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

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

#include <linux/blk-crypto.h>
#include <linux/memblock.h>	/* for max_pfn/max_low_pfn */
#include <xen/xen.h>
#include "blk-crypto-internal.h"

struct elevator_type;

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

extern struct dentry *blk_debugfs_root;

struct blk_flush_queue {
	unsigned int		flush_pending_idx:1;
	unsigned int		flush_running_idx:1;
	blk_status_t 		rq_status;
	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;
};

bool is_flush_rq(struct request *req);

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

void blk_freeze_queue(struct request_queue *q);
void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
void blk_queue_start_drain(struct request_queue *q);
int __bio_queue_enter(struct request_queue *q, struct bio *bio);
void submit_bio_noacct_nocheck(struct bio *bio);

static inline bool blk_try_enter_queue(struct request_queue *q, bool pm)
{
	rcu_read_lock();
	if (!percpu_ref_tryget_live_rcu(&q->q_usage_counter))
		goto fail;

	/*
	 * The code that increments the pm_only counter must ensure that the
	 * counter is globally visible before the queue is unfrozen.
	 */
	if (blk_queue_pm_only(q) &&
	    (!pm || queue_rpm_status(q) == RPM_SUSPENDED))
		goto fail_put;

	rcu_read_unlock();
	return true;

fail_put:
	blk_queue_exit(q);
fail:
	rcu_read_unlock();
	return false;
}

static inline int bio_queue_enter(struct bio *bio)
{
	struct request_queue *q = bdev_get_queue(bio->bi_bdev);

	if (blk_try_enter_queue(q, false))
		return 0;
	return __bio_queue_enter(q, bio);
}

#define BIO_INLINE_VECS 4
struct bio_vec *bvec_alloc(mempool_t *pool, unsigned short *nr_vecs,
		gfp_t gfp_mask);
void bvec_free(mempool_t *pool, struct bio_vec *bv, unsigned short nr_vecs);

static inline bool biovec_phys_mergeable(struct request_queue *q,
		struct bio_vec *vec1, struct bio_vec *vec2)
{
	unsigned long mask = queue_segment_boundary(q);
	phys_addr_t addr1 = page_to_phys(vec1->bv_page) + vec1->bv_offset;
	phys_addr_t addr2 = page_to_phys(vec2->bv_page) + vec2->bv_offset;

	/*
	 * Merging adjacent physical pages may not work correctly under KMSAN
	 * if their metadata pages aren't adjacent. Just disable merging.
	 */
	if (IS_ENABLED(CONFIG_KMSAN))
		return false;

	if (addr1 + vec1->bv_len != addr2)
		return false;
	if (xen_domain() && !xen_biovec_phys_mergeable(vec1, vec2->bv_page))
		return false;
	if ((addr1 | mask) != ((addr2 + vec2->bv_len - 1) | mask))
		return false;
	return true;
}

static inline bool __bvec_gap_to_prev(const struct queue_limits *lim,
		struct bio_vec *bprv, unsigned int offset)
{
	return (offset & lim->virt_boundary_mask) ||
		((bprv->bv_offset + bprv->bv_len) & lim->virt_boundary_mask);
}

/*
 * Check if adding a bio_vec after bprv with offset would create a gap in
 * the SG list. Most drivers don't care about this, but some do.
 */
static inline bool bvec_gap_to_prev(const struct queue_limits *lim,
		struct bio_vec *bprv, unsigned int offset)
{
	if (!lim->virt_boundary_mask)
		return false;
	return __bvec_gap_to_prev(lim, bprv, offset);
}

static inline bool rq_mergeable(struct request *rq)
{
	if (blk_rq_is_passthrough(rq))
		return false;

	if (req_op(rq) == REQ_OP_FLUSH)
		return false;

	if (req_op(rq) == REQ_OP_WRITE_ZEROES)
		return false;

	if (req_op(rq) == REQ_OP_ZONE_APPEND)
		return false;

	if (rq->cmd_flags & REQ_NOMERGE_FLAGS)
		return false;
	if (rq->rq_flags & RQF_NOMERGE_FLAGS)
		return false;

	return true;
}

/*
 * There are two different ways to handle DISCARD merges:
 *  1) If max_discard_segments > 1, the driver treats every bio as a range and
 *     send the bios to controller together. The ranges don't need to be
 *     contiguous.
 *  2) Otherwise, the request will be normal read/write requests.  The ranges
 *     need to be contiguous.
 */
static inline bool blk_discard_mergable(struct request *req)
{
	if (req_op(req) == REQ_OP_DISCARD &&
	    queue_max_discard_segments(req->q) > 1)
		return true;
	return false;
}

static inline unsigned int blk_rq_get_max_segments(struct request *rq)
{
	if (req_op(rq) == REQ_OP_DISCARD)
		return queue_max_discard_segments(rq->q);
	return queue_max_segments(rq->q);
}

