[mirror_ubuntu-hirsute-kernel.git] / include / linux / blk-mq.h

/* SPDX-License-Identifier: GPL-2.0 */
#ifndef BLK_MQ_H
#define BLK_MQ_H

#include <linux/blkdev.h>
#include <linux/sbitmap.h>
#include <linux/srcu.h>

struct blk_mq_tags;
struct blk_flush_queue;

/**
 * struct blk_mq_hw_ctx - State for a hardware queue facing the hardware block device
 */
struct blk_mq_hw_ctx {
	struct {
		spinlock_t		lock;
		struct list_head	dispatch;
		unsigned long		state;		/* BLK_MQ_S_* flags */
	} ____cacheline_aligned_in_smp;

	struct delayed_work	run_work;
	cpumask_var_t		cpumask;
	int			next_cpu;
	int			next_cpu_batch;

	unsigned long		flags;		/* BLK_MQ_F_* flags */

	void			*sched_data;
	struct request_queue	*queue;
	struct blk_flush_queue	*fq;

	void			*driver_data;

	struct sbitmap		ctx_map;

	struct blk_mq_ctx	*dispatch_from;
	unsigned int		dispatch_busy;

	unsigned short		type;
	unsigned short		nr_ctx;
	struct blk_mq_ctx	**ctxs;

	spinlock_t		dispatch_wait_lock;
	wait_queue_entry_t	dispatch_wait;
	atomic_t		wait_index;

	struct blk_mq_tags	*tags;
	struct blk_mq_tags	*sched_tags;

	unsigned long		queued;
	unsigned long		run;
#define BLK_MQ_MAX_DISPATCH_ORDER	7
	unsigned long		dispatched[BLK_MQ_MAX_DISPATCH_ORDER];

	unsigned int		numa_node;
	unsigned int		queue_num;

	atomic_t		nr_active;

	struct hlist_node	cpuhp_dead;
	struct kobject		kobj;

	unsigned long		poll_considered;
	unsigned long		poll_invoked;
	unsigned long		poll_success;

#ifdef CONFIG_BLK_DEBUG_FS
	struct dentry		*debugfs_dir;
	struct dentry		*sched_debugfs_dir;
#endif

	struct list_head	hctx_list;

	/* Must be the last member - see also blk_mq_hw_ctx_size(). */
	struct srcu_struct	srcu[0];
};

struct blk_mq_queue_map {
	unsigned int *mq_map;
	unsigned int nr_queues;
	unsigned int queue_offset;
};

enum hctx_type {
	HCTX_TYPE_DEFAULT,	/* all I/O not otherwise accounted for */
	HCTX_TYPE_READ,		/* just for READ I/O */
	HCTX_TYPE_POLL,		/* polled I/O of any kind */

	HCTX_MAX_TYPES,
};

struct blk_mq_tag_set {
	/*
	 * map[] holds ctx -> hctx mappings, one map exists for each type
	 * that the driver wishes to support. There are no restrictions
	 * on maps being of the same size, and it's perfectly legal to
	 * share maps between types.
	 */
	struct blk_mq_queue_map	map[HCTX_MAX_TYPES];
	unsigned int		nr_maps;	/* nr entries in map[] */
	const struct blk_mq_ops	*ops;
	unsigned int		nr_hw_queues;	/* nr hw queues across maps */
	unsigned int		queue_depth;	/* max hw supported */
	unsigned int		reserved_tags;
	unsigned int		cmd_size;	/* per-request extra data */
	int			numa_node;
	unsigned int		timeout;
	unsigned int		flags;		/* BLK_MQ_F_* */
	void			*driver_data;

	struct blk_mq_tags	**tags;

	struct mutex		tag_list_lock;
	struct list_head	tag_list;
};

struct blk_mq_queue_data {
	struct request *rq;
	bool last;
};

typedef blk_status_t (queue_rq_fn)(struct blk_mq_hw_ctx *,
		const struct blk_mq_queue_data *);
typedef void (commit_rqs_fn)(struct blk_mq_hw_ctx *);
typedef bool (get_budget_fn)(struct blk_mq_hw_ctx *);
typedef void (put_budget_fn)(struct blk_mq_hw_ctx *);
typedef enum blk_eh_timer_return (timeout_fn)(struct request *, bool);
typedef int (init_hctx_fn)(struct blk_mq_hw_ctx *, void *, unsigned int);
typedef void (exit_hctx_fn)(struct blk_mq_hw_ctx *, unsigned int);
typedef int (init_request_fn)(struct blk_mq_tag_set *set, struct request *,
		unsigned int, unsigned int);
typedef void (exit_request_fn)(struct blk_mq_tag_set *set, struct request *,
		unsigned int);

