[linux.git] / block / blk-lib.c

// SPDX-License-Identifier: GPL-2.0
/*
 * Functions related to generic helpers functions
 */
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/bio.h>
#include <linux/blkdev.h>
#include <linux/scatterlist.h>

#include "blk.h"

struct bio *blk_next_bio(struct bio *bio, unsigned int nr_pages, gfp_t gfp)
{
	struct bio *new = bio_alloc(gfp, nr_pages);

	if (bio) {
		bio_chain(bio, new);
		submit_bio(bio);
	}

	return new;
}

int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, int flags,
		struct bio **biop)
{
	struct request_queue *q = bdev_get_queue(bdev);
	struct bio *bio = *biop;
	unsigned int op;
	sector_t bs_mask, part_offset = 0;

	if (!q)
		return -ENXIO;

	if (bdev_read_only(bdev))
		return -EPERM;

	if (flags & BLKDEV_DISCARD_SECURE) {
		if (!blk_queue_secure_erase(q))
			return -EOPNOTSUPP;
		op = REQ_OP_SECURE_ERASE;
	} else {
		if (!blk_queue_discard(q))
			return -EOPNOTSUPP;
		op = REQ_OP_DISCARD;
	}

	/* In case the discard granularity isn't set by buggy device driver */
	if (WARN_ON_ONCE(!q->limits.discard_granularity)) {
		char dev_name[BDEVNAME_SIZE];

		bdevname(bdev, dev_name);
		pr_err_ratelimited("%s: Error: discard_granularity is 0.\n", dev_name);
		return -EOPNOTSUPP;
	}

	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	if ((sector | nr_sects) & bs_mask)
		return -EINVAL;

	if (!nr_sects)
		return -EINVAL;

	/* In case the discard request is in a partition */
	if (bdev_is_partition(bdev))
		part_offset = bdev->bd_start_sect;

	while (nr_sects) {
		sector_t granularity_aligned_lba, req_sects;
		sector_t sector_mapped = sector + part_offset;

		granularity_aligned_lba = round_up(sector_mapped,
				q->limits.discard_granularity >> SECTOR_SHIFT);

		/*
		 * Check whether the discard bio starts at a discard_granularity
		 * aligned LBA,
		 * - If no: set (granularity_aligned_lba - sector_mapped) to
		 *   bi_size of the first split bio, then the second bio will
		 *   start at a discard_granularity aligned LBA on the device.
		 * - If yes: use bio_aligned_discard_max_sectors() as the max
		 *   possible bi_size of the first split bio. Then when this bio
		 *   is split in device drive, the split ones are very probably
		 *   to be aligned to discard_granularity of the device's queue.
		 */
		if (granularity_aligned_lba == sector_mapped)
			req_sects = min_t(sector_t, nr_sects,
					  bio_aligned_discard_max_sectors(q));
		else
			req_sects = min_t(sector_t, nr_sects,
					  granularity_aligned_lba - sector_mapped);

		WARN_ON_ONCE((req_sects << 9) > UINT_MAX);

		bio = blk_next_bio(bio, 0, gfp_mask);
		bio->bi_iter.bi_sector = sector;
		bio_set_dev(bio, bdev);
		bio_set_op_attrs(bio, op, 0);

		bio->bi_iter.bi_size = req_sects << 9;
		sector += req_sects;
		nr_sects -= req_sects;

		/*
		 * We can loop for a long time in here, if someone does
		 * full device discards (like mkfs). Be nice and allow
		 * us to schedule out to avoid softlocking if preempt
		 * is disabled.
		 */
		cond_resched();
	}

	*biop = bio;
	return 0;
}
EXPORT_SYMBOL(__blkdev_issue_discard);

