_queue *q)
{
M
Martin K. Petersen 已提交
1238
	if (q->limits.misaligned)
1239 1240
		return -1;

M
Martin K. Petersen 已提交
1241
	return q->limits.alignment_offset;
1242 1243
}

1244
static inline int queue_limit_alignment_offset(struct queue_limits *lim, sector_t sector)
1245 1246
{
	unsigned int granularity = max(lim->physical_block_size, lim->io_min);
1247
	unsigned int alignment = sector_div(sector, granularity >> 9) << 9;
1248

1249
	return (granularity + lim->alignment_offset - alignment) % granularity;
1250 1251
}

M
Martin K. Petersen 已提交
1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264
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;
}

1265 1266 1267 1268 1269 1270 1271 1272
static inline int queue_discard_alignment(struct request_queue *q)
{
	if (q->limits.discard_misaligned)
		return -1;

	return q->limits.discard_alignment;
}

1273
static inline int queue_limit_discard_alignment(struct queue_limits *lim, sector_t sector)
1274
{
1275
	unsigned int alignment, granularity, offset;
1276

1277 1278 1279
	if (!lim->max_discard_sectors)
		return 0;

1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293
	/* 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;
1294 1295
}

1296 1297 1298 1299 1300 1301 1302 1303 1304 1305
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;
}

1306 1307
static inline unsigned int queue_discard_zeroes_data(struct request_queue *q)
{
1308
	if (q->limits.max_discard_sectors && q->limits.discard_zeroes_data == 1)
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318
		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));
}

1319 1320 1321 1322 1323 1324 1325 1326 1327 1328
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;
}

1329
static inline int queue_dma_alignment(struct request_queue *q)
L
Linus Torvalds 已提交
1330
{
1331
	return q ? q->dma_alignment : 511;
L
Linus Torvalds 已提交
1332 1333
}

1334
static inline int blk_rq_aligned(struct request_queue *q, unsigned long addr,
1335 1336 1337
				 unsigned int len)
{
	unsigned int alignment = queue_dma_alignment(q) | q->dma_pad_mask;
1338
	return !(addr & alignment) && !(len & alignment);
1339 1340
}

L
Linus Torvalds 已提交
1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
/* 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;
}

1352
static inline unsigned int block_size(struct block_device *bdev)
L
Linus Torvalds 已提交
1353 1354 1355 1356
{
	return bdev->bd_block_size;
}

1357 1358 1359 1360 1361
static inline bool queue_flush_queueable(struct request_queue *q)
{
	return !q->flush_not_queueable;
}

L
Linus Torvalds 已提交
1362 1363 1364 1365 1366 1367 1368 1369 1370
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);
}

1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383
/*
 * 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));
}

1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403
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
1404
struct work_struct;
1405 1406
int kblockd_schedule_work(struct work_struct *work);
int kblockd_schedule_delayed_work(struct delayed_work *dwork, unsigned long delay);
1407
int kblockd_schedule_delayed_work_on(int cpu, struct delayed_work *dwork, unsigned long delay);
L
Linus Torvalds 已提交
1408

1409
#ifdef CONFIG_BLK_CGROUP
1410 1411 1412 1413 1414
/*
 * 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().
 */
1415 1416
static inline void set_start_time_ns(struct request *req)
{
1417
	preempt_disable();
1418
	req->start_time_ns = sched_clock();
1419
	preempt_enable();
1420 1421 1422 1423
}

static inline void set_io_start_time_ns(struct request *req)
{
1424
	preempt_disable();
1425
	req->io_start_time_ns = sched_clock();
1426
	preempt_enable();
1427
}
1428 1429 1430 1431 1432 1433 1434 1435 1436 1437

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;
}
1438 1439 1440
#else
static inline void set_start_time_ns(struct request *req) {}
static inline void set_io_start_time_ns(struct request *req) {}
1441 1442 1443 1444 1445 1446 1447 1448
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;
}
1449 1450
#endif

L
Linus Torvalds 已提交
1451 1452 1453 1454 1455
#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) "-*")

1456 1457
#if defined(CONFIG_BLK_DEV_INTEGRITY)

1458 1459 1460
enum blk_integrity_flags {
	BLK_INTEGRITY_VERIFY		= 1 << 0,
	BLK_INTEGRITY_GENERATE		= 1 << 1,
1461
	BLK_INTEGRITY_DEVICE_CAPABLE	= 1 << 2,
1462
	BLK_INTEGRITY_IP_CHECKSUM	= 1 << 3,
1463
};
1464

1465
struct blk_integrity_iter {
1466 1467
	void			*prot_buf;
	void			*data_buf;
1468
	sector_t		seed;
1469
	unsigned int		data_size;
1470
	unsigned short		interval;
1471 1472 1473
	const char		*disk_name;
};

1474
typedef int (integrity_processing_fn) (struct blk_integrity_iter *);
1475

1476 1477 1478 1479 1480
struct blk_integrity_profile {
	integrity_processing_fn		*generate_fn;
	integrity_processing_fn		*verify_fn;
	const char			*name;
};
1481

1482
extern void blk_integrity_register(struct gendisk *, struct blk_integrity *);
1483
extern void blk_integrity_unregister(struct gendisk *);
1484
extern int blk_integrity_compare(struct gendisk *, struct gendisk *);
1485 1486 1487
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 *);
1488 1489 1490 1491
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 *);
1492

