[linux-2.6-block.git] / block / blk.h

#ifndef BLK_INTERNAL_H
#define BLK_INTERNAL_H

#include <linux/idr.h>
#include <linux/blk-mq.h>
#include "blk-mq.h"

/* Amount of time in which a process may batch requests */
#define BLK_BATCH_TIME	(HZ/50UL)

/* Number of requests a "batching" process may submit */
#define BLK_BATCH_REQ	32

/* Max future timer expiry for timeouts */
#define BLK_MAX_TIMEOUT		(5 * HZ)

struct blk_flush_queue {
	unsigned int		flush_queue_delayed:1;
	unsigned int		flush_pending_idx:1;
	unsigned int		flush_running_idx:1;
	unsigned long		flush_pending_since;
	struct list_head	flush_queue[2];
	struct list_head	flush_data_in_flight;
	struct request		*flush_rq;

	/*
	 * flush_rq shares tag with this rq, both can't be active
	 * at the same time
	 */
	struct request		*orig_rq;
	spinlock_t		mq_flush_lock;
};

extern struct kmem_cache *blk_requestq_cachep;
extern struct kmem_cache *request_cachep;
extern struct kobj_type blk_queue_ktype;
extern struct ida blk_queue_ida;

static inline struct blk_flush_queue *blk_get_flush_queue(
		struct request_queue *q, struct blk_mq_ctx *ctx)
{
	struct blk_mq_hw_ctx *hctx;

	if (!q->mq_ops)
		return q->fq;

	hctx = q->mq_ops->map_queue(q, ctx->cpu);

	return hctx->fq;
}

static inline void __blk_get_queue(struct request_queue *q)
{
	kobject_get(&q->kobj);
}

struct blk_flush_queue *blk_alloc_flush_queue(struct request_queue *q,
		int node, int cmd_size);
void blk_free_flush_queue(struct blk_flush_queue *q);

int blk_init_rl(struct request_list *rl, struct request_queue *q,
		gfp_t gfp_mask);
void blk_exit_rl(struct request_list *rl);
void init_request_from_bio(struct request *req, struct bio *bio);
void blk_rq_bio_prep(struct request_queue *q, struct request *rq,
			struct bio *bio);
int blk_rq_append_bio(struct request_queue *q, struct request *rq,
		      struct bio *bio);
void blk_queue_bypass_start(struct request_queue *q);
void blk_queue_bypass_end(struct request_queue *q);
void blk_dequeue_request(struct request *rq);
void __blk_queue_free_tags(struct request_queue *q);
bool __blk_end_bidi_request(struct request *rq, int error,
			    unsigned int nr_bytes, unsigned int bidi_bytes);

void blk_rq_timed_out_timer(unsigned long data);
unsigned long blk_rq_timeout(unsigned long timeout);
void blk_add_timer(struct request *req);
void blk_delete_timer(struct request *);


bool bio_attempt_front_merge(struct request_queue *q, struct request *req,
			     struct bio *bio);
bool bio_attempt_back_merge(struct request_queue *q, struct request *req,
			    struct bio *bio);
bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
			    unsigned int *request_count,
			    struct request **same_queue_rq);
unsigned int blk_plug_queued_count(struct request_queue *q);

void blk_account_io_start(struct request *req, bool new_io);
void blk_account_io_completion(struct request *req, unsigned int bytes);
void blk_account_io_done(struct request *req);

/*
 * Internal atomic flags for request handling
 */
enum rq_atomic_flags {
	REQ_ATOM_COMPLETE = 0,
	REQ_ATOM_STARTED,
};

/*
 * EH timer and IO completion will both attempt to 'grab' the request, make
 * sure that only one of them succeeds
 */
static inline int blk_mark_rq_complete(struct request *rq)
{
	return test_and_set_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
}

static inline void blk_clear_rq_complete(struct request *rq)
{
	clear_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
}

