[GitHub/mt8127/android_kernel_alcatel_ttab.git] / drivers / md / linear.c

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

   Linear mode management functions.

   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2, or (at your option)
   any later version.
   
   You should have received a copy of the GNU General Public License
   (for example /usr/src/linux/COPYING); if not, write to the Free
   Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  
*/

#include <linux/module.h>

#include <linux/raid/md.h>
#include <linux/slab.h>
#include <linux/raid/linear.h>

#define MAJOR_NR MD_MAJOR
#define MD_DRIVER
#define MD_PERSONALITY

/*
 * find which device holds a particular offset 
 */
static inline dev_info_t *which_dev(mddev_t *mddev, sector_t sector)
{
	dev_info_t *hash;
	linear_conf_t *conf = mddev_to_conf(mddev);
	sector_t block = sector >> 1;

	/*
	 * sector_div(a,b) returns the remainer and sets a to a/b
	 */
	block >>= conf->preshift;
	(void)sector_div(block, conf->hash_spacing);
	hash = conf->hash_table[block];

	while ((sector>>1) >= (hash->size + hash->offset))
		hash++;
	return hash;
}

/**
 *	linear_mergeable_bvec -- tell bio layer if two requests can be merged
 *	@q: request queue
 *	@bio: the buffer head that's been built up so far
 *	@biovec: the request that could be merged to it.
 *
 *	Return amount of bytes we can take at this offset
 */
static int linear_mergeable_bvec(struct request_queue *q, struct bio *bio, struct bio_vec *biovec)
{
	mddev_t *mddev = q->queuedata;
	dev_info_t *dev0;
	unsigned long maxsectors, bio_sectors = bio->bi_size >> 9;
	sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);

	dev0 = which_dev(mddev, sector);
	maxsectors = (dev0->size << 1) - (sector - (dev0->offset<<1));

	if (maxsectors < bio_sectors)
		maxsectors = 0;
	else
		maxsectors -= bio_sectors;

	if (maxsectors <= (PAGE_SIZE >> 9 ) && bio_sectors == 0)
		return biovec->bv_len;
	/* The bytes available at this offset could be really big,
	 * so we cap at 2^31 to avoid overflow */
	if (maxsectors > (1 << (31-9)))
		return 1<<31;
	return maxsectors << 9;
}

static void linear_unplug(struct request_queue *q)
{
	mddev_t *mddev = q->queuedata;
	linear_conf_t *conf = mddev_to_conf(mddev);
	int i;

	for (i=0; i < mddev->raid_disks; i++) {
		struct request_queue *r_queue = bdev_get_queue(conf->disks[i].rdev->bdev);
		if (r_queue->unplug_fn)
			r_queue->unplug_fn(r_queue);
	}
}

static int linear_issue_flush(struct request_queue *q, struct gendisk *disk,
			      sector_t *error_sector)
{
	mddev_t *mddev = q->queuedata;
	linear_conf_t *conf = mddev_to_conf(mddev);
	int i, ret = 0;

	for (i=0; i < mddev->raid_disks && ret == 0; i++) {
		struct block_device *bdev = conf->disks[i].rdev->bdev;
		struct request_queue *r_queue = bdev_get_queue(bdev);

		if (!r_queue->issue_flush_fn)
			ret = -EOPNOTSUPP;
		else
			ret = r_queue->issue_flush_fn(r_queue, bdev->bd_disk, error_sector);
	}
	return ret;
}

static int linear_congested(void *data, int bits)
{
	mddev_t *mddev = data;
	linear_conf_t *conf = mddev_to_conf(mddev);
	int i, ret = 0;

	for (i = 0; i < mddev->raid_disks && !ret ; i++) {
		struct request_queue *q = bdev_get_queue(conf->disks[i].rdev->bdev);
		ret |= bdi_congested(&q->backing_dev_info, bits);
	}
	return ret;
}

static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
{
	linear_conf_t *conf;
	dev_info_t **table;
	mdk_rdev_t *rdev;
	int i, nb_zone, cnt;
	sector_t min_spacing;
	sector_t curr_offset;
	struct list_head *tmp;

	conf = kzalloc (sizeof (*conf) + raid_disks*sizeof(dev_info_t),
			GFP_KERNEL);
	if (!conf)
		return NULL;

	cnt = 0;
	conf->array_size = 0;

