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

/*
 * Copyright (C) 2005-2007 Red Hat GmbH
 *
 * A target that delays reads and/or writes and can send
 * them to different devices.
 *
 * This file is released under the GPL.
 */

#include <linux/module.h>
#include <linux/init.h>
#include <linux/blkdev.h>
#include <linux/bio.h>
#include <linux/slab.h>

#include <linux/device-mapper.h>

#define DM_MSG_PREFIX "delay"

struct delay_c {
	struct timer_list delay_timer;
	struct mutex timer_lock;
	struct workqueue_struct *kdelayd_wq;
	struct work_struct flush_expired_bios;
	struct list_head delayed_bios;
	atomic_t may_delay;

	struct dm_dev *dev_read;
	sector_t start_read;
	unsigned read_delay;
	unsigned reads;

	struct dm_dev *dev_write;
	sector_t start_write;
	unsigned write_delay;
	unsigned writes;
};

struct dm_delay_info {
	struct delay_c *context;
	struct list_head list;
	unsigned long expires;
};

static DEFINE_MUTEX(delayed_bios_lock);

static void handle_delayed_timer(unsigned long data)
{
	struct delay_c *dc = (struct delay_c *)data;

	queue_work(dc->kdelayd_wq, &dc->flush_expired_bios);
}

static void queue_timeout(struct delay_c *dc, unsigned long expires)
{
	mutex_lock(&dc->timer_lock);

	if (!timer_pending(&dc->delay_timer) || expires < dc->delay_timer.expires)
		mod_timer(&dc->delay_timer, expires);

	mutex_unlock(&dc->timer_lock);
}

static void flush_bios(struct bio *bio)
{
	struct bio *n;

	while (bio) {
		n = bio->bi_next;
		bio->bi_next = NULL;
		generic_make_request(bio);
		bio = n;
	}
}

static struct bio *flush_delayed_bios(struct delay_c *dc, int flush_all)
{
	struct dm_delay_info *delayed, *next;
	unsigned long next_expires = 0;
	int start_timer = 0;
	struct bio_list flush_bios = { };

	mutex_lock(&delayed_bios_lock);
	list_for_each_entry_safe(delayed, next, &dc->delayed_bios, list) {
		if (flush_all || time_after_eq(jiffies, delayed->expires)) {
			struct bio *bio = dm_bio_from_per_bio_data(delayed,
						sizeof(struct dm_delay_info));
			list_del(&delayed->list);
			bio_list_add(&flush_bios, bio);
			if ((bio_data_dir(bio) == WRITE))
				delayed->context->writes--;
			else
				delayed->context->reads--;
			continue;
		}

		if (!start_timer) {
			start_timer = 1;
			next_expires = delayed->expires;
		} else
			next_expires = min(next_expires, delayed->expires);
	}

	mutex_unlock(&delayed_bios_lock);

	if (start_timer)
		queue_timeout(dc, next_expires);

	return bio_list_get(&flush_bios);
}

static void flush_expired_bios(struct work_struct *work)
{
	struct delay_c *dc;

	dc = container_of(work, struct delay_c, flush_expired_bios);
	flush_bios(flush_delayed_bios(dc, 0));
}

/*
 * Mapping parameters:
 *    <device> <offset> <delay> [<write_device> <write_offset> <write_delay>]
 *
 * With separate write parameters, the first set is only used for reads.
 * Delays are specified in milliseconds.
 */
static int delay_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
	struct delay_c *dc;
	unsigned long long tmpll;
	char dummy;

	if (argc != 3 && argc != 6) {
		ti->error = "requires exactly 3 or 6 arguments";
		return -EINVAL;
	}

	dc = kmalloc(sizeof(*dc), GFP_KERNEL);
	if (!dc) {
		ti->error = "Cannot allocate context";
		return -ENOMEM;
	}

	dc->reads = dc->writes = 0;

	if (sscanf(argv[1], "%llu%c", &tmpll, &dummy) != 1) {
		ti->error = "Invalid device sector";
		goto bad;
	}
	dc->start_read = tmpll;

	if (sscanf(argv[2], "%u%c", &dc->read_delay, &dummy) != 1) {
		ti->error = "Invalid delay";
		goto bad;
	}

