[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"

static struct bio *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 granularity;
	unsigned int op;
	int alignment;
	sector_t bs_mask;

	if (!q)
		return -ENXIO;

	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;
	}

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

	/* Zero-sector (unknown) and one-sector granularities are the same.  */
	granularity = max(q->limits.discard_granularity >> 9, 1U);
	alignment = (bdev_discard_alignment(bdev) >> 9) % granularity;

	while (nr_sects) {
		unsigned int req_sects;
		sector_t end_sect, tmp;

		/* Make sure bi_size doesn't overflow */
		req_sects = min_t(sector_t, nr_sects, UINT_MAX >> 9);

		/**
		 * If splitting a request, and the next starting sector would be
		 * misaligned, stop the discard at the previous aligned sector.
		 */
		end_sect = sector + req_sects;
		tmp = end_sect;
		if (req_sects < nr_sects &&
		    sector_div(tmp, granularity) != alignment) {
			end_sect = end_sect - alignment;
			sector_div(end_sect, granularity);
			end_sect = end_sect * granularity + alignment;
			req_sects = end_sect - sector;
		}

		bio = 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;
		nr_sects -= req_sects;
		sector = end_sect;

		/*
		 * 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;

	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 = UINT_MAX >> 9;

	while (nr_sects) {
		bio = 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;

	/* 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 = 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;

