Commit 79c63eeb authored by Linus Torvalds's avatar Linus Torvalds
Browse files

Merge tag 'dm-3.7-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-dm 

Pull device-mapper changes from Alasdair G Kergon:
 "Remove the power-of-2 block size constraint on discards in dm thin
  provisioning and factor the bio_prison code out into a separate module
  (for sharing with the forthcoming cache target).

  Use struct bio's front_pad to eliminate the use of one separate
  mempool by bio-based devices.

  A few other tiny clean-ups."

* tag 'dm-3.7-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-dm:
  dm: store dm_target_io in bio front_pad
  dm thin: move bio_prison code to separate module
  dm thin: prepare to separate bio_prison code
  dm thin: support discard with non power of two block size
  dm persistent data: convert to use le32_add_cpu
  dm: use ACCESS_ONCE for sysfs values
  dm bufio: use list_move
  dm mpath: fix check for null mpio in end_io fn
parents 6a5a3d6a dba14160

drivers/md/Kconfig

+8 −0
Original line number Diff line number Diff line
@@ -216,6 +216,13 @@ config DM_BUFIO 
	 as a cache, holding recently-read blocks in memory and performing

	 delayed writes.



config DM_BIO_PRISON

       tristate

       depends on BLK_DEV_DM && EXPERIMENTAL

       ---help---

	 Some bio locking schemes used by other device-mapper targets

	 including thin provisioning.



source "drivers/md/persistent-data/Kconfig"



config DM_CRYPT
@@ -247,6 +254,7 @@ config DM_THIN_PROVISIONING 
       tristate "Thin provisioning target (EXPERIMENTAL)"

       depends on BLK_DEV_DM && EXPERIMENTAL

       select DM_PERSISTENT_DATA

       select DM_BIO_PRISON

       ---help---

         Provides thin provisioning and snapshots that share a data store.

drivers/md/Makefile

+1 −0
Original line number Diff line number Diff line
@@ -29,6 +29,7 @@ obj-$(CONFIG_MD_FAULTY) += faulty.o 
obj-$(CONFIG_BLK_DEV_MD)	+= md-mod.o

obj-$(CONFIG_BLK_DEV_DM)	+= dm-mod.o

obj-$(CONFIG_DM_BUFIO)		+= dm-bufio.o

obj-$(CONFIG_DM_BIO_PRISON)	+= dm-bio-prison.o

obj-$(CONFIG_DM_CRYPT)		+= dm-crypt.o

obj-$(CONFIG_DM_DELAY)		+= dm-delay.o

obj-$(CONFIG_DM_FLAKEY)		+= dm-flakey.o

drivers/md/dm-bio-prison.c

0 → 100644
+415 −0
Original line number Diff line number Diff line 
/*

 * Copyright (C) 2012 Red Hat, Inc.

 *

 * This file is released under the GPL.

 */



#include "dm.h"

#include "dm-bio-prison.h"



#include <linux/spinlock.h>

#include <linux/mempool.h>

#include <linux/module.h>

#include <linux/slab.h>



/*----------------------------------------------------------------*/



struct dm_bio_prison_cell {

	struct hlist_node list;

	struct dm_bio_prison *prison;

	struct dm_cell_key key;

	struct bio *holder;

	struct bio_list bios;

};



struct dm_bio_prison {

	spinlock_t lock;

	mempool_t *cell_pool;



	unsigned nr_buckets;

	unsigned hash_mask;

	struct hlist_head *cells;

};



/*----------------------------------------------------------------*/



static uint32_t calc_nr_buckets(unsigned nr_cells)

{

	uint32_t n = 128;



	nr_cells /= 4;

	nr_cells = min(nr_cells, 8192u);



	while (n < nr_cells)

		n <<= 1;



	return n;

}



static struct kmem_cache *_cell_cache;



/*

 * @nr_cells should be the number of cells you want in use _concurrently_.

 * Don't confuse it with the number of distinct keys.