1493
static inline struct blk_integrity *blk_get_integrity(struct gendisk *disk)
1494
{
1495
	struct blk_integrity *bi = &disk->queue->integrity;
1496 1497 1498 1499 1500

	if (!bi->profile)
		return NULL;

	return bi;
1501 1502
}

1503 1504
static inline
struct blk_integrity *bdev_get_integrity(struct block_device *bdev)
1505
{
1506
	return blk_get_integrity(bdev->bd_disk);
1507 1508
}

1509
static inline bool blk_integrity_rq(struct request *rq)
1510
{
1511
	return rq->cmd_flags & REQ_INTEGRITY;
1512 1513
}

1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525
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;
}

1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
static inline bool integrity_req_gap_back_merge(struct request *req,
						struct bio *next)
{
	struct bio_integrity_payload *bip = bio_integrity(req->bio);
	struct bio_integrity_payload *bip_next = bio_integrity(next);

	return bvec_gap_to_prev(req->q, &bip->bip_vec[bip->bip_vcnt - 1],
				bip_next->bip_vec[0].bv_offset);
}

static inline bool integrity_req_gap_front_merge(struct request *req,
						 struct bio *bio)
{
	struct bio_integrity_payload *bip = bio_integrity(bio);
	struct bio_integrity_payload *bip_next = bio_integrity(req->bio);

	return bvec_gap_to_prev(req->q, &bip->bip_vec[bip->bip_vcnt - 1],
				bip_next->bip_vec[0].bv_offset);
}

1546 1547
#else /* CONFIG_BLK_DEV_INTEGRITY */

1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569
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)
{
1570
	return NULL;
1571 1572 1573 1574 1575 1576 1577 1578 1579
}
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;
}
1580
static inline void blk_integrity_register(struct gendisk *d,
1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594
					 struct blk_integrity *b)
{
}
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;
}
1595 1596 1597
static inline bool blk_integrity_merge_rq(struct request_queue *rq,
					  struct request *r1,
					  struct request *r2)
1598
{
1599
	return true;
1600
}
1601 1602 1603
static inline bool blk_integrity_merge_bio(struct request_queue *rq,
					   struct request *r,
					   struct bio *b)
1604
{
1605
	return true;
1606
}
1607

1608 1609 1610 1611 1612 1613 1614 1615 1616 1617
static inline bool integrity_req_gap_back_merge(struct request *req,
						struct bio *next)
{
	return false;
}
static inline bool integrity_req_gap_front_merge(struct request *req,
						 struct bio *bio)
{
	return false;
}
1618 1619 1620

#endif /* CONFIG_BLK_DEV_INTEGRITY */

1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631
/**
 * struct blk_dax_ctl - control and output parameters for ->direct_access
 * @sector: (input) offset relative to a block_device
 * @addr: (output) kernel virtual address for @sector populated by driver
 * @pfn: (output) page frame number for @addr populated by driver
 * @size: (input) number of bytes requested
 */
struct blk_dax_ctl {
	sector_t sector;
	void __pmem *addr;
	long size;
D
Dan Williams 已提交
1632
	pfn_t pfn;
1633 1634
};

1635
struct block_device_operations {
A
Al Viro 已提交
1636
	int (*open) (struct block_device *, fmode_t);
1637
	void (*release) (struct gendisk *, fmode_t);
1638
	int (*rw_page)(struct block_device *, sector_t, struct page *, int rw);
A
Al Viro 已提交
1639 1640
	int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
	int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
1641
	long (*direct_access)(struct block_device *, sector_t, void __pmem **,
D
Dan Williams 已提交
1642
			pfn_t *);
1643 1644 1645
	unsigned int (*check_events) (struct gendisk *disk,
				      unsigned int clearing);
	/* ->media_changed() is DEPRECATED, use ->check_events() instead */
1646
	int (*media_changed) (struct gendisk *);
1647
	void (*unlock_native_capacity) (struct gendisk *);
1648 1649
	int (*revalidate_disk) (struct gendisk *);
	int (*getgeo)(struct block_device *, struct hd_geometry *);
1650 1651
	/* this callback is with swap_lock and sometimes page table lock held */
	void (*swap_slot_free_notify) (struct block_device *, unsigned long);
1652
	struct module *owner;
1653
	const struct pr_ops *pr_ops;
1654 1655
};

1656 1657
extern int __blkdev_driver_ioctl(struct block_device *, fmode_t, unsigned int,
				 unsigned long);
1658 1659 1660
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 *);
1661
extern long bdev_direct_access(struct block_device *, struct blk_dax_ctl *);
1662
#else /* CONFIG_BLOCK */
1663 1664 1665

struct block_device;

1666 1667 1668 1669 1670 1671 1672 1673 1674 1675
/*
 * stubs for when the block layer is configured out
 */
#define buffer_heads_over_limit 0

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

1676 1677 1678 1679
struct blk_plug {
};

static inline void blk_start_plug(struct blk_plug *plug)
1680 1681 1682
{
}

1683
static inline void blk_finish_plug(struct blk_plug *plug)
1684 1685 1686
{
}

1687
static inline void blk_flush_plug(struct task_struct *task)
1688 1689 1690
{
}

1691 1692 1693 1694 1695
static inline void blk_schedule_flush_plug(struct task_struct *task)
{
}


1696 1697 1698 1699 1700
static inline bool blk_needs_flush_plug(struct task_struct *tsk)
{
	return false;
}

1701 1702 1703 1704 1705 1706
static inline int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask,
				     sector_t *error_sector)
{
	return 0;
}

1707 1708
#endif /* CONFIG_BLOCK */

L
Linus Torvalds 已提交
1709
#endif