	ITERATE_RDEV(mddev,rdev,tmp) {
		int j = rdev->raid_disk;
		dev_info_t *disk = conf->disks + j;

		if (j < 0 || j > raid_disks || disk->rdev) {
			printk("linear: disk numbering problem. Aborting!\n");
			goto out;
		}

		disk->rdev = rdev;

		blk_queue_stack_limits(mddev->queue,
				       rdev->bdev->bd_disk->queue);
		/* as we don't honour merge_bvec_fn, we must never risk
		 * violating it, so limit ->max_sector to one PAGE, as
		 * a one page request is never in violation.
		 */
		if (rdev->bdev->bd_disk->queue->merge_bvec_fn &&
		    mddev->queue->max_sectors > (PAGE_SIZE>>9))
			blk_queue_max_sectors(mddev->queue, PAGE_SIZE>>9);

		disk->size = rdev->size;
		conf->array_size += rdev->size;

		cnt++;
	}
	if (cnt != raid_disks) {
		printk("linear: not enough drives present. Aborting!\n");
		goto out;
	}

	min_spacing = conf->array_size;
	sector_div(min_spacing, PAGE_SIZE/sizeof(struct dev_info *));

	/* min_spacing is the minimum spacing that will fit the hash
	 * table in one PAGE.  This may be much smaller than needed.
	 * We find the smallest non-terminal set of consecutive devices
	 * that is larger than min_spacing as use the size of that as
	 * the actual spacing
	 */
	conf->hash_spacing = conf->array_size;
	for (i=0; i < cnt-1 ; i++) {
		sector_t sz = 0;
		int j;
		for (j = i; j < cnt - 1 && sz < min_spacing; j++)
			sz += conf->disks[j].size;
		if (sz >= min_spacing && sz < conf->hash_spacing)
			conf->hash_spacing = sz;
	}

	/* hash_spacing may be too large for sector_div to work with,
	 * so we might need to pre-shift
	 */
	conf->preshift = 0;
	if (sizeof(sector_t) > sizeof(u32)) {
		sector_t space = conf->hash_spacing;
		while (space > (sector_t)(~(u32)0)) {
			space >>= 1;
			conf->preshift++;
		}
	}
	/*
	 * This code was restructured to work around a gcc-2.95.3 internal
	 * compiler error.  Alter it with care.
	 */
	{
		sector_t sz;
		unsigned round;
		unsigned long base;

		sz = conf->array_size >> conf->preshift;
		sz += 1; /* force round-up */
		base = conf->hash_spacing >> conf->preshift;
		round = sector_div(sz, base);
		nb_zone = sz + (round ? 1 : 0);
	}
	BUG_ON(nb_zone > PAGE_SIZE / sizeof(struct dev_info *));

	conf->hash_table = kmalloc (sizeof (struct dev_info *) * nb_zone,
					GFP_KERNEL);
	if (!conf->hash_table)
		goto out;

	/*
	 * Here we generate the linear hash table
	 * First calculate the device offsets.
	 */
	conf->disks[0].offset = 0;
	for (i = 1; i < raid_disks; i++)
		conf->disks[i].offset =
			conf->disks[i-1].offset +
			conf->disks[i-1].size;

	table = conf->hash_table;
	curr_offset = 0;
	i = 0;
	for (curr_offset = 0;
	     curr_offset < conf->array_size;
	     curr_offset += conf->hash_spacing) {

		while (i < raid_disks-1 &&
		       curr_offset >= conf->disks[i+1].offset)
			i++;

		*table ++ = conf->disks + i;
	}

	if (conf->preshift) {
		conf->hash_spacing >>= conf->preshift;
		/* round hash_spacing up so that when we divide by it,
		 * we err on the side of "too-low", which is safest.
		 */
		conf->hash_spacing++;
	}

