[linux-2.6-block.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, int rw, unsigned int nr_pages,
		gfp_t gfp)
{
	struct bio *new = bio_alloc(gfp, nr_pages);

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

	return new;
}

int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
		sector_t nr_sects, gfp_t gfp_mask, int type, struct bio **biop)
{
	struct request_queue *q = bdev_get_queue(bdev);
	struct bio *bio = *biop;
	unsigned int granularity;
	int alignment;

	if (!q)
		return -ENXIO;
	if (!blk_queue_discard(q))
		return -EOPNOTSUPP;
	if ((type & REQ_SECURE) && !blk_queue_secdiscard(q))
		return -EOPNOTSUPP;

	/* 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, type, 1, gfp_mask);
		bio->bi_iter.bi_sector = sector;
		bio->bi_bdev = bdev;

		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_IFL_* 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)
{
	int type = REQ_WRITE | REQ_DISCARD;
	struct bio *bio = NULL;
	struct blk_plug plug;
	int ret;

	if (flags & BLKDEV_DISCARD_SECURE)
		type |= REQ_SECURE;

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

	return ret;
}
EXPORT_SYMBOL(blkdev_issue_discard);

/**
 * 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 to write
 *
 * 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 request_queue *q = bdev_get_queue(bdev);
	unsigned int max_write_same_sectors;
	struct bio *bio = NULL;
	int ret = 0;

	if (!q)
		return -ENXIO;

	/* 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, REQ_WRITE | REQ_WRITE_SAME, 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);

		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;
		}
	}

	if (bio)
		ret = submit_bio_wait(REQ_WRITE | REQ_WRITE_SAME, bio);
	return ret != -EOPNOTSUPP ? ret : 0;
}
EXPORT_SYMBOL(blkdev_issue_write_same);

/**
 * 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)
 *
 * Description:
 *  Generate and issue number of bios with zerofiled pages.
 */

static int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
				  sector_t nr_sects, gfp_t gfp_mask)
{
	int ret;
	struct bio *bio = NULL;
	unsigned int sz;

	while (nr_sects != 0) {
		bio = next_bio(bio, WRITE,
				min(nr_sects, (sector_t)BIO_MAX_PAGES),
				gfp_mask);
		bio->bi_iter.bi_sector = sector;
		bio->bi_bdev   = bdev;

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

	if (bio)
		return submit_bio_wait(WRITE, bio);
	return 0;
}

/**
 * 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)
 * @discard:	whether to discard the block range
 *
 * Description:
 *  Zero-fill a block range.  If the discard flag is set and the block
 *  device guarantees that subsequent READ operations to the block range
 *  in question will return zeroes, the blocks will be discarded. Should
 *  the discard request fail, if the discard flag is not set, or if
 *  discard_zeroes_data is not supported, this function will resort to
 *  zeroing the blocks manually, thus provisioning (allocating,
 *  anchoring) them. If the block device supports the WRITE SAME command
 *  blkdev_issue_zeroout() will use it to optimize the process of
 *  clearing the block range. Otherwise the zeroing will be performed
 *  using regular WRITE calls.
 */

int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
			 sector_t nr_sects, gfp_t gfp_mask, bool discard)
{
	struct request_queue *q = bdev_get_queue(bdev);

	if (discard && blk_queue_discard(q) && q->limits.discard_zeroes_data &&
	    blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, 0) == 0)
		return 0;

	if (bdev_write_same(bdev) &&
	    blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask,
				    ZERO_PAGE(0)) == 0)
		return 0;