	if (dm_get_device(ti, argv[0], dm_table_get_mode(ti->table),
			  &dc->dev_read)) {
		ti->error = "Device lookup failed";
		goto bad;
	}

	dc->dev_write = NULL;
	if (argc == 3)
		goto out;

	if (sscanf(argv[4], "%llu%c", &tmpll, &dummy) != 1) {
		ti->error = "Invalid write device sector";
		goto bad_dev_read;
	}
	dc->start_write = tmpll;

	if (sscanf(argv[5], "%u%c", &dc->write_delay, &dummy) != 1) {
		ti->error = "Invalid write delay";
		goto bad_dev_read;
	}

	if (dm_get_device(ti, argv[3], dm_table_get_mode(ti->table),
			  &dc->dev_write)) {
		ti->error = "Write device lookup failed";
		goto bad_dev_read;
	}

out:
	dc->kdelayd_wq = alloc_workqueue("kdelayd", WQ_MEM_RECLAIM, 0);
	if (!dc->kdelayd_wq) {
		DMERR("Couldn't start kdelayd");
		goto bad_queue;
	}

	setup_timer(&dc->delay_timer, handle_delayed_timer, (unsigned long)dc);

	INIT_WORK(&dc->flush_expired_bios, flush_expired_bios);
	INIT_LIST_HEAD(&dc->delayed_bios);
	mutex_init(&dc->timer_lock);
	atomic_set(&dc->may_delay, 1);

	ti->num_flush_bios = 1;
	ti->num_discard_bios = 1;
	ti->per_bio_data_size = sizeof(struct dm_delay_info);
	ti->private = dc;
	return 0;

bad_queue:
	if (dc->dev_write)
		dm_put_device(ti, dc->dev_write);
bad_dev_read:
	dm_put_device(ti, dc->dev_read);
bad:
	kfree(dc);
	return -EINVAL;
}

static void delay_dtr(struct dm_target *ti)
{
	struct delay_c *dc = ti->private;

	destroy_workqueue(dc->kdelayd_wq);

	dm_put_device(ti, dc->dev_read);

	if (dc->dev_write)
		dm_put_device(ti, dc->dev_write);

	kfree(dc);
}

static int delay_bio(struct delay_c *dc, int delay, struct bio *bio)
{
	struct dm_delay_info *delayed;
	unsigned long expires = 0;

	if (!delay || !atomic_read(&dc->may_delay))
		return 1;

	delayed = dm_per_bio_data(bio, sizeof(struct dm_delay_info));

	delayed->context = dc;
	delayed->expires = expires = jiffies + msecs_to_jiffies(delay);

	mutex_lock(&delayed_bios_lock);

	if (bio_data_dir(bio) == WRITE)
		dc->writes++;
	else
		dc->reads++;

	list_add_tail(&delayed->list, &dc->delayed_bios);

	mutex_unlock(&delayed_bios_lock);

	queue_timeout(dc, expires);

	return 0;
}

static void delay_presuspend(struct dm_target *ti)
{
	struct delay_c *dc = ti->private;

	atomic_set(&dc->may_delay, 0);
	del_timer_sync(&dc->delay_timer);
	flush_bios(flush_delayed_bios(dc, 1));
}

static void delay_resume(struct dm_target *ti)
{
	struct delay_c *dc = ti->private;

	atomic_set(&dc->may_delay, 1);
}

static int delay_map(struct dm_target *ti, struct bio *bio)
{
	struct delay_c *dc = ti->private;

	if ((bio_data_dir(bio) == WRITE) && (dc->dev_write)) {
		bio->bi_bdev = dc->dev_write->bdev;
		if (bio_sectors(bio))
			bio->bi_iter.bi_sector = dc->start_write +
				dm_target_offset(ti, bio->bi_iter.bi_sector);

		return delay_bio(dc, dc->write_delay, bio);
	}

	bio->bi_bdev = dc->dev_read->bdev;
	bio->bi_iter.bi_sector = dc->start_read +
		dm_target_offset(ti, bio->bi_iter.bi_sector);

	return delay_bio(dc, dc->read_delay, bio);
}

static void delay_status(struct dm_target *ti, status_type_t type,
			 unsigned status_flags, char *result, unsigned maxlen)
{
	struct delay_c *dc = ti->private;
	int sz = 0;