	BUG_ON(table - conf->hash_table > nb_zone);

	return conf;

out:
	kfree(conf);
	return NULL;
}

static int linear_run (mddev_t *mddev)
{
	linear_conf_t *conf;

	conf = linear_conf(mddev, mddev->raid_disks);

	if (!conf)
		return 1;
	mddev->private = conf;
	mddev->array_size = conf->array_size;

	blk_queue_merge_bvec(mddev->queue, linear_mergeable_bvec);
	mddev->queue->unplug_fn = linear_unplug;
	mddev->queue->issue_flush_fn = linear_issue_flush;
	mddev->queue->backing_dev_info.congested_fn = linear_congested;
	mddev->queue->backing_dev_info.congested_data = mddev;
	return 0;
}

static int linear_add(mddev_t *mddev, mdk_rdev_t *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.
	 */
	linear_conf_t *newconf;

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

	rdev->raid_disk = rdev->saved_raid_disk;

	newconf = linear_conf(mddev,mddev->raid_disks+1);

	if (!newconf)
		return -ENOMEM;

	newconf->prev = mddev_to_conf(mddev);
	mddev->private = newconf;
	mddev->raid_disks++;
	mddev->array_size = newconf->array_size;
	set_capacity(mddev->gendisk, mddev->array_size << 1);
	return 0;
}

static int linear_stop (mddev_t *mddev)
{
	linear_conf_t *conf = mddev_to_conf(mddev);
  
	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
	do {
		linear_conf_t *t = conf->prev;
		kfree(conf->hash_table);
		kfree(conf);
		conf = t;
	} while (conf);

	return 0;
}

static int linear_make_request (struct request_queue *q, struct bio *bio)
{
	const int rw = bio_data_dir(bio);
	mddev_t *mddev = q->queuedata;
	dev_info_t *tmp_dev;
	sector_t block;

	if (unlikely(bio_barrier(bio))) {
		bio_endio(bio, -EOPNOTSUPP);
		return 0;
	}

	disk_stat_inc(mddev->gendisk, ios[rw]);
	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));

	tmp_dev = which_dev(mddev, bio->bi_sector);
	block = bio->bi_sector >> 1;
    
	if (unlikely(block >= (tmp_dev->size + tmp_dev->offset)
		     || block < tmp_dev->offset)) {
		char b[BDEVNAME_SIZE];

		printk("linear_make_request: Block %llu out of bounds on "
			"dev %s size %llu offset %llu\n",
			(unsigned long long)block,
			bdevname(tmp_dev->rdev->bdev, b),
			(unsigned long long)tmp_dev->size,
		        (unsigned long long)tmp_dev->offset);
		bio_io_error(bio);
		return 0;
	}
	if (unlikely(bio->bi_sector + (bio->bi_size >> 9) >
		     (tmp_dev->offset + tmp_dev->size)<<1)) {
		/* This bio crosses a device boundary, so we have to
		 * split it.
		 */
		struct bio_pair *bp;
		bp = bio_split(bio, bio_split_pool,
			       ((tmp_dev->offset + tmp_dev->size)<<1) - bio->bi_sector);
		if (linear_make_request(q, &bp->bio1))
			generic_make_request(&bp->bio1);
		if (linear_make_request(q, &bp->bio2))
			generic_make_request(&bp->bio2);
		bio_pair_release(bp);
		return 0;
	}
		    
	bio->bi_bdev = tmp_dev->rdev->bdev;
	bio->bi_sector = bio->bi_sector - (tmp_dev->offset << 1) + tmp_dev->rdev->data_offset;

	return 1;
}

static void linear_status (struct seq_file *seq, mddev_t *mddev)
{

#undef MD_DEBUG
#ifdef MD_DEBUG
	int j;
	linear_conf_t *conf = mddev_to_conf(mddev);
	sector_t s = 0;
  