	return __blkdev_issue_zeroout(bdev, sector, nr_sects, gfp_mask);
}
EXPORT_SYMBOL(blkdev_issue_zeroout);
Commit	Line	Data
	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
	12	static struct bio next_bio(struct bio bio, int rw, unsigned int nr_pages,
	13	gfp_t gfp)
	14	{
	15	struct bio *new = bio_alloc(gfp, nr_pages);
	16
	17	if (bio) {
	18	bio_chain(bio, new);
	19	submit_bio(rw, bio);
	20	}
	21
	22	return new;
	23	}
	24
	25	int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
	26	sector_t nr_sects, gfp_t gfp_mask, int type, struct bio **biop)
	27	{
	28	struct request_queue *q = bdev_get_queue(bdev);
	29	struct bio bio = biop;
	30	unsigned int granularity;
	31	int alignment;
	32
	33	if (!q)
	34	return -ENXIO;
	35	if (!blk_queue_discard(q))
	36	return -EOPNOTSUPP;
	37	if ((type & REQ_SECURE) && !blk_queue_secdiscard(q))
	38	return -EOPNOTSUPP;
	39
	40	/* Zero-sector (unknown) and one-sector granularities are the same. */
	41	granularity = max(q->limits.discard_granularity >> 9, 1U);
	42	alignment = (bdev_discard_alignment(bdev) >> 9) % granularity;
	43
	44	while (nr_sects) {
	45	unsigned int req_sects;
	46	sector_t end_sect, tmp;
	47
	48	/* Make sure bi_size doesn't overflow */
	49	req_sects = min_t(sector_t, nr_sects, UINT_MAX >> 9);
	50
	51	/**
	52	* If splitting a request, and the next starting sector would be
	53	* misaligned, stop the discard at the previous aligned sector.
	54	*/
	55	end_sect = sector + req_sects;
	56	tmp = end_sect;
	57	if (req_sects < nr_sects &&
	58	sector_div(tmp, granularity) != alignment) {
	59	end_sect = end_sect - alignment;
	60	sector_div(end_sect, granularity);
	61	end_sect = end_sect * granularity + alignment;
	62	req_sects = end_sect - sector;
	63	}
	64
	65	bio = next_bio(bio, type, 1, gfp_mask);
	66	bio->bi_iter.bi_sector = sector;
	67	bio->bi_bdev = bdev;
	68
	69	bio->bi_iter.bi_size = req_sects << 9;
	70	nr_sects -= req_sects;
	71	sector = end_sect;
	72
	73	/*
	74	* We can loop for a long time in here, if someone does
	75	* full device discards (like mkfs). Be nice and allow
	76	* us to schedule out to avoid softlocking if preempt
	77	* is disabled.
	78	*/
	79	cond_resched();
	80	}
	81
	82	*biop = bio;
	83	return 0;
	84	}
	85	EXPORT_SYMBOL(__blkdev_issue_discard);
	86
	87	/**
	88	* blkdev_issue_discard - queue a discard
	89	* @bdev: blockdev to issue discard for
	90	* @sector: start sector
	91	* @nr_sects: number of sectors to discard
	92	* @gfp_mask: memory allocation flags (for bio_alloc)
	93	* @flags: BLKDEV_IFL_* flags to control behaviour
	94	*
	95	* Description:
	96	* Issue a discard request for the sectors in question.
	97	*/
	98	int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
	99	sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
	100	{
	101	int type = REQ_WRITE \| REQ_DISCARD;
	102	struct bio *bio = NULL;
	103	struct blk_plug plug;
	104	int ret;
	105
	106	if (flags & BLKDEV_DISCARD_SECURE)
	107	type \|= REQ_SECURE;
	108
	109	blk_start_plug(&plug);
	110	ret = __blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, type,
	111	&bio);
	112	if (!ret && bio) {
	113	ret = submit_bio_wait(type, bio);
	114	if (ret == -EOPNOTSUPP)
	115	ret = 0;
	116	}
	117	blk_finish_plug(&plug);
	118
	119	return ret;
	120	}
	121	EXPORT_SYMBOL(blkdev_issue_discard);
	122
	123	/**
	124	* blkdev_issue_write_same - queue a write same operation
	125	* @bdev: target blockdev
	126	* @sector: start sector
	127	* @nr_sects: number of sectors to write
	128	* @gfp_mask: memory allocation flags (for bio_alloc)
	129	* @page: page containing data to write
	130	*
	131	* Description:
	132	* Issue a write same request for the sectors in question.
	133	*/
	134	int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
	135	sector_t nr_sects, gfp_t gfp_mask,
	136	struct page *page)