/*
 * Internal elevator interface
 */
#define ELV_ON_HASH(rq) ((rq)->cmd_flags & REQ_HASHED)

void blk_insert_flush(struct request *rq);

static inline struct request *__elv_next_request(struct request_queue *q)
{
	struct request *rq;
	struct blk_flush_queue *fq = blk_get_flush_queue(q, NULL);

	while (1) {
		if (!list_empty(&q->queue_head)) {
			rq = list_entry_rq(q->queue_head.next);
			return rq;
		}

		/*
		 * Flush request is running and flush request isn't queueable
		 * in the drive, we can hold the queue till flush request is
		 * finished. Even we don't do this, driver can't dispatch next
		 * requests and will requeue them. And this can improve
		 * throughput too. For example, we have request flush1, write1,
		 * flush 2. flush1 is dispatched, then queue is hold, write1
		 * isn't inserted to queue. After flush1 is finished, flush2
		 * will be dispatched. Since disk cache is already clean,
		 * flush2 will be finished very soon, so looks like flush2 is
		 * folded to flush1.
		 * Since the queue is hold, a flag is set to indicate the queue
		 * should be restarted later. Please see flush_end_io() for
		 * details.
		 */
		if (fq->flush_pending_idx != fq->flush_running_idx &&
				!queue_flush_queueable(q)) {
			fq->flush_queue_delayed = 1;
			return NULL;
		}
		if (unlikely(blk_queue_bypass(q)) ||
		    !q->elevator->type->ops.elevator_dispatch_fn(q, 0))
			return NULL;
	}
}

static inline void elv_activate_rq(struct request_queue *q, struct request *rq)
{
	struct elevator_queue *e = q->elevator;

	if (e->type->ops.elevator_activate_req_fn)
		e->type->ops.elevator_activate_req_fn(q, rq);
}

static inline void elv_deactivate_rq(struct request_queue *q, struct request *rq)
{
	struct elevator_queue *e = q->elevator;

	if (e->type->ops.elevator_deactivate_req_fn)
		e->type->ops.elevator_deactivate_req_fn(q, rq);
}

#ifdef CONFIG_FAIL_IO_TIMEOUT
int blk_should_fake_timeout(struct request_queue *);
ssize_t part_timeout_show(struct device *, struct device_attribute *, char *);
ssize_t part_timeout_store(struct device *, struct device_attribute *,
				const char *, size_t);
#else
static inline int blk_should_fake_timeout(struct request_queue *q)
{
	return 0;
}
#endif

int ll_back_merge_fn(struct request_queue *q, struct request *req,
		     struct bio *bio);
int ll_front_merge_fn(struct request_queue *q, struct request *req, 
		      struct bio *bio);
int attempt_back_merge(struct request_queue *q, struct request *rq);
int attempt_front_merge(struct request_queue *q, struct request *rq);
int blk_attempt_req_merge(struct request_queue *q, struct request *rq,
				struct request *next);
void blk_recalc_rq_segments(struct request *rq);
void blk_rq_set_mixed_merge(struct request *rq);
bool blk_rq_merge_ok(struct request *rq, struct bio *bio);
int blk_try_merge(struct request *rq, struct bio *bio);

void blk_queue_congestion_threshold(struct request_queue *q);

int blk_dev_init(void);


/*
 * Return the threshold (number of used requests) at which the queue is
 * considered to be congested.  It include a little hysteresis to keep the
 * context switch rate down.
 */
static inline int queue_congestion_on_threshold(struct request_queue *q)
{
	return q->nr_congestion_on;
}

/*
 * The threshold at which a queue is considered to be uncongested
 */
static inline int queue_congestion_off_threshold(struct request_queue *q)
{
	return q->nr_congestion_off;
}

extern int blk_update_nr_requests(struct request_queue *, unsigned int);

/*
 * Contribute to IO statistics IFF:
 *
 *	a) it's attached to a gendisk, and
 *	b) the queue had IO stats enabled when this request was started, and
 *	c) it's a file system request
 */
static inline int blk_do_io_stat(struct request *rq)
{
	return rq->rq_disk &&
	       (rq->cmd_flags & REQ_IO_STAT) &&
		(rq->cmd_type == REQ_TYPE_FS);
}

