[linux.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 {
	spinlock_t		mq_flush_lock;
	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];
	unsigned long		flush_data_in_flight;
	struct request		*flush_rq;
};

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);

bool bvec_try_merge_hw_page(struct request_queue *q, struct bio_vec *bv,
		struct page *page, unsigned len, unsigned offset,
		bool *same_page);

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);

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)

bool 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);
int blkdev_report_zones_ioctl(struct block_device *bdev, unsigned int cmd,
		unsigned long arg);
int blkdev_zone_mgmt_ioctl(struct block_device *bdev, blk_mode_t mode,
		unsigned int cmd, unsigned long arg);
#else /* CONFIG_BLK_DEV_ZONED */
static inline void disk_free_zone_bitmaps(struct gendisk *disk) {}
static inline int blkdev_report_zones_ioctl(struct block_device *bdev,
		unsigned int cmd, unsigned long arg)
{
	return -ENOTTY;
}
static inline int blkdev_zone_mgmt_ioctl(struct block_device *bdev,
		blk_mode_t mode, unsigned int cmd, unsigned long arg)
{
	return -ENOTTY;
}
#endif /* CONFIG_BLK_DEV_ZONED */

struct block_device *bdev_alloc(struct gendisk *disk, u8 partno);
void bdev_add(struct block_device *bdev, dev_t dev);

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 drop_partition(struct block_device *part);

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);

/*
 * Clean up a page appropriately, where the page may be pinned, may have a
 * ref taken on it or neither.
 */
static inline void bio_release_page(struct bio *bio, struct page *page)
{
	if (bio_flagged(bio, BIO_PAGE_PINNED))
		unpin_user_page(page);
}

struct request_queue *blk_alloc_queue(int node_id);

int disk_scan_partitions(struct gendisk *disk, blk_mode_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;