	while (nr_sects != 0) {
		bio = 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
4e49ea4a	13	static struct bio next_bio(struct bio bio, unsigned int nr_pages,
9082e87b	14	gfp_t gfp)
f31e7e40	15	{
9082e87b CH	16	struct bio *new = bio_alloc(gfp, nr_pages);
	17
	18	if (bio) {
	19	bio_chain(bio, new);
4e49ea4a	20	submit_bio(bio);
9082e87b	21	}
5dba3089	22
9082e87b	23	return new;
f31e7e40 DM	24	}
f31e7e40 DM	25
38f25255	26	int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
288dab8a	27	sector_t nr_sects, gfp_t gfp_mask, int flags,
469e3216	28	struct bio **biop)
f31e7e40	29	{
f31e7e40	30	struct request_queue *q = bdev_get_queue(bdev);
38f25255	31	struct bio bio = biop;
a22c4d7e	32	unsigned int granularity;
ef295ecf	33	unsigned int op;
a22c4d7e	34	int alignment;
28b2be20	35	sector_t bs_mask;
f31e7e40 DM	36
	37	if (!q)
	38	return -ENXIO;
288dab8a CH	39
	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
28b2be20 DW	50	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	51	if ((sector \| nr_sects) & bs_mask)
	52	return -EINVAL;
	53
a22c4d7e ML	54	/* Zero-sector (unknown) and one-sector granularities are the same. */
	55	granularity = max(q->limits.discard_granularity >> 9, 1U);
	56	alignment = (bdev_discard_alignment(bdev) >> 9) % granularity;
	57
5dba3089	58	while (nr_sects) {
c6e66634	59	unsigned int req_sects;
a22c4d7e	60	sector_t end_sect, tmp;
c6e66634	61
a22c4d7e ML	62	/* Make sure bi_size doesn't overflow */
	63	req_sects = min_t(sector_t, nr_sects, UINT_MAX >> 9);
	64
9082e87b	65	/**
a22c4d7e ML	66	* If splitting a request, and the next starting sector would be
	67	* misaligned, stop the discard at the previous aligned sector.
	68	*/
c6e66634	69	end_sect = sector + req_sects;
a22c4d7e ML	70	tmp = end_sect;
	71	if (req_sects < nr_sects &&
	72	sector_div(tmp, granularity) != alignment) {
	73	end_sect = end_sect - alignment;
	74	sector_div(end_sect, granularity);
	75	end_sect = end_sect * granularity + alignment;
	76	req_sects = end_sect - sector;
	77	}
c6e66634	78
f9d03f96	79	bio = next_bio(bio, 0, gfp_mask);
4f024f37	80	bio->bi_iter.bi_sector = sector;
74d46992	81	bio_set_dev(bio, bdev);
288dab8a	82	bio_set_op_attrs(bio, op, 0);
f31e7e40	83
4f024f37	84	bio->bi_iter.bi_size = req_sects << 9;
c6e66634 PB	85	nr_sects -= req_sects;
c6e66634 PB	86	sector = end_sect;
f31e7e40	87
c8123f8c JA	88	/*
	89	* We can loop for a long time in here, if someone does
	90	* full device discards (like mkfs). Be nice and allow
	91	* us to schedule out to avoid softlocking if preempt
	92	* is disabled.
	93	*/
	94	cond_resched();
5dba3089	95	}
38f25255 CH	96
	97	*biop = bio;
	98	return 0;
	99	}
	100	EXPORT_SYMBOL(__blkdev_issue_discard);
	101
	102	/**
	103	* blkdev_issue_discard - queue a discard
	104	* @bdev: blockdev to issue discard for
	105	* @sector: start sector
	106	* @nr_sects: number of sectors to discard
	107	* @gfp_mask: memory allocation flags (for bio_alloc)
e554911c	108	* @flags: BLKDEV_DISCARD_* flags to control behaviour
38f25255 CH	109	*
	110	* Description:
	111	* Issue a discard request for the sectors in question.
	112	*/
	113	int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
	114	sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
	115	{
38f25255 CH	116	struct bio *bio = NULL;
	117	struct blk_plug plug;
	118	int ret;
	119
38f25255	120	blk_start_plug(&plug);
288dab8a	121	ret = __blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, flags,
38f25255	122	&bio);
bbd848e0	123	if (!ret && bio) {
4e49ea4a	124	ret = submit_bio_wait(bio);
48920ff2	125	if (ret == -EOPNOTSUPP)
bbd848e0	126	ret = 0;
05bd92dd	127	bio_put(bio);
bbd848e0	128	}
0cfbcafc	129	blk_finish_plug(&plug);
f31e7e40	130
bbd848e0	131	return ret;
f31e7e40 DM	132	}
f31e7e40 DM	133	EXPORT_SYMBOL(blkdev_issue_discard);
3f14d792	134
4363ac7c	135	/**
e73c23ff	136	* __blkdev_issue_write_same - generate number of bios with same page
4363ac7c MP	137	* @bdev: target blockdev
	138	* @sector: start sector
	139	* @nr_sects: number of sectors to write
	140	* @gfp_mask: memory allocation flags (for bio_alloc)
	141	* @page: page containing data to write
e73c23ff	142	* @biop: pointer to anchor bio
4363ac7c MP	143	*
4363ac7c MP	144	* Description:
e73c23ff	145	* Generate and issue number of bios(REQ_OP_WRITE_SAME) with same page.
4363ac7c	146	*/
e73c23ff CK	147	static int __blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
	148	sector_t nr_sects, gfp_t gfp_mask, struct page *page,
	149	struct bio **biop)