static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
						     enum req_op op)
{
	if (unlikely(op == REQ_OP_DISCARD || op == REQ_OP_SECURE_ERASE))
		return min(q->limits.max_discard_sectors,
			   UINT_MAX >> SECTOR_SHIFT);

	if (unlikely(op == REQ_OP_WRITE_ZEROES))
		return q->limits.max_write_zeroes_sectors;

	return q->limits.max_sectors;
}

#ifdef CONFIG_BLK_DEV_INTEGRITY
void blk_flush_integrity(void);
bool __bio_integrity_endio(struct bio *);
void bio_integrity_free(struct bio *bio);
static inline bool bio_integrity_endio(struct bio *bio)
{
	if (bio_integrity(bio))
		return __bio_integrity_endio(bio);
	return true;
}

bool blk_integrity_merge_rq(struct request_queue *, struct request *,
		struct request *);
bool blk_integrity_merge_bio(struct request_queue *, struct request *,
		struct bio *);

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

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

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

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

extern const struct attribute_group blk_integrity_attr_group;
#else /* CONFIG_BLK_DEV_INTEGRITY */
static inline bool blk_integrity_merge_rq(struct request_queue *rq,
		struct request *r1, struct request *r2)
{
	return true;
}
static inline bool blk_integrity_merge_bio(struct request_queue *rq,
		struct request *r, struct bio *b)
{
	return true;
}
static inline bool integrity_req_gap_back_merge(struct request *req,
		struct bio *next)
{
	return false;
}
static inline bool integrity_req_gap_front_merge(struct request *req,
		struct bio *bio)
{
	return false;
}

static inline void blk_flush_integrity(void)
{
}
static inline bool bio_integrity_endio(struct bio *bio)
{
	return true;
}
static inline void bio_integrity_free(struct bio *bio)
{
}
#endif /* CONFIG_BLK_DEV_INTEGRITY */

unsigned long blk_rq_timeout(unsigned long timeout);
void blk_add_timer(struct request *req);
const char *blk_status_to_str(blk_status_t status);

bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
		unsigned int nr_segs);
bool blk_bio_list_merge(struct request_queue *q, struct list_head *list,
			struct bio *bio, unsigned int nr_segs);

/*
 * Plug flush limits
 */
#define BLK_MAX_REQUEST_COUNT	32
#define BLK_PLUG_FLUSH_SIZE	(128 * 1024)

/*
 * Internal elevator interface
 */
#define ELV_ON_HASH(rq) ((rq)->rq_flags & RQF_HASHED)

void blk_insert_flush(struct request *rq);

int elevator_switch(struct request_queue *q, struct elevator_type *new_e);
void elevator_disable(struct request_queue *q);
void elevator_exit(struct request_queue *q);
int elv_register_queue(struct request_queue *q, bool uevent);
void elv_unregister_queue(struct request_queue *q);

ssize_t part_size_show(struct device *dev, struct device_attribute *attr,
		char *buf);
ssize_t part_stat_show(struct device *dev, struct device_attribute *attr,
		char *buf);
ssize_t part_inflight_show(struct device *dev, struct device_attribute *attr,
		char *buf);
ssize_t part_fail_show(struct device *dev, struct device_attribute *attr,
		char *buf);
ssize_t part_fail_store(struct device *dev, struct device_attribute *attr,
		const char *buf, size_t count);
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);

static inline bool bio_may_exceed_limits(struct bio *bio,
					 const struct queue_limits *lim)
{
	switch (bio_op(bio)) {
	case REQ_OP_DISCARD:
	case REQ_OP_SECURE_ERASE:
	case REQ_OP_WRITE_ZEROES:
		return true; /* non-trivial splitting decisions */
	default:
		break;
	}

	/*
	 * All drivers must accept single-segments bios that are <= PAGE_SIZE.
	 * This is a quick and dirty check that relies on the fact that
	 * bi_io_vec[0] is always valid if a bio has data.  The check might
	 * lead to occasional false negatives when bios are cloned, but compared
	 * to the performance impact of cloned bios themselves the loop below
	 * doesn't matter anyway.
	 */
	return lim->chunk_sectors || bio->bi_vcnt != 1 ||
		bio->bi_io_vec->bv_len + bio->bi_io_vec->bv_offset > PAGE_SIZE;
}

struct bio *__bio_split_to_limits(struct bio *bio,
				  const struct queue_limits *lim,
				  unsigned int *nr_segs);
int ll_back_merge_fn(struct request *req, struct bio *bio,
		unsigned int nr_segs);
bool blk_attempt_req_merge(struct request_queue *q, struct request *rq,
				struct request *next);
unsigned int 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);
enum elv_merge blk_try_merge(struct request *rq, struct bio *bio);

void blk_set_default_limits(struct queue_limits *lim);
int blk_dev_init(void);