typedef bool (busy_iter_fn)(struct blk_mq_hw_ctx *, struct request *, void *,
		bool);
typedef bool (busy_tag_iter_fn)(struct request *, void *, bool);
typedef int (poll_fn)(struct blk_mq_hw_ctx *);
typedef int (map_queues_fn)(struct blk_mq_tag_set *set);
typedef bool (busy_fn)(struct request_queue *);
typedef void (complete_fn)(struct request *);
typedef void (cleanup_rq_fn)(struct request *);


struct blk_mq_ops {
	/*
	 * Queue request
	 */
	queue_rq_fn		*queue_rq;

	/*
	 * If a driver uses bd->last to judge when to submit requests to
	 * hardware, it must define this function. In case of errors that
	 * make us stop issuing further requests, this hook serves the
	 * purpose of kicking the hardware (which the last request otherwise
	 * would have done).
	 */
	commit_rqs_fn		*commit_rqs;

	/*
	 * Reserve budget before queue request, once .queue_rq is
	 * run, it is driver's responsibility to release the
	 * reserved budget. Also we have to handle failure case
	 * of .get_budget for avoiding I/O deadlock.
	 */
	get_budget_fn		*get_budget;
	put_budget_fn		*put_budget;

	/*
	 * Called on request timeout
	 */
	timeout_fn		*timeout;

	/*
	 * Called to poll for completion of a specific tag.
	 */
	poll_fn			*poll;

	complete_fn		*complete;

	/*
	 * Called when the block layer side of a hardware queue has been
	 * set up, allowing the driver to allocate/init matching structures.
	 * Ditto for exit/teardown.
	 */
	init_hctx_fn		*init_hctx;
	exit_hctx_fn		*exit_hctx;

	/*
	 * Called for every command allocated by the block layer to allow
	 * the driver to set up driver specific data.
	 *
	 * Tag greater than or equal to queue_depth is for setting up
	 * flush request.
	 *
	 * Ditto for exit/teardown.
	 */
	init_request_fn		*init_request;
	exit_request_fn		*exit_request;
	/* Called from inside blk_get_request() */
	void (*initialize_rq_fn)(struct request *rq);

	/*
	 * Called before freeing one request which isn't completed yet,
	 * and usually for freeing the driver private data
	 */
	cleanup_rq_fn		*cleanup_rq;

	/*
	 * If set, returns whether or not this queue currently is busy
	 */
	busy_fn			*busy;

	map_queues_fn		*map_queues;

#ifdef CONFIG_BLK_DEBUG_FS
	/*
	 * Used by the debugfs implementation to show driver-specific
	 * information about a request.
	 */
	void (*show_rq)(struct seq_file *m, struct request *rq);
#endif
};

enum {
	BLK_MQ_F_SHOULD_MERGE	= 1 << 0,
	BLK_MQ_F_TAG_SHARED	= 1 << 1,
	BLK_MQ_F_BLOCKING	= 1 << 5,
	BLK_MQ_F_NO_SCHED	= 1 << 6,
	BLK_MQ_F_ALLOC_POLICY_START_BIT = 8,
	BLK_MQ_F_ALLOC_POLICY_BITS = 1,

	BLK_MQ_S_STOPPED	= 0,
	BLK_MQ_S_TAG_ACTIVE	= 1,
	BLK_MQ_S_SCHED_RESTART	= 2,

	BLK_MQ_MAX_DEPTH	= 10240,

	BLK_MQ_CPU_WORK_BATCH	= 8,
};
#define BLK_MQ_FLAG_TO_ALLOC_POLICY(flags) \
	((flags >> BLK_MQ_F_ALLOC_POLICY_START_BIT) & \
		((1 << BLK_MQ_F_ALLOC_POLICY_BITS) - 1))
#define BLK_ALLOC_POLICY_TO_MQ_FLAG(policy) \
	((policy & ((1 << BLK_MQ_F_ALLOC_POLICY_BITS) - 1)) \
		<< BLK_MQ_F_ALLOC_POLICY_START_BIT)

