[linux-2.6-block.git] / drivers / md / md-linear.c

// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * linear.c : Multiple Devices driver for Linux Copyright (C) 1994-96 Marc
 * ZYNGIER <zyngier@ufr-info-p7.ibp.fr> or <maz@gloups.fdn.fr>
 */

#include <linux/blkdev.h>
#include <linux/seq_file.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <trace/events/block.h>
#include "md.h"

struct dev_info {
	struct md_rdev	*rdev;
	sector_t	end_sector;
};

struct linear_conf {
	struct rcu_head         rcu;
	sector_t                array_sectors;
	/* a copy of mddev->raid_disks */
	int                     raid_disks;
	struct dev_info         disks[] __counted_by(raid_disks);
};

/*
 * find which device holds a particular offset
 */
static inline struct dev_info *which_dev(struct mddev *mddev, sector_t sector)
{
	int lo, mid, hi;
	struct linear_conf *conf;

	lo = 0;
	hi = mddev->raid_disks - 1;
	conf = mddev->private;

	/*
	 * Binary Search
	 */

	while (hi > lo) {

		mid = (hi + lo) / 2;
		if (sector < conf->disks[mid].end_sector)
			hi = mid;
		else
			lo = mid + 1;
	}

	return conf->disks + lo;
}

static sector_t linear_size(struct mddev *mddev, sector_t sectors, int raid_disks)
{
	struct linear_conf *conf;
	sector_t array_sectors;

	conf = mddev->private;
	WARN_ONCE(sectors || raid_disks,
		  "%s does not support generic reshape\n", __func__);
	array_sectors = conf->array_sectors;

	return array_sectors;
}

static int linear_set_limits(struct mddev *mddev)
{
	struct queue_limits lim;
	int err;

	md_init_stacking_limits(&lim);
	lim.max_hw_sectors = mddev->chunk_sectors;
	lim.max_write_zeroes_sectors = mddev->chunk_sectors;
	lim.io_min = mddev->chunk_sectors << 9;
	err = mddev_stack_rdev_limits(mddev, &lim, MDDEV_STACK_INTEGRITY);
	if (err)
		return err;

	return queue_limits_set(mddev->gendisk->queue, &lim);
}

static struct linear_conf *linear_conf(struct mddev *mddev, int raid_disks)
{
	struct linear_conf *conf;
	struct md_rdev *rdev;
	int ret = -EINVAL;
	int cnt;
	int i;

	conf = kzalloc(struct_size(conf, disks, raid_disks), GFP_KERNEL);
	if (!conf)
		return ERR_PTR(-ENOMEM);

	/*
	 * conf->raid_disks is copy of mddev->raid_disks. The reason to
	 * keep a copy of mddev->raid_disks in struct linear_conf is,
	 * mddev->raid_disks may not be consistent with pointers number of
	 * conf->disks[] when it is updated in linear_add() and used to
	 * iterate old conf->disks[] earray in linear_congested().
	 * Here conf->raid_disks is always consitent with number of
	 * pointers in conf->disks[] array, and mddev->private is updated
	 * with rcu_assign_pointer() in linear_addr(), such race can be
	 * avoided.
	 */
	conf->raid_disks = raid_disks;

	cnt = 0;
	conf->array_sectors = 0;

	rdev_for_each(rdev, mddev) {
		int j = rdev->raid_disk;
		struct dev_info *disk = conf->disks + j;
		sector_t sectors;

		if (j < 0 || j >= raid_disks || disk->rdev) {
			pr_warn("md/linear:%s: disk numbering problem. Aborting!\n",
				mdname(mddev));
			goto out;
		}

		disk->rdev = rdev;
		if (mddev->chunk_sectors) {
			sectors = rdev->sectors;
			sector_div(sectors, mddev->chunk_sectors);
			rdev->sectors = sectors * mddev->chunk_sectors;
		}

		conf->array_sectors += rdev->sectors;
		cnt++;
	}
	if (cnt != raid_disks) {
		pr_warn("md/linear:%s: not enough drives present. Aborting!\n",
			mdname(mddev));
		goto out;
	}