/*
 * 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
 */
static inline bool blk_do_io_stat(struct request *rq)
{
	return (rq->rq_flags & RQF_IO_STAT) && !blk_rq_is_passthrough(rq);
}

void update_io_ticks(struct block_device *part, unsigned long now, bool end);

static inline void req_set_nomerge(struct request_queue *q, struct request *req)
{
	req->cmd_flags |= REQ_NOMERGE;
	if (req == q->last_merge)
		q->last_merge = NULL;
}

/*
 * Internal io_context interface
 */
struct io_cq *ioc_find_get_icq(struct request_queue *q);
struct io_cq *ioc_lookup_icq(struct request_queue *q);
#ifdef CONFIG_BLK_ICQ
void ioc_clear_queue(struct request_queue *q);
#else
static inline void ioc_clear_queue(struct request_queue *q)
{
}
#endif /* CONFIG_BLK_ICQ */

#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
extern ssize_t blk_throtl_sample_time_show(struct request_queue *q, char *page);
extern ssize_t blk_throtl_sample_time_store(struct request_queue *q,
	const char *page, size_t count);
extern void blk_throtl_bio_endio(struct bio *bio);
extern void blk_throtl_stat_add(struct request *rq, u64 time);
#else
static inline void blk_throtl_bio_endio(struct bio *bio) { }
static inline void blk_throtl_stat_add(struct request *rq, u64 time) { }
#endif

struct bio *__blk_queue_bounce(struct bio *bio, struct request_queue *q);

static inline bool blk_queue_may_bounce(struct request_queue *q)
{
	return IS_ENABLED(CONFIG_BOUNCE) &&
		q->limits.bounce == BLK_BOUNCE_HIGH &&
		max_low_pfn >= max_pfn;
}

static inline struct bio *blk_queue_bounce(struct bio *bio,
		struct request_queue *q)
{
	if (unlikely(blk_queue_may_bounce(q) && bio_has_data(bio)))
		return __blk_queue_bounce(bio, q);
	return bio;
}

#ifdef CONFIG_BLK_DEV_ZONED
void disk_free_zone_bitmaps(struct gendisk *disk);
void disk_clear_zone_settings(struct gendisk *disk);
#else
static inline void disk_free_zone_bitmaps(struct gendisk *disk) {}
static inline void disk_clear_zone_settings(struct gendisk *disk) {}
#endif

int blk_alloc_ext_minor(void);
void blk_free_ext_minor(unsigned int minor);
#define ADDPART_FLAG_NONE	0
#define ADDPART_FLAG_RAID	1
#define ADDPART_FLAG_WHOLEDISK	2
int bdev_add_partition(struct gendisk *disk, int partno, sector_t start,
		sector_t length);
int bdev_del_partition(struct gendisk *disk, int partno);
int bdev_resize_partition(struct gendisk *disk, int partno, sector_t start,
		sector_t length);
void blk_drop_partitions(struct gendisk *disk);

void bdev_set_nr_sectors(struct block_device *bdev, sector_t sectors);

struct gendisk *__alloc_disk_node(struct request_queue *q, int node_id,
		struct lock_class_key *lkclass);

int bio_add_hw_page(struct request_queue *q, struct bio *bio,
		struct page *page, unsigned int len, unsigned int offset,
		unsigned int max_sectors, bool *same_page);

struct request_queue *blk_alloc_queue(int node_id);

int disk_scan_partitions(struct gendisk *disk, fmode_t mode);

int disk_alloc_events(struct gendisk *disk);
void disk_add_events(struct gendisk *disk);
void disk_del_events(struct gendisk *disk);
void disk_release_events(struct gendisk *disk);
void disk_block_events(struct gendisk *disk);
void disk_unblock_events(struct gendisk *disk);
void disk_flush_events(struct gendisk *disk, unsigned int mask);
extern struct device_attribute dev_attr_events;
extern struct device_attribute dev_attr_events_async;
extern struct device_attribute dev_attr_events_poll_msecs;

extern struct attribute_group blk_trace_attr_group;

long blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg);
long compat_blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg);

extern const struct address_space_operations def_blk_aops;

int disk_register_independent_access_ranges(struct gendisk *disk);
void disk_unregister_independent_access_ranges(struct gendisk *disk);

#ifdef CONFIG_FAIL_MAKE_REQUEST
bool should_fail_request(struct block_device *part, unsigned int bytes);
#else /* CONFIG_FAIL_MAKE_REQUEST */
static inline bool should_fail_request(struct block_device *part,
					unsigned int bytes)
{
	return false;
}
#endif /* CONFIG_FAIL_MAKE_REQUEST */