struct request_queue *blk_mq_init_queue(struct blk_mq_tag_set *);
struct request_queue *blk_mq_init_allocated_queue(struct blk_mq_tag_set *set,
						  struct request_queue *q,
						  bool elevator_init);
struct request_queue *blk_mq_init_sq_queue(struct blk_mq_tag_set *set,
						const struct blk_mq_ops *ops,
						unsigned int queue_depth,
						unsigned int set_flags);
void blk_mq_unregister_dev(struct device *, struct request_queue *);

int blk_mq_alloc_tag_set(struct blk_mq_tag_set *set);
void blk_mq_free_tag_set(struct blk_mq_tag_set *set);

void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule);

void blk_mq_free_request(struct request *rq);
bool blk_mq_can_queue(struct blk_mq_hw_ctx *);

bool blk_mq_queue_inflight(struct request_queue *q);

enum {
	/* return when out of requests */
	BLK_MQ_REQ_NOWAIT	= (__force blk_mq_req_flags_t)(1 << 0),
	/* allocate from reserved pool */
	BLK_MQ_REQ_RESERVED	= (__force blk_mq_req_flags_t)(1 << 1),
	/* allocate internal/sched tag */
	BLK_MQ_REQ_INTERNAL	= (__force blk_mq_req_flags_t)(1 << 2),
	/* set RQF_PREEMPT */
	BLK_MQ_REQ_PREEMPT	= (__force blk_mq_req_flags_t)(1 << 3),
};

struct request *blk_mq_alloc_request(struct request_queue *q, unsigned int op,
		blk_mq_req_flags_t flags);
struct request *blk_mq_alloc_request_hctx(struct request_queue *q,
		unsigned int op, blk_mq_req_flags_t flags,
		unsigned int hctx_idx);
struct request *blk_mq_tag_to_rq(struct blk_mq_tags *tags, unsigned int tag);

enum {
	BLK_MQ_UNIQUE_TAG_BITS = 16,
	BLK_MQ_UNIQUE_TAG_MASK = (1 << BLK_MQ_UNIQUE_TAG_BITS) - 1,
};

u32 blk_mq_unique_tag(struct request *rq);

static inline u16 blk_mq_unique_tag_to_hwq(u32 unique_tag)
{
	return unique_tag >> BLK_MQ_UNIQUE_TAG_BITS;
}

static inline u16 blk_mq_unique_tag_to_tag(u32 unique_tag)
{
	return unique_tag & BLK_MQ_UNIQUE_TAG_MASK;
}


int blk_mq_request_started(struct request *rq);
int blk_mq_request_completed(struct request *rq);
void blk_mq_start_request(struct request *rq);
void blk_mq_end_request(struct request *rq, blk_status_t error);
void __blk_mq_end_request(struct request *rq, blk_status_t error);

void blk_mq_requeue_request(struct request *rq, bool kick_requeue_list);
void blk_mq_kick_requeue_list(struct request_queue *q);
void blk_mq_delay_kick_requeue_list(struct request_queue *q, unsigned long msecs);
bool blk_mq_complete_request(struct request *rq);
bool blk_mq_bio_list_merge(struct request_queue *q, struct list_head *list,
			   struct bio *bio, unsigned int nr_segs);
bool blk_mq_queue_stopped(struct request_queue *q);
void blk_mq_stop_hw_queue(struct blk_mq_hw_ctx *hctx);
void blk_mq_start_hw_queue(struct blk_mq_hw_ctx *hctx);
void blk_mq_stop_hw_queues(struct request_queue *q);
void blk_mq_start_hw_queues(struct request_queue *q);
void blk_mq_start_stopped_hw_queue(struct blk_mq_hw_ctx *hctx, bool async);
void blk_mq_start_stopped_hw_queues(struct request_queue *q, bool async);
void blk_mq_quiesce_queue(struct request_queue *q);
void blk_mq_unquiesce_queue(struct request_queue *q);
void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs);
bool blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async);
void blk_mq_run_hw_queues(struct request_queue *q, bool async);
void blk_mq_tagset_busy_iter(struct blk_mq_tag_set *tagset,
		busy_tag_iter_fn *fn, void *priv);
void blk_mq_tagset_wait_completed_request(struct blk_mq_tag_set *tagset);
void blk_mq_freeze_queue(struct request_queue *q);
void blk_mq_unfreeze_queue(struct request_queue *q);
void blk_freeze_queue_start(struct request_queue *q);
void blk_mq_freeze_queue_wait(struct request_queue *q);
int blk_mq_freeze_queue_wait_timeout(struct request_queue *q,
				     unsigned long timeout);

int blk_mq_map_queues(struct blk_mq_queue_map *qmap);
void blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set, int nr_hw_queues);

void blk_mq_quiesce_queue_nowait(struct request_queue *q);

unsigned int blk_mq_rq_cpu(struct request *rq);