	/*
	 * Here we calculate the device offsets.
	 */
	conf->disks[0].end_sector = conf->disks[0].rdev->sectors;

	for (i = 1; i < raid_disks; i++)
		conf->disks[i].end_sector =
			conf->disks[i-1].end_sector +
			conf->disks[i].rdev->sectors;

	if (!mddev_is_dm(mddev)) {
		ret = linear_set_limits(mddev);
		if (ret)
			goto out;
	}

	return conf;

out:
	kfree(conf);
	return ERR_PTR(ret);
}

static int linear_run(struct mddev *mddev)
{
	struct linear_conf *conf;
	int ret;

	if (md_check_no_bitmap(mddev))
		return -EINVAL;

	conf = linear_conf(mddev, mddev->raid_disks);
	if (IS_ERR(conf))
		return PTR_ERR(conf);

	mddev->private = conf;
	md_set_array_sectors(mddev, linear_size(mddev, 0, 0));

	ret =  md_integrity_register(mddev);
	if (ret) {
		kfree(conf);
		mddev->private = NULL;
	}
	return ret;
}

static int linear_add(struct mddev *mddev, struct md_rdev *rdev)
{
	/* Adding a drive to a linear array allows the array to grow.
	 * It is permitted if the new drive has a matching superblock
	 * already on it, with raid_disk equal to raid_disks.
	 * It is achieved by creating a new linear_private_data structure
	 * and swapping it in in-place of the current one.
	 * The current one is never freed until the array is stopped.
	 * This avoids races.
	 */
	struct linear_conf *newconf, *oldconf;

	if (rdev->saved_raid_disk != mddev->raid_disks)
		return -EINVAL;

	rdev->raid_disk = rdev->saved_raid_disk;
	rdev->saved_raid_disk = -1;

	newconf = linear_conf(mddev, mddev->raid_disks + 1);
	if (IS_ERR(newconf))
		return PTR_ERR(newconf);

	/* newconf->raid_disks already keeps a copy of * the increased
	 * value of mddev->raid_disks, WARN_ONCE() is just used to make
	 * sure of this. It is possible that oldconf is still referenced
	 * in linear_congested(), therefore kfree_rcu() is used to free
	 * oldconf until no one uses it anymore.
	 */
	oldconf = rcu_dereference_protected(mddev->private,
			lockdep_is_held(&mddev->reconfig_mutex));
	mddev->raid_disks++;
	WARN_ONCE(mddev->raid_disks != newconf->raid_disks,
		"copied raid_disks doesn't match mddev->raid_disks");
	rcu_assign_pointer(mddev->private, newconf);
	md_set_array_sectors(mddev, linear_size(mddev, 0, 0));
	set_capacity_and_notify(mddev->gendisk, mddev->array_sectors);
	kfree_rcu(oldconf, rcu);
	return 0;
}

static void linear_free(struct mddev *mddev, void *priv)
{
	struct linear_conf *conf = priv;

	kfree(conf);
}

static bool linear_make_request(struct mddev *mddev, struct bio *bio)
{
	struct dev_info *tmp_dev;
	sector_t start_sector, end_sector, data_offset;
	sector_t bio_sector = bio->bi_iter.bi_sector;

	if (unlikely(bio->bi_opf & REQ_PREFLUSH)
	    && md_flush_request(mddev, bio))
		return true;

	tmp_dev = which_dev(mddev, bio_sector);
	start_sector = tmp_dev->end_sector - tmp_dev->rdev->sectors;
	end_sector = tmp_dev->end_sector;
	data_offset = tmp_dev->rdev->data_offset;

	if (unlikely(bio_sector >= end_sector ||
		     bio_sector < start_sector))
		goto out_of_bounds;

	if (unlikely(is_rdev_broken(tmp_dev->rdev))) {
		md_error(mddev, tmp_dev->rdev);
		bio_io_error(bio);
		return true;
	}

	if (unlikely(bio_end_sector(bio) > end_sector)) {
		/* This bio crosses a device boundary, so we have to split it */
		struct bio *split = bio_split(bio, end_sector - bio_sector,
					      GFP_NOIO, &mddev->bio_set);

