blkdev.h 46.4 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 15 16 17 18 19
#include <linux/timer.h>
#include <linux/workqueue.h>
#include <linux/pagemap.h>
#include <linux/backing-dev.h>
#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>
L
Linus Torvalds 已提交
25 26 27

#include <asm/scatterlist.h>

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;
L
Linus Torvalds 已提交
39 40 41 42

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

T
Tejun Heo 已提交
43 44 45 46 47 48
/*
 * 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
49
struct request;
50
typedef void (rq_end_io_fn)(struct request *, int);
L
Linus Torvalds 已提交
51

52 53 54
#define BLK_RL_SYNCFULL		(1U << 0)
#define BLK_RL_ASYNCFULL	(1U << 1)

L
Linus Torvalds 已提交
55
struct request_list {
56
	struct request_queue	*q;	/* the queue this rl belongs to */
57 58 59
#ifdef CONFIG_BLK_CGROUP
	struct blkcg_gq		*blkg;	/* blkg this request pool belongs to */
#endif
60 61 62 63
	/*
	 * count[], starved[], and wait[] are indexed by
	 * BLK_RW_SYNC/BLK_RW_ASYNC
	 */
64 65 66 67
	int			count[2];
	int			starved[2];
	mempool_t		*rq_pool;
	wait_queue_head_t	wait[2];
68
	unsigned int		flags;
L
Linus Torvalds 已提交
69 70
};

71 72 73 74 75 76
/*
 * request command types
 */
enum rq_cmd_type_bits {
	REQ_TYPE_FS		= 1,	/* fs request */
	REQ_TYPE_BLOCK_PC,		/* scsi command */
77
	REQ_TYPE_DRV_PRIV,		/* driver defined types from here */
78 79
};

L
Linus Torvalds 已提交
80 81 82
#define BLK_MAX_CDB	16

/*
83 84 85 86
 * 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
87 88
 */
struct request {
89
	struct list_head queuelist;
90 91
	union {
		struct call_single_data csd;
92
		unsigned long fifo_time;
93
	};
94

95
	struct request_queue *q;
96
	struct blk_mq_ctx *mq_ctx;
97

J
Jens Axboe 已提交
98
	u64 cmd_flags;
99
	unsigned cmd_type;
J
Jens Axboe 已提交
100
	unsigned long atomic_flags;
L
Linus Torvalds 已提交
101

102 103
	int cpu;

104 105
	/* the following two fields are internal, NEVER access directly */
	unsigned int __data_len;	/* total data len */
106
	sector_t __sector;		/* sector cursor */
L
Linus Torvalds 已提交
107 108 109 110

	struct bio *bio;
	struct bio *biotail;

111 112 113 114 115 116 117 118 119 120 121 122
	/*
	 * 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;
	};

123 124 125
	/*
	 * The rb_node is only used inside the io scheduler, requests
	 * are pruned when moved to the dispatch queue. So let the
126
	 * completion_data share space with the rb_node.
127 128 129
	 */
	union {
		struct rb_node rb_node;	/* sort/lookup */
130
		void *completion_data;
131
	};
132

133
	/*
134
	 * Three pointers are available for the IO schedulers, if they need
135 136
	 * more they have to dynamically allocate it.  Flush requests are
	 * never put on the IO scheduler. So let the flush fields share
137
	 * space with the elevator data.
138
	 */
139
	union {
140 141 142 143 144
		struct {
			struct io_cq		*icq;
			void			*priv[2];
		} elv;

145 146 147
		struct {
			unsigned int		seq;
			struct list_head	list;
148
			rq_end_io_fn		*saved_end_io;
149 150
		} flush;
	};
151

152
	struct gendisk *rq_disk;
153
	struct hd_struct *part;
L
Linus Torvalds 已提交
154
	unsigned long start_time;
155
#ifdef CONFIG_BLK_CGROUP
156
	struct request_list *rl;		/* rl this rq is alloced from */
157 158 159
	unsigned long long start_time_ns;
	unsigned long long io_start_time_ns;    /* when passed to hardware */
#endif
L
Linus Torvalds 已提交
160 161 162 163
	/* Number of scatter-gather DMA addr+len pairs after
	 * physical address coalescing is performed.
	 */
	unsigned short nr_phys_segments;
164 165 166
#if defined(CONFIG_BLK_DEV_INTEGRITY)
	unsigned short nr_integrity_segments;
#endif
L
Linus Torvalds 已提交
167

168 169
	unsigned short ioprio;

T
Tejun Heo 已提交
170
	void *special;		/* opaque pointer available for LLD use */
L
Linus Torvalds 已提交
171

172 173 174
	int tag;
	int errors;

L
Linus Torvalds 已提交
175 176 177
	/*
	 * when request is used as a packet command carrier
	 */
178 179
	unsigned char __cmd[BLK_MAX_CDB];
	unsigned char *cmd;
180
	unsigned short cmd_len;
L
Linus Torvalds 已提交
181

182
	unsigned int extra_len;	/* length of alignment and padding */
L
Linus Torvalds 已提交
183
	unsigned int sense_len;
T
Tejun Heo 已提交
184
	unsigned int resid_len;	/* residual count */
L
Linus Torvalds 已提交
185 186
	void *sense;

J
Jens Axboe 已提交
187 188
	unsigned long deadline;
	struct list_head timeout_list;
L
Linus Torvalds 已提交
189
	unsigned int timeout;
190
	int retries;
L
Linus Torvalds 已提交
191 192

	/*
193
	 * completion callback.
L
Linus Torvalds 已提交
194 195 196
	 */
	rq_end_io_fn *end_io;
	void *end_io_data;
197 198 199

	/* for bidi */
	struct request *next_rq;
L
Linus Torvalds 已提交
200 201
};

202 203 204 205 206
static inline unsigned short req_get_ioprio(struct request *req)
{
	return req->ioprio;
}

L
Linus Torvalds 已提交
207 208
#include <linux/elevator.h>

209 210
struct blk_queue_ctx;