	seq_printf(seq, "      ");
	for (j = 0; j < mddev->raid_disks; j++)
	{
		char b[BDEVNAME_SIZE];
		s += conf->smallest_size;
		seq_printf(seq, "[%s",
			   bdevname(conf->hash_table[j][0].rdev->bdev,b));

		while (s > conf->hash_table[j][0].offset +
		           conf->hash_table[j][0].size)
			seq_printf(seq, "/%s] ",
				   bdevname(conf->hash_table[j][1].rdev->bdev,b));
		else
			seq_printf(seq, "] ");
	}
	seq_printf(seq, "\n");
#endif
	seq_printf(seq, " %dk rounding", mddev->chunk_size/1024);
}


static struct mdk_personality linear_personality =
{
	.name		= "linear",
	.level		= LEVEL_LINEAR,
	.owner		= THIS_MODULE,
	.make_request	= linear_make_request,
	.run		= linear_run,
	.stop		= linear_stop,
	.status		= linear_status,
	.hot_add_disk	= linear_add,
};

static int __init linear_init (void)
{
	return register_md_personality (&linear_personality);
}

static void linear_exit (void)
{
	unregister_md_personality (&linear_personality);
}


module_init(linear_init);
module_exit(linear_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS("md-personality-1"); /* LINEAR - deprecated*/
MODULE_ALIAS("md-linear");
MODULE_ALIAS("md-level--1");
Commit	Line	Data
1da177e4 LT	1	/*
	2	linear.c : Multiple Devices driver for Linux
	3	Copyright (C) 1994-96 Marc ZYNGIER
	4	<zyngier@ufr-info-p7.ibp.fr> or
	5	<maz@gloups.fdn.fr>
	6
	7	Linear mode management functions.
	8
	9	This program is free software; you can redistribute it and/or modify
	10	it under the terms of the GNU General Public License as published by
	11	the Free Software Foundation; either version 2, or (at your option)
	12	any later version.
	13
	14	You should have received a copy of the GNU General Public License
	15	(for example /usr/src/linux/COPYING); if not, write to the Free
	16	Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
	17	*/
	18
	19	#include <linux/module.h>
	20
	21	#include <linux/raid/md.h>
	22	#include <linux/slab.h>
	23	#include <linux/raid/linear.h>
	24
	25	#define MAJOR_NR MD_MAJOR
	26	#define MD_DRIVER
	27	#define MD_PERSONALITY
	28
	29	/*
	30	* find which device holds a particular offset
	31	*/
	32	static inline dev_info_t which_dev(mddev_t mddev, sector_t sector)
	33	{
	34	dev_info_t *hash;
	35	linear_conf_t *conf = mddev_to_conf(mddev);
	36	sector_t block = sector >> 1;
	37
	38	/*
	39	* sector_div(a,b) returns the remainer and sets a to a/b
	40	*/
15945fee N	41	block >>= conf->preshift;
15945fee N	42	(void)sector_div(block, conf->hash_spacing);
1da177e4 LT	43	hash = conf->hash_table[block];
	44
	45	while ((sector>>1) >= (hash->size + hash->offset))
	46	hash++;
	47	return hash;
	48	}
	49
	50	/**
15945fee	51	* linear_mergeable_bvec -- tell bio layer if two requests can be merged
1da177e4 LT	52	* @q: request queue
	53	* @bio: the buffer head that's been built up so far
	54	* @biovec: the request that could be merged to it.
	55	*
	56	* Return amount of bytes we can take at this offset
	57	*/
165125e1	58	static int linear_mergeable_bvec(struct request_queue q, struct bio bio, struct bio_vec *biovec)