/**
 * blkdev_issue_discard - queue a discard
 * @bdev:	blockdev to issue discard for
 * @sector:	start sector
 * @nr_sects:	number of sectors to discard
 * @gfp_mask:	memory allocation flags (for bio_alloc)
 * @flags:	BLKDEV_DISCARD_* flags to control behaviour
 *
 * Description:
 *    Issue a discard request for the sectors in question.
 */
int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
{
	struct bio *bio = NULL;
	struct blk_plug plug;
	int ret;

	blk_start_plug(&plug);
	ret = __blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, flags,
			&bio);
	if (!ret && bio) {
		ret = submit_bio_wait(bio);
		if (ret == -EOPNOTSUPP)
			ret = 0;
		bio_put(bio);
	}
	blk_finish_plug(&plug);

	return ret;
}
EXPORT_SYMBOL(blkdev_issue_discard);

/**
 * __blkdev_issue_write_same - generate number of bios with same page
 * @bdev:	target blockdev
 * @sector:	start sector
 * @nr_sects:	number of sectors to write
 * @gfp_mask:	memory allocation flags (for bio_alloc)
 * @page:	page containing data to write
 * @biop:	pointer to anchor bio
 *
 * Description:
 *  Generate and issue number of bios(REQ_OP_WRITE_SAME) with same page.
 */
static int __blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, struct page *page,
		struct bio **biop)
{
	struct request_queue *q = bdev_get_queue(bdev);
	unsigned int max_write_same_sectors;
	struct bio *bio = *biop;
	sector_t bs_mask;

	if (!q)
		return -ENXIO;

	if (bdev_read_only(bdev))
		return -EPERM;

	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	if ((sector | nr_sects) & bs_mask)
		return -EINVAL;

	if (!bdev_write_same(bdev))
		return -EOPNOTSUPP;

	/* Ensure that max_write_same_sectors doesn't overflow bi_size */
	max_write_same_sectors = bio_allowed_max_sectors(q);

	while (nr_sects) {
		bio = blk_next_bio(bio, 1, gfp_mask);
		bio->bi_iter.bi_sector = sector;
		bio_set_dev(bio, bdev);
		bio->bi_vcnt = 1;
		bio->bi_io_vec->bv_page = page;
		bio->bi_io_vec->bv_offset = 0;
		bio->bi_io_vec->bv_len = bdev_logical_block_size(bdev);
		bio_set_op_attrs(bio, REQ_OP_WRITE_SAME, 0);

		if (nr_sects > max_write_same_sectors) {
			bio->bi_iter.bi_size = max_write_same_sectors << 9;
			nr_sects -= max_write_same_sectors;
			sector += max_write_same_sectors;
		} else {
			bio->bi_iter.bi_size = nr_sects << 9;
			nr_sects = 0;
		}
		cond_resched();
	}

	*biop = bio;
	return 0;
}

/**
 * blkdev_issue_write_same - queue a write same operation
 * @bdev:	target blockdev
 * @sector:	start sector
 * @nr_sects:	number of sectors to write
 * @gfp_mask:	memory allocation flags (for bio_alloc)
 * @page:	page containing data
 *
 * Description:
 *    Issue a write same request for the sectors in question.
 */
int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
				sector_t nr_sects, gfp_t gfp_mask,
				struct page *page)
{
	struct bio *bio = NULL;
	struct blk_plug plug;
	int ret;

	blk_start_plug(&plug);
	ret = __blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask, page,
			&bio);
	if (ret == 0 && bio) {
		ret = submit_bio_wait(bio);
		bio_put(bio);
	}
	blk_finish_plug(&plug);
	return ret;
}
EXPORT_SYMBOL(blkdev_issue_write_same);

static int __blkdev_issue_write_zeroes(struct block_device *bdev,
		sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
		struct bio **biop, unsigned flags)
{
	struct bio *bio = *biop;
	unsigned int max_write_zeroes_sectors;
	struct request_queue *q = bdev_get_queue(bdev);

	if (!q)
		return -ENXIO;

	if (bdev_read_only(bdev))
		return -EPERM;