211
typedef void (request_fn_proc) (struct request_queue *q);
212
typedef void (make_request_fn) (struct request_queue *q, struct bio *bio);
213
typedef int (prep_rq_fn) (struct request_queue *, struct request *);
214
typedef void (unprep_rq_fn) (struct request_queue *, struct request *);
L
Linus Torvalds 已提交
215 216

struct bio_vec;
217 218 219 220 221 222 223 224
struct bvec_merge_data {
	struct block_device *bi_bdev;
	sector_t bi_sector;
	unsigned bi_size;
	unsigned long bi_rw;
};
typedef int (merge_bvec_fn) (struct request_queue *, struct bvec_merge_data *,
			     struct bio_vec *);
225
typedef void (softirq_done_fn)(struct request *);
226
typedef int (dma_drain_needed_fn)(struct request *);
227
typedef int (lld_busy_fn) (struct request_queue *q);
M
Mike Christie 已提交
228
typedef int (bsg_job_fn) (struct bsg_job *);
L
Linus Torvalds 已提交
229

J
Jens Axboe 已提交
230 231 232 233 234 235 236 237
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
238 239 240 241 242 243 244 245 246 247
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 */
248
	int real_max_depth;		/* what the array can hold */
L
Linus Torvalds 已提交
249
	atomic_t refcnt;		/* map can be shared */
250 251
	int alloc_policy;		/* tag allocation policy */
	int next_tag;			/* next tag */
L
Linus Torvalds 已提交
252
};
253 254
#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
255

256 257 258
#define BLK_SCSI_MAX_CMDS	(256)
#define BLK_SCSI_CMD_PER_LONG	(BLK_SCSI_MAX_CMDS / (sizeof(long) * 8))

259 260 261 262 263
struct queue_limits {
	unsigned long		bounce_pfn;
	unsigned long		seg_boundary_mask;

	unsigned int		max_hw_sectors;
264
	unsigned int		chunk_sectors;
265 266
	unsigned int		max_sectors;
	unsigned int		max_segment_size;
267 268 269 270
	unsigned int		physical_block_size;
	unsigned int		alignment_offset;
	unsigned int		io_min;
	unsigned int		io_opt;
271
	unsigned int		max_discard_sectors;
272
	unsigned int		max_write_same_sectors;
273 274
	unsigned int		discard_granularity;
	unsigned int		discard_alignment;
275 276

	unsigned short		logical_block_size;
277
	unsigned short		max_segments;
278
	unsigned short		max_integrity_segments;
279

280
	unsigned char		misaligned;
281
	unsigned char		discard_misaligned;
282
	unsigned char		cluster;
283
	unsigned char		discard_zeroes_data;
284
	unsigned char		raid_partial_stripes_expensive;
285 286
};

287
struct request_queue {
L
Linus Torvalds 已提交
288 289 290 291 292
	/*
	 * Together with queue_head for cacheline sharing
	 */
	struct list_head	queue_head;
	struct request		*last_merge;
J
Jens Axboe 已提交
293
	struct elevator_queue	*elevator;
294 295
	int			nr_rqs[2];	/* # allocated [a]sync rqs */
	int			nr_rqs_elvpriv;	/* # allocated rqs w/ elvpriv */
L
Linus Torvalds 已提交
296 297

	/*
298 299 300 301
	 * 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
302
	 */
303
	struct request_list	root_rl;
L
Linus Torvalds 已提交
304 305 306 307

	request_fn_proc		*request_fn;
	make_request_fn		*make_request_fn;
	prep_rq_fn		*prep_rq_fn;
308
	unprep_rq_fn		*unprep_rq_fn;
L
Linus Torvalds 已提交
309
	merge_bvec_fn		*merge_bvec_fn;
310
	softirq_done_fn		*softirq_done_fn;
J
Jens Axboe 已提交
311
	rq_timed_out_fn		*rq_timed_out_fn;
312
	dma_drain_needed_fn	*dma_drain_needed;
313
	lld_busy_fn		*lld_busy_fn;
L
Linus Torvalds 已提交
314

315 316 317 318 319
	struct blk_mq_ops	*mq_ops;

	unsigned int		*mq_map;

	/* sw queues */
320
	struct blk_mq_ctx __percpu	*queue_ctx;
321 322 323 324 325 326
	unsigned int		nr_queues;

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

327 328 329
	/*
	 * Dispatch queue sorting
	 */
J
Jens Axboe 已提交
330
	sector_t		end_sector;
331 332
	struct request		*boundary_rq;

L
Linus Torvalds 已提交
333
	/*
334
	 * Delayed queue handling
L
Linus Torvalds 已提交
335
	 */
336
	struct delayed_work	delay_work;
L
Linus Torvalds 已提交
337 338 339 340 341 342 343 344 345 346

	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;

	/*
347
	 * various queue flags, see QUEUE_* below
L
Linus Torvalds 已提交
348
	 */
349
	unsigned long		queue_flags;
L
Linus Torvalds 已提交
350

351 352 353 354 355 356
	/*
	 * ida allocated id for this queue.  Used to index queues from
	 * ioctx.
	 */
	int			id;

L
Linus Torvalds 已提交
357
	/*
358
	 * queue needs bounce pages for pages above this limit
L
Linus Torvalds 已提交
359
	 */
360
	gfp_t			bounce_gfp;
L
Linus Torvalds 已提交
361 362

	/*
已提交
363 364 365
	 * 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
366
	 */
已提交
367
	spinlock_t		__queue_lock;
L
Linus Torvalds 已提交
368 369 370 371 372 373 374
	spinlock_t		*queue_lock;

	/*
	 * queue kobject
	 */
	struct kobject kobj;

375 376 377 378 379
	/*
	 * mq queue kobject
	 */
	struct kobject mq_kobj;

380
#ifdef CONFIG_PM
L
Lin Ming 已提交
381 382 383 384 385
	struct device		*dev;
	int			rpm_status;
	unsigned int		nr_pending;
#endif

L
Linus Torvalds 已提交
386 387 388 389 390 391 392 393
	/*
	 * 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
394
	unsigned int		dma_drain_size;
395
	void			*dma_drain_buffer;
396
	unsigned int		dma_pad_mask;
L
Linus Torvalds 已提交
397 398 399
	unsigned int		dma_alignment;

	struct blk_queue_tag	*queue_tags;
400
	struct list_head	tag_busy_list;
L
Linus Torvalds 已提交
401

402
	unsigned int		nr_sorted;
403
	unsigned int		in_flight[2];
404 405 406 407 408 409
	/*
	 * 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
410

J
Jens Axboe 已提交
411 412 413 414
	unsigned int		rq_timeout;
	struct timer_list	timeout;
	struct list_head	timeout_list;

415
	struct list_head	icq_list;
416
#ifdef CONFIG_BLK_CGROUP
417
	DECLARE_BITMAP		(blkcg_pols, BLKCG_MAX_POLS);
T
Tejun Heo 已提交
418
	struct blkcg_gq		*root_blkg;
419
	struct list_head	blkg_list;
420
#endif
421