blk_mode_t file_to_blk_mode(struct file *file);
int truncate_bdev_range(struct block_device *bdev, blk_mode_t mode,
		loff_t lstart, loff_t lend);
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 {
b175c867	18	spinlock_t mq_flush_lock;
7c94e1c1 ML	19	unsigned int flush_pending_idx:1;
7c94e1c1 ML	20	unsigned int flush_running_idx:1;
8d699663	21	blk_status_t rq_status;
7c94e1c1 ML	22	unsigned long flush_pending_since;
7c94e1c1 ML	23	struct list_head flush_queue[2];
b175c867	24	unsigned long flush_data_in_flight;
7c94e1c1	25	struct request *flush_rq;
7c94e1c1 ML	26	};
7c94e1c1 ML	27
a9ed27a7	28	bool is_flush_rq(struct request *req);
8d699663	29
754a1572 GJ	30	struct blk_flush_queue *blk_alloc_flush_queue(int node, int cmd_size,
754a1572 GJ	31	gfp_t flags);
f70ced09	32	void blk_free_flush_queue(struct blk_flush_queue *q);
f3552655	33
3ef28e83	34	void blk_freeze_queue(struct request_queue *q);
aec89dc5	35	void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
8e141f9e	36	void blk_queue_start_drain(struct request_queue *q);
c98cb5bb	37	int __bio_queue_enter(struct request_queue q, struct bio bio);
3f98c753	38	void submit_bio_noacct_nocheck(struct bio *bio);
c98cb5bb JA	39
	40	static inline bool blk_try_enter_queue(struct request_queue *q, bool pm)
	41	{
	42	rcu_read_lock();
	43	if (!percpu_ref_tryget_live_rcu(&q->q_usage_counter))
	44	goto fail;
	45
	46	/*
	47	* The code that increments the pm_only counter must ensure that the
	48	* counter is globally visible before the queue is unfrozen.
	49	*/
	50	if (blk_queue_pm_only(q) &&
	51	(!pm \|\| queue_rpm_status(q) == RPM_SUSPENDED))
	52	goto fail_put;
	53
	54	rcu_read_unlock();
	55	return true;
	56
	57	fail_put:
	58	blk_queue_exit(q);
	59	fail:
	60	rcu_read_unlock();
	61	return false;
	62	}
	63
	64	static inline int bio_queue_enter(struct bio *bio)
	65	{
	66	struct request_queue *q = bdev_get_queue(bio->bi_bdev);
	67
	68	if (blk_try_enter_queue(q, false))
	69	return 0;
	70	return __bio_queue_enter(q, bio);
	71	}
3ef28e83	72
dc0b8a57	73	#define BIO_INLINE_VECS 4
7a800a20 CH	74	struct bio_vec bvec_alloc(mempool_t pool, unsigned short *nr_vecs,
	75	gfp_t gfp_mask);
	76	void bvec_free(mempool_t pool, struct bio_vec bv, unsigned short nr_vecs);
eec716a1	77
7c8998f7 JC	78	bool bvec_try_merge_hw_page(struct request_queue q, struct bio_vec bv,
	79	struct page *page, unsigned len, unsigned offset,
	80	bool *same_page);
	81
3dccdae5 CH	82	static inline bool biovec_phys_mergeable(struct request_queue *q,
3dccdae5 CH	83	struct bio_vec vec1, struct bio_vec vec2)
6a9f5f24	84	{
3dccdae5	85	unsigned long mask = queue_segment_boundary(q);
6e768461 CH	86	phys_addr_t addr1 = page_to_phys(vec1->bv_page) + vec1->bv_offset;
6e768461 CH	87	phys_addr_t addr2 = page_to_phys(vec2->bv_page) + vec2->bv_offset;
3dccdae5	88
f630a5d0 AP	89	/*
	90	* Merging adjacent physical pages may not work correctly under KMSAN
	91	* if their metadata pages aren't adjacent. Just disable merging.
	92	*/
	93	if (IS_ENABLED(CONFIG_KMSAN))
	94	return false;
	95
3dccdae5	96	if (addr1 + vec1->bv_len != addr2)
6a9f5f24	97	return false;
0383ad43	98	if (xen_domain() && !xen_biovec_phys_mergeable(vec1, vec2->bv_page))
6a9f5f24	99	return false;
3dccdae5 CH	100	if ((addr1 \| mask) != ((addr2 + vec2->bv_len - 1) \| mask))
3dccdae5 CH	101	return false;
6a9f5f24 CH	102	return true;
	103	}
	104
aa261f20	105	static inline bool __bvec_gap_to_prev(const struct queue_limits *lim,
27ca1d4e CH	106	struct bio_vec *bprv, unsigned int offset)
27ca1d4e CH	107	{
c55ddd90 CH	108	return (offset & lim->virt_boundary_mask) \|\|
c55ddd90 CH	109	((bprv->bv_offset + bprv->bv_len) & lim->virt_boundary_mask);
27ca1d4e CH	110	}
	111
	112	/*
	113	* Check if adding a bio_vec after bprv with offset would create a gap in
	114	* the SG list. Most drivers don't care about this, but some do.
	115	*/
aa261f20	116	static inline bool bvec_gap_to_prev(const struct queue_limits *lim,