	/* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
	max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev);

	if (max_write_zeroes_sectors == 0)
		return -EOPNOTSUPP;

	while (nr_sects) {
		bio = blk_next_bio(bio, 0, gfp_mask);
		bio->bi_iter.bi_sector = sector;
		bio_set_dev(bio, bdev);
		bio->bi_opf = REQ_OP_WRITE_ZEROES;
		if (flags & BLKDEV_ZERO_NOUNMAP)
			bio->bi_opf |= REQ_NOUNMAP;

		if (nr_sects > max_write_zeroes_sectors) {
			bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
			nr_sects -= max_write_zeroes_sectors;
			sector += max_write_zeroes_sectors;
		} else {
			bio->bi_iter.bi_size = nr_sects << 9;
			nr_sects = 0;
		}
		cond_resched();
	}

	*biop = bio;
	return 0;
}

/*
 * Convert a number of 512B sectors to a number of pages.
 * The result is limited to a number of pages that can fit into a BIO.
 * Also make sure that the result is always at least 1 (page) for the cases
 * where nr_sects is lower than the number of sectors in a page.
 */
static unsigned int __blkdev_sectors_to_bio_pages(sector_t nr_sects)
{
	sector_t pages = DIV_ROUND_UP_SECTOR_T(nr_sects, PAGE_SIZE / 512);

	return min(pages, (sector_t)BIO_MAX_PAGES);
}

static int __blkdev_issue_zero_pages(struct block_device *bdev,
		sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
		struct bio **biop)
{
	struct request_queue *q = bdev_get_queue(bdev);
	struct bio *bio = *biop;
	int bi_size = 0;
	unsigned int sz;

	if (!q)
		return -ENXIO;

	if (bdev_read_only(bdev))
		return -EPERM;

	while (nr_sects != 0) {
		bio = blk_next_bio(bio, __blkdev_sectors_to_bio_pages(nr_sects),
				   gfp_mask);
		bio->bi_iter.bi_sector = sector;
		bio_set_dev(bio, bdev);
		bio_set_op_attrs(bio, REQ_OP_WRITE, 0);

		while (nr_sects != 0) {
			sz = min((sector_t) PAGE_SIZE, nr_sects << 9);
			bi_size = bio_add_page(bio, ZERO_PAGE(0), sz, 0);
			nr_sects -= bi_size >> 9;
			sector += bi_size >> 9;
			if (bi_size < sz)
				break;
		}
		cond_resched();
	}

	*biop = bio;
	return 0;
}

/**
 * __blkdev_issue_zeroout - generate number of zero filed write bios
 * @bdev:	blockdev to issue
 * @sector:	start sector
 * @nr_sects:	number of sectors to write
 * @gfp_mask:	memory allocation flags (for bio_alloc)
 * @biop:	pointer to anchor bio
 * @flags:	controls detailed behavior
 *
 * Description:
 *  Zero-fill a block range, either using hardware offload or by explicitly
 *  writing zeroes to the device.
 *
 *  If a device is using logical block provisioning, the underlying space will
 *  not be released if %flags contains BLKDEV_ZERO_NOUNMAP.
 *
 *  If %flags contains BLKDEV_ZERO_NOFALLBACK, the function will return
 *  -EOPNOTSUPP if no explicit hardware offload for zeroing is provided.
 */
int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
		unsigned flags)
{
	int ret;
	sector_t bs_mask;

	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	if ((sector | nr_sects) & bs_mask)
		return -EINVAL;

	ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects, gfp_mask,
			biop, flags);
	if (ret != -EOPNOTSUPP || (flags & BLKDEV_ZERO_NOFALLBACK))
		return ret;

	return __blkdev_issue_zero_pages(bdev, sector, nr_sects, gfp_mask,
					 biop);
}
EXPORT_SYMBOL(__blkdev_issue_zeroout);

/**
 * blkdev_issue_zeroout - zero-fill a block range
 * @bdev:	blockdev to write
 * @sector:	start sector
 * @nr_sects:	number of sectors to write
 * @gfp_mask:	memory allocation flags (for bio_alloc)
 * @flags:	controls detailed behavior
 *
 * Description:
 *  Zero-fill a block range, either using hardware offload or by explicitly
 *  writing zeroes to the device.  See __blkdev_issue_zeroout() for the
 *  valid values for %flags.
 */
int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, unsigned flags)
{
	int ret = 0;
	sector_t bs_mask;
	struct bio *bio;
	struct blk_plug plug;
	bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev);