	switch (type) {
	case STATUSTYPE_INFO:
		DMEMIT("%u %u", dc->reads, dc->writes);
		break;

	case STATUSTYPE_TABLE:
		DMEMIT("%s %llu %u", dc->dev_read->name,
		       (unsigned long long) dc->start_read,
		       dc->read_delay);
		if (dc->dev_write)
			DMEMIT(" %s %llu %u", dc->dev_write->name,
			       (unsigned long long) dc->start_write,
			       dc->write_delay);
		break;
	}
}

static int delay_iterate_devices(struct dm_target *ti,
				 iterate_devices_callout_fn fn, void *data)
{
	struct delay_c *dc = ti->private;
	int ret = 0;

	ret = fn(ti, dc->dev_read, dc->start_read, ti->len, data);
	if (ret)
		goto out;

	if (dc->dev_write)
		ret = fn(ti, dc->dev_write, dc->start_write, ti->len, data);

out:
	return ret;
}

static struct target_type delay_target = {
	.name	     = "delay",
	.version     = {1, 2, 1},
	.module      = THIS_MODULE,
	.ctr	     = delay_ctr,
	.dtr	     = delay_dtr,
	.map	     = delay_map,
	.presuspend  = delay_presuspend,
	.resume	     = delay_resume,
	.status	     = delay_status,
	.iterate_devices = delay_iterate_devices,
};

static int __init dm_delay_init(void)
{
	int r;

	r = dm_register_target(&delay_target);
	if (r < 0) {
		DMERR("register failed %d", r);
		goto bad_register;
	}

	return 0;

bad_register:
	return r;
}

static void __exit dm_delay_exit(void)
{
	dm_unregister_target(&delay_target);
}

/* Module hooks */
module_init(dm_delay_init);
module_exit(dm_delay_exit);