/*
 * Optimized request reference counting. Ideally we'd make timeouts be more
 * clever, as that's the only reason we need references at all... But until
 * this happens, this is faster than using refcount_t. Also see:
 *
 * abc54d634334 ("io_uring: switch to atomic_t for io_kiocb reference count")
 */
#define req_ref_zero_or_close_to_overflow(req)	\
	((unsigned int) atomic_read(&(req->ref)) + 127u <= 127u)

static inline bool req_ref_inc_not_zero(struct request *req)
{
	return atomic_inc_not_zero(&req->ref);
}

static inline bool req_ref_put_and_test(struct request *req)
{
	WARN_ON_ONCE(req_ref_zero_or_close_to_overflow(req));
	return atomic_dec_and_test(&req->ref);
}

static inline void req_ref_set(struct request *req, int value)
{
	atomic_set(&req->ref, value);
}

static inline int req_ref_read(struct request *req)
{
	return atomic_read(&req->ref);
}

#endif /* BLK_INTERNAL_H */
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
8324aa91 JA	2	#ifndef BLK_INTERNAL_H
	3	#define BLK_INTERNAL_H
	4
a892c8d5	5	#include <linux/blk-crypto.h>
9bb33f24	6	#include <linux/memblock.h> /* for max_pfn/max_low_pfn */
c39ae60d	7	#include <xen/xen.h>
a892c8d5	8	#include "blk-crypto-internal.h"
a73f730d	9
2e9bc346 CH	10	struct elevator_type;
2e9bc346 CH	11
0d2602ca JA	12	/* Max future timer expiry for timeouts */
	13	#define BLK_MAX_TIMEOUT (5 * HZ)
	14
18fbda91	15	extern struct dentry *blk_debugfs_root;
18fbda91	16
7c94e1c1	17	struct blk_flush_queue {
7c94e1c1 ML	18	unsigned int flush_pending_idx:1;
7c94e1c1 ML	19	unsigned int flush_running_idx:1;
8d699663	20	blk_status_t rq_status;
7c94e1c1 ML	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	25
7c94e1c1 ML	26	spinlock_t mq_flush_lock;
	27	};
	28
a9ed27a7	29	bool is_flush_rq(struct request *req);
8d699663	30
754a1572 GJ	31	struct blk_flush_queue *blk_alloc_flush_queue(int node, int cmd_size,
754a1572 GJ	32	gfp_t flags);
f70ced09	33	void blk_free_flush_queue(struct blk_flush_queue *q);
f3552655	34
3ef28e83	35	void blk_freeze_queue(struct request_queue *q);
aec89dc5	36	void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
8e141f9e	37	void blk_queue_start_drain(struct request_queue *q);
c98cb5bb	38	int __bio_queue_enter(struct request_queue q, struct bio bio);
3f98c753	39	void submit_bio_noacct_nocheck(struct bio *bio);
c98cb5bb JA	40
	41	static inline bool blk_try_enter_queue(struct request_queue *q, bool pm)
	42	{
	43	rcu_read_lock();
	44	if (!percpu_ref_tryget_live_rcu(&q->q_usage_counter))
	45	goto fail;
	46
	47	/*
	48	* The code that increments the pm_only counter must ensure that the
	49	* counter is globally visible before the queue is unfrozen.
	50	*/
	51	if (blk_queue_pm_only(q) &&
	52	(!pm \|\| queue_rpm_status(q) == RPM_SUSPENDED))
	53	goto fail_put;
	54
	55	rcu_read_unlock();
	56	return true;
	57
	58	fail_put:
	59	blk_queue_exit(q);
	60	fail:
	61	rcu_read_unlock();
	62	return false;
	63	}
	64
	65	static inline int bio_queue_enter(struct bio *bio)
	66	{
	67	struct request_queue *q = bdev_get_queue(bio->bi_bdev);
	68
	69	if (blk_try_enter_queue(q, false))
	70	return 0;
	71	return __bio_queue_enter(q, bio);
	72	}
3ef28e83	73
dc0b8a57	74	#define BIO_INLINE_VECS 4
7a800a20 CH	75	struct bio_vec bvec_alloc(mempool_t pool, unsigned short *nr_vecs,
	76	gfp_t gfp_mask);
	77	void bvec_free(mempool_t pool, struct bio_vec bv, unsigned short nr_vecs);
eec716a1	78
3dccdae5 CH	79	static inline bool biovec_phys_mergeable(struct request_queue *q,
3dccdae5 CH	80	struct bio_vec vec1, struct bio_vec vec2)
6a9f5f24	81	{
3dccdae5	82	unsigned long mask = queue_segment_boundary(q);
6e768461 CH	83	phys_addr_t addr1 = page_to_phys(vec1->bv_page) + vec1->bv_offset;
6e768461 CH	84	phys_addr_t addr2 = page_to_phys(vec2->bv_page) + vec2->bv_offset;
3dccdae5	85
f630a5d0 AP	86	/*
	87	* Merging adjacent physical pages may not work correctly under KMSAN
	88	* if their metadata pages aren't adjacent. Just disable merging.