	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	if ((sector | nr_sects) & bs_mask)
		return -EINVAL;

retry:
	bio = NULL;
	blk_start_plug(&plug);
	if (try_write_zeroes) {
		ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects,
						  gfp_mask, &bio, flags);
	} else if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
		ret = __blkdev_issue_zero_pages(bdev, sector, nr_sects,
						gfp_mask, &bio);
	} else {
		/* No zeroing offload support */
		ret = -EOPNOTSUPP;
	}
	if (ret == 0 && bio) {
		ret = submit_bio_wait(bio);
		bio_put(bio);
	}
	blk_finish_plug(&plug);
	if (ret && try_write_zeroes) {
		if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
			try_write_zeroes = false;
			goto retry;
		}
		if (!bdev_write_zeroes_sectors(bdev)) {
			/*
			 * Zeroing offload support was indicated, but the
			 * device reported ILLEGAL REQUEST (for some devices
			 * there is no non-destructive way to verify whether
			 * WRITE ZEROES is actually supported).
			 */
			ret = -EOPNOTSUPP;
		}
	}

	return ret;
}
EXPORT_SYMBOL(blkdev_issue_zeroout);
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
f31e7e40 DM	2	/*
	3	* Functions related to generic helpers functions
	4	*/
	5	#include <linux/kernel.h>
	6	#include <linux/module.h>
	7	#include <linux/bio.h>
	8	#include <linux/blkdev.h>
	9	#include <linux/scatterlist.h>
	10
	11	#include "blk.h"
	12
a2d6b3a2	13	struct bio blk_next_bio(struct bio bio, unsigned int nr_pages, gfp_t gfp)
f31e7e40	14	{
9082e87b CH	15	struct bio *new = bio_alloc(gfp, nr_pages);
	16
	17	if (bio) {
	18	bio_chain(bio, new);
4e49ea4a	19	submit_bio(bio);
9082e87b	20	}
5dba3089	21
9082e87b	22	return new;
f31e7e40 DM	23	}
f31e7e40 DM	24
38f25255	25	int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
288dab8a	26	sector_t nr_sects, gfp_t gfp_mask, int flags,
469e3216	27	struct bio **biop)
f31e7e40	28	{
f31e7e40	29	struct request_queue *q = bdev_get_queue(bdev);
38f25255	30	struct bio bio = biop;
ef295ecf	31	unsigned int op;
9b15d109	32	sector_t bs_mask, part_offset = 0;
f31e7e40 DM	33
	34	if (!q)
	35	return -ENXIO;
288dab8a	36
a13553c7 ID	37	if (bdev_read_only(bdev))
	38	return -EPERM;
	39
288dab8a CH	40	if (flags & BLKDEV_DISCARD_SECURE) {
	41	if (!blk_queue_secure_erase(q))
	42	return -EOPNOTSUPP;
	43	op = REQ_OP_SECURE_ERASE;
	44	} else {
	45	if (!blk_queue_discard(q))
	46	return -EOPNOTSUPP;
	47	op = REQ_OP_DISCARD;
	48	}
f31e7e40	49
b35fd742 CL	50	/* In case the discard granularity isn't set by buggy device driver */
	51	if (WARN_ON_ONCE(!q->limits.discard_granularity)) {
	52	char dev_name[BDEVNAME_SIZE];
	53
	54	bdevname(bdev, dev_name);
	55	pr_err_ratelimited("%s: Error: discard_granularity is 0.\n", dev_name);
	56	return -EOPNOTSUPP;
	57	}
	58
28b2be20 DW	59	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	60	if ((sector \| nr_sects) & bs_mask)
	61	return -EINVAL;
	62
ba5d7385 ML	63	if (!nr_sects)
ba5d7385 ML	64	return -EINVAL;
a22c4d7e	65
9b15d109	66	/* In case the discard request is in a partition */
fa01b1e9	67	if (bdev_is_partition(bdev))
29ff57c6	68	part_offset = bdev->bd_start_sect;
9b15d109	69
ba5d7385	70	while (nr_sects) {
9b15d109 CL	71	sector_t granularity_aligned_lba, req_sects;