		if (IS_ERR(split)) {
			bio->bi_status = errno_to_blk_status(PTR_ERR(split));
			bio_endio(bio);
			return true;
		}

		bio_chain(split, bio);
		submit_bio_noacct(bio);
		bio = split;
	}

	md_account_bio(mddev, &bio);
	bio_set_dev(bio, tmp_dev->rdev->bdev);
	bio->bi_iter.bi_sector = bio->bi_iter.bi_sector -
		start_sector + data_offset;

	if (unlikely((bio_op(bio) == REQ_OP_DISCARD) &&
		     !bdev_max_discard_sectors(bio->bi_bdev))) {
		/* Just ignore it */
		bio_endio(bio);
	} else {
		if (mddev->gendisk)
			trace_block_bio_remap(bio, disk_devt(mddev->gendisk),
					      bio_sector);
		mddev_check_write_zeroes(mddev, bio);
		submit_bio_noacct(bio);
	}
	return true;

out_of_bounds:
	pr_err("md/linear:%s: make_request: Sector %llu out of bounds on dev %pg: %llu sectors, offset %llu\n",
	       mdname(mddev),
	       (unsigned long long)bio->bi_iter.bi_sector,
	       tmp_dev->rdev->bdev,
	       (unsigned long long)tmp_dev->rdev->sectors,
	       (unsigned long long)start_sector);
	bio_io_error(bio);
	return true;
}

static void linear_status(struct seq_file *seq, struct mddev *mddev)
{
	seq_printf(seq, " %dk rounding", mddev->chunk_sectors / 2);
}

static void linear_error(struct mddev *mddev, struct md_rdev *rdev)
{
	if (!test_and_set_bit(MD_BROKEN, &mddev->flags)) {
		char *md_name = mdname(mddev);

		pr_crit("md/linear%s: Disk failure on %pg detected, failing array.\n",
			md_name, rdev->bdev);
	}
}

static void linear_quiesce(struct mddev *mddev, int state)
{
}

static struct md_personality linear_personality = {
	.head = {
		.type	= MD_PERSONALITY,
		.id	= ID_LINEAR,
		.name	= "linear",
		.owner	= THIS_MODULE,
	},

	.make_request	= linear_make_request,
	.run		= linear_run,
	.free		= linear_free,
	.status		= linear_status,
	.hot_add_disk	= linear_add,
	.size		= linear_size,
	.quiesce	= linear_quiesce,
	.error_handler	= linear_error,
};

static int __init linear_init(void)
{
	return register_md_submodule(&linear_personality.head);
}

static void linear_exit(void)
{
	unregister_md_submodule(&linear_personality.head);
}