4363ac7c	150	{
4363ac7c MP	151	struct request_queue *q = bdev_get_queue(bdev);
4363ac7c MP	152	unsigned int max_write_same_sectors;
e73c23ff	153	struct bio bio = biop;
28b2be20	154	sector_t bs_mask;
4363ac7c MP	155
	156	if (!q)
	157	return -ENXIO;
	158
28b2be20 DW	159	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	160	if ((sector \| nr_sects) & bs_mask)
	161	return -EINVAL;
	162
e73c23ff CK	163	if (!bdev_write_same(bdev))
	164	return -EOPNOTSUPP;
	165
b49a0871 ML	166	/* Ensure that max_write_same_sectors doesn't overflow bi_size */
b49a0871 ML	167	max_write_same_sectors = UINT_MAX >> 9;
4363ac7c	168
4363ac7c	169	while (nr_sects) {
4e49ea4a	170	bio = next_bio(bio, 1, gfp_mask);
4f024f37	171	bio->bi_iter.bi_sector = sector;
74d46992	172	bio_set_dev(bio, bdev);
4363ac7c MP	173	bio->bi_vcnt = 1;
	174	bio->bi_io_vec->bv_page = page;
	175	bio->bi_io_vec->bv_offset = 0;
	176	bio->bi_io_vec->bv_len = bdev_logical_block_size(bdev);
95fe6c1a	177	bio_set_op_attrs(bio, REQ_OP_WRITE_SAME, 0);
4363ac7c MP	178
4363ac7c MP	179	if (nr_sects > max_write_same_sectors) {
4f024f37	180	bio->bi_iter.bi_size = max_write_same_sectors << 9;
4363ac7c MP	181	nr_sects -= max_write_same_sectors;
	182	sector += max_write_same_sectors;
	183	} else {
4f024f37	184	bio->bi_iter.bi_size = nr_sects << 9;
4363ac7c MP	185	nr_sects = 0;
4363ac7c MP	186	}
e73c23ff	187	cond_resched();
4363ac7c MP	188	}
4363ac7c MP	189
e73c23ff CK	190	*biop = bio;
	191	return 0;
	192	}
	193
	194	/**
	195	* blkdev_issue_write_same - queue a write same operation
	196	* @bdev: target blockdev
	197	* @sector: start sector
	198	* @nr_sects: number of sectors to write
	199	* @gfp_mask: memory allocation flags (for bio_alloc)
	200	* @page: page containing data
	201	*
	202	* Description:
	203	* Issue a write same request for the sectors in question.
	204	*/
	205	int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
	206	sector_t nr_sects, gfp_t gfp_mask,
	207	struct page *page)
	208	{
	209	struct bio *bio = NULL;
	210	struct blk_plug plug;
	211	int ret;
	212
	213	blk_start_plug(&plug);
	214	ret = __blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask, page,
	215	&bio);
	216	if (ret == 0 && bio) {
4e49ea4a	217	ret = submit_bio_wait(bio);
05bd92dd ST	218	bio_put(bio);
05bd92dd ST	219	}
e73c23ff	220	blk_finish_plug(&plug);
3f40bf2c	221	return ret;
4363ac7c MP	222	}
	223	EXPORT_SYMBOL(blkdev_issue_write_same);
	224
a6f0788e CK	225	static int __blkdev_issue_write_zeroes(struct block_device *bdev,
a6f0788e CK	226	sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
d928be9f	227	struct bio **biop, unsigned flags)
a6f0788e CK	228	{
	229	struct bio bio = biop;
	230	unsigned int max_write_zeroes_sectors;
	231	struct request_queue *q = bdev_get_queue(bdev);
	232
	233	if (!q)
	234	return -ENXIO;
	235
	236	/* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
	237	max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev);
	238
	239	if (max_write_zeroes_sectors == 0)
	240	return -EOPNOTSUPP;
	241
	242	while (nr_sects) {
	243	bio = next_bio(bio, 0, gfp_mask);
	244	bio->bi_iter.bi_sector = sector;
74d46992	245	bio_set_dev(bio, bdev);
d928be9f CH	246	bio->bi_opf = REQ_OP_WRITE_ZEROES;
	247	if (flags & BLKDEV_ZERO_NOUNMAP)
	248	bio->bi_opf \|= REQ_NOUNMAP;
a6f0788e CK	249
	250	if (nr_sects > max_write_zeroes_sectors) {
	251	bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
	252	nr_sects -= max_write_zeroes_sectors;
	253	sector += max_write_zeroes_sectors;
	254	} else {
	255	bio->bi_iter.bi_size = nr_sects << 9;
	256	nr_sects = 0;
	257	}
	258	cond_resched();
	259	}
	260
	261	*biop = bio;
	262	return 0;
	263	}
	264
615d22a5 DLM	265	/*
	266	* Convert a number of 512B sectors to a number of pages.
	267	* The result is limited to a number of pages that can fit into a BIO.
	268	* Also make sure that the result is always at least 1 (page) for the cases
	269	* where nr_sects is lower than the number of sectors in a page.
	270	*/
	271	static unsigned int __blkdev_sectors_to_bio_pages(sector_t nr_sects)
	272	{
09c2c359	273	sector_t pages = DIV_ROUND_UP_SECTOR_T(nr_sects, PAGE_SIZE / 512);
615d22a5	274
09c2c359	275	return min(pages, (sector_t)BIO_MAX_PAGES);
615d22a5 DLM	276	}
615d22a5 DLM	277
425a4dba ID	278	static int __blkdev_issue_zero_pages(struct block_device *bdev,
	279	sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
	280	struct bio **biop)
	281	{
	282	struct request_queue *q = bdev_get_queue(bdev);
	283	struct bio bio = biop;