1da177e4 LT	59	{
	60	mddev_t *mddev = q->queuedata;
	61	dev_info_t *dev0;
	62	unsigned long maxsectors, bio_sectors = bio->bi_size >> 9;
	63	sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);
	64
	65	dev0 = which_dev(mddev, sector);
	66	maxsectors = (dev0->size << 1) - (sector - (dev0->offset<<1));
	67
	68	if (maxsectors < bio_sectors)
	69	maxsectors = 0;
	70	else
	71	maxsectors -= bio_sectors;
	72
	73	if (maxsectors <= (PAGE_SIZE >> 9 ) && bio_sectors == 0)
	74	return biovec->bv_len;
	75	/* The bytes available at this offset could be really big,
	76	* so we cap at 2^31 to avoid overflow */
	77	if (maxsectors > (1 << (31-9)))
	78	return 1<<31;
	79	return maxsectors << 9;
	80	}
	81
165125e1	82	static void linear_unplug(struct request_queue *q)
1da177e4 LT	83	{
	84	mddev_t *mddev = q->queuedata;
	85	linear_conf_t *conf = mddev_to_conf(mddev);
	86	int i;
	87
	88	for (i=0; i < mddev->raid_disks; i++) {
165125e1	89	struct request_queue *r_queue = bdev_get_queue(conf->disks[i].rdev->bdev);
1da177e4 LT	90	if (r_queue->unplug_fn)
	91	r_queue->unplug_fn(r_queue);
	92	}
	93	}
	94
165125e1	95	static int linear_issue_flush(struct request_queue q, struct gendisk disk,
1da177e4 LT	96	sector_t *error_sector)
	97	{
	98	mddev_t *mddev = q->queuedata;
	99	linear_conf_t *conf = mddev_to_conf(mddev);
	100	int i, ret = 0;
	101
	102	for (i=0; i < mddev->raid_disks && ret == 0; i++) {
	103	struct block_device *bdev = conf->disks[i].rdev->bdev;
165125e1	104	struct request_queue *r_queue = bdev_get_queue(bdev);
1da177e4 LT	105
	106	if (!r_queue->issue_flush_fn)
	107	ret = -EOPNOTSUPP;
	108	else
	109	ret = r_queue->issue_flush_fn(r_queue, bdev->bd_disk, error_sector);
	110	}
	111	return ret;
	112	}
	113
26be34dc N	114	static int linear_congested(void *data, int bits)
	115	{
	116	mddev_t *mddev = data;
	117	linear_conf_t *conf = mddev_to_conf(mddev);
	118	int i, ret = 0;
	119
	120	for (i = 0; i < mddev->raid_disks && !ret ; i++) {
165125e1	121	struct request_queue *q = bdev_get_queue(conf->disks[i].rdev->bdev);
26be34dc N	122	ret \|= bdi_congested(&q->backing_dev_info, bits);
	123	}
	124	return ret;
	125	}
	126
7c7546cc	127	static linear_conf_t linear_conf(mddev_t mddev, int raid_disks)
1da177e4 LT	128	{
	129	linear_conf_t *conf;
	130	dev_info_t **table;
	131	mdk_rdev_t *rdev;
	132	int i, nb_zone, cnt;
15945fee	133	sector_t min_spacing;
1da177e4 LT	134	sector_t curr_offset;
	135	struct list_head *tmp;
	136
7c7546cc	137	conf = kzalloc (sizeof (conf) + raid_diskssizeof(dev_info_t),
1da177e4 LT	138	GFP_KERNEL);
1da177e4 LT	139	if (!conf)
7c7546cc N	140	return NULL;
7c7546cc N	141
1da177e4	142	cnt = 0;
7c7546cc	143	conf->array_size = 0;
1da177e4 LT	144
	145	ITERATE_RDEV(mddev,rdev,tmp) {
	146	int j = rdev->raid_disk;
	147	dev_info_t *disk = conf->disks + j;
	148
7c7546cc	149	if (j < 0 \|\| j > raid_disks \|\| disk->rdev) {
1da177e4 LT	150	printk("linear: disk numbering problem. Aborting!\n");
	151	goto out;
	152	}
	153
	154	disk->rdev = rdev;
	155
	156	blk_queue_stack_limits(mddev->queue,