	137	{
	138	struct request_queue *q = bdev_get_queue(bdev);
	139	unsigned int max_write_same_sectors;
	140	struct bio *bio = NULL;
	141	int ret = 0;
	142
	143	if (!q)
	144	return -ENXIO;
	145
	146	/* Ensure that max_write_same_sectors doesn't overflow bi_size */
	147	max_write_same_sectors = UINT_MAX >> 9;
	148
	149	while (nr_sects) {
	150	bio = next_bio(bio, REQ_WRITE \| REQ_WRITE_SAME, 1, gfp_mask);
	151	bio->bi_iter.bi_sector = sector;
	152	bio->bi_bdev = bdev;
	153	bio->bi_vcnt = 1;
	154	bio->bi_io_vec->bv_page = page;
	155	bio->bi_io_vec->bv_offset = 0;
	156	bio->bi_io_vec->bv_len = bdev_logical_block_size(bdev);
	157
	158	if (nr_sects > max_write_same_sectors) {
	159	bio->bi_iter.bi_size = max_write_same_sectors << 9;
	160	nr_sects -= max_write_same_sectors;
	161	sector += max_write_same_sectors;
	162	} else {
	163	bio->bi_iter.bi_size = nr_sects << 9;
	164	nr_sects = 0;
	165	}
	166	}
	167
	168	if (bio)
	169	ret = submit_bio_wait(REQ_WRITE \| REQ_WRITE_SAME, bio);
	170	return ret != -EOPNOTSUPP ? ret : 0;
	171	}
	172	EXPORT_SYMBOL(blkdev_issue_write_same);
	173
	174	/**
	175	* blkdev_issue_zeroout - generate number of zero filed write bios
	176	* @bdev: blockdev to issue
	177	* @sector: start sector
	178	* @nr_sects: number of sectors to write
	179	* @gfp_mask: memory allocation flags (for bio_alloc)
	180	*
	181	* Description:
	182	* Generate and issue number of bios with zerofiled pages.
	183	*/
	184
	185	static int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
	186	sector_t nr_sects, gfp_t gfp_mask)
	187	{
	188	int ret;
	189	struct bio *bio = NULL;
	190	unsigned int sz;
	191
	192	while (nr_sects != 0) {
	193	bio = next_bio(bio, WRITE,
	194	min(nr_sects, (sector_t)BIO_MAX_PAGES),
	195	gfp_mask);
	196	bio->bi_iter.bi_sector = sector;
	197	bio->bi_bdev = bdev;
	198
	199	while (nr_sects != 0) {
	200	sz = min((sector_t) PAGE_SIZE >> 9 , nr_sects);
	201	ret = bio_add_page(bio, ZERO_PAGE(0), sz << 9, 0);
	202	nr_sects -= ret >> 9;
	203	sector += ret >> 9;
	204	if (ret < (sz << 9))
	205	break;
	206	}
	207	}
	208
	209	if (bio)
	210	return submit_bio_wait(WRITE, bio);
	211	return 0;
	212	}
	213
	214	/**
	215	* blkdev_issue_zeroout - zero-fill a block range
	216	* @bdev: blockdev to write
	217	* @sector: start sector
	218	* @nr_sects: number of sectors to write
	219	* @gfp_mask: memory allocation flags (for bio_alloc)
	220	* @discard: whether to discard the block range
	221	*
	222	* Description:
	223	* Zero-fill a block range. If the discard flag is set and the block
	224	* device guarantees that subsequent READ operations to the block range
	225	* in question will return zeroes, the blocks will be discarded. Should
	226	* the discard request fail, if the discard flag is not set, or if
	227	* discard_zeroes_data is not supported, this function will resort to
	228	* zeroing the blocks manually, thus provisioning (allocating,
	229	* anchoring) them. If the block device supports the WRITE SAME command
	230	* blkdev_issue_zeroout() will use it to optimize the process of
	231	* clearing the block range. Otherwise the zeroing will be performed
	232	* using regular WRITE calls.
	233	*/
	234
	235	int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
	236	sector_t nr_sects, gfp_t gfp_mask, bool discard)
	237	{
	238	struct request_queue *q = bdev_get_queue(bdev);
	239
	240	if (discard && blk_queue_discard(q) && q->limits.discard_zeroes_data &&
	241	blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, 0) == 0)
	242	return 0;
	243
	244	if (bdev_write_same(bdev) &&
	245	blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask,
	246	ZERO_PAGE(0)) == 0)
	247	return 0;
	248
	249	return __blkdev_issue_zeroout(bdev, sector, nr_sects, gfp_mask);
	250	}
	251	EXPORT_SYMBOL(blkdev_issue_zeroout);