module_init(linear_init);
module_exit(linear_exit);
MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("Linear device concatenation personality for MD (deprecated)");
MODULE_ALIAS("md-personality-1"); /* LINEAR - deprecated*/
MODULE_ALIAS("md-linear");
MODULE_ALIAS("md-level--1");
Commit	Line	Data
127186cf YK	1	// SPDX-License-Identifier: GPL-2.0-or-later
	2	/*
	3	* linear.c : Multiple Devices driver for Linux Copyright (C) 1994-96 Marc
	4	* ZYNGIER <zyngier@ufr-info-p7.ibp.fr> or <maz@gloups.fdn.fr>
	5	*/
	6
	7	#include <linux/blkdev.h>
127186cf YK	8	#include <linux/seq_file.h>
	9	#include <linux/module.h>
	10	#include <linux/slab.h>
	11	#include <trace/events/block.h>
	12	#include "md.h"
	13
	14	struct dev_info {
	15	struct md_rdev *rdev;
	16	sector_t end_sector;
	17	};
	18
	19	struct linear_conf {
	20	struct rcu_head rcu;
	21	sector_t array_sectors;
	22	/* a copy of mddev->raid_disks */
	23	int raid_disks;
	24	struct dev_info disks[] __counted_by(raid_disks);
	25	};
	26
	27	/*
	28	* find which device holds a particular offset
	29	*/
	30	static inline struct dev_info which_dev(struct mddev mddev, sector_t sector)
	31	{
	32	int lo, mid, hi;
	33	struct linear_conf *conf;
	34
	35	lo = 0;
	36	hi = mddev->raid_disks - 1;
	37	conf = mddev->private;
	38
	39	/*
	40	* Binary Search
	41	*/
	42
	43	while (hi > lo) {
	44
	45	mid = (hi + lo) / 2;
	46	if (sector < conf->disks[mid].end_sector)
	47	hi = mid;
	48	else
	49	lo = mid + 1;
	50	}
	51
	52	return conf->disks + lo;
	53	}
	54
	55	static sector_t linear_size(struct mddev *mddev, sector_t sectors, int raid_disks)
	56	{
	57	struct linear_conf *conf;
	58	sector_t array_sectors;
	59
	60	conf = mddev->private;
	61	WARN_ONCE(sectors \|\| raid_disks,
	62	"%s does not support generic reshape\n", __func__);
	63	array_sectors = conf->array_sectors;
	64
	65	return array_sectors;
	66	}
	67
	68	static int linear_set_limits(struct mddev *mddev)
	69	{
	70	struct queue_limits lim;
	71	int err;
72
73	md_init_stacking_limits(&lim);
74	lim.max_hw_sectors = mddev->chunk_sectors;
75	lim.max_write_zeroes_sectors = mddev->chunk_sectors;
76	lim.io_min = mddev->chunk_sectors << 9;
77	err = mddev_stack_rdev_limits(mddev, &lim, MDDEV_STACK_INTEGRITY);
a572593a	78	if (err)
127186cf	79	return err;
127186cf YK	80
	81	return queue_limits_set(mddev->gendisk->queue, &lim);
	82	}
	83
	84	static struct linear_conf linear_conf(struct mddev mddev, int raid_disks)
	85	{
	86	struct linear_conf *conf;
	87	struct md_rdev *rdev;
	88	int ret = -EINVAL;
	89	int cnt;
	90	int i;
	91
	92	conf = kzalloc(struct_size(conf, disks, raid_disks), GFP_KERNEL);
	93	if (!conf)
	94	return ERR_PTR(-ENOMEM);
	95
	96	/*
	97	* conf->raid_disks is copy of mddev->raid_disks. The reason to
	98	* keep a copy of mddev->raid_disks in struct linear_conf is,
	99	* mddev->raid_disks may not be consistent with pointers number of
	100	* conf->disks[] when it is updated in linear_add() and used to
	101	* iterate old conf->disks[] earray in linear_congested().
	102	* Here conf->raid_disks is always consitent with number of
	103	* pointers in conf->disks[] array, and mddev->private is updated
	104	* with rcu_assign_pointer() in linear_addr(), such race can be
	105	* avoided.
	106	*/
	107	conf->raid_disks = raid_disks;
	108
	109	cnt = 0;
	110	conf->array_sectors = 0;
	111
	112	rdev_for_each(rdev, mddev) {
	113	int j = rdev->raid_disk;
	114	struct dev_info *disk = conf->disks + j;
	115	sector_t sectors;
	116
	117	if (j < 0 \|\| j >= raid_disks \|\| disk->rdev) {