MODULE_DESCRIPTION(DM_NAME " delay target");
MODULE_AUTHOR("Heinz Mauelshagen <mauelshagen@redhat.com>");
MODULE_LICENSE("GPL");
Commit	Line	Data
26b9f228 HM	1	/*
	2	* Copyright (C) 2005-2007 Red Hat GmbH
	3	*
	4	* A target that delays reads and/or writes and can send
	5	* them to different devices.
	6	*
	7	* This file is released under the GPL.
	8	*/
	9
	10	#include <linux/module.h>
	11	#include <linux/init.h>
	12	#include <linux/blkdev.h>
	13	#include <linux/bio.h>
	14	#include <linux/slab.h>
	15
586e80e6 MP	16	#include <linux/device-mapper.h>
586e80e6 MP	17
26b9f228 HM	18	#define DM_MSG_PREFIX "delay"
	19
	20	struct delay_c {
	21	struct timer_list delay_timer;
ac818646	22	struct mutex timer_lock;
718822c1	23	struct workqueue_struct *kdelayd_wq;
26b9f228 HM	24	struct work_struct flush_expired_bios;
	25	struct list_head delayed_bios;
	26	atomic_t may_delay;
26b9f228 HM	27
	28	struct dm_dev *dev_read;
	29	sector_t start_read;
	30	unsigned read_delay;
	31	unsigned reads;
	32
	33	struct dm_dev *dev_write;
	34	sector_t start_write;
	35	unsigned write_delay;
	36	unsigned writes;
	37	};
	38
028867ac	39	struct dm_delay_info {
26b9f228 HM	40	struct delay_c *context;
26b9f228 HM	41	struct list_head list;
26b9f228 HM	42	unsigned long expires;
	43	};
	44
	45	static DEFINE_MUTEX(delayed_bios_lock);
	46
26b9f228 HM	47	static void handle_delayed_timer(unsigned long data)
	48	{
	49	struct delay_c dc = (struct delay_c )data;
	50
718822c1	51	queue_work(dc->kdelayd_wq, &dc->flush_expired_bios);
26b9f228 HM	52	}
	53
	54	static void queue_timeout(struct delay_c *dc, unsigned long expires)
	55	{
ac818646	56	mutex_lock(&dc->timer_lock);
26b9f228 HM	57
	58	if (!timer_pending(&dc->delay_timer) \|\| expires < dc->delay_timer.expires)
	59	mod_timer(&dc->delay_timer, expires);
	60
ac818646	61	mutex_unlock(&dc->timer_lock);
26b9f228 HM	62	}
	63
	64	static void flush_bios(struct bio *bio)
	65	{
	66	struct bio *n;
	67
	68	while (bio) {
	69	n = bio->bi_next;
	70	bio->bi_next = NULL;
	71	generic_make_request(bio);
	72	bio = n;
	73	}
	74	}
	75
	76	static struct bio flush_delayed_bios(struct delay_c dc, int flush_all)
	77	{
028867ac	78	struct dm_delay_info delayed, next;
26b9f228 HM	79	unsigned long next_expires = 0;
26b9f228 HM	80	int start_timer = 0;
051814c6	81	struct bio_list flush_bios = { };
26b9f228 HM	82
	83	mutex_lock(&delayed_bios_lock);
	84	list_for_each_entry_safe(delayed, next, &dc->delayed_bios, list) {
	85	if (flush_all \|\| time_after_eq(jiffies, delayed->expires)) {
42065460 MP	86	struct bio *bio = dm_bio_from_per_bio_data(delayed,
42065460 MP	87	sizeof(struct dm_delay_info));
26b9f228	88	list_del(&delayed->list);
42065460 MP	89	bio_list_add(&flush_bios, bio);
42065460 MP	90	if ((bio_data_dir(bio) == WRITE))
26b9f228 HM	91	delayed->context->writes--;
	92	else
	93	delayed->context->reads--;
26b9f228 HM	94	continue;
	95	}
	96
	97	if (!start_timer) {
	98	start_timer = 1;
	99	next_expires = delayed->expires;
	100	} else
	101	next_expires = min(next_expires, delayed->expires);
	102	}
	103
	104	mutex_unlock(&delayed_bios_lock);
	105
	106	if (start_timer)
	107	queue_timeout(dc, next_expires);
	108
	109	return bio_list_get(&flush_bios);
	110	}
	111
	112	static void flush_expired_bios(struct work_struct *work)
	113	{
	114	struct delay_c *dc;
	115
	116	dc = container_of(work, struct delay_c, flush_expired_bios);
	117	flush_bios(flush_delayed_bios(dc, 0));
	118	}
	119
	120	/*
	121	* Mapping parameters:
	122	* <device> <offset> <delay> [<write_device> <write_offset> <write_delay>]
	123	*