	72	sector_t sector_mapped = sector + part_offset;
	73
	74	granularity_aligned_lba = round_up(sector_mapped,
	75	q->limits.discard_granularity >> SECTOR_SHIFT);
	76
	77	/*
	78	* Check whether the discard bio starts at a discard_granularity
	79	* aligned LBA,
	80	* - If no: set (granularity_aligned_lba - sector_mapped) to
	81	* bi_size of the first split bio, then the second bio will
	82	* start at a discard_granularity aligned LBA on the device.
	83	* - If yes: use bio_aligned_discard_max_sectors() as the max
	84	* possible bi_size of the first split bio. Then when this bio
	85	* is split in device drive, the split ones are very probably
	86	* to be aligned to discard_granularity of the device's queue.
	87	*/
	88	if (granularity_aligned_lba == sector_mapped)
	89	req_sects = min_t(sector_t, nr_sects,
	90	bio_aligned_discard_max_sectors(q));
	91	else
	92	req_sects = min_t(sector_t, nr_sects,
	93	granularity_aligned_lba - sector_mapped);
c6e66634	94
4800bf7b DC	95	WARN_ON_ONCE((req_sects << 9) > UINT_MAX);
4800bf7b DC	96
a2d6b3a2	97	bio = blk_next_bio(bio, 0, gfp_mask);
4f024f37	98	bio->bi_iter.bi_sector = sector;
74d46992	99	bio_set_dev(bio, bdev);
288dab8a	100	bio_set_op_attrs(bio, op, 0);
f31e7e40	101
4f024f37	102	bio->bi_iter.bi_size = req_sects << 9;
ba5d7385	103	sector += req_sects;
c6e66634	104	nr_sects -= req_sects;
f31e7e40	105
c8123f8c JA	106	/*
	107	* We can loop for a long time in here, if someone does
	108	* full device discards (like mkfs). Be nice and allow
	109	* us to schedule out to avoid softlocking if preempt
	110	* is disabled.
	111	*/
	112	cond_resched();
5dba3089	113	}
38f25255 CH	114
	115	*biop = bio;
	116	return 0;
	117	}
	118	EXPORT_SYMBOL(__blkdev_issue_discard);
	119
	120	/**
	121	* blkdev_issue_discard - queue a discard
	122	* @bdev: blockdev to issue discard for
	123	* @sector: start sector
	124	* @nr_sects: number of sectors to discard
	125	* @gfp_mask: memory allocation flags (for bio_alloc)
e554911c	126	* @flags: BLKDEV_DISCARD_* flags to control behaviour
38f25255 CH	127	*
	128	* Description:
	129	* Issue a discard request for the sectors in question.
	130	*/
	131	int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
	132	sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
	133	{
38f25255 CH	134	struct bio *bio = NULL;
	135	struct blk_plug plug;
	136	int ret;
	137
38f25255	138	blk_start_plug(&plug);
288dab8a	139	ret = __blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, flags,
38f25255	140	&bio);
bbd848e0	141	if (!ret && bio) {
4e49ea4a	142	ret = submit_bio_wait(bio);
48920ff2	143	if (ret == -EOPNOTSUPP)
bbd848e0	144	ret = 0;
05bd92dd	145	bio_put(bio);
bbd848e0	146	}
0cfbcafc	147	blk_finish_plug(&plug);
f31e7e40	148
bbd848e0	149	return ret;
f31e7e40 DM	150	}
f31e7e40 DM	151	EXPORT_SYMBOL(blkdev_issue_discard);
3f14d792	152
4363ac7c	153	/**
e73c23ff	154	* __blkdev_issue_write_same - generate number of bios with same page
4363ac7c MP	155	* @bdev: target blockdev
	156	* @sector: start sector
	157	* @nr_sects: number of sectors to write
	158	* @gfp_mask: memory allocation flags (for bio_alloc)
	159	* @page: page containing data to write
e73c23ff	160	* @biop: pointer to anchor bio