	118	pr_warn("md/linear:%s: disk numbering problem. Aborting!\n",
	119	mdname(mddev));
	120	goto out;
	121	}
	122
	123	disk->rdev = rdev;
	124	if (mddev->chunk_sectors) {
	125	sectors = rdev->sectors;
	126	sector_div(sectors, mddev->chunk_sectors);
	127	rdev->sectors = sectors * mddev->chunk_sectors;
	128	}
	129
	130	conf->array_sectors += rdev->sectors;
	131	cnt++;
	132	}
	133	if (cnt != raid_disks) {
	134	pr_warn("md/linear:%s: not enough drives present. Aborting!\n",
	135	mdname(mddev));
	136	goto out;
	137	}
	138
	139	/*
	140	* Here we calculate the device offsets.
	141	*/
	142	conf->disks[0].end_sector = conf->disks[0].rdev->sectors;
	143
144	for (i = 1; i < raid_disks; i++)
145	conf->disks[i].end_sector =
146	conf->disks[i-1].end_sector +
147	conf->disks[i].rdev->sectors;
148
149	if (!mddev_is_dm(mddev)) {
150	ret = linear_set_limits(mddev);
151	if (ret)
152	goto out;
153	}
154
155	return conf;
156
157	out:
158	kfree(conf);
159	return ERR_PTR(ret);
160	}
161
162	static int linear_run(struct mddev *mddev)
163	{
164	struct linear_conf *conf;
165	int ret;
166
167	if (md_check_no_bitmap(mddev))
168	return -EINVAL;
169
170	conf = linear_conf(mddev, mddev->raid_disks);
171	if (IS_ERR(conf))
172	return PTR_ERR(conf);
173
174	mddev->private = conf;
175	md_set_array_sectors(mddev, linear_size(mddev, 0, 0));
176
177	ret = md_integrity_register(mddev);
178	if (ret) {
179	kfree(conf);
180	mddev->private = NULL;
181	}
182	return ret;
183	}
184
185	static int linear_add(struct mddev mddev, struct md_rdev rdev)
186	{
187	/* Adding a drive to a linear array allows the array to grow.
188	* It is permitted if the new drive has a matching superblock
189	* already on it, with raid_disk equal to raid_disks.
190	* It is achieved by creating a new linear_private_data structure
191	* and swapping it in in-place of the current one.
192	* The current one is never freed until the array is stopped.
193	* This avoids races.
194	*/
195	struct linear_conf newconf, oldconf;
196
197	if (rdev->saved_raid_disk != mddev->raid_disks)
198	return -EINVAL;
199
200	rdev->raid_disk = rdev->saved_raid_disk;
201	rdev->saved_raid_disk = -1;
202
203	newconf = linear_conf(mddev, mddev->raid_disks + 1);
62c55207 DC	204	if (IS_ERR(newconf))
62c55207 DC	205	return PTR_ERR(newconf);
127186cf YK	206
	207	/* newconf->raid_disks already keeps a copy of * the increased
	208	* value of mddev->raid_disks, WARN_ONCE() is just used to make
	209	* sure of this. It is possible that oldconf is still referenced
	210	* in linear_congested(), therefore kfree_rcu() is used to free
	211	* oldconf until no one uses it anymore.
	212	*/
	213	oldconf = rcu_dereference_protected(mddev->private,
	214	lockdep_is_held(&mddev->reconfig_mutex));
	215	mddev->raid_disks++;
	216	WARN_ONCE(mddev->raid_disks != newconf->raid_disks,
	217	"copied raid_disks doesn't match mddev->raid_disks");
	218	rcu_assign_pointer(mddev->private, newconf);
	219	md_set_array_sectors(mddev, linear_size(mddev, 0, 0));
	220	set_capacity_and_notify(mddev->gendisk, mddev->array_sectors);
	221	kfree_rcu(oldconf, rcu);
	222	return 0;
	223	}
	224
	225	static void linear_free(struct mddev mddev, void priv)
	226	{
	227	struct linear_conf *conf = priv;
	228
	229	kfree(conf);
	230	}
	231
	232	static bool linear_make_request(struct mddev mddev, struct bio bio)
	233	{
	234	struct dev_info *tmp_dev;
	235	sector_t start_sector, end_sector, data_offset;
	236	sector_t bio_sector = bio->bi_iter.bi_sector;