27ca1d4e CH	117	struct bio_vec *bprv, unsigned int offset)
27ca1d4e CH	118	{
c55ddd90	119	if (!lim->virt_boundary_mask)
27ca1d4e	120	return false;
c55ddd90	121	return __bvec_gap_to_prev(lim, bprv, offset);
27ca1d4e CH	122	}
27ca1d4e CH	123
badf7f64 CH	124	static inline bool rq_mergeable(struct request *rq)
	125	{
	126	if (blk_rq_is_passthrough(rq))
	127	return false;
	128
	129	if (req_op(rq) == REQ_OP_FLUSH)
	130	return false;
	131
	132	if (req_op(rq) == REQ_OP_WRITE_ZEROES)
	133	return false;
	134
	135	if (req_op(rq) == REQ_OP_ZONE_APPEND)
	136	return false;
	137
	138	if (rq->cmd_flags & REQ_NOMERGE_FLAGS)
	139	return false;
	140	if (rq->rq_flags & RQF_NOMERGE_FLAGS)
	141	return false;
	142
	143	return true;
	144	}
	145
	146	/*
	147	* There are two different ways to handle DISCARD merges:
	148	* 1) If max_discard_segments > 1, the driver treats every bio as a range and
	149	* send the bios to controller together. The ranges don't need to be
	150	* contiguous.
	151	* 2) Otherwise, the request will be normal read/write requests. The ranges
	152	* need to be contiguous.
	153	*/
	154	static inline bool blk_discard_mergable(struct request *req)
	155	{
	156	if (req_op(req) == REQ_OP_DISCARD &&
	157	queue_max_discard_segments(req->q) > 1)
	158	return true;
	159	return false;
	160	}
	161
49d24398 US	162	static inline unsigned int blk_rq_get_max_segments(struct request *rq)
	163	{
	164	if (req_op(rq) == REQ_OP_DISCARD)
	165	return queue_max_discard_segments(rq->q);
	166	return queue_max_segments(rq->q);
	167	}
	168
2a9336c4	169	static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
77e7ffd7	170	enum req_op op)
2a9336c4 CH	171	{
	172	if (unlikely(op == REQ_OP_DISCARD \|\| op == REQ_OP_SECURE_ERASE))
	173	return min(q->limits.max_discard_sectors,
	174	UINT_MAX >> SECTOR_SHIFT);
	175
	176	if (unlikely(op == REQ_OP_WRITE_ZEROES))
	177	return q->limits.max_write_zeroes_sectors;
	178
	179	return q->limits.max_sectors;
	180	}
	181
5a48fc14 DW	182	#ifdef CONFIG_BLK_DEV_INTEGRITY
5a48fc14 DW	183	void blk_flush_integrity(void);
7c20f116	184	bool __bio_integrity_endio(struct bio *);
ece841ab	185	void bio_integrity_free(struct bio *bio);
7c20f116 CH	186	static inline bool bio_integrity_endio(struct bio *bio)
	187	{
	188	if (bio_integrity(bio))
	189	return __bio_integrity_endio(bio);
	190	return true;
	191	}
43b729bf	192
92cf2fd1 CH	193	bool blk_integrity_merge_rq(struct request_queue , struct request ,
92cf2fd1 CH	194	struct request *);
d59da419 CH	195	bool blk_integrity_merge_bio(struct request_queue , struct request ,
d59da419 CH	196	struct bio *);
92cf2fd1	197
43b729bf CH	198	static inline bool integrity_req_gap_back_merge(struct request *req,
	199	struct bio *next)
	200	{
	201	struct bio_integrity_payload *bip = bio_integrity(req->bio);
	202	struct bio_integrity_payload *bip_next = bio_integrity(next);
	203
c55ddd90 CH	204	return bvec_gap_to_prev(&req->q->limits,
c55ddd90 CH	205	&bip->bip_vec[bip->bip_vcnt - 1],
43b729bf CH	206	bip_next->bip_vec[0].bv_offset);
	207	}
	208
	209	static inline bool integrity_req_gap_front_merge(struct request *req,
	210	struct bio *bio)
	211	{
	212	struct bio_integrity_payload *bip = bio_integrity(bio);
	213	struct bio_integrity_payload *bip_next = bio_integrity(req->bio);
	214
c55ddd90 CH	215	return bvec_gap_to_prev(&req->q->limits,
c55ddd90 CH	216	&bip->bip_vec[bip->bip_vcnt - 1],
43b729bf CH	217	bip_next->bip_vec[0].bv_offset);
43b729bf CH	218	}
581e2600	219
ff53cd52	220	extern const struct attribute_group blk_integrity_attr_group;
43b729bf	221	#else /* CONFIG_BLK_DEV_INTEGRITY */
92cf2fd1 CH	222	static inline bool blk_integrity_merge_rq(struct request_queue *rq,
	223	struct request r1, struct request r2)
	224	{
	225	return true;
	226	}
d59da419 CH	227	static inline bool blk_integrity_merge_bio(struct request_queue *rq,
	228	struct request r, struct bio b)
	229	{
	230	return true;
	231	}
43b729bf CH	232	static inline bool integrity_req_gap_back_merge(struct request *req,