422 423
	struct queue_limits	limits;

L
Linus Torvalds 已提交
424 425 426 427 428
	/*
	 * sg stuff
	 */
	unsigned int		sg_timeout;
	unsigned int		sg_reserved_size;
429
	int			node;
430
#ifdef CONFIG_BLK_DEV_IO_TRACE
431
	struct blk_trace	*blk_trace;
432
#endif
L
Linus Torvalds 已提交
433
	/*
434
	 * for flush operations
L
Linus Torvalds 已提交
435
	 */
436
	unsigned int		flush_flags;
437
	unsigned int		flush_not_queueable:1;
438
	struct blk_flush_queue	*fq;
439

440 441 442 443
	struct list_head	requeue_list;
	spinlock_t		requeue_lock;
	struct work_struct	requeue_work;

444
	struct mutex		sysfs_lock;
445

446
	int			bypass_depth;
447
	atomic_t		mq_freeze_depth;
448

449
#if defined(CONFIG_BLK_DEV_BSG)
M
Mike Christie 已提交
450 451
	bsg_job_fn		*bsg_job_fn;
	int			bsg_job_size;
452 453
	struct bsg_class_device bsg_dev;
#endif
454 455 456 457 458

#ifdef CONFIG_BLK_DEV_THROTTLING
	/* Throttle data */
	struct throtl_data *td;
#endif
T
Tejun Heo 已提交
459
	struct rcu_head		rcu_head;
460
	wait_queue_head_t	mq_freeze_wq;
461
	struct percpu_ref	mq_usage_counter;
462
	struct list_head	all_q_node;
463 464 465

	struct blk_mq_tag_set	*tag_set;
	struct list_head	tag_set_list;
L
Linus Torvalds 已提交
466 467 468 469
};

#define QUEUE_FLAG_QUEUED	1	/* uses generic tag queueing */
#define QUEUE_FLAG_STOPPED	2	/* queue is stopped */
470 471
#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
472
#define QUEUE_FLAG_DYING	5	/* queue being torn down */
473
#define QUEUE_FLAG_BYPASS	6	/* act as dumb FIFO queue */
474 475
#define QUEUE_FLAG_BIDI		7	/* queue supports bidi requests */
#define QUEUE_FLAG_NOMERGES     8	/* disable merge attempts */
D
Dan Williams 已提交
476
#define QUEUE_FLAG_SAME_COMP	9	/* complete on same CPU-group */
477 478 479
#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) */
480
#define QUEUE_FLAG_VIRT        QUEUE_FLAG_NONROT /* paravirt device */
481 482 483 484 485
#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
486
#define QUEUE_FLAG_SAME_FORCE  18	/* force complete on same CPU */
487
#define QUEUE_FLAG_DEAD        19	/* queue tear-down finished */
488
#define QUEUE_FLAG_INIT_DONE   20	/* queue is initialized */
489
#define QUEUE_FLAG_NO_SG_MERGE 21	/* don't attempt to merge SG segments*/
490
#define QUEUE_FLAG_SG_GAPS     22	/* queue doesn't support SG gaps */
491 492

#define QUEUE_FLAG_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
493
				 (1 << QUEUE_FLAG_STACKABLE)	|	\
494 495
				 (1 << QUEUE_FLAG_SAME_COMP)	|	\
				 (1 << QUEUE_FLAG_ADD_RANDOM))
496

497
#define QUEUE_FLAG_MQ_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
498
				 (1 << QUEUE_FLAG_STACKABLE)	|	\
499 500
				 (1 << QUEUE_FLAG_SAME_COMP))

501
static inline void queue_lockdep_assert_held(struct request_queue *q)
502
{
503 504
	if (q->queue_lock)
		lockdep_assert_held(q->queue_lock);
505 506
}

N
Nick Piggin 已提交
507 508 509 510 511 512
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
513 514 515
static inline int queue_flag_test_and_clear(unsigned int flag,
					    struct request_queue *q)
{
516
	queue_lockdep_assert_held(q);
J
Jens Axboe 已提交
517 518 519 520 521 522 523 524 525 526 527 528

	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)
{
529
	queue_lockdep_assert_held(q);
J
Jens Axboe 已提交
530 531 532 533 534 535 536 537 538

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

	return 1;
}

N
Nick Piggin 已提交
539 540
static inline void queue_flag_set(unsigned int flag, struct request_queue *q)
{
541
	queue_lockdep_assert_held(q);
N
Nick Piggin 已提交
542 543 544 545 546 547 548 549 550
	__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);
}

551 552 553 554 555
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
556 557
static inline void queue_flag_clear(unsigned int flag, struct request_queue *q)
{
558
	queue_lockdep_assert_held(q);
N
Nick Piggin 已提交
559 560 561
	__clear_bit(flag, &q->queue_flags);
}

L
Linus Torvalds 已提交
562 563
#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
564
#define blk_queue_dying(q)	test_bit(QUEUE_FLAG_DYING, &(q)->queue_flags)
565
#define blk_queue_dead(q)	test_bit(QUEUE_FLAG_DEAD, &(q)->queue_flags)
566
#define blk_queue_bypass(q)	test_bit(QUEUE_FLAG_BYPASS, &(q)->queue_flags)
567
#define blk_queue_init_done(q)	test_bit(QUEUE_FLAG_INIT_DONE, &(q)->queue_flags)
568
#define blk_queue_nomerges(q)	test_bit(QUEUE_FLAG_NOMERGES, &(q)->queue_flags)
569 570
#define blk_queue_noxmerges(q)	\
	test_bit(QUEUE_FLAG_NOXMERGES, &(q)->queue_flags)
571
#define blk_queue_nonrot(q)	test_bit(QUEUE_FLAG_NONROT, &(q)->queue_flags)
572
#define blk_queue_io_stat(q)	test_bit(QUEUE_FLAG_IO_STAT, &(q)->queue_flags)
573
#define blk_queue_add_random(q)	test_bit(QUEUE_FLAG_ADD_RANDOM, &(q)->queue_flags)
574 575
#define blk_queue_stackable(q)	\
	test_bit(QUEUE_FLAG_STACKABLE, &(q)->queue_flags)
576
#define blk_queue_discard(q)	test_bit(QUEUE_FLAG_DISCARD, &(q)->queue_flags)
A
Adrian Hunter 已提交
577 578
#define blk_queue_secdiscard(q)	(blk_queue_discard(q) && \
	test_bit(QUEUE_FLAG_SECDISCARD, &(q)->queue_flags))
L
Linus Torvalds 已提交
579