	157	rdev->bdev->bd_disk->queue);
	158	/* as we don't honour merge_bvec_fn, we must never risk
	159	* violating it, so limit ->max_sector to one PAGE, as
	160	* a one page request is never in violation.
	161	*/
	162	if (rdev->bdev->bd_disk->queue->merge_bvec_fn &&
	163	mddev->queue->max_sectors > (PAGE_SIZE>>9))
	164	blk_queue_max_sectors(mddev->queue, PAGE_SIZE>>9);
	165
	166	disk->size = rdev->size;
7c7546cc	167	conf->array_size += rdev->size;
1da177e4	168
1da177e4 LT	169	cnt++;
1da177e4 LT	170	}
7c7546cc	171	if (cnt != raid_disks) {
1da177e4 LT	172	printk("linear: not enough drives present. Aborting!\n");
	173	goto out;
	174	}
	175
f9abd1ac	176	min_spacing = conf->array_size;
15945fee N	177	sector_div(min_spacing, PAGE_SIZE/sizeof(struct dev_info *));
	178
	179	/* min_spacing is the minimum spacing that will fit the hash
	180	* table in one PAGE. This may be much smaller than needed.
	181	* We find the smallest non-terminal set of consecutive devices
	182	* that is larger than min_spacing as use the size of that as
	183	* the actual spacing
	184	*/
f9abd1ac	185	conf->hash_spacing = conf->array_size;
15945fee N	186	for (i=0; i < cnt-1 ; i++) {
	187	sector_t sz = 0;
	188	int j;
bed31ed9	189	for (j = i; j < cnt - 1 && sz < min_spacing; j++)
15945fee N	190	sz += conf->disks[j].size;
	191	if (sz >= min_spacing && sz < conf->hash_spacing)
	192	conf->hash_spacing = sz;
	193	}
	194
	195	/* hash_spacing may be too large for sector_div to work with,
	196	* so we might need to pre-shift
	197	*/
	198	conf->preshift = 0;
	199	if (sizeof(sector_t) > sizeof(u32)) {
	200	sector_t space = conf->hash_spacing;
	201	while (space > (sector_t)(~(u32)0)) {
	202	space >>= 1;
	203	conf->preshift++;
	204	}
	205	}
1da177e4 LT	206	/*
	207	* This code was restructured to work around a gcc-2.95.3 internal
	208	* compiler error. Alter it with care.
	209	*/
	210	{
	211	sector_t sz;
	212	unsigned round;
	213	unsigned long base;
	214
7c7546cc	215	sz = conf->array_size >> conf->preshift;
15945fee N	216	sz += 1; /* force round-up */
15945fee N	217	base = conf->hash_spacing >> conf->preshift;
1da177e4	218	round = sector_div(sz, base);
15945fee	219	nb_zone = sz + (round ? 1 : 0);
1da177e4	220	}
15945fee N	221	BUG_ON(nb_zone > PAGE_SIZE / sizeof(struct dev_info *));
	222
	223	conf->hash_table = kmalloc (sizeof (struct dev_info ) nb_zone,
1da177e4 LT	224	GFP_KERNEL);
	225	if (!conf->hash_table)
	226	goto out;
	227
	228	/*
	229	* Here we generate the linear hash table
15945fee	230	* First calculate the device offsets.
1da177e4	231	*/
15945fee	232	conf->disks[0].offset = 0;
a778b73f	233	for (i = 1; i < raid_disks; i++)
15945fee N	234	conf->disks[i].offset =
	235	conf->disks[i-1].offset +
	236	conf->disks[i-1].size;
	237
1da177e4	238	table = conf->hash_table;
1da177e4	239	curr_offset = 0;
15945fee N	240	i = 0;
15945fee N	241	for (curr_offset = 0;
f9abd1ac	242	curr_offset < conf->array_size;
15945fee	243	curr_offset += conf->hash_spacing) {
1da177e4	244
a778b73f	245	while (i < raid_disks-1 &&
