[mirror_ubuntu-artful-kernel.git] / block / blk-lib.c

/*
 * 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->bi_bdev = 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->bi_bdev = 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->bi_bdev = 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 bytes = (nr_sects << 9) + PAGE_SIZE - 1;

	return min(bytes >> PAGE_SHIFT, (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->bi_bdev = 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
f31e7e40 DM	1	/*
	2	* Functions related to generic helpers functions
	3	*/
	4	#include <linux/kernel.h>
	5	#include <linux/module.h>
	6	#include <linux/bio.h>
	7	#include <linux/blkdev.h>
	8	#include <linux/scatterlist.h>
	9
	10	#include "blk.h"
	11
4e49ea4a	12	static struct bio next_bio(struct bio bio, unsigned int nr_pages,
9082e87b	13	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;
a22c4d7e	31	unsigned int granularity;
ef295ecf	32	unsigned int op;
a22c4d7e	33	int alignment;
28b2be20	34	sector_t bs_mask;
f31e7e40 DM	35
	36	if (!q)
	37	return -ENXIO;
288dab8a CH	38
	39	if (flags & BLKDEV_DISCARD_SECURE) {
	40	if (!blk_queue_secure_erase(q))
	41	return -EOPNOTSUPP;
	42	op = REQ_OP_SECURE_ERASE;
	43	} else {
	44	if (!blk_queue_discard(q))
	45	return -EOPNOTSUPP;
	46	op = REQ_OP_DISCARD;
	47	}
f31e7e40	48
28b2be20 DW	49	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	50	if ((sector \| nr_sects) & bs_mask)
	51	return -EINVAL;
	52
a22c4d7e ML	53	/* Zero-sector (unknown) and one-sector granularities are the same. */
	54	granularity = max(q->limits.discard_granularity >> 9, 1U);
	55	alignment = (bdev_discard_alignment(bdev) >> 9) % granularity;
	56
5dba3089	57	while (nr_sects) {
c6e66634	58	unsigned int req_sects;
a22c4d7e	59	sector_t end_sect, tmp;
c6e66634	60
a22c4d7e ML	61	/* Make sure bi_size doesn't overflow */
	62	req_sects = min_t(sector_t, nr_sects, UINT_MAX >> 9);
	63
9082e87b	64	/**
a22c4d7e ML	65	* If splitting a request, and the next starting sector would be
	66	* misaligned, stop the discard at the previous aligned sector.
	67	*/
c6e66634	68	end_sect = sector + req_sects;
a22c4d7e ML	69	tmp = end_sect;
	70	if (req_sects < nr_sects &&
	71	sector_div(tmp, granularity) != alignment) {
	72	end_sect = end_sect - alignment;
	73	sector_div(end_sect, granularity);
	74	end_sect = end_sect * granularity + alignment;
	75	req_sects = end_sect - sector;
	76	}
c6e66634	77
f9d03f96	78	bio = next_bio(bio, 0, gfp_mask);
4f024f37	79	bio->bi_iter.bi_sector = sector;
f31e7e40	80	bio->bi_bdev = bdev;
288dab8a	81	bio_set_op_attrs(bio, op, 0);
f31e7e40	82
4f024f37	83	bio->bi_iter.bi_size = req_sects << 9;
c6e66634 PB	84	nr_sects -= req_sects;
c6e66634 PB	85	sector = end_sect;
f31e7e40	86
c8123f8c JA	87	/*
	88	* We can loop for a long time in here, if someone does
	89	* full device discards (like mkfs). Be nice and allow
	90	* us to schedule out to avoid softlocking if preempt
	91	* is disabled.
	92	*/
	93	cond_resched();
5dba3089	94	}
38f25255 CH	95
	96	*biop = bio;
	97	return 0;
	98	}
	99	EXPORT_SYMBOL(__blkdev_issue_discard);
	100
	101	/**
	102	* blkdev_issue_discard - queue a discard
	103	* @bdev: blockdev to issue discard for
	104	* @sector: start sector
	105	* @nr_sects: number of sectors to discard
	106	* @gfp_mask: memory allocation flags (for bio_alloc)
e554911c	107	* @flags: BLKDEV_DISCARD_* flags to control behaviour
38f25255 CH	108	*
	109	* Description:
	110	* Issue a discard request for the sectors in question.