580 581 582 583 584 585
#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) && \
586
	 ((rq)->cmd_type == REQ_TYPE_FS))
587

588
#define blk_rq_cpu_valid(rq)	((rq)->cpu != -1)
589
#define blk_bidi_rq(rq)		((rq)->next_rq != NULL)
590 591
/* rq->queuelist of dequeued request must be list_empty() */
#define blk_queued_rq(rq)	(!list_empty(&(rq)->queuelist))
L
Linus Torvalds 已提交
592 593 594

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

J
Jens Axboe 已提交
595
#define rq_data_dir(rq)		(((rq)->cmd_flags & 1) != 0)
L
Linus Torvalds 已提交
596

597 598 599 600 601 602 603 604 605
/*
 * 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;
}

606 607 608 609 610
static inline unsigned int blk_queue_cluster(struct request_queue *q)
{
	return q->limits.cluster;
}

611
/*
612
 * We regard a request as sync, if either a read or a sync write
613
 */
614 615
static inline bool rw_is_sync(unsigned int rw_flags)
{
616
	return !(rw_flags & REQ_WRITE) || (rw_flags & REQ_SYNC);
617 618 619 620 621 622 623
}

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

624
static inline bool blk_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
625
{
626 627 628
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	return rl->flags & flag;
L
Linus Torvalds 已提交
629 630
}

631
static inline void blk_set_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
632
{
633 634 635
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	rl->flags |= flag;
L
Linus Torvalds 已提交
636 637
}

638
static inline void blk_clear_rl_full(struct request_list *rl, bool sync)
L
Linus Torvalds 已提交
639
{
640 641 642
	unsigned int flag = sync ? BLK_RL_SYNCFULL : BLK_RL_ASYNCFULL;

	rl->flags &= ~flag;
L
Linus Torvalds 已提交
643 644
}

645 646 647 648
static inline bool rq_mergeable(struct request *rq)
{
	if (rq->cmd_type != REQ_TYPE_FS)
		return false;
L
Linus Torvalds 已提交
649

650 651 652 653 654
	if (rq->cmd_flags & REQ_NOMERGE_FLAGS)
		return false;

	return true;
}
L
Linus Torvalds 已提交
655

656 657 658 659 660 661 662 663 664
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;

665 666 667
	if ((flags1 & REQ_WRITE_SAME) != (flags2 & REQ_WRITE_SAME))
		return false;

668 669 670
	return true;
}

671 672 673 674 675 676 677 678
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
679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694
/*
 * 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
 */
695 696

#if BITS_PER_LONG == 32
L
Linus Torvalds 已提交
697
#define BLK_BOUNCE_HIGH		((u64)blk_max_low_pfn << PAGE_SHIFT)
698 699 700 701
#else
#define BLK_BOUNCE_HIGH		-1ULL
#endif
#define BLK_BOUNCE_ANY		(-1ULL)
702
#define BLK_BOUNCE_ISA		(DMA_BIT_MASK(24))
L
Linus Torvalds 已提交
703

704 705 706 707
/*
 * default timeout for SG_IO if none specified
 */
#define BLK_DEFAULT_SG_TIMEOUT	(60 * HZ)
L
Linus Torvalds 已提交
708
#define BLK_MIN_SG_TIMEOUT	(7 * HZ)
709

710
#ifdef CONFIG_BOUNCE
L
Linus Torvalds 已提交
711
extern int init_emergency_isa_pool(void);
712
extern void blk_queue_bounce(struct request_queue *q, struct bio **bio);
L
Linus Torvalds 已提交
713 714 715 716 717
#else
static inline int init_emergency_isa_pool(void)
{
	return 0;
}
718
static inline void blk_queue_bounce(struct request_queue *q, struct bio **bio)
L
Linus Torvalds 已提交
719 720 721 722
{
}
#endif /* CONFIG_MMU */

723 724 725 726
struct rq_map_data {
	struct page **pages;
	int page_order;
	int nr_entries;
727
	unsigned long offset;
728
	int null_mapped;
729
	int from_user;
730 731
};

732
struct req_iterator {
733
	struct bvec_iter iter;
734 735 736 737
	struct bio *bio;
};

/* This should not be used directly - use rq_for_each_segment */
738 739
#define for_each_bio(_bio)		\
	for (; _bio; _bio = _bio->bi_next)
740
#define __rq_for_each_bio(_bio, rq)	\
L
Linus Torvalds 已提交
741 742 743
	if ((rq->bio))			\
		for (_bio = (rq)->bio; _bio; _bio = _bio->bi_next)

744 745
#define rq_for_each_segment(bvl, _rq, _iter)			\
	__rq_for_each_bio(_iter.bio, _rq)			\
746
		bio_for_each_segment(bvl, _iter.bio, _iter.iter)
747

K
Kent Overstreet 已提交
748
#define rq_iter_last(bvec, _iter)				\
749
		(_iter.bio->bi_next == NULL &&			\
K
Kent Overstreet 已提交
750
		 bio_iter_last(bvec, _iter.iter))
751

752 753 754 755 756 757 758 759 760 761 762
#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
763 764 765
extern int blk_register_queue(struct gendisk *disk);
extern void blk_unregister_queue(struct gendisk *disk);
extern void generic_make_request(struct bio *bio);
766
extern void blk_rq_init(struct request_queue *q, struct request *rq);
L
Linus Torvalds 已提交
767
extern void blk_put_request(struct request *);
768 769
extern void __blk_put_request(struct request_queue *, struct request *);
extern struct request *blk_get_request(struct request_queue *, int, gfp_t);
770 771
extern struct request *blk_make_request(struct request_queue *, struct bio *,
					gfp_t);
J
Jens Axboe 已提交
772
extern void blk_rq_set_block_pc(struct request *);
773
extern void blk_requeue_request(struct request_queue *, struct request *);
774 775
extern void blk_add_request_payload(struct request *rq, struct page *page,
		unsigned int len);
776
extern int blk_rq_check_limits(struct request_queue *q, struct request *rq);
777
extern int blk_lld_busy(struct request_queue *q);
778
extern void blk_rq_prep_clone(struct request *rq, struct request *rq_src);
779 780
extern int blk_insert_cloned_request(struct request_queue *q,
				     struct request *rq);
781
extern void blk_delay_queue(struct request_queue *, unsigned long);
782
extern void blk_recount_segments(struct request_queue *, struct bio *);
783
extern int scsi_verify_blk_ioctl(struct block_device *, unsigned int);
784 785
extern int scsi_cmd_blk_ioctl(struct block_device *, fmode_t,
			      unsigned int, void __user *);
786 787
extern int scsi_cmd_ioctl(struct request_queue *, struct gendisk *, fmode_t,
			  unsigned int, void __user *);
788 789
extern int sg_scsi_ioctl(struct request_queue *, struct gendisk *, fmode_t,
			 struct scsi_ioctl_command __user *);
790