	124	* With separate write parameters, the first set is only used for reads.
	125	* Delays are specified in milliseconds.
	126	*/
	127	static int delay_ctr(struct dm_target ti, unsigned int argc, char *argv)
	128	{
	129	struct delay_c *dc;
	130	unsigned long long tmpll;
31998ef1	131	char dummy;
26b9f228 HM	132
	133	if (argc != 3 && argc != 6) {
	134	ti->error = "requires exactly 3 or 6 arguments";
	135	return -EINVAL;
	136	}
	137
	138	dc = kmalloc(sizeof(*dc), GFP_KERNEL);
	139	if (!dc) {
	140	ti->error = "Cannot allocate context";
	141	return -ENOMEM;
	142	}
	143
	144	dc->reads = dc->writes = 0;
	145
31998ef1	146	if (sscanf(argv[1], "%llu%c", &tmpll, &dummy) != 1) {
26b9f228 HM	147	ti->error = "Invalid device sector";
	148	goto bad;
	149	}
	150	dc->start_read = tmpll;
	151
31998ef1	152	if (sscanf(argv[2], "%u%c", &dc->read_delay, &dummy) != 1) {
26b9f228 HM	153	ti->error = "Invalid delay";
	154	goto bad;
	155	}
	156
8215d6ec NK	157	if (dm_get_device(ti, argv[0], dm_table_get_mode(ti->table),
8215d6ec NK	158	&dc->dev_read)) {
26b9f228 HM	159	ti->error = "Device lookup failed";
	160	goto bad;
	161	}
	162
2e64a0f9 DM	163	dc->dev_write = NULL;
2e64a0f9 DM	164	if (argc == 3)
26b9f228	165	goto out;
26b9f228	166
31998ef1	167	if (sscanf(argv[4], "%llu%c", &tmpll, &dummy) != 1) {
26b9f228	168	ti->error = "Invalid write device sector";
2e64a0f9	169	goto bad_dev_read;
26b9f228 HM	170	}
	171	dc->start_write = tmpll;
	172
31998ef1	173	if (sscanf(argv[5], "%u%c", &dc->write_delay, &dummy) != 1) {
26b9f228	174	ti->error = "Invalid write delay";
2e64a0f9	175	goto bad_dev_read;
26b9f228 HM	176	}
26b9f228 HM	177
8215d6ec NK	178	if (dm_get_device(ti, argv[3], dm_table_get_mode(ti->table),
8215d6ec NK	179	&dc->dev_write)) {
26b9f228	180	ti->error = "Write device lookup failed";
2e64a0f9	181	goto bad_dev_read;
26b9f228 HM	182	}
	183
	184	out:
718822c1 MP	185	dc->kdelayd_wq = alloc_workqueue("kdelayd", WQ_MEM_RECLAIM, 0);
	186	if (!dc->kdelayd_wq) {
	187	DMERR("Couldn't start kdelayd");
	188	goto bad_queue;
	189	}
	190
ac818646	191	setup_timer(&dc->delay_timer, handle_delayed_timer, (unsigned long)dc);
26b9f228 HM	192
	193	INIT_WORK(&dc->flush_expired_bios, flush_expired_bios);
	194	INIT_LIST_HEAD(&dc->delayed_bios);
ac818646	195	mutex_init(&dc->timer_lock);
26b9f228 HM	196	atomic_set(&dc->may_delay, 1);
26b9f228 HM	197
55a62eef AK	198	ti->num_flush_bios = 1;
55a62eef AK	199	ti->num_discard_bios = 1;
42065460	200	ti->per_bio_data_size = sizeof(struct dm_delay_info);
26b9f228 HM	201	ti->private = dc;
	202	return 0;
	203
718822c1	204	bad_queue:
2e64a0f9 DM	205	if (dc->dev_write)
	206	dm_put_device(ti, dc->dev_write);
	207	bad_dev_read:
	208	dm_put_device(ti, dc->dev_read);
26b9f228 HM	209	bad:
	210	kfree(dc);
	211	return -EINVAL;
	212	}
	213
	214	static void delay_dtr(struct dm_target *ti)
	215	{
	216	struct delay_c *dc = ti->private;
	217
718822c1	218	destroy_workqueue(dc->kdelayd_wq);
26b9f228 HM	219
	220	dm_put_device(ti, dc->dev_read);
	221
	222	if (dc->dev_write)
	223	dm_put_device(ti, dc->dev_write);
	224
26b9f228 HM	225	kfree(dc);
	226	}
	227
	228	static int delay_bio(struct delay_c dc, int delay, struct bio bio)
	229	{
028867ac	230	struct dm_delay_info *delayed;
26b9f228 HM	231	unsigned long expires = 0;
	232
	233	if (!delay \|\| !atomic_read(&dc->may_delay))
	234	return 1;
	235
42065460	236	delayed = dm_per_bio_data(bio, sizeof(struct dm_delay_info));
26b9f228 HM	237
26b9f228 HM	238	delayed->context = dc;
aca607ba	239	delayed->expires = expires = jiffies + msecs_to_jiffies(delay);
26b9f228 HM	240
	241	mutex_lock(&delayed_bios_lock);
	242