	284	int bi_size = 0;
	285	unsigned int sz;
	286
	287	if (!q)
	288	return -ENXIO;
	289
	290	while (nr_sects != 0) {
	291	bio = next_bio(bio, __blkdev_sectors_to_bio_pages(nr_sects),
	292	gfp_mask);
	293	bio->bi_iter.bi_sector = sector;
	294	bio_set_dev(bio, bdev);
	295	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
	296
	297	while (nr_sects != 0) {
	298	sz = min((sector_t) PAGE_SIZE, nr_sects << 9);
	299	bi_size = bio_add_page(bio, ZERO_PAGE(0), sz, 0);
	300	nr_sects -= bi_size >> 9;
	301	sector += bi_size >> 9;
	302	if (bi_size < sz)
	303	break;
	304	}
	305	cond_resched();
	306	}
	307
	308	*biop = bio;
	309	return 0;
	310	}
	311
3f14d792	312	/**
e73c23ff	313	* __blkdev_issue_zeroout - generate number of zero filed write bios
3f14d792 DM	314	* @bdev: blockdev to issue
	315	* @sector: start sector
	316	* @nr_sects: number of sectors to write
	317	* @gfp_mask: memory allocation flags (for bio_alloc)
e73c23ff	318	* @biop: pointer to anchor bio
ee472d83	319	* @flags: controls detailed behavior
3f14d792 DM	320	*
3f14d792 DM	321	* Description:
ee472d83 CH	322	* Zero-fill a block range, either using hardware offload or by explicitly
	323	* writing zeroes to the device.
	324	*
	325	* If a device is using logical block provisioning, the underlying space will
	326	* not be released if %flags contains BLKDEV_ZERO_NOUNMAP.
cb365b96 CH	327	*
	328	* If %flags contains BLKDEV_ZERO_NOFALLBACK, the function will return
	329	* -EOPNOTSUPP if no explicit hardware offload for zeroing is provided.
3f14d792	330	*/
e73c23ff CK	331	int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
e73c23ff CK	332	sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
ee472d83	333	unsigned flags)
3f14d792	334	{
18edc8ea	335	int ret;
28b2be20 DW	336	sector_t bs_mask;
	337
	338	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	339	if ((sector \| nr_sects) & bs_mask)
	340	return -EINVAL;
3f14d792	341
a6f0788e	342	ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects, gfp_mask,
d928be9f	343	biop, flags);
cb365b96	344	if (ret != -EOPNOTSUPP \|\| (flags & BLKDEV_ZERO_NOFALLBACK))
425a4dba	345	return ret;
3f14d792	346
425a4dba ID	347	return __blkdev_issue_zero_pages(bdev, sector, nr_sects, gfp_mask,
425a4dba ID	348	biop);
3f14d792	349	}
e73c23ff	350	EXPORT_SYMBOL(__blkdev_issue_zeroout);
579e8f3c MP	351
	352	/**
	353	* blkdev_issue_zeroout - zero-fill a block range
	354	* @bdev: blockdev to write
	355	* @sector: start sector
	356	* @nr_sects: number of sectors to write
	357	* @gfp_mask: memory allocation flags (for bio_alloc)
ee472d83	358	* @flags: controls detailed behavior
579e8f3c MP	359	*
579e8f3c MP	360	* Description:
ee472d83 CH	361	* Zero-fill a block range, either using hardware offload or by explicitly
	362	* writing zeroes to the device. See __blkdev_issue_zeroout() for the
	363	* valid values for %flags.
579e8f3c	364	*/
579e8f3c	365	int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
ee472d83	366	sector_t nr_sects, gfp_t gfp_mask, unsigned flags)
579e8f3c	367	{
d5ce4c31 ID	368	int ret = 0;
	369	sector_t bs_mask;
	370	struct bio *bio;
e73c23ff	371	struct blk_plug plug;
d5ce4c31	372	bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev);
d93ba7a5	373
d5ce4c31 ID	374	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	375	if ((sector \| nr_sects) & bs_mask)
	376	return -EINVAL;
	377
	378	retry:
	379	bio = NULL;
e73c23ff	380	blk_start_plug(&plug);
d5ce4c31 ID	381	if (try_write_zeroes) {
	382	ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects,
	383	gfp_mask, &bio, flags);
	384	} else if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
	385	ret = __blkdev_issue_zero_pages(bdev, sector, nr_sects,
	386	gfp_mask, &bio);
	387	} else {
	388	/* No zeroing offload support */
	389	ret = -EOPNOTSUPP;
	390	}
e73c23ff CK	391	if (ret == 0 && bio) {
	392	ret = submit_bio_wait(bio);
	393	bio_put(bio);
	394	}
	395	blk_finish_plug(&plug);
d5ce4c31 ID	396	if (ret && try_write_zeroes) {
	397	if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
	398	try_write_zeroes = false;
	399	goto retry;
	400	}
	401	if (!bdev_write_zeroes_sectors(bdev)) {
	402	/*
	403	* Zeroing offload support was indicated, but the
	404	* device reported ILLEGAL REQUEST (for some devices
	405	* there is no non-destructive way to verify whether
	406	* WRITE ZEROES is actually supported).
	407	*/
	408	ret = -EOPNOTSUPP;
	409	}
	410	}
579e8f3c	411
e73c23ff	412	return ret;
579e8f3c	413	}
3f14d792	414	EXPORT_SYMBOL(blkdev_issue_zeroout);