791
extern void blk_queue_bio(struct request_queue *q, struct bio *bio);
C
Christoph Hellwig 已提交
792

793 794 795 796 797
/*
 * A queue has just exitted congestion.  Note this in the global counter of
 * congested queues, and wake up anyone who was waiting for requests to be
 * put back.
 */
798
static inline void blk_clear_queue_congested(struct request_queue *q, int sync)
799
{
800
	clear_bdi_congested(&q->backing_dev_info, sync);
801 802 803 804 805 806
}

/*
 * A queue has just entered congestion.  Flag that in the queue's VM-visible
 * state flags and increment the global gounter of congested queues.
 */
807
static inline void blk_set_queue_congested(struct request_queue *q, int sync)
808
{
809
	set_bdi_congested(&q->backing_dev_info, sync);
810 811
}

812 813
extern void blk_start_queue(struct request_queue *q);
extern void blk_stop_queue(struct request_queue *q);
L
Linus Torvalds 已提交
814
extern void blk_sync_queue(struct request_queue *q);
815
extern void __blk_stop_queue(struct request_queue *q);
816
extern void __blk_run_queue(struct request_queue *q);
817
extern void __blk_run_queue_uncond(struct request_queue *q);
818
extern void blk_run_queue(struct request_queue *);
819
extern void blk_run_queue_async(struct request_queue *q);
820
extern int blk_rq_map_user(struct request_queue *, struct request *,
821 822
			   struct rq_map_data *, void __user *, unsigned long,
			   gfp_t);
J
Jens Axboe 已提交
823
extern int blk_rq_unmap_user(struct bio *);
824 825
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 *,
826 827
			       struct rq_map_data *, const struct iov_iter *,
			       gfp_t);
828
extern int blk_execute_rq(struct request_queue *, struct gendisk *,
829
			  struct request *, int);
830
extern void blk_execute_rq_nowait(struct request_queue *, struct gendisk *,
831
				  struct request *, int, rq_end_io_fn *);
832

833
static inline struct request_queue *bdev_get_queue(struct block_device *bdev)
L
Linus Torvalds 已提交
834
{
835
	return bdev->bd_disk->queue;	/* this is never NULL */
L
Linus Torvalds 已提交
836 837
}

838
/*
839 840 841 842 843 844
 * 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
845
 */
846 847
static inline sector_t blk_rq_pos(const struct request *rq)
{
848
	return rq->__sector;
849 850 851 852
}

static inline unsigned int blk_rq_bytes(const struct request *rq)
{
853
	return rq->__data_len;
854 855
}

856 857 858 859
static inline int blk_rq_cur_bytes(const struct request *rq)
{
	return rq->bio ? bio_cur_bytes(rq->bio) : 0;
}
860

861 862
extern unsigned int blk_rq_err_bytes(const struct request *rq);

863 864
static inline unsigned int blk_rq_sectors(const struct request *rq)
{
865
	return blk_rq_bytes(rq) >> 9;
866 867 868 869
}

static inline unsigned int blk_rq_cur_sectors(const struct request *rq)
{
870
	return blk_rq_cur_bytes(rq) >> 9;
871 872
}

873 874 875 876
static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
						     unsigned int cmd_flags)
{
	if (unlikely(cmd_flags & REQ_DISCARD))
877
		return min(q->limits.max_discard_sectors, UINT_MAX >> 9);
878

879 880 881
	if (unlikely(cmd_flags & REQ_WRITE_SAME))
		return q->limits.max_write_same_sectors;

882 883 884
	return q->limits.max_sectors;
}

885 886 887 888 889 890 891 892
/*
 * 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)
893
		return q->limits.max_sectors;
894 895 896 897 898

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

899 900 901 902 903 904 905
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;

906
	if (!q->limits.chunk_sectors || (rq->cmd_flags & REQ_DISCARD))
907 908 909 910
		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));
911 912
}

913 914 915 916 917 918 919 920 921 922 923
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;
}

924 925 926 927 928 929 930
/*
 * 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
931
/*
932 933 934 935 936
 * Request completion related functions.
 *
 * blk_update_request() completes given number of bytes and updates
 * the request without completing it.
 *
937 938
 * blk_end_request() and friends.  __blk_end_request() must be called
 * with the request queue spinlock acquired.
L
Linus Torvalds 已提交
939 940
 *
 * Several drivers define their own end_request and call
941 942
 * blk_end_request() for parts of the original function.
 * This prevents code duplication in drivers.
L
Linus Torvalds 已提交
943
 */
944 945
extern bool blk_update_request(struct request *rq, int error,
			       unsigned int nr_bytes);
946
extern void blk_finish_request(struct request *rq, int error);
947 948 949 950
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);
951
extern bool blk_end_request_err(struct request *rq, int error);
952 953 954 955
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);
956
extern bool __blk_end_request_err(struct request *rq, int error);
957

958
extern void blk_complete_request(struct request *);
J
Jens Axboe 已提交
959 960
extern void __blk_complete_request(struct request *);
extern void blk_abort_request(struct request *);
961
extern void blk_unprep_request(struct request *);
962