	111	*/
	112	int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
	113	sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
	114	{
38f25255 CH	115	struct bio *bio = NULL;
	116	struct blk_plug plug;
	117	int ret;
	118
38f25255	119	blk_start_plug(&plug);
288dab8a	120	ret = __blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, flags,
38f25255	121	&bio);
bbd848e0	122	if (!ret && bio) {
4e49ea4a	123	ret = submit_bio_wait(bio);
48920ff2	124	if (ret == -EOPNOTSUPP)
bbd848e0	125	ret = 0;
05bd92dd	126	bio_put(bio);
bbd848e0	127	}
0cfbcafc	128	blk_finish_plug(&plug);
f31e7e40	129
bbd848e0	130	return ret;
f31e7e40 DM	131	}
f31e7e40 DM	132	EXPORT_SYMBOL(blkdev_issue_discard);
3f14d792	133
4363ac7c	134	/**
e73c23ff	135	* __blkdev_issue_write_same - generate number of bios with same page
4363ac7c MP	136	* @bdev: target blockdev
	137	* @sector: start sector
	138	* @nr_sects: number of sectors to write
	139	* @gfp_mask: memory allocation flags (for bio_alloc)
	140	* @page: page containing data to write
e73c23ff	141	* @biop: pointer to anchor bio
4363ac7c MP	142	*
4363ac7c MP	143	* Description:
e73c23ff	144	* Generate and issue number of bios(REQ_OP_WRITE_SAME) with same page.
4363ac7c	145	*/
e73c23ff CK	146	static int __blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
	147	sector_t nr_sects, gfp_t gfp_mask, struct page *page,
	148	struct bio **biop)
4363ac7c	149	{
4363ac7c MP	150	struct request_queue *q = bdev_get_queue(bdev);
4363ac7c MP	151	unsigned int max_write_same_sectors;
e73c23ff	152	struct bio bio = biop;
28b2be20	153	sector_t bs_mask;
4363ac7c MP	154
	155	if (!q)
	156	return -ENXIO;
	157
28b2be20 DW	158	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	159	if ((sector \| nr_sects) & bs_mask)
	160	return -EINVAL;
	161
e73c23ff CK	162	if (!bdev_write_same(bdev))
	163	return -EOPNOTSUPP;
	164
b49a0871 ML	165	/* Ensure that max_write_same_sectors doesn't overflow bi_size */
b49a0871 ML	166	max_write_same_sectors = UINT_MAX >> 9;
4363ac7c	167
4363ac7c	168	while (nr_sects) {
4e49ea4a	169	bio = next_bio(bio, 1, gfp_mask);
4f024f37	170	bio->bi_iter.bi_sector = sector;
4363ac7c	171	bio->bi_bdev = bdev;
4363ac7c MP	172	bio->bi_vcnt = 1;
	173	bio->bi_io_vec->bv_page = page;
	174	bio->bi_io_vec->bv_offset = 0;
	175	bio->bi_io_vec->bv_len = bdev_logical_block_size(bdev);
95fe6c1a	176	bio_set_op_attrs(bio, REQ_OP_WRITE_SAME, 0);
4363ac7c MP	177
4363ac7c MP	178	if (nr_sects > max_write_same_sectors) {
4f024f37	179	bio->bi_iter.bi_size = max_write_same_sectors << 9;
4363ac7c MP	180	nr_sects -= max_write_same_sectors;
	181	sector += max_write_same_sectors;
	182	} else {
4f024f37	183	bio->bi_iter.bi_size = nr_sects << 9;
4363ac7c MP	184	nr_sects = 0;
4363ac7c MP	185	}
e73c23ff	186	cond_resched();
4363ac7c MP	187	}
4363ac7c MP	188
e73c23ff CK	189	*biop = bio;
	190	return 0;
	191	}
	192
	193	/**
	194	* blkdev_issue_write_same - queue a write same operation
	195	* @bdev: target blockdev
	196	* @sector: start sector
	197	* @nr_sects: number of sectors to write
	198	* @gfp_mask: memory allocation flags (for bio_alloc)
	199	* @page: page containing data
	200	*
	201	* Description:
	202	* Issue a write same request for the sectors in question.