/*
 * Driver command data is immediately after the request. So subtract request
 * size to get back to the original request, add request size to get the PDU.
 */
static inline struct request *blk_mq_rq_from_pdu(void *pdu)
{
	return pdu - sizeof(struct request);
}
static inline void *blk_mq_rq_to_pdu(struct request *rq)
{
	return rq + 1;
}

#define queue_for_each_hw_ctx(q, hctx, i)				\
	for ((i) = 0; (i) < (q)->nr_hw_queues &&			\
	     ({ hctx = (q)->queue_hw_ctx[i]; 1; }); (i)++)

#define hctx_for_each_ctx(hctx, ctx, i)					\
	for ((i) = 0; (i) < (hctx)->nr_ctx &&				\
	     ({ ctx = (hctx)->ctxs[(i)]; 1; }); (i)++)

static inline blk_qc_t request_to_qc_t(struct blk_mq_hw_ctx *hctx,
		struct request *rq)
{
	if (rq->tag != -1)
		return rq->tag | (hctx->queue_num << BLK_QC_T_SHIFT);

	return rq->internal_tag | (hctx->queue_num << BLK_QC_T_SHIFT) |
			BLK_QC_T_INTERNAL;
}

static inline void blk_mq_cleanup_rq(struct request *rq)
{
	if (rq->q->mq_ops->cleanup_rq)
		rq->q->mq_ops->cleanup_rq(rq);
}