	89	*/
	90	if (IS_ENABLED(CONFIG_KMSAN))
	91	return false;
	92
3dccdae5	93	if (addr1 + vec1->bv_len != addr2)
6a9f5f24	94	return false;
0383ad43	95	if (xen_domain() && !xen_biovec_phys_mergeable(vec1, vec2->bv_page))
6a9f5f24	96	return false;
3dccdae5 CH	97	if ((addr1 \| mask) != ((addr2 + vec2->bv_len - 1) \| mask))
3dccdae5 CH	98	return false;
6a9f5f24 CH	99	return true;
	100	}
	101
aa261f20	102	static inline bool __bvec_gap_to_prev(const struct queue_limits *lim,
27ca1d4e CH	103	struct bio_vec *bprv, unsigned int offset)
27ca1d4e CH	104	{
c55ddd90 CH	105	return (offset & lim->virt_boundary_mask) \|\|
c55ddd90 CH	106	((bprv->bv_offset + bprv->bv_len) & lim->virt_boundary_mask);
27ca1d4e CH	107	}
	108
	109	/*
	110	* Check if adding a bio_vec after bprv with offset would create a gap in
	111	* the SG list. Most drivers don't care about this, but some do.
	112	*/
aa261f20	113	static inline bool bvec_gap_to_prev(const struct queue_limits *lim,
27ca1d4e CH	114	struct bio_vec *bprv, unsigned int offset)
27ca1d4e CH	115	{
c55ddd90	116	if (!lim->virt_boundary_mask)
27ca1d4e	117	return false;
c55ddd90	118	return __bvec_gap_to_prev(lim, bprv, offset);
27ca1d4e CH	119	}
27ca1d4e CH	120
badf7f64 CH	121	static inline bool rq_mergeable(struct request *rq)
	122	{
	123	if (blk_rq_is_passthrough(rq))
	124	return false;
	125
	126	if (req_op(rq) == REQ_OP_FLUSH)
	127	return false;
	128
	129	if (req_op(rq) == REQ_OP_WRITE_ZEROES)
	130	return false;
	131
	132	if (req_op(rq) == REQ_OP_ZONE_APPEND)
	133	return false;
	134
	135	if (rq->cmd_flags & REQ_NOMERGE_FLAGS)
	136	return false;
	137	if (rq->rq_flags & RQF_NOMERGE_FLAGS)
	138	return false;
	139
	140	return true;
	141	}
	142
	143	/*
	144	* There are two different ways to handle DISCARD merges:
	145	* 1) If max_discard_segments > 1, the driver treats every bio as a range and
	146	* send the bios to controller together. The ranges don't need to be
	147	* contiguous.
	148	* 2) Otherwise, the request will be normal read/write requests. The ranges
	149	* need to be contiguous.
	150	*/
	151	static inline bool blk_discard_mergable(struct request *req)
	152	{
	153	if (req_op(req) == REQ_OP_DISCARD &&
	154	queue_max_discard_segments(req->q) > 1)
	155	return true;
	156	return false;
	157	}
	158
49d24398 US	159	static inline unsigned int blk_rq_get_max_segments(struct request *rq)
	160	{
	161	if (req_op(rq) == REQ_OP_DISCARD)
	162	return queue_max_discard_segments(rq->q);
	163	return queue_max_segments(rq->q);
	164	}
	165
2a9336c4	166	static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
77e7ffd7	167	enum req_op op)
2a9336c4 CH	168	{
	169	if (unlikely(op == REQ_OP_DISCARD \|\| op == REQ_OP_SECURE_ERASE))
	170	return min(q->limits.max_discard_sectors,
	171	UINT_MAX >> SECTOR_SHIFT);
	172
	173	if (unlikely(op == REQ_OP_WRITE_ZEROES))
	174	return q->limits.max_write_zeroes_sectors;
	175
	176	return q->limits.max_sectors;
	177	}
	178
5a48fc14 DW	179	#ifdef CONFIG_BLK_DEV_INTEGRITY
5a48fc14 DW	180	void blk_flush_integrity(void);
7c20f116	181	bool __bio_integrity_endio(struct bio *);
ece841ab	182	void bio_integrity_free(struct bio *bio);
7c20f116 CH	183	static inline bool bio_integrity_endio(struct bio *bio)
	184	{
	185	if (bio_integrity(bio))
	186	return __bio_integrity_endio(bio);
	187	return true;
	188	}
43b729bf	189
92cf2fd1 CH	190	bool blk_integrity_merge_rq(struct request_queue , struct request ,
92cf2fd1 CH	191	struct request *);
d59da419 CH	192	bool blk_integrity_merge_bio(struct request_queue , struct request ,
d59da419 CH	193	struct bio *);
92cf2fd1	194
43b729bf CH	195	static inline bool integrity_req_gap_back_merge(struct request *req,
	196	struct bio *next)
	197	{
	198	struct bio_integrity_payload *bip = bio_integrity(req->bio);
	199	struct bio_integrity_payload *bip_next = bio_integrity(next);
	200
c55ddd90 CH	201	return bvec_gap_to_prev(&req->q->limits,
c55ddd90 CH	202	&bip->bip_vec[bip->bip_vcnt - 1],