15945fee N	246	curr_offset >= conf->disks[i+1].offset)
15945fee N	247	i++;
1da177e4	248
15945fee N	249	*table ++ = conf->disks + i;
	250	}
	251
	252	if (conf->preshift) {
	253	conf->hash_spacing >>= conf->preshift;
	254	/* round hash_spacing up so that when we divide by it,
	255	* we err on the side of "too-low", which is safest.
1da177e4	256	*/
15945fee	257	conf->hash_spacing++;
1da177e4	258	}
15945fee N	259
15945fee N	260	BUG_ON(table - conf->hash_table > nb_zone);
1da177e4	261
7c7546cc N	262	return conf;
	263
	264	out:
	265	kfree(conf);
	266	return NULL;
	267	}
	268
	269	static int linear_run (mddev_t *mddev)
	270	{
	271	linear_conf_t *conf;
	272
	273	conf = linear_conf(mddev, mddev->raid_disks);
	274
	275	if (!conf)
	276	return 1;
	277	mddev->private = conf;
	278	mddev->array_size = conf->array_size;
	279
1da177e4 LT	280	blk_queue_merge_bvec(mddev->queue, linear_mergeable_bvec);
	281	mddev->queue->unplug_fn = linear_unplug;
	282	mddev->queue->issue_flush_fn = linear_issue_flush;
26be34dc N	283	mddev->queue->backing_dev_info.congested_fn = linear_congested;
26be34dc N	284	mddev->queue->backing_dev_info.congested_data = mddev;
1da177e4	285	return 0;
7c7546cc	286	}
1da177e4	287
7c7546cc N	288	static int linear_add(mddev_t mddev, mdk_rdev_t rdev)
	289	{
	290	/* Adding a drive to a linear array allows the array to grow.
	291	* It is permitted if the new drive has a matching superblock
	292	* already on it, with raid_disk equal to raid_disks.
	293	* It is achieved by creating a new linear_private_data structure
	294	* and swapping it in in-place of the current one.
	295	* The current one is never freed until the array is stopped.
	296	* This avoids races.
	297	*/
	298	linear_conf_t *newconf;
	299
a778b73f	300	if (rdev->saved_raid_disk != mddev->raid_disks)
7c7546cc N	301	return -EINVAL;
7c7546cc N	302
a778b73f N	303	rdev->raid_disk = rdev->saved_raid_disk;
a778b73f N	304
7c7546cc N	305	newconf = linear_conf(mddev,mddev->raid_disks+1);
	306
	307	if (!newconf)
	308	return -ENOMEM;
	309
	310	newconf->prev = mddev_to_conf(mddev);
	311	mddev->private = newconf;
	312	mddev->raid_disks++;
	313	mddev->array_size = newconf->array_size;
	314	set_capacity(mddev->gendisk, mddev->array_size << 1);
	315	return 0;
1da177e4 LT	316	}
	317
	318	static int linear_stop (mddev_t *mddev)
	319	{
	320	linear_conf_t *conf = mddev_to_conf(mddev);
	321
	322	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
7c7546cc N	323	do {
	324	linear_conf_t *t = conf->prev;
	325	kfree(conf->hash_table);
	326	kfree(conf);
	327	conf = t;
	328	} while (conf);
1da177e4 LT	329
	330	return 0;
	331	}
	332
165125e1	333	static int linear_make_request (struct request_queue q, struct bio bio)
1da177e4	334	{
a362357b	335	const int rw = bio_data_dir(bio);
1da177e4 LT	336	mddev_t *mddev = q->queuedata;
	337	dev_info_t *tmp_dev;
	338	sector_t block;
	339
e5dcdd80	340	if (unlikely(bio_barrier(bio))) {
6712ecf8	341	bio_endio(bio, -EOPNOTSUPP);
e5dcdd80 N	342	return 0;
	343	}
	344
a362357b JA	345	disk_stat_inc(mddev->gendisk, ios[rw]);