#endif
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
320ae51f JA	2	#ifndef BLK_MQ_H
	3	#define BLK_MQ_H
	4
	5	#include <linux/blkdev.h>
88459642	6	#include <linux/sbitmap.h>
6a83e74d	7	#include <linux/srcu.h>
320ae51f JA	8
320ae51f JA	9	struct blk_mq_tags;
f70ced09	10	struct blk_flush_queue;
320ae51f	11
fe644072 LW	12	/**
	13	* struct blk_mq_hw_ctx - State for a hardware queue facing the hardware block device
	14	*/
320ae51f JA	15	struct blk_mq_hw_ctx {
	16	struct {
	17	spinlock_t lock;
	18	struct list_head dispatch;
8d354f13	19	unsigned long state; /* BLK_MQ_S_* flags */
320ae51f JA	20	} ____cacheline_aligned_in_smp;
320ae51f JA	21
9f993737	22	struct delayed_work run_work;
e4043dcf	23	cpumask_var_t cpumask;
506e931f JA	24	int next_cpu;
506e931f JA	25	int next_cpu_batch;
320ae51f JA	26
	27	unsigned long flags; /* BLK_MQ_F_* flags */
	28
bd166ef1	29	void *sched_data;
320ae51f	30	struct request_queue *queue;
f70ced09	31	struct blk_flush_queue *fq;
320ae51f JA	32
	33	void *driver_data;
	34
88459642	35	struct sbitmap ctx_map;
1429d7c9	36
b347689f	37	struct blk_mq_ctx *dispatch_from;
6e768717	38	unsigned int dispatch_busy;
b347689f	39
f31967f0 JA	40	unsigned short type;
f31967f0 JA	41	unsigned short nr_ctx;
6e768717	42	struct blk_mq_ctx **ctxs;
4bb659b1	43
5815839b	44	spinlock_t dispatch_wait_lock;
eb619fdb	45	wait_queue_entry_t dispatch_wait;
8537b120	46	atomic_t wait_index;
320ae51f	47
320ae51f	48	struct blk_mq_tags *tags;
bd166ef1	49	struct blk_mq_tags *sched_tags;
320ae51f JA	50
	51	unsigned long queued;
	52	unsigned long run;
8d354f13	53	#define BLK_MQ_MAX_DISPATCH_ORDER 7
320ae51f JA	54	unsigned long dispatched[BLK_MQ_MAX_DISPATCH_ORDER];
320ae51f JA	55
320ae51f	56	unsigned int numa_node;
17ded320	57	unsigned int queue_num;
320ae51f	58
0d2602ca JA	59	atomic_t nr_active;
0d2602ca JA	60
9467f859	61	struct hlist_node cpuhp_dead;
320ae51f	62	struct kobject kobj;
05229bee	63
6e219353	64	unsigned long poll_considered;
05229bee JA	65	unsigned long poll_invoked;
05229bee JA	66	unsigned long poll_success;
9c1051aa OS	67
	68	#ifdef CONFIG_BLK_DEBUG_FS
	69	struct dentry *debugfs_dir;
d332ce09	70	struct dentry *sched_debugfs_dir;
9c1051aa	71	#endif
07319678	72
2f8f1336 ML	73	struct list_head hctx_list;
2f8f1336 ML	74
07319678	75	/* Must be the last member - see also blk_mq_hw_ctx_size(). */
05707b64	76	struct srcu_struct srcu[0];
320ae51f JA	77	};
320ae51f JA	78
ed76e329 JA	79	struct blk_mq_queue_map {
	80	unsigned int *mq_map;
	81	unsigned int nr_queues;
843477d4	82	unsigned int queue_offset;
ed76e329 JA	83	};
ed76e329 JA	84
e20ba6e1 CH	85	enum hctx_type {
	86	HCTX_TYPE_DEFAULT, /* all I/O not otherwise accounted for */
	87	HCTX_TYPE_READ, /* just for READ I/O */
	88	HCTX_TYPE_POLL, /* polled I/O of any kind */
	89
	90	HCTX_MAX_TYPES,
ed76e329 JA	91	};
ed76e329 JA	92
24d2f903	93	struct blk_mq_tag_set {
b3c661b1 JA	94	/*
	95	* map[] holds ctx -> hctx mappings, one map exists for each type
	96	* that the driver wishes to support. There are no restrictions
	97	* on maps being of the same size, and it's perfectly legal to
	98	* share maps between types.
	99	*/
ed76e329	100	struct blk_mq_queue_map map[HCTX_MAX_TYPES];
b3c661b1	101	unsigned int nr_maps; /* nr entries in map[] */
f8a5b122	102	const struct blk_mq_ops *ops;
ed76e329	103	unsigned int nr_hw_queues; /* nr hw queues across maps */
e3a2b3f9	104	unsigned int queue_depth; /* max hw supported */
320ae51f JA	105	unsigned int reserved_tags;
	106	unsigned int cmd_size; /* per-request extra data */
	107	int numa_node;
	108	unsigned int timeout;
	109	unsigned int flags; /* BLK_MQ_F_* */
24d2f903 CH	110	void *driver_data;
	111
	112	struct blk_mq_tags **tags;
0d2602ca JA	113
	114	struct mutex tag_list_lock;