43b729bf CH	203	bip_next->bip_vec[0].bv_offset);
	204	}
	205
	206	static inline bool integrity_req_gap_front_merge(struct request *req,
	207	struct bio *bio)
	208	{
	209	struct bio_integrity_payload *bip = bio_integrity(bio);
	210	struct bio_integrity_payload *bip_next = bio_integrity(req->bio);
	211
c55ddd90 CH	212	return bvec_gap_to_prev(&req->q->limits,
c55ddd90 CH	213	&bip->bip_vec[bip->bip_vcnt - 1],
43b729bf CH	214	bip_next->bip_vec[0].bv_offset);
43b729bf CH	215	}
581e2600	216
ff53cd52	217	extern const struct attribute_group blk_integrity_attr_group;
43b729bf	218	#else /* CONFIG_BLK_DEV_INTEGRITY */
92cf2fd1 CH	219	static inline bool blk_integrity_merge_rq(struct request_queue *rq,
	220	struct request r1, struct request r2)
	221	{
	222	return true;
	223	}
d59da419 CH	224	static inline bool blk_integrity_merge_bio(struct request_queue *rq,
	225	struct request r, struct bio b)
	226	{
	227	return true;
	228	}
43b729bf CH	229	static inline bool integrity_req_gap_back_merge(struct request *req,
	230	struct bio *next)
	231	{
	232	return false;
	233	}
	234	static inline bool integrity_req_gap_front_merge(struct request *req,
	235	struct bio *bio)
	236	{
	237	return false;
	238	}
	239
5a48fc14 DW	240	static inline void blk_flush_integrity(void)
	241	{
	242	}
7c20f116 CH	243	static inline bool bio_integrity_endio(struct bio *bio)
	244	{
	245	return true;
	246	}
ece841ab JT	247	static inline void bio_integrity_free(struct bio *bio)
	248	{
	249	}
43b729bf	250	#endif /* CONFIG_BLK_DEV_INTEGRITY */
8324aa91	251
0d2602ca	252	unsigned long blk_rq_timeout(unsigned long timeout);
87ee7b11	253	void blk_add_timer(struct request *req);
0d7a29a2	254	const char *blk_status_to_str(blk_status_t status);
320ae51f	255
320ae51f	256	bool blk_attempt_plug_merge(struct request_queue q, struct bio bio,
0c5bcc92	257	unsigned int nr_segs);
bdc6a287 BW	258	bool blk_bio_list_merge(struct request_queue q, struct list_head list,
bdc6a287 BW	259	struct bio *bio, unsigned int nr_segs);
320ae51f	260
ba0ffdd8 JA	261	/*
	262	* Plug flush limits
	263	*/
	264	#define BLK_MAX_REQUEST_COUNT 32
	265	#define BLK_PLUG_FLUSH_SIZE (128 * 1024)
	266
158dbda0 TH	267	/*
	268	* Internal elevator interface
	269	*/
e8064021	270	#define ELV_ON_HASH(rq) ((rq)->rq_flags & RQF_HASHED)
158dbda0	271
2b504bd4	272	void blk_insert_flush(struct request *rq);
dd831006	273
8237c01f	274	int elevator_switch(struct request_queue q, struct elevator_type new_e);
64b36075	275	void elevator_disable(struct request_queue *q);
0c6cb3a2	276	void elevator_exit(struct request_queue *q);
cecf5d87	277	int elv_register_queue(struct request_queue *q, bool uevent);
83d016ac BVA	278	void elv_unregister_queue(struct request_queue *q);
83d016ac BVA	279
3ad5cee5 CH	280	ssize_t part_size_show(struct device dev, struct device_attribute attr,
	281	char *buf);
	282	ssize_t part_stat_show(struct device dev, struct device_attribute attr,
	283	char *buf);
	284	ssize_t part_inflight_show(struct device dev, struct device_attribute attr,
	285	char *buf);
	286	ssize_t part_fail_show(struct device dev, struct device_attribute attr,
	287	char *buf);
	288	ssize_t part_fail_store(struct device dev, struct device_attribute attr,
	289	const char *buf, size_t count);
581d4e28 JA	290	ssize_t part_timeout_show(struct device , struct device_attribute , char *);
	291	ssize_t part_timeout_store(struct device , struct device_attribute ,
	292	const char *, size_t);
581d4e28	293
c55ddd90	294	static inline bool bio_may_exceed_limits(struct bio *bio,
aa261f20	295	const struct queue_limits *lim)
abd45c15 JA	296	{
	297	switch (bio_op(bio)) {
	298	case REQ_OP_DISCARD:
	299	case REQ_OP_SECURE_ERASE:
	300	case REQ_OP_WRITE_ZEROES:
abd45c15 JA	301	return true; /* non-trivial splitting decisions */
	302	default:
	303	break;
	304	}
	305
	306	/*
	307	* All drivers must accept single-segments bios that are <= PAGE_SIZE.