L
Linus Torvalds 已提交
963 964 965
/*
 * Access functions for manipulating queue properties
 */
966
extern struct request_queue *blk_init_queue_node(request_fn_proc *rfn,
967
					spinlock_t *lock, int node_id);
968
extern struct request_queue *blk_init_queue(request_fn_proc *, spinlock_t *);
969 970
extern struct request_queue *blk_init_allocated_queue(struct request_queue *,
						      request_fn_proc *, spinlock_t *);
971 972 973
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);
974
extern void blk_limits_max_hw_sectors(struct queue_limits *, unsigned int);
975
extern void blk_queue_max_hw_sectors(struct request_queue *, unsigned int);
976
extern void blk_queue_chunk_sectors(struct request_queue *, unsigned int);
977
extern void blk_queue_max_segments(struct request_queue *, unsigned short);
978
extern void blk_queue_max_segment_size(struct request_queue *, unsigned int);
979 980
extern void blk_queue_max_discard_sectors(struct request_queue *q,
		unsigned int max_discard_sectors);
981 982
extern void blk_queue_max_write_same_sectors(struct request_queue *q,
		unsigned int max_write_same_sectors);
983
extern void blk_queue_logical_block_size(struct request_queue *, unsigned short);
984
extern void blk_queue_physical_block_size(struct request_queue *, unsigned int);
985 986
extern void blk_queue_alignment_offset(struct request_queue *q,
				       unsigned int alignment);
987
extern void blk_limits_io_min(struct queue_limits *limits, unsigned int min);
988
extern void blk_queue_io_min(struct request_queue *q, unsigned int min);
989
extern void blk_limits_io_opt(struct queue_limits *limits, unsigned int opt);
990
extern void blk_queue_io_opt(struct request_queue *q, unsigned int opt);
991
extern void blk_set_default_limits(struct queue_limits *lim);
992
extern void blk_set_stacking_limits(struct queue_limits *lim);
993 994
extern int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
			    sector_t offset);
995 996
extern int bdev_stack_limits(struct queue_limits *t, struct block_device *bdev,
			    sector_t offset);
997 998
extern void disk_stack_limits(struct gendisk *disk, struct block_device *bdev,
			      sector_t offset);
999
extern void blk_queue_stack_limits(struct request_queue *t, struct request_queue *b);
1000
extern void blk_queue_dma_pad(struct request_queue *, unsigned int);
1001
extern void blk_queue_update_dma_pad(struct request_queue *, unsigned int);
1002 1003 1004
extern int blk_queue_dma_drain(struct request_queue *q,
			       dma_drain_needed_fn *dma_drain_needed,
			       void *buf, unsigned int size);
1005
extern void blk_queue_lld_busy(struct request_queue *q, lld_busy_fn *fn);
1006 1007
extern void blk_queue_segment_boundary(struct request_queue *, unsigned long);
extern void blk_queue_prep_rq(struct request_queue *, prep_rq_fn *pfn);
1008
extern void blk_queue_unprep_rq(struct request_queue *, unprep_rq_fn *ufn);
1009 1010
extern void blk_queue_merge_bvec(struct request_queue *, merge_bvec_fn *);
extern void blk_queue_dma_alignment(struct request_queue *, int);
1011
extern void blk_queue_update_dma_alignment(struct request_queue *, int);
1012
extern void blk_queue_softirq_done(struct request_queue *, softirq_done_fn *);
J
Jens Axboe 已提交
1013 1014
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);
1015
extern void blk_queue_flush(struct request_queue *q, unsigned int flush);
1016
extern void blk_queue_flush_queueable(struct request_queue *q, bool queueable);
L
Linus Torvalds 已提交
1017 1018
extern struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev);

1019
extern int blk_rq_map_sg(struct request_queue *, struct request *, struct scatterlist *);
L
Linus Torvalds 已提交
1020 1021 1022
extern void blk_dump_rq_flags(struct request *, char *);
extern long nr_blockdev_pages(void);

T
Tejun Heo 已提交
1023
bool __must_check blk_get_queue(struct request_queue *);
1024 1025 1026
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 *);
L
Linus Torvalds 已提交
1027

L
Lin Ming 已提交
1028 1029 1030
/*
 * block layer runtime pm functions
 */
1031
#ifdef CONFIG_PM
L
Lin Ming 已提交
1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048
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

1049
/*
S
Suresh Jayaraman 已提交
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059
 * 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.
1060
 */
1061
struct blk_plug {
S
Suresh Jayaraman 已提交
1062
	struct list_head list; /* requests */
1063
	struct list_head mq_list; /* blk-mq requests */
S
Suresh Jayaraman 已提交
1064
	struct list_head cb_list; /* md requires an unplug callback */
1065
};
1066 1067
#define BLK_MAX_REQUEST_COUNT 16

1068
struct blk_plug_cb;
1069
typedef void (*blk_plug_cb_fn)(struct blk_plug_cb *, bool);
1070 1071
struct blk_plug_cb {
	struct list_head list;
1072 1073
	blk_plug_cb_fn callback;
	void *data;
1074
};
1075 1076
extern struct blk_plug_cb *blk_check_plugged(blk_plug_cb_fn unplug,
					     void *data, int size);
1077 1078
extern void blk_start_plug(struct blk_plug *);
extern void blk_finish_plug(struct blk_plug *);
1079
extern void blk_flush_plug_list(struct blk_plug *, bool);
1080 1081 1082 1083 1084

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

1085 1086 1087 1088 1089 1090 1091 1092
	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;

1093
	if (plug)
1094
		blk_flush_plug_list(plug, true);
1095 1096 1097 1098 1099 1100
}

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

1101 1102 1103 1104
	return plug &&
		(!list_empty(&plug->list) ||
		 !list_empty(&plug->mq_list) ||
		 !list_empty(&plug->cb_list));
1105 1106
}

L
Linus Torvalds 已提交
1107 1108 1109
/*
 * tag stuff
 */
1110 1111 1112
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 *);
1113
extern int blk_queue_init_tags(struct request_queue *, int, struct blk_queue_tag *, int);
1114 1115 1116
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 *);
1117
extern struct blk_queue_tag *blk_init_tags(int, int);
1118
extern void blk_free_tags(struct blk_queue_tag *);
L
Linus Torvalds 已提交
1119

1120 1121 1122 1123 1124 1125 1126
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];
}
1127 1128 1129 1130

#define BLKDEV_DISCARD_SECURE  0x01    /* secure discard */

extern int blkdev_issue_flush(struct block_device *, gfp_t, sector_t *);
1131 1132
extern int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, unsigned long flags);
1133 1134
extern int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, struct page *page);
1135
extern int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
1136
		sector_t nr_sects, gfp_t gfp_mask, bool discard);
1137 1138
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
1139
{
1140 1141 1142
	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
1143
}
1144
static inline int sb_issue_zeroout(struct super_block *sb, sector_t block,
1145
		sector_t nr_blocks, gfp_t gfp_mask)
1146 1147 1148 1149
{
	return blkdev_issue_zeroout(sb->s_bdev,
				    block << (sb->s_blocksize_bits - 9),
				    nr_blocks << (sb->s_blocksize_bits - 9),
1150
				    gfp_mask, true);
1151
}
L
Linus Torvalds 已提交
1152