	233	struct bio *next)
	234	{
	235	return false;
	236	}
	237	static inline bool integrity_req_gap_front_merge(struct request *req,
	238	struct bio *bio)
	239	{
	240	return false;
	241	}
	242
5a48fc14 DW	243	static inline void blk_flush_integrity(void)
	244	{
	245	}
7c20f116 CH	246	static inline bool bio_integrity_endio(struct bio *bio)
	247	{
	248	return true;
	249	}
ece841ab JT	250	static inline void bio_integrity_free(struct bio *bio)
	251	{
	252	}
43b729bf	253	#endif /* CONFIG_BLK_DEV_INTEGRITY */
8324aa91	254
0d2602ca	255	unsigned long blk_rq_timeout(unsigned long timeout);
87ee7b11	256	void blk_add_timer(struct request *req);
320ae51f	257
320ae51f	258	bool blk_attempt_plug_merge(struct request_queue q, struct bio bio,
0c5bcc92	259	unsigned int nr_segs);
bdc6a287 BW	260	bool blk_bio_list_merge(struct request_queue q, struct list_head list,
bdc6a287 BW	261	struct bio *bio, unsigned int nr_segs);
320ae51f	262
ba0ffdd8 JA	263	/*
	264	* Plug flush limits
	265	*/
	266	#define BLK_MAX_REQUEST_COUNT 32
	267	#define BLK_PLUG_FLUSH_SIZE (128 * 1024)
	268
158dbda0 TH	269	/*
	270	* Internal elevator interface
	271	*/
e8064021	272	#define ELV_ON_HASH(rq) ((rq)->rq_flags & RQF_HASHED)
158dbda0	273
360f2648	274	bool blk_insert_flush(struct request *rq);
dd831006	275
8237c01f	276	int elevator_switch(struct request_queue q, struct elevator_type new_e);
64b36075	277	void elevator_disable(struct request_queue *q);
0c6cb3a2	278	void elevator_exit(struct request_queue *q);
cecf5d87	279	int elv_register_queue(struct request_queue *q, bool uevent);
83d016ac BVA	280	void elv_unregister_queue(struct request_queue *q);
83d016ac BVA	281
3ad5cee5 CH	282	ssize_t part_size_show(struct device dev, struct device_attribute attr,
	283	char *buf);
	284	ssize_t part_stat_show(struct device dev, struct device_attribute attr,
	285	char *buf);
	286	ssize_t part_inflight_show(struct device dev, struct device_attribute attr,
	287	char *buf);
	288	ssize_t part_fail_show(struct device dev, struct device_attribute attr,
	289	char *buf);
	290	ssize_t part_fail_store(struct device dev, struct device_attribute attr,
	291	const char *buf, size_t count);
581d4e28 JA	292	ssize_t part_timeout_show(struct device , struct device_attribute , char *);
	293	ssize_t part_timeout_store(struct device , struct device_attribute ,
	294	const char *, size_t);
581d4e28	295
c55ddd90	296	static inline bool bio_may_exceed_limits(struct bio *bio,
aa261f20	297	const struct queue_limits *lim)
abd45c15 JA	298	{
	299	switch (bio_op(bio)) {
	300	case REQ_OP_DISCARD:
	301	case REQ_OP_SECURE_ERASE:
	302	case REQ_OP_WRITE_ZEROES:
abd45c15 JA	303	return true; /* non-trivial splitting decisions */
	304	default:
	305	break;
	306	}
	307
	308	/*
	309	* All drivers must accept single-segments bios that are <= PAGE_SIZE.
	310	* This is a quick and dirty check that relies on the fact that
	311	* bi_io_vec[0] is always valid if a bio has data. The check might
	312	* lead to occasional false negatives when bios are cloned, but compared
	313	* to the performance impact of cloned bios themselves the loop below
	314	* doesn't matter anyway.
	315	*/
c55ddd90	316	return lim->chunk_sectors \|\| bio->bi_vcnt != 1 \|\|
abd45c15 JA	317	bio->bi_io_vec->bv_len + bio->bi_io_vec->bv_offset > PAGE_SIZE;
	318	}
	319
aa261f20 BVA	320	struct bio __bio_split_to_limits(struct bio bio,
	321	const struct queue_limits *lim,
	322	unsigned int *nr_segs);
14ccb66b CH	323	int ll_back_merge_fn(struct request req, struct bio bio,
14ccb66b CH	324	unsigned int nr_segs);
fd2ef39c	325	bool blk_attempt_req_merge(struct request_queue q, struct request rq,
5e84ea3a	326	struct request *next);
e9cd19c0	327	unsigned int blk_recalc_rq_segments(struct request *rq);
80a761fd	328	void blk_rq_set_mixed_merge(struct request *rq);
050c8ea8	329	bool blk_rq_merge_ok(struct request rq, struct bio bio);
34fe7c05	330	enum elv_merge blk_try_merge(struct request rq, struct bio bio);