	308	* This is a quick and dirty check that relies on the fact that
	309	* bi_io_vec[0] is always valid if a bio has data. The check might
	310	* lead to occasional false negatives when bios are cloned, but compared
	311	* to the performance impact of cloned bios themselves the loop below
	312	* doesn't matter anyway.
	313	*/
c55ddd90	314	return lim->chunk_sectors \|\| bio->bi_vcnt != 1 \|\|
abd45c15 JA	315	bio->bi_io_vec->bv_len + bio->bi_io_vec->bv_offset > PAGE_SIZE;
	316	}
	317
aa261f20 BVA	318	struct bio __bio_split_to_limits(struct bio bio,
	319	const struct queue_limits *lim,
	320	unsigned int *nr_segs);
14ccb66b CH	321	int ll_back_merge_fn(struct request req, struct bio bio,
14ccb66b CH	322	unsigned int nr_segs);
fd2ef39c	323	bool blk_attempt_req_merge(struct request_queue q, struct request rq,
5e84ea3a	324	struct request *next);
e9cd19c0	325	unsigned int blk_recalc_rq_segments(struct request *rq);
80a761fd	326	void blk_rq_set_mixed_merge(struct request *rq);
050c8ea8	327	bool blk_rq_merge_ok(struct request rq, struct bio bio);
34fe7c05	328	enum elv_merge blk_try_merge(struct request rq, struct bio bio);
d6d48196	329
b3228254	330	void blk_set_default_limits(struct queue_limits *lim);
ff88972c AB	331	int blk_dev_init(void);
ff88972c AB	332
c2553b58 JA	333	/*
	334	* Contribute to IO statistics IFF:
	335	*
	336	* a) it's attached to a gendisk, and
48d9b0d4	337	* b) the queue had IO stats enabled when this request was started
c2553b58	338	*/
599d067d	339	static inline bool blk_do_io_stat(struct request *rq)
fb8ec18c	340	{
41fa7222	341	return (rq->rq_flags & RQF_IO_STAT) && !blk_rq_is_passthrough(rq);
be6bfe36 PB	342	}
be6bfe36 PB	343
450b7879	344	void update_io_ticks(struct block_device *part, unsigned long now, bool end);
fb8ec18c	345
6cf7677f CH	346	static inline void req_set_nomerge(struct request_queue q, struct request req)
	347	{
	348	req->cmd_flags \|= REQ_NOMERGE;
	349	if (req == q->last_merge)
	350	q->last_merge = NULL;
	351	}
	352
f2dbd76a TH	353	/*
	354	* Internal io_context interface
	355	*/
87dd1d63	356	struct io_cq ioc_find_get_icq(struct request_queue q);
eca5892a	357	struct io_cq ioc_lookup_icq(struct request_queue q);
5ef16305	358	#ifdef CONFIG_BLK_ICQ
7e5a8794	359	void ioc_clear_queue(struct request_queue *q);
5ef16305 CH	360	#else
	361	static inline void ioc_clear_queue(struct request_queue *q)
	362	{
	363	}
	364	#endif /* CONFIG_BLK_ICQ */
f2dbd76a	365
297e3d85 SL	366	#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
	367	extern ssize_t blk_throtl_sample_time_show(struct request_queue q, char page);
	368	extern ssize_t blk_throtl_sample_time_store(struct request_queue *q,
	369	const char *page, size_t count);
9e234eea	370	extern void blk_throtl_bio_endio(struct bio *bio);
b9147dd1	371	extern void blk_throtl_stat_add(struct request *rq, u64 time);
9e234eea SL	372	#else
9e234eea SL	373	static inline void blk_throtl_bio_endio(struct bio *bio) { }
b9147dd1	374	static inline void blk_throtl_stat_add(struct request *rq, u64 time) { }
297e3d85	375	#endif
bc9fcbf9	376
51d798cd	377	struct bio __blk_queue_bounce(struct bio bio, struct request_queue *q);
9bb33f24 CH	378
	379	static inline bool blk_queue_may_bounce(struct request_queue *q)
	380	{
	381	return IS_ENABLED(CONFIG_BOUNCE) &&
	382	q->limits.bounce == BLK_BOUNCE_HIGH &&
	383	max_low_pfn >= max_pfn;
	384	}
	385
51d798cd CH	386	static inline struct bio blk_queue_bounce(struct bio bio,
51d798cd CH	387	struct request_queue *q)
3bce016a	388	{
51d798cd CH	389	if (unlikely(blk_queue_may_bounce(q) && bio_has_data(bio)))
	390	return __blk_queue_bounce(bio, q);
	391	return bio;
3bce016a	392	}
3bce016a	393
bf505456	394	#ifdef CONFIG_BLK_DEV_ZONED
5d400665	395	void disk_free_zone_bitmaps(struct gendisk *disk);
b3c72f81	396	void disk_clear_zone_settings(struct gendisk *disk);
bf505456	397	#else
5d400665	398	static inline void disk_free_zone_bitmaps(struct gendisk *disk) {}
