[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;
	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);

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);

void __blk_run_queue_uncond(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 bool blk_throtl_bio(struct request_queue *q, struct bio *bio);
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 bool blk_throtl_bio(struct request_queue *q, struct bio *bio)
{
	return false;
}
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;
	25	spinlock_t mq_flush_lock;
	26	};
	27
8324aa91	28	extern struct kmem_cache *blk_requestq_cachep;
320ae51f	29	extern struct kmem_cache *request_cachep;
8324aa91	30	extern struct kobj_type blk_queue_ktype;
a73f730d	31	extern struct ida blk_queue_ida;
8324aa91	32
7c94e1c1	33	static inline struct blk_flush_queue *blk_get_flush_queue(
e97c293c	34	struct request_queue q, struct blk_mq_ctx ctx)
7c94e1c1	35	{
f70ced09 ML	36	struct blk_mq_hw_ctx *hctx;
	37
	38	if (!q->mq_ops)
	39	return q->fq;
	40
	41	hctx = q->mq_ops->map_queue(q, ctx->cpu);
	42
	43	return hctx->fq;
7c94e1c1 ML	44	}
7c94e1c1 ML	45
09ac46c4 TH	46	static inline void __blk_get_queue(struct request_queue *q)
	47	{
	48	kobject_get(&q->kobj);
	49	}
	50
f70ced09 ML	51	struct blk_flush_queue blk_alloc_flush_queue(struct request_queue q,
	52	int node, int cmd_size);
	53	void blk_free_flush_queue(struct blk_flush_queue *q);
f3552655	54
5b788ce3 TH	55	int blk_init_rl(struct request_list rl, struct request_queue q,
	56	gfp_t gfp_mask);
	57	void blk_exit_rl(struct request_list *rl);
86db1e29 JA	58	void init_request_from_bio(struct request req, struct bio bio);
	59	void blk_rq_bio_prep(struct request_queue q, struct request rq,
	60	struct bio *bio);
a411f4bb BH	61	int blk_rq_append_bio(struct request_queue q, struct request rq,
a411f4bb BH	62	struct bio *bio);
d732580b TH	63	void blk_queue_bypass_start(struct request_queue *q);
d732580b TH	64	void blk_queue_bypass_end(struct request_queue *q);
9934c8c0	65	void blk_dequeue_request(struct request *rq);
8324aa91	66	void __blk_queue_free_tags(struct request_queue *q);
4853abaa JM	67	bool __blk_end_bidi_request(struct request *rq, int error,
4853abaa JM	68	unsigned int nr_bytes, unsigned int bidi_bytes);
8324aa91	69
242f9dcb	70	void blk_rq_timed_out_timer(unsigned long data);
0d2602ca	71	unsigned long blk_rq_timeout(unsigned long timeout);
87ee7b11	72	void blk_add_timer(struct request *req);
242f9dcb	73	void blk_delete_timer(struct request *);
242f9dcb	74
320ae51f JA	75
	76	bool bio_attempt_front_merge(struct request_queue q, struct request req,
	77	struct bio *bio);
	78	bool bio_attempt_back_merge(struct request_queue q, struct request req,
	79	struct bio *bio);
	80	bool blk_attempt_plug_merge(struct request_queue q, struct bio bio,
	81	unsigned int *request_count);
	82
	83	void blk_account_io_start(struct request *req, bool new_io);
	84	void blk_account_io_completion(struct request *req, unsigned int bytes);
	85	void blk_account_io_done(struct request *req);
	86
242f9dcb JA	87	/*
	88	* Internal atomic flags for request handling
	89	*/
	90	enum rq_atomic_flags {
	91	REQ_ATOM_COMPLETE = 0,
320ae51f	92	REQ_ATOM_STARTED,
242f9dcb JA	93	};
	94
	95	/*
	96	* EH timer and IO completion will both attempt to 'grab' the request, make
25985edc	97	* sure that only one of them succeeds
242f9dcb JA	98	*/
	99	static inline int blk_mark_rq_complete(struct request *rq)
	100	{
	101	return test_and_set_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
	102	}
	103
	104	static inline void blk_clear_rq_complete(struct request *rq)
	105	{
	106	clear_bit(REQ_ATOM_COMPLETE, &rq->atomic_flags);
	107	}
86db1e29	108
158dbda0 TH	109	/*
	110	* Internal elevator interface
	111	*/
360f92c2	112	#define ELV_ON_HASH(rq) ((rq)->cmd_flags & REQ_HASHED)