4363ac7c MP	161	*
4363ac7c MP	162	* Description:
e73c23ff	163	* Generate and issue number of bios(REQ_OP_WRITE_SAME) with same page.
4363ac7c	164	*/
e73c23ff CK	165	static int __blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
	166	sector_t nr_sects, gfp_t gfp_mask, struct page *page,
	167	struct bio **biop)
4363ac7c	168	{
4363ac7c MP	169	struct request_queue *q = bdev_get_queue(bdev);
4363ac7c MP	170	unsigned int max_write_same_sectors;
e73c23ff	171	struct bio bio = biop;
28b2be20	172	sector_t bs_mask;
4363ac7c MP	173
	174	if (!q)
	175	return -ENXIO;
	176
a13553c7 ID	177	if (bdev_read_only(bdev))
	178	return -EPERM;
	179
28b2be20 DW	180	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	181	if ((sector \| nr_sects) & bs_mask)
	182	return -EINVAL;
	183
e73c23ff CK	184	if (!bdev_write_same(bdev))
	185	return -EOPNOTSUPP;
	186
b49a0871	187	/* Ensure that max_write_same_sectors doesn't overflow bi_size */
34ffec60	188	max_write_same_sectors = bio_allowed_max_sectors(q);
4363ac7c	189
4363ac7c	190	while (nr_sects) {
a2d6b3a2	191	bio = blk_next_bio(bio, 1, gfp_mask);
4f024f37	192	bio->bi_iter.bi_sector = sector;
74d46992	193	bio_set_dev(bio, bdev);
4363ac7c MP	194	bio->bi_vcnt = 1;
	195	bio->bi_io_vec->bv_page = page;
	196	bio->bi_io_vec->bv_offset = 0;
	197	bio->bi_io_vec->bv_len = bdev_logical_block_size(bdev);
95fe6c1a	198	bio_set_op_attrs(bio, REQ_OP_WRITE_SAME, 0);
4363ac7c MP	199
4363ac7c MP	200	if (nr_sects > max_write_same_sectors) {
4f024f37	201	bio->bi_iter.bi_size = max_write_same_sectors << 9;
4363ac7c MP	202	nr_sects -= max_write_same_sectors;
	203	sector += max_write_same_sectors;
	204	} else {
4f024f37	205	bio->bi_iter.bi_size = nr_sects << 9;
4363ac7c MP	206	nr_sects = 0;
4363ac7c MP	207	}
e73c23ff	208	cond_resched();
4363ac7c MP	209	}
4363ac7c MP	210
e73c23ff CK	211	*biop = bio;
	212	return 0;
	213	}
	214
	215	/**
	216	* blkdev_issue_write_same - queue a write same operation
	217	* @bdev: target blockdev
	218	* @sector: start sector
	219	* @nr_sects: number of sectors to write
	220	* @gfp_mask: memory allocation flags (for bio_alloc)
	221	* @page: page containing data
	222	*
	223	* Description:
	224	* Issue a write same request for the sectors in question.
	225	*/
	226	int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
	227	sector_t nr_sects, gfp_t gfp_mask,
	228	struct page *page)
	229	{
	230	struct bio *bio = NULL;
	231	struct blk_plug plug;
	232	int ret;
	233
	234	blk_start_plug(&plug);
	235	ret = __blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask, page,
	236	&bio);
	237	if (ret == 0 && bio) {
4e49ea4a	238	ret = submit_bio_wait(bio);
05bd92dd ST	239	bio_put(bio);
05bd92dd ST	240	}
e73c23ff	241	blk_finish_plug(&plug);
3f40bf2c	242	return ret;
4363ac7c MP	243	}
	244	EXPORT_SYMBOL(blkdev_issue_write_same);
	245
a6f0788e CK	246	static int __blkdev_issue_write_zeroes(struct block_device *bdev,
a6f0788e CK	247	sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
d928be9f	248	struct bio **biop, unsigned flags)
a6f0788e CK	249	{
	250	struct bio bio = biop;
	251	unsigned int max_write_zeroes_sectors;
	252	struct request_queue *q = bdev_get_queue(bdev);
	253
	254	if (!q)
	255	return -ENXIO;