	115	struct list_head tag_list;
320ae51f JA	116	};
320ae51f JA	117
74c45052 JA	118	struct blk_mq_queue_data {
74c45052 JA	119	struct request *rq;
74c45052 JA	120	bool last;
	121	};
	122
fc17b653 CH	123	typedef blk_status_t (queue_rq_fn)(struct blk_mq_hw_ctx *,
fc17b653 CH	124	const struct blk_mq_queue_data *);
d666ba98	125	typedef void (commit_rqs_fn)(struct blk_mq_hw_ctx *);
88022d72	126	typedef bool (get_budget_fn)(struct blk_mq_hw_ctx *);
de148297	127	typedef void (put_budget_fn)(struct blk_mq_hw_ctx *);
0152fb6b	128	typedef enum blk_eh_timer_return (timeout_fn)(struct request *, bool);
320ae51f JA	129	typedef int (init_hctx_fn)(struct blk_mq_hw_ctx , void , unsigned int);
320ae51f JA	130	typedef void (exit_hctx_fn)(struct blk_mq_hw_ctx *, unsigned int);
d6296d39	131	typedef int (init_request_fn)(struct blk_mq_tag_set set, struct request ,
24d2f903	132	unsigned int, unsigned int);
d6296d39	133	typedef void (exit_request_fn)(struct blk_mq_tag_set set, struct request ,
24d2f903	134	unsigned int);
320ae51f	135
7baa8572	136	typedef bool (busy_iter_fn)(struct blk_mq_hw_ctx , struct request , void *,
81481eb4	137	bool);
7baa8572	138	typedef bool (busy_tag_iter_fn)(struct request , void , bool);
9743139c	139	typedef int (poll_fn)(struct blk_mq_hw_ctx *);
da695ba2	140	typedef int (map_queues_fn)(struct blk_mq_tag_set *set);
9ba20527	141	typedef bool (busy_fn)(struct request_queue *);
c7bb9ad1	142	typedef void (complete_fn)(struct request *);
226b4fc7	143	typedef void (cleanup_rq_fn)(struct request *);
05229bee	144
81481eb4	145
320ae51f JA	146	struct blk_mq_ops {
	147	/*
	148	* Queue request
	149	*/
	150	queue_rq_fn *queue_rq;
	151
d666ba98 JA	152	/*
	153	* If a driver uses bd->last to judge when to submit requests to
	154	* hardware, it must define this function. In case of errors that
	155	* make us stop issuing further requests, this hook serves the
	156	* purpose of kicking the hardware (which the last request otherwise
	157	* would have done).
	158	*/
	159	commit_rqs_fn *commit_rqs;
	160
de148297 ML	161	/*
	162	* Reserve budget before queue request, once .queue_rq is
	163	* run, it is driver's responsibility to release the
	164	* reserved budget. Also we have to handle failure case
	165	* of .get_budget for avoiding I/O deadlock.
	166	*/
	167	get_budget_fn *get_budget;
	168	put_budget_fn *put_budget;
	169
320ae51f JA	170	/*
	171	* Called on request timeout
	172	*/
0152fb6b	173	timeout_fn *timeout;
320ae51f	174
05229bee JA	175	/*
	176	* Called to poll for completion of a specific tag.
	177	*/
	178	poll_fn *poll;
	179
c7bb9ad1	180	complete_fn *complete;
30a91cb4	181
320ae51f JA	182	/*
	183	* Called when the block layer side of a hardware queue has been
	184	* set up, allowing the driver to allocate/init matching structures.
	185	* Ditto for exit/teardown.
	186	*/
	187	init_hctx_fn *init_hctx;
	188	exit_hctx_fn *exit_hctx;
e9b267d9 CH	189
	190	/*
	191	* Called for every command allocated by the block layer to allow
	192	* the driver to set up driver specific data.
f70ced09 ML	193	*
	194	* Tag greater than or equal to queue_depth is for setting up
	195	* flush request.
	196	*
e9b267d9 CH	197	* Ditto for exit/teardown.
	198	*/
	199	init_request_fn *init_request;
	200	exit_request_fn *exit_request;
d280bab3 BVA	201	/* Called from inside blk_get_request() */
d280bab3 BVA	202	void (initialize_rq_fn)(struct request rq);
da695ba2	203
226b4fc7 ML	204	/*
	205	* Called before freeing one request which isn't completed yet,
	206	* and usually for freeing the driver private data
	207	*/
	208	cleanup_rq_fn *cleanup_rq;
	209
9ba20527 JA	210	/*
	211	* If set, returns whether or not this queue currently is busy
	212	*/
	213	busy_fn *busy;
	214
da695ba2	215	map_queues_fn *map_queues;