/*
 * Internal io_context interface
 */
void get_io_context(struct io_context *ioc);
struct io_cq *ioc_lookup_icq(struct io_context *ioc, struct request_queue *q);
struct io_cq *ioc_create_icq(struct io_context *ioc, struct request_queue *q,
			     gfp_t gfp_mask);
void ioc_clear_queue(struct request_queue *q);

int create_task_io_context(struct task_struct *task, gfp_t gfp_mask, int node);

/**
 * create_io_context - try to create task->io_context
 * @gfp_mask: allocation mask
 * @node: allocation node
 *
 * If %current->io_context is %NULL, allocate a new io_context and install
 * it.  Returns the current %current->io_context which may be %NULL if
 * allocation failed.
 *
 * Note that this function can't be called with IRQ disabled because
 * task_lock which protects %current->io_context is IRQ-unsafe.
 */
static inline struct io_context *create_io_context(gfp_t gfp_mask, int node)
{
	WARN_ON_ONCE(irqs_disabled());
	if (unlikely(!current->io_context))
		create_task_io_context(current, gfp_mask, node);
	return current->io_context;
}

/*
 * Internal throttling interface
 */
#ifdef CONFIG_BLK_DEV_THROTTLING
extern void blk_throtl_drain(struct request_queue *q);
extern int blk_throtl_init(struct request_queue *q);
extern void blk_throtl_exit(struct request_queue *q);
#else /* CONFIG_BLK_DEV_THROTTLING */
static inline void blk_throtl_drain(struct request_queue *q) { }
static inline int blk_throtl_init(struct request_queue *q) { return 0; }
static inline void blk_throtl_exit(struct request_queue *q) { }
#endif /* CONFIG_BLK_DEV_THROTTLING */