	243	if (bio_data_dir(bio) == WRITE)
	244	dc->writes++;
	245	else
	246	dc->reads++;
	247
	248	list_add_tail(&delayed->list, &dc->delayed_bios);
	249
	250	mutex_unlock(&delayed_bios_lock);
	251
	252	queue_timeout(dc, expires);
	253
	254	return 0;
	255	}
	256
	257	static void delay_presuspend(struct dm_target *ti)
	258	{
	259	struct delay_c *dc = ti->private;
	260
	261	atomic_set(&dc->may_delay, 0);
	262	del_timer_sync(&dc->delay_timer);
	263	flush_bios(flush_delayed_bios(dc, 1));
	264	}
	265
	266	static void delay_resume(struct dm_target *ti)
	267	{
	268	struct delay_c *dc = ti->private;
	269
	270	atomic_set(&dc->may_delay, 1);
	271	}
	272
7de3ee57	273	static int delay_map(struct dm_target ti, struct bio bio)
26b9f228 HM	274	{
	275	struct delay_c *dc = ti->private;
	276
	277	if ((bio_data_dir(bio) == WRITE) && (dc->dev_write)) {
	278	bio->bi_bdev = dc->dev_write->bdev;
c927259e	279	if (bio_sectors(bio))
4f024f37 KO	280	bio->bi_iter.bi_sector = dc->start_write +
4f024f37 KO	281	dm_target_offset(ti, bio->bi_iter.bi_sector);
26b9f228 HM	282
	283	return delay_bio(dc, dc->write_delay, bio);
	284	}
	285
	286	bio->bi_bdev = dc->dev_read->bdev;
4f024f37 KO	287	bio->bi_iter.bi_sector = dc->start_read +
4f024f37 KO	288	dm_target_offset(ti, bio->bi_iter.bi_sector);
26b9f228 HM	289
	290	return delay_bio(dc, dc->read_delay, bio);
	291	}
	292
fd7c092e MP	293	static void delay_status(struct dm_target *ti, status_type_t type,
fd7c092e MP	294	unsigned status_flags, char *result, unsigned maxlen)
26b9f228 HM	295	{
	296	struct delay_c *dc = ti->private;
	297	int sz = 0;
	298
	299	switch (type) {
	300	case STATUSTYPE_INFO:
	301	DMEMIT("%u %u", dc->reads, dc->writes);
	302	break;
	303
	304	case STATUSTYPE_TABLE:
	305	DMEMIT("%s %llu %u", dc->dev_read->name,
	306	(unsigned long long) dc->start_read,
	307	dc->read_delay);
	308	if (dc->dev_write)
79662d1e	309	DMEMIT(" %s %llu %u", dc->dev_write->name,
26b9f228 HM	310	(unsigned long long) dc->start_write,
	311	dc->write_delay);
	312	break;
	313	}
26b9f228 HM	314	}
26b9f228 HM	315
af4874e0 MS	316	static int delay_iterate_devices(struct dm_target *ti,
	317	iterate_devices_callout_fn fn, void *data)
	318	{
	319	struct delay_c *dc = ti->private;
	320	int ret = 0;
	321
5dea271b	322	ret = fn(ti, dc->dev_read, dc->start_read, ti->len, data);
af4874e0 MS	323	if (ret)
	324	goto out;
	325
	326	if (dc->dev_write)
5dea271b	327	ret = fn(ti, dc->dev_write, dc->start_write, ti->len, data);
af4874e0 MS	328
	329	out:
	330	return ret;
	331	}
	332
26b9f228 HM	333	static struct target_type delay_target = {
26b9f228 HM	334	.name = "delay",
fd7c092e	335	.version = {1, 2, 1},
26b9f228 HM	336	.module = THIS_MODULE,
	337	.ctr = delay_ctr,
	338	.dtr = delay_dtr,
	339	.map = delay_map,
	340	.presuspend = delay_presuspend,
	341	.resume = delay_resume,
	342	.status = delay_status,
af4874e0	343	.iterate_devices = delay_iterate_devices,
26b9f228 HM	344	};
	345
	346	static int __init dm_delay_init(void)
	347	{
42065460	348	int r;
26b9f228 HM	349
	350	r = dm_register_target(&delay_target);
	351	if (r < 0) {
	352	DMERR("register failed %d", r);
	353	goto bad_register;
	354	}
	355
	356	return 0;
	357
	358	bad_register:
26b9f228 HM	359	return r;
	360	}
	361
	362	static void __exit dm_delay_exit(void)
	363	{
10d3bd09	364	dm_unregister_target(&delay_target);
26b9f228 HM	365	}
	366
	367	/* Module hooks */
	368	module_init(dm_delay_init);
	369	module_exit(dm_delay_exit);
	370
	371	MODULE_DESCRIPTION(DM_NAME " delay target");
	372	MODULE_AUTHOR("Heinz Mauelshagen <mauelshagen@redhat.com>");
	373	MODULE_LICENSE("GPL");