2836ee4b BVA	216
	217	#ifdef CONFIG_BLK_DEBUG_FS
	218	/*
	219	* Used by the debugfs implementation to show driver-specific
	220	* information about a request.
	221	*/
	222	void (show_rq)(struct seq_file m, struct request *rq);
	223	#endif
320ae51f JA	224	};
	225
	226	enum {
320ae51f	227	BLK_MQ_F_SHOULD_MERGE = 1 << 0,
8a58d1f1	228	BLK_MQ_F_TAG_SHARED = 1 << 1,
1b792f2f	229	BLK_MQ_F_BLOCKING = 1 << 5,
d3484991	230	BLK_MQ_F_NO_SCHED = 1 << 6,
24391c0d SL	231	BLK_MQ_F_ALLOC_POLICY_START_BIT = 8,
24391c0d SL	232	BLK_MQ_F_ALLOC_POLICY_BITS = 1,
320ae51f	233
5d12f905	234	BLK_MQ_S_STOPPED = 0,
0d2602ca	235	BLK_MQ_S_TAG_ACTIVE = 1,
bd166ef1	236	BLK_MQ_S_SCHED_RESTART = 2,
320ae51f	237
a4391c64	238	BLK_MQ_MAX_DEPTH = 10240,
506e931f JA	239
506e931f JA	240	BLK_MQ_CPU_WORK_BATCH = 8,
320ae51f	241	};
24391c0d SL	242	#define BLK_MQ_FLAG_TO_ALLOC_POLICY(flags) \
	243	((flags >> BLK_MQ_F_ALLOC_POLICY_START_BIT) & \
	244	((1 << BLK_MQ_F_ALLOC_POLICY_BITS) - 1))
	245	#define BLK_ALLOC_POLICY_TO_MQ_FLAG(policy) \
	246	((policy & ((1 << BLK_MQ_F_ALLOC_POLICY_BITS) - 1)) \
	247	<< BLK_MQ_F_ALLOC_POLICY_START_BIT)
320ae51f	248
24d2f903	249	struct request_queue blk_mq_init_queue(struct blk_mq_tag_set );
b62c21b7	250	struct request_queue blk_mq_init_allocated_queue(struct blk_mq_tag_set set,
737eb78e DLM	251	struct request_queue *q,
737eb78e DLM	252	bool elevator_init);
9316a9ed JA	253	struct request_queue blk_mq_init_sq_queue(struct blk_mq_tag_set set,
	254	const struct blk_mq_ops *ops,
	255	unsigned int queue_depth,
	256	unsigned int set_flags);
b21d5b30	257	void blk_mq_unregister_dev(struct device , struct request_queue );
320ae51f	258
24d2f903 CH	259	int blk_mq_alloc_tag_set(struct blk_mq_tag_set *set);
	260	void blk_mq_free_tag_set(struct blk_mq_tag_set *set);
	261
320ae51f JA	262	void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule);
320ae51f JA	263
320ae51f JA	264	void blk_mq_free_request(struct request *rq);
320ae51f JA	265	bool blk_mq_can_queue(struct blk_mq_hw_ctx *);
6f3b0e8b	266
3c94d83c	267	bool blk_mq_queue_inflight(struct request_queue *q);
ae879912	268
6f3b0e8b	269	enum {
9a95e4ef BVA	270	/* return when out of requests */
	271	BLK_MQ_REQ_NOWAIT = (__force blk_mq_req_flags_t)(1 << 0),
	272	/* allocate from reserved pool */
	273	BLK_MQ_REQ_RESERVED = (__force blk_mq_req_flags_t)(1 << 1),
	274	/* allocate internal/sched tag */
	275	BLK_MQ_REQ_INTERNAL = (__force blk_mq_req_flags_t)(1 << 2),
	276	/* set RQF_PREEMPT */
	277	BLK_MQ_REQ_PREEMPT = (__force blk_mq_req_flags_t)(1 << 3),
6f3b0e8b CH	278	};
6f3b0e8b CH	279
cd6ce148	280	struct request blk_mq_alloc_request(struct request_queue q, unsigned int op,
9a95e4ef	281	blk_mq_req_flags_t flags);
cd6ce148	282	struct request blk_mq_alloc_request_hctx(struct request_queue q,
9a95e4ef BVA	283	unsigned int op, blk_mq_req_flags_t flags,
9a95e4ef BVA	284	unsigned int hctx_idx);
0e62f51f	285	struct request blk_mq_tag_to_rq(struct blk_mq_tags tags, unsigned int tag);
320ae51f	286
205fb5f5 BVA	287	enum {
	288	BLK_MQ_UNIQUE_TAG_BITS = 16,
	289	BLK_MQ_UNIQUE_TAG_MASK = (1 << BLK_MQ_UNIQUE_TAG_BITS) - 1,
	290	};
	291
	292	u32 blk_mq_unique_tag(struct request *rq);
	293
	294	static inline u16 blk_mq_unique_tag_to_hwq(u32 unique_tag)
	295	{
	296	return unique_tag >> BLK_MQ_UNIQUE_TAG_BITS;
	297	}
	298
	299	static inline u16 blk_mq_unique_tag_to_tag(u32 unique_tag)
	300	{
	301	return unique_tag & BLK_MQ_UNIQUE_TAG_MASK;
	302	}
	303
320ae51f	304
973c0191	305	int blk_mq_request_started(struct request *rq);
aa306ab7	306	int blk_mq_request_completed(struct request *rq);
e2490073	307	void blk_mq_start_request(struct request *rq);
2a842aca CH	308	void blk_mq_end_request(struct request *rq, blk_status_t error);