#endif /* BLK_INTERNAL_H */
Commit	Line	Data
8324aa91 JA	1	#ifndef BLK_INTERNAL_H
	2	#define BLK_INTERNAL_H
	3
a73f730d	4	#include <linux/idr.h>
f70ced09 ML	5	#include <linux/blk-mq.h>
f70ced09 ML	6	#include "blk-mq.h"
a73f730d	7
86db1e29 JA	8	/* Amount of time in which a process may batch requests */
	9	#define BLK_BATCH_TIME (HZ/50UL)
	10
	11	/* Number of requests a "batching" process may submit */
	12	#define BLK_BATCH_REQ 32
	13
0d2602ca JA	14	/* Max future timer expiry for timeouts */
	15	#define BLK_MAX_TIMEOUT (5 * HZ)
	16
7c94e1c1 ML	17	struct blk_flush_queue {
	18	unsigned int flush_queue_delayed:1;
	19	unsigned int flush_pending_idx:1;
	20	unsigned int flush_running_idx:1;
	21	unsigned long flush_pending_since;
	22	struct list_head flush_queue[2];
	23	struct list_head flush_data_in_flight;
	24	struct request *flush_rq;
0048b483 ML	25
	26	/*
	27	* flush_rq shares tag with this rq, both can't be active
	28	* at the same time
	29	*/
	30	struct request *orig_rq;
7c94e1c1 ML	31	spinlock_t mq_flush_lock;
	32	};
	33
8324aa91	34	extern struct kmem_cache *blk_requestq_cachep;
320ae51f	35	extern struct kmem_cache *request_cachep;
8324aa91	36	extern struct kobj_type blk_queue_ktype;
a73f730d	37	extern struct ida blk_queue_ida;
8324aa91	38
7c94e1c1	39	static inline struct blk_flush_queue *blk_get_flush_queue(
e97c293c	40	struct request_queue q, struct blk_mq_ctx ctx)
7c94e1c1	41	{
f70ced09 ML	42	struct blk_mq_hw_ctx *hctx;
	43
	44	if (!q->mq_ops)
	45	return q->fq;
	46
	47	hctx = q->mq_ops->map_queue(q, ctx->cpu);
	48
	49	return hctx->fq;
7c94e1c1 ML	50	}
7c94e1c1 ML	51
09ac46c4 TH	52	static inline void __blk_get_queue(struct request_queue *q)
	53	{
	54	kobject_get(&q->kobj);
	55	}
	56
f70ced09 ML	57	struct blk_flush_queue blk_alloc_flush_queue(struct request_queue q,
	58	int node, int cmd_size);
	59	void blk_free_flush_queue(struct blk_flush_queue *q);
f3552655	60
5b788ce3 TH	61	int blk_init_rl(struct request_list rl, struct request_queue q,
	62	gfp_t gfp_mask);
	63	void blk_exit_rl(struct request_list *rl);
86db1e29 JA	64	void init_request_from_bio(struct request req, struct bio bio);
	65	void blk_rq_bio_prep(struct request_queue q, struct request rq,
	66	struct bio *bio);
a411f4bb BH	67	int blk_rq_append_bio(struct request_queue q, struct request rq,
a411f4bb BH	68	struct bio *bio);
d732580b TH	69	void blk_queue_bypass_start(struct request_queue *q);
d732580b TH	70	void blk_queue_bypass_end(struct request_queue *q);
9934c8c0	71	void blk_dequeue_request(struct request *rq);
8324aa91	72	void __blk_queue_free_tags(struct request_queue *q);
4853abaa JM	73	bool __blk_end_bidi_request(struct request *rq, int error,
4853abaa JM	74	unsigned int nr_bytes, unsigned int bidi_bytes);
8324aa91	75
242f9dcb	76	void blk_rq_timed_out_timer(unsigned long data);
0d2602ca	77	unsigned long blk_rq_timeout(unsigned long timeout);
87ee7b11	78	void blk_add_timer(struct request *req);
242f9dcb	79	void blk_delete_timer(struct request *);
242f9dcb	80
320ae51f JA	81
	82	bool bio_attempt_front_merge(struct request_queue q, struct request req,
	83	struct bio *bio);
	84	bool bio_attempt_back_merge(struct request_queue q, struct request req,
	85	struct bio *bio);
	86	bool blk_attempt_plug_merge(struct request_queue q, struct bio bio,
5b3f341f SL	87	unsigned int *request_count,
5b3f341f SL	88	struct request **same_queue_rq);
0809e3ac	89	unsigned int blk_plug_queued_count(struct request_queue *q);
320ae51f JA	90
	91	void blk_account_io_start(struct request *req, bool new_io);
	92	void blk_account_io_completion(struct request *req, unsigned int bytes);
	93	void blk_account_io_done(struct request *req);
	94
242f9dcb JA	95	/*
	96	* Internal atomic flags for request handling
	97	*/