	256
a13553c7 ID	257	if (bdev_read_only(bdev))
	258	return -EPERM;
	259
a6f0788e CK	260	/* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
	261	max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev);
	262
	263	if (max_write_zeroes_sectors == 0)
	264	return -EOPNOTSUPP;
	265
	266	while (nr_sects) {
a2d6b3a2	267	bio = blk_next_bio(bio, 0, gfp_mask);
a6f0788e	268	bio->bi_iter.bi_sector = sector;
74d46992	269	bio_set_dev(bio, bdev);
d928be9f CH	270	bio->bi_opf = REQ_OP_WRITE_ZEROES;
	271	if (flags & BLKDEV_ZERO_NOUNMAP)
	272	bio->bi_opf \|= REQ_NOUNMAP;
a6f0788e CK	273
	274	if (nr_sects > max_write_zeroes_sectors) {
	275	bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
	276	nr_sects -= max_write_zeroes_sectors;
	277	sector += max_write_zeroes_sectors;
	278	} else {
	279	bio->bi_iter.bi_size = nr_sects << 9;
	280	nr_sects = 0;
	281	}
	282	cond_resched();
	283	}
	284
	285	*biop = bio;
	286	return 0;
	287	}
	288
615d22a5 DLM	289	/*
	290	* Convert a number of 512B sectors to a number of pages.
	291	* The result is limited to a number of pages that can fit into a BIO.
	292	* Also make sure that the result is always at least 1 (page) for the cases
	293	* where nr_sects is lower than the number of sectors in a page.
	294	*/
	295	static unsigned int __blkdev_sectors_to_bio_pages(sector_t nr_sects)
	296	{
09c2c359	297	sector_t pages = DIV_ROUND_UP_SECTOR_T(nr_sects, PAGE_SIZE / 512);
615d22a5	298
09c2c359	299	return min(pages, (sector_t)BIO_MAX_PAGES);
615d22a5 DLM	300	}
615d22a5 DLM	301
425a4dba ID	302	static int __blkdev_issue_zero_pages(struct block_device *bdev,
	303	sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
	304	struct bio **biop)
	305	{
	306	struct request_queue *q = bdev_get_queue(bdev);
	307	struct bio bio = biop;
	308	int bi_size = 0;
	309	unsigned int sz;
	310
	311	if (!q)
	312	return -ENXIO;
	313
a13553c7 ID	314	if (bdev_read_only(bdev))
	315	return -EPERM;
	316
425a4dba	317	while (nr_sects != 0) {
a2d6b3a2 DLM	318	bio = blk_next_bio(bio, __blkdev_sectors_to_bio_pages(nr_sects),
a2d6b3a2 DLM	319	gfp_mask);
425a4dba ID	320	bio->bi_iter.bi_sector = sector;
	321	bio_set_dev(bio, bdev);
	322	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
	323
	324	while (nr_sects != 0) {
	325	sz = min((sector_t) PAGE_SIZE, nr_sects << 9);
	326	bi_size = bio_add_page(bio, ZERO_PAGE(0), sz, 0);
	327	nr_sects -= bi_size >> 9;
	328	sector += bi_size >> 9;
	329	if (bi_size < sz)
	330	break;
	331	}
	332	cond_resched();
	333	}
	334
	335	*biop = bio;
	336	return 0;
	337	}
	338
3f14d792	339	/**
e73c23ff	340	* __blkdev_issue_zeroout - generate number of zero filed write bios
3f14d792 DM	341	* @bdev: blockdev to issue
	342	* @sector: start sector
	343	* @nr_sects: number of sectors to write
	344	* @gfp_mask: memory allocation flags (for bio_alloc)
e73c23ff	345	* @biop: pointer to anchor bio
ee472d83	346	* @flags: controls detailed behavior
3f14d792 DM	347	*
3f14d792 DM	348	* Description:
ee472d83 CH	349	* Zero-fill a block range, either using hardware offload or by explicitly
	350	* writing zeroes to the device.
	351	*
	352	* If a device is using logical block provisioning, the underlying space will