2a842aca CH	309	void __blk_mq_end_request(struct request *rq, blk_status_t error);
320ae51f	310
2b053aca	311	void blk_mq_requeue_request(struct request *rq, bool kick_requeue_list);
6fca6a61	312	void blk_mq_kick_requeue_list(struct request_queue *q);
2849450a	313	void blk_mq_delay_kick_requeue_list(struct request_queue *q, unsigned long msecs);
16c15eb1	314	bool blk_mq_complete_request(struct request *rq);
9c558734	315	bool blk_mq_bio_list_merge(struct request_queue q, struct list_head list,
14ccb66b	316	struct bio *bio, unsigned int nr_segs);
fd001443	317	bool blk_mq_queue_stopped(struct request_queue *q);
320ae51f JA	318	void blk_mq_stop_hw_queue(struct blk_mq_hw_ctx *hctx);
320ae51f JA	319	void blk_mq_start_hw_queue(struct blk_mq_hw_ctx *hctx);
280d45f6	320	void blk_mq_stop_hw_queues(struct request_queue *q);
2f268556	321	void blk_mq_start_hw_queues(struct request_queue *q);
ae911c5e	322	void blk_mq_start_stopped_hw_queue(struct blk_mq_hw_ctx *hctx, bool async);
1b4a3258	323	void blk_mq_start_stopped_hw_queues(struct request_queue *q, bool async);
97e01209	324	void blk_mq_quiesce_queue(struct request_queue *q);
e4e73913	325	void blk_mq_unquiesce_queue(struct request_queue *q);
7587a5ae	326	void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs);
79f720a7	327	bool blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async);
b94ec296	328	void blk_mq_run_hw_queues(struct request_queue *q, bool async);
e0489487 SG	329	void blk_mq_tagset_busy_iter(struct blk_mq_tag_set *tagset,
e0489487 SG	330	busy_tag_iter_fn fn, void priv);
f9934a80	331	void blk_mq_tagset_wait_completed_request(struct blk_mq_tag_set *tagset);
c761d96b	332	void blk_mq_freeze_queue(struct request_queue *q);
b4c6a028	333	void blk_mq_unfreeze_queue(struct request_queue *q);
1671d522	334	void blk_freeze_queue_start(struct request_queue *q);
6bae363e	335	void blk_mq_freeze_queue_wait(struct request_queue *q);
f91328c4 KB	336	int blk_mq_freeze_queue_wait_timeout(struct request_queue *q,
f91328c4 KB	337	unsigned long timeout);
320ae51f	338
ed76e329	339	int blk_mq_map_queues(struct blk_mq_queue_map *qmap);
868f2f0b KB	340	void blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set, int nr_hw_queues);
868f2f0b KB	341
852ec809	342	void blk_mq_quiesce_queue_nowait(struct request_queue *q);
4f084b41	343
9cf2bab6 JA	344	unsigned int blk_mq_rq_cpu(struct request *rq);
9cf2bab6 JA	345
320ae51f JA	346	/*
320ae51f JA	347	* Driver command data is immediately after the request. So subtract request
2963e3f7	348	* size to get back to the original request, add request size to get the PDU.
320ae51f JA	349	*/
	350	static inline struct request blk_mq_rq_from_pdu(void pdu)
	351	{
	352	return pdu - sizeof(struct request);
	353	}
	354	static inline void blk_mq_rq_to_pdu(struct request rq)
	355	{
2963e3f7	356	return rq + 1;
320ae51f JA	357	}
320ae51f JA	358
320ae51f	359	#define queue_for_each_hw_ctx(q, hctx, i) \
0d0b7d42 JA	360	for ((i) = 0; (i) < (q)->nr_hw_queues && \
0d0b7d42 JA	361	({ hctx = (q)->queue_hw_ctx[i]; 1; }); (i)++)
320ae51f	362
320ae51f	363	#define hctx_for_each_ctx(hctx, ctx, i) \
0d0b7d42 JA	364	for ((i) = 0; (i) < (hctx)->nr_ctx && \
0d0b7d42 JA	365	({ ctx = (hctx)->ctxs[(i)]; 1; }); (i)++)
320ae51f	366
7b7ab780 SG	367	static inline blk_qc_t request_to_qc_t(struct blk_mq_hw_ctx *hctx,
	368	struct request *rq)
	369	{
	370	if (rq->tag != -1)
	371	return rq->tag \| (hctx->queue_num << BLK_QC_T_SHIFT);
	372
	373	return rq->internal_tag \| (hctx->queue_num << BLK_QC_T_SHIFT) \|
	374	BLK_QC_T_INTERNAL;
	375	}
	376
226b4fc7 ML	377	static inline void blk_mq_cleanup_rq(struct request *rq)
	378	{
	379	if (rq->q->mq_ops->cleanup_rq)
	380	rq->q->mq_ops->cleanup_rq(rq);
	381	}
	382
320ae51f	383	#endif