d6d48196	331
b3228254	332	void blk_set_default_limits(struct queue_limits *lim);
ff88972c AB	333	int blk_dev_init(void);
ff88972c AB	334
c2553b58 JA	335	/*
	336	* Contribute to IO statistics IFF:
	337	*
	338	* a) it's attached to a gendisk, and
48d9b0d4	339	* b) the queue had IO stats enabled when this request was started
c2553b58	340	*/
599d067d	341	static inline bool blk_do_io_stat(struct request *rq)
fb8ec18c	342	{
41fa7222	343	return (rq->rq_flags & RQF_IO_STAT) && !blk_rq_is_passthrough(rq);
be6bfe36 PB	344	}
be6bfe36 PB	345
450b7879	346	void update_io_ticks(struct block_device *part, unsigned long now, bool end);
fb8ec18c	347
6cf7677f CH	348	static inline void req_set_nomerge(struct request_queue q, struct request req)
	349	{
	350	req->cmd_flags \|= REQ_NOMERGE;
	351	if (req == q->last_merge)
	352	q->last_merge = NULL;
	353	}
	354
f2dbd76a TH	355	/*
	356	* Internal io_context interface
	357	*/
87dd1d63	358	struct io_cq ioc_find_get_icq(struct request_queue q);
eca5892a	359	struct io_cq ioc_lookup_icq(struct request_queue q);
5ef16305	360	#ifdef CONFIG_BLK_ICQ
7e5a8794	361	void ioc_clear_queue(struct request_queue *q);
5ef16305 CH	362	#else
	363	static inline void ioc_clear_queue(struct request_queue *q)
	364	{
	365	}
	366	#endif /* CONFIG_BLK_ICQ */
f2dbd76a	367
297e3d85 SL	368	#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
	369	extern ssize_t blk_throtl_sample_time_show(struct request_queue q, char page);
	370	extern ssize_t blk_throtl_sample_time_store(struct request_queue *q,
	371	const char *page, size_t count);
9e234eea	372	extern void blk_throtl_bio_endio(struct bio *bio);
b9147dd1	373	extern void blk_throtl_stat_add(struct request *rq, u64 time);
9e234eea SL	374	#else
9e234eea SL	375	static inline void blk_throtl_bio_endio(struct bio *bio) { }
b9147dd1	376	static inline void blk_throtl_stat_add(struct request *rq, u64 time) { }
297e3d85	377	#endif
bc9fcbf9	378
51d798cd	379	struct bio __blk_queue_bounce(struct bio bio, struct request_queue *q);
9bb33f24 CH	380
	381	static inline bool blk_queue_may_bounce(struct request_queue *q)
	382	{
	383	return IS_ENABLED(CONFIG_BOUNCE) &&
	384	q->limits.bounce == BLK_BOUNCE_HIGH &&
	385	max_low_pfn >= max_pfn;
	386	}
	387
51d798cd CH	388	static inline struct bio blk_queue_bounce(struct bio bio,
51d798cd CH	389	struct request_queue *q)
3bce016a	390	{
51d798cd CH	391	if (unlikely(blk_queue_may_bounce(q) && bio_has_data(bio)))
	392	return __blk_queue_bounce(bio, q);
	393	return bio;
3bce016a	394	}
3bce016a	395
bf505456	396	#ifdef CONFIG_BLK_DEV_ZONED
5d400665	397	void disk_free_zone_bitmaps(struct gendisk *disk);
5e4ea834 CH	398	int blkdev_report_zones_ioctl(struct block_device *bdev, unsigned int cmd,
5e4ea834 CH	399	unsigned long arg);
05bdb996	400	int blkdev_zone_mgmt_ioctl(struct block_device *bdev, blk_mode_t mode,
cfb42576 CH	401	unsigned int cmd, unsigned long arg);
cfb42576 CH	402	#else /* CONFIG_BLK_DEV_ZONED */
5d400665	403	static inline void disk_free_zone_bitmaps(struct gendisk *disk) {}
cfb42576	404	static inline int blkdev_report_zones_ioctl(struct block_device *bdev,
5e4ea834	405	unsigned int cmd, unsigned long arg)
cfb42576 CH	406	{
	407	return -ENOTTY;
	408	}
	409	static inline int blkdev_zone_mgmt_ioctl(struct block_device *bdev,
05bdb996	410	blk_mode_t mode, unsigned int cmd, unsigned long arg)
cfb42576 CH	411	{
	412	return -ENOTTY;
	413	}
	414	#endif /* CONFIG_BLK_DEV_ZONED */
	415
	416	struct block_device bdev_alloc(struct gendisk disk, u8 partno);
	417	void bdev_add(struct block_device *bdev, dev_t dev);
bf505456	418
7c3f828b CH	419	int blk_alloc_ext_minor(void);
7c3f828b CH	420	void blk_free_ext_minor(unsigned int minor);
581e2600 CH	421	#define ADDPART_FLAG_NONE 0
	422	#define ADDPART_FLAG_RAID 1
	423	#define ADDPART_FLAG_WHOLEDISK 2
7f6be376 CH	424	int bdev_add_partition(struct gendisk *disk, int partno, sector_t start,
7f6be376 CH	425	sector_t length);