158dbda0	113
ae1b1539	114	void blk_insert_flush(struct request *rq);
dd831006	115
158dbda0 TH	116	static inline struct request __elv_next_request(struct request_queue q)
	117	{
	118	struct request *rq;
e97c293c	119	struct blk_flush_queue *fq = blk_get_flush_queue(q, NULL);
158dbda0 TH	120
158dbda0 TH	121	while (1) {
ae1b1539	122	if (!list_empty(&q->queue_head)) {
158dbda0	123	rq = list_entry_rq(q->queue_head.next);
ae1b1539	124	return rq;
158dbda0 TH	125	}
158dbda0 TH	126
3ac0cc45	127	/*
	128	* Flush request is running and flush request isn't queueable
	129	* in the drive, we can hold the queue till flush request is
	130	* finished. Even we don't do this, driver can't dispatch next
	131	* requests and will requeue them. And this can improve
	132	* throughput too. For example, we have request flush1, write1,
	133	* flush 2. flush1 is dispatched, then queue is hold, write1
	134	* isn't inserted to queue. After flush1 is finished, flush2
	135	* will be dispatched. Since disk cache is already clean,
	136	* flush2 will be finished very soon, so looks like flush2 is
	137	* folded to flush1.
	138	* Since the queue is hold, a flag is set to indicate the queue
	139	* should be restarted later. Please see flush_end_io() for
	140	* details.
	141	*/
7c94e1c1	142	if (fq->flush_pending_idx != fq->flush_running_idx &&
3ac0cc45	143	!queue_flush_queueable(q)) {
7c94e1c1	144	fq->flush_queue_delayed = 1;
3ac0cc45	145	return NULL;
3ac0cc45	146	}
556ee818	147	if (unlikely(blk_queue_bypass(q)) \|\|
22f746e2	148	!q->elevator->type->ops.elevator_dispatch_fn(q, 0))
158dbda0 TH	149	return NULL;
	150	}
	151	}
	152
	153	static inline void elv_activate_rq(struct request_queue q, struct request rq)
	154	{
	155	struct elevator_queue *e = q->elevator;
	156
22f746e2 TH	157	if (e->type->ops.elevator_activate_req_fn)
22f746e2 TH	158	e->type->ops.elevator_activate_req_fn(q, rq);
158dbda0 TH	159	}
	160
	161	static inline void elv_deactivate_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_deactivate_req_fn)
22f746e2 TH	166	e->type->ops.elevator_deactivate_req_fn(q, rq);
158dbda0 TH	167	}
158dbda0 TH	168
581d4e28 JA	169	#ifdef CONFIG_FAIL_IO_TIMEOUT
	170	int blk_should_fake_timeout(struct request_queue *);
	171	ssize_t part_timeout_show(struct device , struct device_attribute , char *);
	172	ssize_t part_timeout_store(struct device , struct device_attribute ,
	173	const char *, size_t);
	174	#else
	175	static inline int blk_should_fake_timeout(struct request_queue *q)
	176	{
	177	return 0;
	178	}
	179	#endif
	180
d6d48196 JA	181	int ll_back_merge_fn(struct request_queue q, struct request req,
	182	struct bio *bio);
	183	int ll_front_merge_fn(struct request_queue q, struct request req,
	184	struct bio *bio);
	185	int attempt_back_merge(struct request_queue q, struct request rq);
	186	int attempt_front_merge(struct request_queue q, struct request rq);
5e84ea3a JA	187	int blk_attempt_req_merge(struct request_queue q, struct request rq,
5e84ea3a JA	188	struct request *next);
d6d48196	189	void blk_recalc_rq_segments(struct request *rq);
80a761fd	190	void blk_rq_set_mixed_merge(struct request *rq);
050c8ea8 TH	191	bool blk_rq_merge_ok(struct request rq, struct bio bio);
050c8ea8 TH	192	int blk_try_merge(struct request rq, struct bio bio);
d6d48196	193
8324aa91 JA	194	void blk_queue_congestion_threshold(struct request_queue *q);
8324aa91 JA	195
c246e80d BVA	196	void __blk_run_queue_uncond(struct request_queue *q);
c246e80d BVA	197
ff88972c AB	198	int blk_dev_init(void);
ff88972c AB	199
f253b86b	200
8324aa91 JA	201	/*
	202	* Return the threshold (number of used requests) at which the queue is
	203	* considered to be congested. It include a little hysteresis to keep the
	204	* context switch rate down.