	353	* not be released if %flags contains BLKDEV_ZERO_NOUNMAP.
cb365b96 CH	354	*
	355	* If %flags contains BLKDEV_ZERO_NOFALLBACK, the function will return
	356	* -EOPNOTSUPP if no explicit hardware offload for zeroing is provided.
3f14d792	357	*/
e73c23ff CK	358	int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
e73c23ff CK	359	sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
ee472d83	360	unsigned flags)
3f14d792	361	{
18edc8ea	362	int ret;
28b2be20 DW	363	sector_t bs_mask;
	364
	365	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	366	if ((sector \| nr_sects) & bs_mask)
	367	return -EINVAL;
3f14d792	368
a6f0788e	369	ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects, gfp_mask,
d928be9f	370	biop, flags);
cb365b96	371	if (ret != -EOPNOTSUPP \|\| (flags & BLKDEV_ZERO_NOFALLBACK))
425a4dba	372	return ret;
3f14d792	373
425a4dba ID	374	return __blkdev_issue_zero_pages(bdev, sector, nr_sects, gfp_mask,
425a4dba ID	375	biop);
3f14d792	376	}
e73c23ff	377	EXPORT_SYMBOL(__blkdev_issue_zeroout);
579e8f3c MP	378
	379	/**
	380	* blkdev_issue_zeroout - zero-fill a block range
	381	* @bdev: blockdev to write
	382	* @sector: start sector
	383	* @nr_sects: number of sectors to write
	384	* @gfp_mask: memory allocation flags (for bio_alloc)
ee472d83	385	* @flags: controls detailed behavior
579e8f3c MP	386	*
579e8f3c MP	387	* Description:
ee472d83 CH	388	* Zero-fill a block range, either using hardware offload or by explicitly
	389	* writing zeroes to the device. See __blkdev_issue_zeroout() for the
	390	* valid values for %flags.
579e8f3c	391	*/
579e8f3c	392	int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
ee472d83	393	sector_t nr_sects, gfp_t gfp_mask, unsigned flags)
579e8f3c	394	{
d5ce4c31 ID	395	int ret = 0;
	396	sector_t bs_mask;
	397	struct bio *bio;
e73c23ff	398	struct blk_plug plug;
d5ce4c31	399	bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev);
d93ba7a5	400
d5ce4c31 ID	401	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	402	if ((sector \| nr_sects) & bs_mask)
	403	return -EINVAL;
	404
	405	retry:
	406	bio = NULL;
e73c23ff	407	blk_start_plug(&plug);
d5ce4c31 ID	408	if (try_write_zeroes) {
	409	ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects,
	410	gfp_mask, &bio, flags);
	411	} else if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
	412	ret = __blkdev_issue_zero_pages(bdev, sector, nr_sects,
	413	gfp_mask, &bio);
	414	} else {
	415	/* No zeroing offload support */
	416	ret = -EOPNOTSUPP;
	417	}
e73c23ff CK	418	if (ret == 0 && bio) {
	419	ret = submit_bio_wait(bio);
	420	bio_put(bio);
	421	}
	422	blk_finish_plug(&plug);
d5ce4c31 ID	423	if (ret && try_write_zeroes) {
	424	if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
	425	try_write_zeroes = false;
	426	goto retry;
	427	}
	428	if (!bdev_write_zeroes_sectors(bdev)) {
	429	/*
	430	* Zeroing offload support was indicated, but the
	431	* device reported ILLEGAL REQUEST (for some devices
	432	* there is no non-destructive way to verify whether
	433	* WRITE ZEROES is actually supported).
	434	*/
	435	ret = -EOPNOTSUPP;
	436	}
	437	}
579e8f3c	438
e73c23ff	439	return ret;
579e8f3c	440	}
3f14d792	441	EXPORT_SYMBOL(blkdev_issue_zeroout);