926fbb16	426	int bdev_del_partition(struct gendisk *disk, int partno);
3d2e7989 CH	427	int bdev_resize_partition(struct gendisk *disk, int partno, sector_t start,
3d2e7989 CH	428	sector_t length);
eec1be4c	429	void drop_partition(struct block_device *part);
581e2600	430
83794367 DLM	431	void bdev_set_nr_sectors(struct block_device *bdev, sector_t sectors);
83794367 DLM	432
6f8191fd CH	433	struct gendisk __alloc_disk_node(struct request_queue q, int node_id,
	434	struct lock_class_key *lkclass);
	435
e4581105	436	int bio_add_hw_page(struct request_queue q, struct bio bio,
130879f1	437	struct page *page, unsigned int len, unsigned int offset,
e4581105	438	unsigned int max_sectors, bool *same_page);
130879f1	439
fd363244 DH	440	/*
	441	* Clean up a page appropriately, where the page may be pinned, may have a
	442	* ref taken on it or neither.
	443	*/
	444	static inline void bio_release_page(struct bio bio, struct page page)
	445	{
	446	if (bio_flagged(bio, BIO_PAGE_PINNED))
	447	unpin_user_page(page);
fd363244 DH	448	}
fd363244 DH	449
80bd4a7a	450	struct request_queue *blk_alloc_queue(int node_id);
704b914f	451
05bdb996	452	int disk_scan_partitions(struct gendisk *disk, blk_mode_t mode);
da7ba729	453
92e7755e	454	int disk_alloc_events(struct gendisk *disk);
d5870edf CH	455	void disk_add_events(struct gendisk *disk);
	456	void disk_del_events(struct gendisk *disk);
	457	void disk_release_events(struct gendisk *disk);
926597ff CH	458	void disk_block_events(struct gendisk *disk);
	459	void disk_unblock_events(struct gendisk *disk);
	460	void disk_flush_events(struct gendisk *disk, unsigned int mask);
2bc8cda5 CH	461	extern struct device_attribute dev_attr_events;
	462	extern struct device_attribute dev_attr_events_async;
	463	extern struct device_attribute dev_attr_events_poll_msecs;
d5870edf	464
cc5c516d CH	465	extern struct attribute_group blk_trace_attr_group;
cc5c516d CH	466
05bdb996 CH	467	blk_mode_t file_to_blk_mode(struct file *file);
	468	int truncate_bdev_range(struct block_device *bdev, blk_mode_t mode,
	469	loff_t lstart, loff_t lend);
8a709512	470	long blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg);
84b8514b CH	471	long compat_blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg);
84b8514b CH	472
cd82cca7 CH	473	extern const struct address_space_operations def_blk_aops;
cd82cca7 CH	474
22d0c408	475	int disk_register_independent_access_ranges(struct gendisk *disk);
a2247f19 DLM	476	void disk_unregister_independent_access_ranges(struct gendisk *disk);
a2247f19 DLM	477
06c8c691 CH	478	#ifdef CONFIG_FAIL_MAKE_REQUEST
	479	bool should_fail_request(struct block_device *part, unsigned int bytes);
	480	#else /* CONFIG_FAIL_MAKE_REQUEST */
	481	static inline bool should_fail_request(struct block_device *part,
	482	unsigned int bytes)
	483	{
	484	return false;
	485	}
	486	#endif /* CONFIG_FAIL_MAKE_REQUEST */
	487
0a467d0f JA	488	/*
	489	* Optimized request reference counting. Ideally we'd make timeouts be more
	490	* clever, as that's the only reason we need references at all... But until
	491	* this happens, this is faster than using refcount_t. Also see:
	492	*
	493	* abc54d634334 ("io_uring: switch to atomic_t for io_kiocb reference count")
	494	*/
	495	#define req_ref_zero_or_close_to_overflow(req) \
	496	((unsigned int) atomic_read(&(req->ref)) + 127u <= 127u)
	497
	498	static inline bool req_ref_inc_not_zero(struct request *req)
	499	{
	500	return atomic_inc_not_zero(&req->ref);
	501	}
	502
	503	static inline bool req_ref_put_and_test(struct request *req)
	504	{
	505	WARN_ON_ONCE(req_ref_zero_or_close_to_overflow(req));
	506	return atomic_dec_and_test(&req->ref);
	507	}
	508
	509	static inline void req_ref_set(struct request *req, int value)
	510	{
	511	atomic_set(&req->ref, value);
	512	}
	513
	514	static inline int req_ref_read(struct request *req)
	515	{
	516	return atomic_read(&req->ref);
	517	}
	518
bc9fcbf9	519	#endif /* BLK_INTERNAL_H */