b3c72f81	399	static inline void disk_clear_zone_settings(struct gendisk *disk) {}
bf505456 DLM	400	#endif
bf505456 DLM	401
7c3f828b CH	402	int blk_alloc_ext_minor(void);
7c3f828b CH	403	void blk_free_ext_minor(unsigned int minor);
581e2600 CH	404	#define ADDPART_FLAG_NONE 0
	405	#define ADDPART_FLAG_RAID 1
	406	#define ADDPART_FLAG_WHOLEDISK 2
7f6be376 CH	407	int bdev_add_partition(struct gendisk *disk, int partno, sector_t start,
7f6be376 CH	408	sector_t length);
926fbb16	409	int bdev_del_partition(struct gendisk *disk, int partno);
3d2e7989 CH	410	int bdev_resize_partition(struct gendisk *disk, int partno, sector_t start,
3d2e7989 CH	411	sector_t length);
e7243285	412	void blk_drop_partitions(struct gendisk *disk);
581e2600	413
83794367 DLM	414	void bdev_set_nr_sectors(struct block_device *bdev, sector_t sectors);
83794367 DLM	415
6f8191fd CH	416	struct gendisk __alloc_disk_node(struct request_queue q, int node_id,
	417	struct lock_class_key *lkclass);
	418
e4581105	419	int bio_add_hw_page(struct request_queue q, struct bio bio,
130879f1	420	struct page *page, unsigned int len, unsigned int offset,
e4581105	421	unsigned int max_sectors, bool *same_page);
130879f1	422
80bd4a7a	423	struct request_queue *blk_alloc_queue(int node_id);
704b914f	424
0f77b29a	425	int disk_scan_partitions(struct gendisk *disk, fmode_t mode);
da7ba729	426
92e7755e	427	int disk_alloc_events(struct gendisk *disk);
d5870edf CH	428	void disk_add_events(struct gendisk *disk);
	429	void disk_del_events(struct gendisk *disk);
	430	void disk_release_events(struct gendisk *disk);
926597ff CH	431	void disk_block_events(struct gendisk *disk);
	432	void disk_unblock_events(struct gendisk *disk);
	433	void disk_flush_events(struct gendisk *disk, unsigned int mask);
2bc8cda5 CH	434	extern struct device_attribute dev_attr_events;
	435	extern struct device_attribute dev_attr_events_async;
	436	extern struct device_attribute dev_attr_events_poll_msecs;
d5870edf	437
cc5c516d CH	438	extern struct attribute_group blk_trace_attr_group;
cc5c516d CH	439
8a709512	440	long blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg);
84b8514b CH	441	long compat_blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg);
84b8514b CH	442
cd82cca7 CH	443	extern const struct address_space_operations def_blk_aops;
cd82cca7 CH	444
22d0c408	445	int disk_register_independent_access_ranges(struct gendisk *disk);
a2247f19 DLM	446	void disk_unregister_independent_access_ranges(struct gendisk *disk);
a2247f19 DLM	447
06c8c691 CH	448	#ifdef CONFIG_FAIL_MAKE_REQUEST
	449	bool should_fail_request(struct block_device *part, unsigned int bytes);
	450	#else /* CONFIG_FAIL_MAKE_REQUEST */
	451	static inline bool should_fail_request(struct block_device *part,
	452	unsigned int bytes)
	453	{
	454	return false;
	455	}
	456	#endif /* CONFIG_FAIL_MAKE_REQUEST */
	457
0a467d0f JA	458	/*
	459	* Optimized request reference counting. Ideally we'd make timeouts be more
	460	* clever, as that's the only reason we need references at all... But until
	461	* this happens, this is faster than using refcount_t. Also see:
	462	*
	463	* abc54d634334 ("io_uring: switch to atomic_t for io_kiocb reference count")
	464	*/
	465	#define req_ref_zero_or_close_to_overflow(req) \
	466	((unsigned int) atomic_read(&(req->ref)) + 127u <= 127u)
	467
	468	static inline bool req_ref_inc_not_zero(struct request *req)
	469	{
	470	return atomic_inc_not_zero(&req->ref);
	471	}
	472
	473	static inline bool req_ref_put_and_test(struct request *req)
	474	{
	475	WARN_ON_ONCE(req_ref_zero_or_close_to_overflow(req));
	476	return atomic_dec_and_test(&req->ref);
	477	}
	478
	479	static inline void req_ref_set(struct request *req, int value)
	480	{
	481	atomic_set(&req->ref, value);
	482	}
	483
	484	static inline int req_ref_read(struct request *req)
	485	{
	486	return atomic_read(&req->ref);
	487	}
	488
bc9fcbf9	489	#endif /* BLK_INTERNAL_H */