	237
	238	if (unlikely(bio->bi_opf & REQ_PREFLUSH)
	239	&& md_flush_request(mddev, bio))
	240	return true;
	241
	242	tmp_dev = which_dev(mddev, bio_sector);
	243	start_sector = tmp_dev->end_sector - tmp_dev->rdev->sectors;
	244	end_sector = tmp_dev->end_sector;
	245	data_offset = tmp_dev->rdev->data_offset;
	246
	247	if (unlikely(bio_sector >= end_sector \|\|
	248	bio_sector < start_sector))
	249	goto out_of_bounds;
	250
	251	if (unlikely(is_rdev_broken(tmp_dev->rdev))) {
	252	md_error(mddev, tmp_dev->rdev);
	253	bio_io_error(bio);
	254	return true;
	255	}
	256
	257	if (unlikely(bio_end_sector(bio) > end_sector)) {
	258	/* This bio crosses a device boundary, so we have to split it */
	259	struct bio *split = bio_split(bio, end_sector - bio_sector,
	260	GFP_NOIO, &mddev->bio_set);
	261
	262	if (IS_ERR(split)) {
	263	bio->bi_status = errno_to_blk_status(PTR_ERR(split));
	264	bio_endio(bio);
	265	return true;
	266	}
	267
	268	bio_chain(split, bio);
	269	submit_bio_noacct(bio);
270	bio = split;
271	}
272
273	md_account_bio(mddev, &bio);
274	bio_set_dev(bio, tmp_dev->rdev->bdev);
275	bio->bi_iter.bi_sector = bio->bi_iter.bi_sector -
276	start_sector + data_offset;
277
278	if (unlikely((bio_op(bio) == REQ_OP_DISCARD) &&
279	!bdev_max_discard_sectors(bio->bi_bdev))) {
280	/* Just ignore it */
281	bio_endio(bio);
282	} else {
283	if (mddev->gendisk)
284	trace_block_bio_remap(bio, disk_devt(mddev->gendisk),
285	bio_sector);
286	mddev_check_write_zeroes(mddev, bio);
287	submit_bio_noacct(bio);
288	}
289	return true;
290
291	out_of_bounds:
292	pr_err("md/linear:%s: make_request: Sector %llu out of bounds on dev %pg: %llu sectors, offset %llu\n",
293	mdname(mddev),
294	(unsigned long long)bio->bi_iter.bi_sector,
295	tmp_dev->rdev->bdev,
296	(unsigned long long)tmp_dev->rdev->sectors,
297	(unsigned long long)start_sector);
298	bio_io_error(bio);
299	return true;
300	}
301
302	static void linear_status(struct seq_file seq, struct mddev mddev)
303	{
304	seq_printf(seq, " %dk rounding", mddev->chunk_sectors / 2);
305	}
306
307	static void linear_error(struct mddev mddev, struct md_rdev rdev)
308	{
309	if (!test_and_set_bit(MD_BROKEN, &mddev->flags)) {
310	char *md_name = mdname(mddev);
311
312	pr_crit("md/linear%s: Disk failure on %pg detected, failing array.\n",
313	md_name, rdev->bdev);
314	}
315	}
316
317	static void linear_quiesce(struct mddev *mddev, int state)
318	{
319	}
320
321	static struct md_personality linear_personality = {
3d44e1d1 YK	322	.head = {
	323	.type = MD_PERSONALITY,
	324	.id = ID_LINEAR,
	325	.name = "linear",
	326	.owner = THIS_MODULE,
	327	},
	328
127186cf YK	329	.make_request = linear_make_request,
	330	.run = linear_run,
	331	.free = linear_free,
	332	.status = linear_status,
	333	.hot_add_disk = linear_add,
	334	.size = linear_size,
	335	.quiesce = linear_quiesce,
	336	.error_handler = linear_error,
	337	};
	338
	339	static int __init linear_init(void)
	340	{
3d44e1d1	341	return register_md_submodule(&linear_personality.head);
127186cf YK	342	}
	343
	344	static void linear_exit(void)
	345	{
3d44e1d1	346	unregister_md_submodule(&linear_personality.head);
127186cf YK	347	}
	348
	349	module_init(linear_init);
	350	module_exit(linear_exit);
	351	MODULE_LICENSE("GPL");
	352	MODULE_DESCRIPTION("Linear device concatenation personality for MD (deprecated)");
	353	MODULE_ALIAS("md-personality-1"); /* LINEAR - deprecated*/
	354	MODULE_ALIAS("md-linear");
	355	MODULE_ALIAS("md-level--1");