	203	*/
	204	int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
	205	sector_t nr_sects, gfp_t gfp_mask,
	206	struct page *page)
	207	{
	208	struct bio *bio = NULL;
	209	struct blk_plug plug;
	210	int ret;
	211
	212	blk_start_plug(&plug);
	213	ret = __blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask, page,
	214	&bio);
	215	if (ret == 0 && bio) {
4e49ea4a	216	ret = submit_bio_wait(bio);
05bd92dd ST	217	bio_put(bio);
05bd92dd ST	218	}
e73c23ff	219	blk_finish_plug(&plug);
3f40bf2c	220	return ret;
4363ac7c MP	221	}
	222	EXPORT_SYMBOL(blkdev_issue_write_same);
	223
a6f0788e CK	224	static int __blkdev_issue_write_zeroes(struct block_device *bdev,
a6f0788e CK	225	sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
d928be9f	226	struct bio **biop, unsigned flags)
a6f0788e CK	227	{
	228	struct bio bio = biop;
	229	unsigned int max_write_zeroes_sectors;
	230	struct request_queue *q = bdev_get_queue(bdev);
	231
	232	if (!q)
	233	return -ENXIO;
	234
	235	/* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
	236	max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev);
	237
	238	if (max_write_zeroes_sectors == 0)
	239	return -EOPNOTSUPP;
	240
	241	while (nr_sects) {
	242	bio = next_bio(bio, 0, gfp_mask);
	243	bio->bi_iter.bi_sector = sector;
	244	bio->bi_bdev = bdev;
d928be9f CH	245	bio->bi_opf = REQ_OP_WRITE_ZEROES;
	246	if (flags & BLKDEV_ZERO_NOUNMAP)
	247	bio->bi_opf \|= REQ_NOUNMAP;
a6f0788e CK	248
	249	if (nr_sects > max_write_zeroes_sectors) {
	250	bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
	251	nr_sects -= max_write_zeroes_sectors;
	252	sector += max_write_zeroes_sectors;
	253	} else {
	254	bio->bi_iter.bi_size = nr_sects << 9;
	255	nr_sects = 0;
	256	}
	257	cond_resched();
	258	}
	259
	260	*biop = bio;
	261	return 0;
	262	}
	263
615d22a5 DLM	264	/*
	265	* Convert a number of 512B sectors to a number of pages.
	266	* The result is limited to a number of pages that can fit into a BIO.
	267	* Also make sure that the result is always at least 1 (page) for the cases
	268	* where nr_sects is lower than the number of sectors in a page.
	269	*/
	270	static unsigned int __blkdev_sectors_to_bio_pages(sector_t nr_sects)
	271	{
	272	sector_t bytes = (nr_sects << 9) + PAGE_SIZE - 1;
	273
	274	return min(bytes >> PAGE_SHIFT, (sector_t)BIO_MAX_PAGES);
	275	}
	276
9c99c989 CIK	277	static int __blkdev_issue_zero_pages(struct block_device *bdev,
	278	sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
	279	struct bio **biop)
	280	{
	281	struct request_queue *q = bdev_get_queue(bdev);
	282	struct bio bio = biop;
	283	int bi_size = 0;
	284	unsigned int sz;
	285
	286	if (!q)
	287	return -ENXIO;
	288
	289	while (nr_sects != 0) {
	290	bio = next_bio(bio, __blkdev_sectors_to_bio_pages(nr_sects),
	291	gfp_mask);
	292	bio->bi_iter.bi_sector = sector;
	293	bio->bi_bdev = bdev;
	294	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
	295
	296	while (nr_sects != 0) {
	297	sz = min((sector_t) PAGE_SIZE, nr_sects << 9);
	298	bi_size = bio_add_page(bio, ZERO_PAGE(0), sz, 0);
	299	nr_sects -= bi_size >> 9;
	300	sector += bi_size >> 9;
	301	if (bi_size < sz)
	302	break;
	303	}
	304	cond_resched();
	305	}
	306
	307	*biop = bio;
	308	return 0;
	309	}
	310
3f14d792	311	/**
e73c23ff	312	* __blkdev_issue_zeroout - generate number of zero filed write bios
3f14d792 DM	313	* @bdev: blockdev to issue
	314	* @sector: start sector
	315	* @nr_sects: number of sectors to write
	316	* @gfp_mask: memory allocation flags (for bio_alloc)
e73c23ff	317	* @biop: pointer to anchor bio
ee472d83	318	* @flags: controls detailed behavior
3f14d792 DM	319	*
3f14d792 DM	320	* Description:
ee472d83 CH	321	* Zero-fill a block range, either using hardware offload or by explicitly
	322	* writing zeroes to the device.