1153
extern int blk_verify_command(unsigned char *cmd, fmode_t has_write_perm);
1154

1155 1156 1157 1158 1159 1160
enum blk_default_limits {
	BLK_MAX_SEGMENTS	= 128,
	BLK_SAFE_MAX_SECTORS	= 255,
	BLK_MAX_SEGMENT_SIZE	= 65536,
	BLK_SEG_BOUNDARY_MASK	= 0xFFFFFFFFUL,
};
1161

L
Linus Torvalds 已提交
1162 1163
#define blkdev_entry_to_request(entry) list_entry((entry), struct request, queuelist)

1164 1165
static inline unsigned long queue_bounce_pfn(struct request_queue *q)
{
1166
	return q->limits.bounce_pfn;
1167 1168 1169 1170
}

static inline unsigned long queue_segment_boundary(struct request_queue *q)
{
1171
	return q->limits.seg_boundary_mask;
1172 1173 1174 1175
}

static inline unsigned int queue_max_sectors(struct request_queue *q)
{
1176
	return q->limits.max_sectors;
1177 1178 1179 1180
}

static inline unsigned int queue_max_hw_sectors(struct request_queue *q)
{
1181
	return q->limits.max_hw_sectors;
1182 1183
}

1184
static inline unsigned short queue_max_segments(struct request_queue *q)
1185
{
1186
	return q->limits.max_segments;
1187 1188 1189 1190
}

static inline unsigned int queue_max_segment_size(struct request_queue *q)
{
1191
	return q->limits.max_segment_size;
1192 1193
}

1194
static inline unsigned short queue_logical_block_size(struct request_queue *q)
L
Linus Torvalds 已提交
1195 1196 1197
{
	int retval = 512;

1198 1199
	if (q && q->limits.logical_block_size)
		retval = q->limits.logical_block_size;
L
Linus Torvalds 已提交
1200 1201 1202 1203

	return retval;
}

1204
static inline unsigned short bdev_logical_block_size(struct block_device *bdev)
L
Linus Torvalds 已提交
1205
{
1206
	return queue_logical_block_size(bdev_get_queue(bdev));
L
Linus Torvalds 已提交
1207 1208
}

1209 1210 1211 1212 1213
static inline unsigned int queue_physical_block_size(struct request_queue *q)
{
	return q->limits.physical_block_size;
}

1214
static inline unsigned int bdev_physical_block_size(struct block_device *bdev)
M
Martin K. Petersen 已提交
1215 1216 1217 1218
{
	return queue_physical_block_size(bdev_get_queue(bdev));
}

1219 1220 1221 1222 1223
static inline unsigned int queue_io_min(struct request_queue *q)
{
	return q->limits.io_min;
}

M
Martin K. Petersen 已提交
1224 1225 1226 1227 1228
static inline int bdev_io_min(struct block_device *bdev)
{
	return queue_io_min(bdev_get_queue(bdev));
}

1229 1230 1231 1232 1233
static inline unsigned int queue_io_opt(struct request_queue *q)
{
	return q->limits.io_opt;
}

M
Martin K. Petersen 已提交
1234 1235 1236 1237 1238
static inline int bdev_io_opt(struct block_device *bdev)
{
	return queue_io_opt(bdev_get_queue(bdev));
}

1239 1240
static inline int queue_alignment_offset(struct request_queue *q)
{
M
Martin K. Petersen 已提交
1241
	if (q->limits.misaligned)
1242 1243
		return -1;

M
Martin K. Petersen 已提交
1244
	return q->limits.alignment_offset;
1245 1246
}

1247
static inline int queue_limit_alignment_offset(struct queue_limits *lim, sector_t sector)
1248 1249
{
	unsigned int granularity = max(lim->physical_block_size, lim->io_min);
1250
	unsigned int alignment = sector_div(sector, granularity >> 9) << 9;
1251

1252
	return (granularity + lim->alignment_offset - alignment) % granularity;
1253 1254
}

M
Martin K. Petersen 已提交
1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267
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;
}

1268 1269 1270 1271 1272 1273 1274 1275
static inline int queue_discard_alignment(struct request_queue *q)
{
	if (q->limits.discard_misaligned)
		return -1;

	return q->limits.discard_alignment;
}

1276
static inline int queue_limit_discard_alignment(struct queue_limits *lim, sector_t sector)
1277
{
1278
	unsigned int alignment, granularity, offset;
1279

1280 1281 1282
	if (!lim->max_discard_sectors)
		return 0;

1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
	/* 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;
1297 1298
}

1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
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;
}

1309 1310
static inline unsigned int queue_discard_zeroes_data(struct request_queue *q)
{
1311
	if (q->limits.max_discard_sectors && q->limits.discard_zeroes_data == 1)
1312 1313 1314 1315 1316 1317 1318 1319 1320 1321
		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));
}

1322 1323 1324 1325 1326 1327 1328 1329 1330 1331
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;
}

1332
static inline int queue_dma_alignment(struct request_queue *q)
L
Linus Torvalds 已提交
1333
{
1334
	return q ? q->dma_alignment : 511;
L
Linus Torvalds 已提交
1335 1336
}

1337
static inline int blk_rq_aligned(struct request_queue *q, unsigned long addr,
1338 1339 1340
				 unsigned int len)
{
	unsigned int alignment = queue_dma_alignment(q) | q->dma_pad_mask;
1341
	return !(addr & alignment) && !(len & alignment);
1342 1343
}

L
Linus Torvalds 已提交
1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354
/* 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;
}

1355
static inline unsigned int block_size(struct block_device *bdev)
L
Linus Torvalds 已提交
1356 1357 1358 1359
{
	return bdev->bd_block_size;
}

1360 1361 1362 1363 1364
static inline bool queue_flush_queueable(struct request_queue *q)
{
	return !q->flush_not_queueable;
}

L
Linus Torvalds 已提交
1365 1366 1367 1368 1369 1370 1371 1372 1373 1374
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);
}

struct work_struct;
1375 1376
int kblockd_schedule_work(struct work_struct *work);
int kblockd_schedule_delayed_work(struct delayed_work *dwork, unsigned long delay);
1377
int kblockd_schedule_delayed_work_on(int cpu, struct delayed_work *dwork, unsigned long delay);
L
Linus Torvalds 已提交
1378