	98	enum rq_atomic_flags {
	99	REQ_ATOM_COMPLETE = 0,
320ae51f	100	REQ_ATOM_STARTED,
242f9dcb JA	101	};
	102
	103	/*
	104	* EH timer and IO completion will both attempt to 'grab' the request, make
25985edc	105	* sure that only one of them succeeds
242f9dcb JA	106	*/
	107	static inline int blk_mark_rq_complete(struct request *rq)
	108	{
	109	return test_and_set_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
	110	}
	111
	112	static inline void blk_clear_rq_complete(struct request *rq)
	113	{
	114	clear_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
	115	}
86db1e29	116
158dbda0 TH	117	/*
	118	* Internal elevator interface
	119	*/
360f92c2	120	#define ELV_ON_HASH(rq) ((rq)->cmd_flags & REQ_HASHED)
158dbda0	121
ae1b1539	122	void blk_insert_flush(struct request *rq);
dd831006	123
158dbda0 TH	124	static inline struct request __elv_next_request(struct request_queue q)
	125	{
	126	struct request *rq;
e97c293c	127	struct blk_flush_queue *fq = blk_get_flush_queue(q, NULL);
158dbda0 TH	128
158dbda0 TH	129	while (1) {
ae1b1539	130	if (!list_empty(&q->queue_head)) {
158dbda0	131	rq = list_entry_rq(q->queue_head.next);
ae1b1539	132	return rq;
158dbda0 TH	133	}
158dbda0 TH	134
3ac0cc45	135	/*
	136	* Flush request is running and flush request isn't queueable
	137	* in the drive, we can hold the queue till flush request is
	138	* finished. Even we don't do this, driver can't dispatch next
	139	* requests and will requeue them. And this can improve
	140	* throughput too. For example, we have request flush1, write1,
	141	* flush 2. flush1 is dispatched, then queue is hold, write1
	142	* isn't inserted to queue. After flush1 is finished, flush2
	143	* will be dispatched. Since disk cache is already clean,
	144	* flush2 will be finished very soon, so looks like flush2 is
	145	* folded to flush1.
	146	* Since the queue is hold, a flag is set to indicate the queue
	147	* should be restarted later. Please see flush_end_io() for
	148	* details.
	149	*/
7c94e1c1	150	if (fq->flush_pending_idx != fq->flush_running_idx &&
3ac0cc45	151	!queue_flush_queueable(q)) {
7c94e1c1	152	fq->flush_queue_delayed = 1;
3ac0cc45	153	return NULL;
3ac0cc45	154	}
556ee818	155	if (unlikely(blk_queue_bypass(q)) \|\|
22f746e2	156	!q->elevator->type->ops.elevator_dispatch_fn(q, 0))
158dbda0 TH	157	return NULL;
	158	}
	159	}
	160
	161	static inline void elv_activate_rq(struct request_queue q, struct request rq)
	162	{
	163	struct elevator_queue *e = q->elevator;
	164
22f746e2 TH	165	if (e->type->ops.elevator_activate_req_fn)
22f746e2 TH	166	e->type->ops.elevator_activate_req_fn(q, rq);
158dbda0 TH	167	}
	168
	169	static inline void elv_deactivate_rq(struct request_queue q, struct request rq)
	170	{
	171	struct elevator_queue *e = q->elevator;
	172
22f746e2 TH	173	if (e->type->ops.elevator_deactivate_req_fn)
22f746e2 TH	174	e->type->ops.elevator_deactivate_req_fn(q, rq);
158dbda0 TH	175	}
158dbda0 TH	176
581d4e28 JA	177	#ifdef CONFIG_FAIL_IO_TIMEOUT
	178	int blk_should_fake_timeout(struct request_queue *);
	179	ssize_t part_timeout_show(struct device , struct device_attribute , char *);
	180	ssize_t part_timeout_store(struct device , struct device_attribute ,
	181	const char *, size_t);
	182	#else
	183	static inline int blk_should_fake_timeout(struct request_queue *q)
	184	{
	185	return 0;
	186	}
	187	#endif
	188
d6d48196 JA	189	int ll_back_merge_fn(struct request_queue q, struct request req,
	190	struct bio *bio);
	191	int ll_front_merge_fn(struct request_queue q, struct request req,
	192	struct bio *bio);
	193	int attempt_back_merge(struct request_queue q, struct request rq);
	194	int attempt_front_merge(struct request_queue q, struct request rq);