	205	*/
	206	static inline int queue_congestion_on_threshold(struct request_queue *q)
	207	{
	208	return q->nr_congestion_on;
	209	}
	210
	211	/*
	212	* The threshold at which a queue is considered to be uncongested
	213	*/
	214	static inline int queue_congestion_off_threshold(struct request_queue *q)
	215	{
	216	return q->nr_congestion_off;
	217	}
	218
e3a2b3f9 JA	219	extern int blk_update_nr_requests(struct request_queue *, unsigned int);
e3a2b3f9 JA	220
c2553b58 JA	221	/*
	222	* Contribute to IO statistics IFF:
	223	*
	224	* a) it's attached to a gendisk, and
	225	* b) the queue had IO stats enabled when this request was started, and
e2a60da7	226	* c) it's a file system request
c2553b58	227	*/
26308eab	228	static inline int blk_do_io_stat(struct request *rq)
fb8ec18c	229	{
33659ebb CH	230	return rq->rq_disk &&
33659ebb CH	231	(rq->cmd_flags & REQ_IO_STAT) &&
e2a60da7	232	(rq->cmd_type == REQ_TYPE_FS);
fb8ec18c JA	233	}
fb8ec18c JA	234
f2dbd76a TH	235	/*
	236	* Internal io_context interface
	237	*/
	238	void get_io_context(struct io_context *ioc);
47fdd4ca	239	struct io_cq ioc_lookup_icq(struct io_context ioc, struct request_queue *q);
24acfc34 TH	240	struct io_cq ioc_create_icq(struct io_context ioc, struct request_queue *q,
24acfc34 TH	241	gfp_t gfp_mask);
7e5a8794	242	void ioc_clear_queue(struct request_queue *q);
f2dbd76a	243
24acfc34	244	int create_task_io_context(struct task_struct *task, gfp_t gfp_mask, int node);
f2dbd76a TH	245
	246	/**
	247	* create_io_context - try to create task->io_context
f2dbd76a TH	248	* @gfp_mask: allocation mask
	249	* @node: allocation node
	250	*
24acfc34 TH	251	* If %current->io_context is %NULL, allocate a new io_context and install
	252	* it. Returns the current %current->io_context which may be %NULL if
	253	* allocation failed.
f2dbd76a TH	254	*
f2dbd76a TH	255	* Note that this function can't be called with IRQ disabled because
24acfc34	256	* task_lock which protects %current->io_context is IRQ-unsafe.
f2dbd76a	257	*/
24acfc34	258	static inline struct io_context *create_io_context(gfp_t gfp_mask, int node)
f2dbd76a TH	259	{
f2dbd76a TH	260	WARN_ON_ONCE(irqs_disabled());
24acfc34 TH	261	if (unlikely(!current->io_context))
	262	create_task_io_context(current, gfp_mask, node);
	263	return current->io_context;
f2dbd76a TH	264	}
	265
	266	/*
	267	* Internal throttling interface
	268	*/
bc9fcbf9	269	#ifdef CONFIG_BLK_DEV_THROTTLING
bc16a4f9	270	extern bool blk_throtl_bio(struct request_queue q, struct bio bio);
c9a929dd	271	extern void blk_throtl_drain(struct request_queue *q);
bc9fcbf9 TH	272	extern int blk_throtl_init(struct request_queue *q);
	273	extern void blk_throtl_exit(struct request_queue *q);
	274	#else /* CONFIG_BLK_DEV_THROTTLING */
bc16a4f9	275	static inline bool blk_throtl_bio(struct request_queue q, struct bio bio)
bc9fcbf9	276	{
bc16a4f9	277	return false;
bc9fcbf9	278	}
c9a929dd	279	static inline void blk_throtl_drain(struct request_queue *q) { }
bc9fcbf9 TH	280	static inline int blk_throtl_init(struct request_queue *q) { return 0; }
	281	static inline void blk_throtl_exit(struct request_queue *q) { }
	282	#endif /* CONFIG_BLK_DEV_THROTTLING */
	283
	284	#endif /* BLK_INTERNAL_H */