a362357b JA	346	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
1da177e4 LT	347
	348	tmp_dev = which_dev(mddev, bio->bi_sector);
	349	block = bio->bi_sector >> 1;
	350
	351	if (unlikely(block >= (tmp_dev->size + tmp_dev->offset)
	352	\|\| block < tmp_dev->offset)) {
	353	char b[BDEVNAME_SIZE];
	354
	355	printk("linear_make_request: Block %llu out of bounds on "
	356	"dev %s size %llu offset %llu\n",
	357	(unsigned long long)block,
	358	bdevname(tmp_dev->rdev->bdev, b),
	359	(unsigned long long)tmp_dev->size,
	360	(unsigned long long)tmp_dev->offset);
6712ecf8	361	bio_io_error(bio);
1da177e4 LT	362	return 0;
	363	}
	364	if (unlikely(bio->bi_sector + (bio->bi_size >> 9) >
	365	(tmp_dev->offset + tmp_dev->size)<<1)) {
	366	/* This bio crosses a device boundary, so we have to
	367	* split it.
	368	*/
	369	struct bio_pair *bp;
29ac8e05 N	370	bp = bio_split(bio, bio_split_pool,
29ac8e05 N	371	((tmp_dev->offset + tmp_dev->size)<<1) - bio->bi_sector);
1da177e4 LT	372	if (linear_make_request(q, &bp->bio1))
	373	generic_make_request(&bp->bio1);
	374	if (linear_make_request(q, &bp->bio2))
	375	generic_make_request(&bp->bio2);
	376	bio_pair_release(bp);
	377	return 0;
	378	}
	379
	380	bio->bi_bdev = tmp_dev->rdev->bdev;
	381	bio->bi_sector = bio->bi_sector - (tmp_dev->offset << 1) + tmp_dev->rdev->data_offset;
	382
	383	return 1;
	384	}
	385
	386	static void linear_status (struct seq_file seq, mddev_t mddev)
	387	{
	388
	389	#undef MD_DEBUG
	390	#ifdef MD_DEBUG
	391	int j;
	392	linear_conf_t *conf = mddev_to_conf(mddev);
	393	sector_t s = 0;
	394
	395	seq_printf(seq, " ");
15945fee	396	for (j = 0; j < mddev->raid_disks; j++)
1da177e4 LT	397	{
	398	char b[BDEVNAME_SIZE];
	399	s += conf->smallest_size;
	400	seq_printf(seq, "[%s",
	401	bdevname(conf->hash_table[j][0].rdev->bdev,b));
	402
	403	while (s > conf->hash_table[j][0].offset +
	404	conf->hash_table[j][0].size)
	405	seq_printf(seq, "/%s] ",
	406	bdevname(conf->hash_table[j][1].rdev->bdev,b));
	407	else
	408	seq_printf(seq, "] ");
	409	}
	410	seq_printf(seq, "\n");
	411	#endif
	412	seq_printf(seq, " %dk rounding", mddev->chunk_size/1024);
	413	}
	414
	415
2604b703	416	static struct mdk_personality linear_personality =
1da177e4 LT	417	{
1da177e4 LT	418	.name = "linear",
2604b703	419	.level = LEVEL_LINEAR,
1da177e4 LT	420	.owner = THIS_MODULE,
	421	.make_request = linear_make_request,
	422	.run = linear_run,
	423	.stop = linear_stop,
	424	.status = linear_status,
7c7546cc	425	.hot_add_disk = linear_add,
1da177e4 LT	426	};
	427
	428	static int __init linear_init (void)
	429	{
2604b703	430	return register_md_personality (&linear_personality);
1da177e4 LT	431	}
	432
	433	static void linear_exit (void)
	434	{
2604b703	435	unregister_md_personality (&linear_personality);
1da177e4 LT	436	}
	437
	438
	439	module_init(linear_init);
	440	module_exit(linear_exit);
	441	MODULE_LICENSE("GPL");
d9d166c2 N	442	MODULE_ALIAS("md-personality-1"); /* LINEAR - deprecated*/
d9d166c2 N	443	MODULE_ALIAS("md-linear");
2604b703	444	MODULE_ALIAS("md-level--1");