	323	*
	324	* If a device is using logical block provisioning, the underlying space will
	325	* not be released if %flags contains BLKDEV_ZERO_NOUNMAP.
cb365b96 CH	326	*
	327	* If %flags contains BLKDEV_ZERO_NOFALLBACK, the function will return
	328	* -EOPNOTSUPP if no explicit hardware offload for zeroing is provided.
3f14d792	329	*/
e73c23ff CK	330	int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
e73c23ff CK	331	sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
ee472d83	332	unsigned flags)
3f14d792	333	{
18edc8ea	334	int ret;
28b2be20 DW	335	sector_t bs_mask;
	336
	337	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	338	if ((sector \| nr_sects) & bs_mask)
	339	return -EINVAL;
3f14d792	340
a6f0788e	341	ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects, gfp_mask,
d928be9f	342	biop, flags);
cb365b96	343	if (ret != -EOPNOTSUPP \|\| (flags & BLKDEV_ZERO_NOFALLBACK))
9c99c989	344	return ret;
3f14d792	345
9c99c989 CIK	346	return __blkdev_issue_zero_pages(bdev, sector, nr_sects, gfp_mask,
9c99c989 CIK	347	biop);
3f14d792	348	}
e73c23ff	349	EXPORT_SYMBOL(__blkdev_issue_zeroout);
579e8f3c MP	350
	351	/**
	352	* blkdev_issue_zeroout - zero-fill a block range
	353	* @bdev: blockdev to write
	354	* @sector: start sector
	355	* @nr_sects: number of sectors to write
	356	* @gfp_mask: memory allocation flags (for bio_alloc)
ee472d83	357	* @flags: controls detailed behavior
579e8f3c MP	358	*
579e8f3c MP	359	* Description:
ee472d83 CH	360	* Zero-fill a block range, either using hardware offload or by explicitly
	361	* writing zeroes to the device. See __blkdev_issue_zeroout() for the
	362	* valid values for %flags.
579e8f3c	363	*/
579e8f3c	364	int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
ee472d83	365	sector_t nr_sects, gfp_t gfp_mask, unsigned flags)
579e8f3c	366	{
c37bc414 ID	367	int ret = 0;
	368	sector_t bs_mask;
	369	struct bio *bio;
e73c23ff	370	struct blk_plug plug;
c37bc414	371	bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev);
d93ba7a5	372
c37bc414 ID	373	bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
	374	if ((sector \| nr_sects) & bs_mask)
	375	return -EINVAL;
	376
	377	retry:
	378	bio = NULL;
e73c23ff	379	blk_start_plug(&plug);
c37bc414 ID	380	if (try_write_zeroes) {
	381	ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects,
	382	gfp_mask, &bio, flags);
	383	} else if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
	384	ret = __blkdev_issue_zero_pages(bdev, sector, nr_sects,
	385	gfp_mask, &bio);
	386	} else {
	387	/* No zeroing offload support */
	388	ret = -EOPNOTSUPP;
	389	}
e73c23ff CK	390	if (ret == 0 && bio) {
	391	ret = submit_bio_wait(bio);
	392	bio_put(bio);
	393	}
	394	blk_finish_plug(&plug);
c37bc414 ID	395	if (ret && try_write_zeroes) {
	396	if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
	397	try_write_zeroes = false;
	398	goto retry;
	399	}
	400	if (!bdev_write_zeroes_sectors(bdev)) {
	401	/*
	402	* Zeroing offload support was indicated, but the
	403	* device reported ILLEGAL REQUEST (for some devices
	404	* there is no non-destructive way to verify whether
	405	* WRITE ZEROES is actually supported).
	406	*/
	407	ret = -EOPNOTSUPP;
	408	}
	409	}
579e8f3c	410
e73c23ff	411	return ret;
579e8f3c	412	}
3f14d792	413	EXPORT_SYMBOL(blkdev_issue_zeroout);