1379
#ifdef CONFIG_BLK_CGROUP
1380 1381 1382 1383 1384
/*
 * 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().
 */
1385 1386
static inline void set_start_time_ns(struct request *req)
{
1387
	preempt_disable();
1388
	req->start_time_ns = sched_clock();
1389
	preempt_enable();
1390 1391 1392 1393
}

static inline void set_io_start_time_ns(struct request *req)
{
1394
	preempt_disable();
1395
	req->io_start_time_ns = sched_clock();
1396
	preempt_enable();
1397
}
1398 1399 1400 1401 1402 1403 1404 1405 1406 1407

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;
}
1408 1409 1410
#else
static inline void set_start_time_ns(struct request *req) {}
static inline void set_io_start_time_ns(struct request *req) {}
1411 1412 1413 1414 1415 1416 1417 1418
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;
}
1419 1420
#endif

L
Linus Torvalds 已提交
1421 1422 1423 1424 1425
#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) "-*")

1426 1427
#if defined(CONFIG_BLK_DEV_INTEGRITY)

1428 1429 1430
enum blk_integrity_flags {
	BLK_INTEGRITY_VERIFY		= 1 << 0,
	BLK_INTEGRITY_GENERATE		= 1 << 1,
1431
	BLK_INTEGRITY_DEVICE_CAPABLE	= 1 << 2,
1432
	BLK_INTEGRITY_IP_CHECKSUM	= 1 << 3,
1433
};
1434

1435
struct blk_integrity_iter {
1436 1437
	void			*prot_buf;
	void			*data_buf;
1438
	sector_t		seed;
1439
	unsigned int		data_size;
1440
	unsigned short		interval;
1441 1442 1443
	const char		*disk_name;
};

1444
typedef int (integrity_processing_fn) (struct blk_integrity_iter *);
1445 1446

struct blk_integrity {
1447 1448
	integrity_processing_fn	*generate_fn;
	integrity_processing_fn	*verify_fn;
1449 1450 1451

	unsigned short		flags;
	unsigned short		tuple_size;
1452
	unsigned short		interval;
1453 1454 1455 1456 1457 1458 1459
	unsigned short		tag_size;

	const char		*name;

	struct kobject		kobj;
};

1460
extern bool blk_integrity_is_initialized(struct gendisk *);
1461 1462
extern int blk_integrity_register(struct gendisk *, struct blk_integrity *);
extern void blk_integrity_unregister(struct gendisk *);
1463
extern int blk_integrity_compare(struct gendisk *, struct gendisk *);
1464 1465 1466
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 *);
1467 1468 1469 1470
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 *);
1471

1472 1473 1474 1475 1476 1477
static inline
struct blk_integrity *bdev_get_integrity(struct block_device *bdev)
{
	return bdev->bd_disk->integrity;
}

1478 1479 1480 1481 1482
static inline struct blk_integrity *blk_get_integrity(struct gendisk *disk)
{
	return disk->integrity;
}

1483
static inline bool blk_integrity_rq(struct request *rq)
1484
{
1485
	return rq->cmd_flags & REQ_INTEGRITY;
1486 1487
}

1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499
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;
}

1500 1501
#else /* CONFIG_BLK_DEV_INTEGRITY */

1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523
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)
{
1524
	return NULL;
1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549
}
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;
}
1550 1551 1552
static inline bool blk_integrity_merge_rq(struct request_queue *rq,
					  struct request *r1,
					  struct request *r2)
1553
{
1554
	return true;
1555
}
1556 1557 1558
static inline bool blk_integrity_merge_bio(struct request_queue *rq,
					   struct request *r,
					   struct bio *b)
1559
{
1560
	return true;
1561 1562 1563 1564 1565
}
static inline bool blk_integrity_is_initialized(struct gendisk *g)
{
	return 0;
}
1566 1567 1568

#endif /* CONFIG_BLK_DEV_INTEGRITY */

1569
struct block_device_operations {
A
Al Viro 已提交
1570
	int (*open) (struct block_device *, fmode_t);
1571
	void (*release) (struct gendisk *, fmode_t);
1572
	int (*rw_page)(struct block_device *, sector_t, struct page *, int rw);
A
Al Viro 已提交
1573 1574
	int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
	int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long);
1575 1576
	long (*direct_access)(struct block_device *, sector_t,
					void **, unsigned long *pfn, long size);
1577 1578 1579
	unsigned int (*check_events) (struct gendisk *disk,
				      unsigned int clearing);
	/* ->media_changed() is DEPRECATED, use ->check_events() instead */
1580
	int (*media_changed) (struct gendisk *);
1581
	void (*unlock_native_capacity) (struct gendisk *);
1582 1583
	int (*revalidate_disk) (struct gendisk *);
	int (*getgeo)(struct block_device *, struct hd_geometry *);
1584 1585
	/* this callback is with swap_lock and sometimes page table lock held */
	void (*swap_slot_free_notify) (struct block_device *, unsigned long);
1586 1587 1588
	struct module *owner;
};

1589 1590
extern int __blkdev_driver_ioctl(struct block_device *, fmode_t, unsigned int,
				 unsigned long);
1591 1592 1593
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 *);
1594 1595
extern long bdev_direct_access(struct block_device *, sector_t, void **addr,
						unsigned long *pfn, long size);
1596
#else /* CONFIG_BLOCK */
1597 1598 1599

struct block_device;

1600 1601 1602 1603 1604 1605 1606 1607 1608 1609
/*
 * stubs for when the block layer is configured out
 */
#define buffer_heads_over_limit 0

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

1610 1611 1612 1613
struct blk_plug {
};

static inline void blk_start_plug(struct blk_plug *plug)
1614 1615 1616
{
}

1617
static inline void blk_finish_plug(struct blk_plug *plug)
1618 1619 1620
{
}

1621
static inline void blk_flush_plug(struct task_struct *task)
1622 1623 1624
{
}

1625 1626 1627 1628 1629
static inline void blk_schedule_flush_plug(struct task_struct *task)
{
}


1630 1631 1632 1633 1634
static inline bool blk_needs_flush_plug(struct task_struct *tsk)
{
	return false;
}

1635 1636 1637 1638 1639 1640
static inline int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask,
				     sector_t *error_sector)
{
	return 0;
}

1641 1642
#endif /* CONFIG_BLOCK */

L
Linus Torvalds 已提交
1643
#endif