5e84ea3a JA	195	int blk_attempt_req_merge(struct request_queue q, struct request rq,
5e84ea3a JA	196	struct request *next);
d6d48196	197	void blk_recalc_rq_segments(struct request *rq);
80a761fd	198	void blk_rq_set_mixed_merge(struct request *rq);
050c8ea8 TH	199	bool blk_rq_merge_ok(struct request rq, struct bio bio);
050c8ea8 TH	200	int blk_try_merge(struct request rq, struct bio bio);
d6d48196	201
8324aa91 JA	202	void blk_queue_congestion_threshold(struct request_queue *q);
8324aa91 JA	203
ff88972c AB	204	int blk_dev_init(void);
ff88972c AB	205
f253b86b	206
8324aa91 JA	207	/*
	208	* Return the threshold (number of used requests) at which the queue is
	209	* considered to be congested. It include a little hysteresis to keep the
	210	* context switch rate down.
	211	*/
	212	static inline int queue_congestion_on_threshold(struct request_queue *q)
	213	{
	214	return q->nr_congestion_on;
	215	}
	216
	217	/*
	218	* The threshold at which a queue is considered to be uncongested
	219	*/
	220	static inline int queue_congestion_off_threshold(struct request_queue *q)
	221	{
	222	return q->nr_congestion_off;
	223	}
	224
e3a2b3f9 JA	225	extern int blk_update_nr_requests(struct request_queue *, unsigned int);
e3a2b3f9 JA	226
c2553b58 JA	227	/*
	228	* Contribute to IO statistics IFF:
	229	*
	230	* a) it's attached to a gendisk, and
	231	* b) the queue had IO stats enabled when this request was started, and
e2a60da7	232	* c) it's a file system request
c2553b58	233	*/
26308eab	234	static inline int blk_do_io_stat(struct request *rq)
fb8ec18c	235	{
33659ebb CH	236	return rq->rq_disk &&
33659ebb CH	237	(rq->cmd_flags & REQ_IO_STAT) &&
e2a60da7	238	(rq->cmd_type == REQ_TYPE_FS);
fb8ec18c JA	239	}
fb8ec18c JA	240
f2dbd76a TH	241	/*
	242	* Internal io_context interface
	243	*/
	244	void get_io_context(struct io_context *ioc);
47fdd4ca	245	struct io_cq ioc_lookup_icq(struct io_context ioc, struct request_queue *q);
24acfc34 TH	246	struct io_cq ioc_create_icq(struct io_context ioc, struct request_queue *q,
24acfc34 TH	247	gfp_t gfp_mask);
7e5a8794	248	void ioc_clear_queue(struct request_queue *q);
f2dbd76a	249
24acfc34	250	int create_task_io_context(struct task_struct *task, gfp_t gfp_mask, int node);
f2dbd76a TH	251
	252	/**
	253	* create_io_context - try to create task->io_context
f2dbd76a TH	254	* @gfp_mask: allocation mask
	255	* @node: allocation node
	256	*
24acfc34 TH	257	* If %current->io_context is %NULL, allocate a new io_context and install
	258	* it. Returns the current %current->io_context which may be %NULL if
	259	* allocation failed.
f2dbd76a TH	260	*
f2dbd76a TH	261	* Note that this function can't be called with IRQ disabled because
24acfc34	262	* task_lock which protects %current->io_context is IRQ-unsafe.
f2dbd76a	263	*/
24acfc34	264	static inline struct io_context *create_io_context(gfp_t gfp_mask, int node)
f2dbd76a TH	265	{
f2dbd76a TH	266	WARN_ON_ONCE(irqs_disabled());
24acfc34 TH	267	if (unlikely(!current->io_context))
	268	create_task_io_context(current, gfp_mask, node);
	269	return current->io_context;
f2dbd76a TH	270	}
	271
	272	/*
	273	* Internal throttling interface
	274	*/
bc9fcbf9	275	#ifdef CONFIG_BLK_DEV_THROTTLING
c9a929dd	276	extern void blk_throtl_drain(struct request_queue *q);
bc9fcbf9 TH	277	extern int blk_throtl_init(struct request_queue *q);
	278	extern void blk_throtl_exit(struct request_queue *q);
	279	#else /* CONFIG_BLK_DEV_THROTTLING */
c9a929dd	280	static inline void blk_throtl_drain(struct request_queue *q) { }
bc9fcbf9 TH	281	static inline int blk_throtl_init(struct request_queue *q) { return 0; }
	282	static inline void blk_throtl_exit(struct request_queue *q) { }
	283	#endif /* CONFIG_BLK_DEV_THROTTLING */
	284
	285	#endif /* BLK_INTERNAL_H */