 */

struct dm_bio_prison *dm_bio_prison_create(unsigned nr_cells)

{

	unsigned i;

	uint32_t nr_buckets = calc_nr_buckets(nr_cells);

	size_t len = sizeof(struct dm_bio_prison) +

		(sizeof(struct hlist_head) * nr_buckets);

	struct dm_bio_prison *prison = kmalloc(len, GFP_KERNEL);



	if (!prison)

		return NULL;



	spin_lock_init(&prison->lock);

	prison->cell_pool = mempool_create_slab_pool(nr_cells, _cell_cache);

	if (!prison->cell_pool) {

		kfree(prison);

		return NULL;

	}



	prison->nr_buckets = nr_buckets;

	prison->hash_mask = nr_buckets - 1;

	prison->cells = (struct hlist_head *) (prison + 1);

	for (i = 0; i < nr_buckets; i++)

		INIT_HLIST_HEAD(prison->cells + i);



	return prison;

}

EXPORT_SYMBOL_GPL(dm_bio_prison_create);



void dm_bio_prison_destroy(struct dm_bio_prison *prison)

{

	mempool_destroy(prison->cell_pool);

	kfree(prison);

}

EXPORT_SYMBOL_GPL(dm_bio_prison_destroy);



static uint32_t hash_key(struct dm_bio_prison *prison, struct dm_cell_key *key)

{

	const unsigned long BIG_PRIME = 4294967291UL;

	uint64_t hash = key->block * BIG_PRIME;



	return (uint32_t) (hash & prison->hash_mask);

}



static int keys_equal(struct dm_cell_key *lhs, struct dm_cell_key *rhs)

{

	       return (lhs->virtual == rhs->virtual) &&

		       (lhs->dev == rhs->dev) &&

		       (lhs->block == rhs->block);

}



static struct dm_bio_prison_cell *__search_bucket(struct hlist_head *bucket,

						  struct dm_cell_key *key)

{

	struct dm_bio_prison_cell *cell;

	struct hlist_node *tmp;



	hlist_for_each_entry(cell, tmp, bucket, list)

		if (keys_equal(&cell->key, key))

			return cell;



	return NULL;

}



/*

 * This may block if a new cell needs allocating.  You must ensure that

 * cells will be unlocked even if the calling thread is blocked.

 *

 * Returns 1 if the cell was already held, 0 if @inmate is the new holder.

 */

int dm_bio_detain(struct dm_bio_prison *prison, struct dm_cell_key *key,

		  struct bio *inmate, struct dm_bio_prison_cell **ref)

{

	int r = 1;

	unsigned long flags;

	uint32_t hash = hash_key(prison, key);

	struct dm_bio_prison_cell *cell, *cell2;



	BUG_ON(hash > prison->nr_buckets);



	spin_lock_irqsave(&prison->lock, flags);



	cell = __search_bucket(prison->cells + hash, key);

	if (cell) {

		bio_list_add(&cell->bios, inmate);

		goto out;

	}



	/*

	 * Allocate a new cell

	 */

	spin_unlock_irqrestore(&prison->lock, flags);

	cell2 = mempool_alloc(prison->cell_pool, GFP_NOIO);

	spin_lock_irqsave(&prison->lock, flags);



	/*

	 * We've been unlocked, so we have to double check that

	 * nobody else has inserted this cell in the meantime.

	 */

	cell = __search_bucket(prison->cells + hash, key);

	if (cell) {

		mempool_free(cell2, prison->cell_pool);

		bio_list_add(&cell->bios, inmate);

		goto out;

	}



	/*

	 * Use new cell.

	 */

	cell = cell2;



	cell->prison = prison;

	memcpy(&cell->key, key, sizeof(cell->key));

	cell->holder = inmate;

	bio_list_init(&cell->bios);

	hlist_add_head(&cell->list, prison->cells + hash);



	r = 0;



out:

	spin_unlock_irqrestore(&prison->lock, flags);



	*ref = cell;



	return r;

}

EXPORT_SYMBOL_GPL(dm_bio_detain);



/*

 * @inmates must have been initialised prior to this call

 */

static void __cell_release(struct dm_bio_prison_cell *cell, struct bio_list *inmates)

{

	struct dm_bio_prison *prison = cell->prison;



	hlist_del(&cell->list);



	if (inmates) {

		bio_list_add(inmates, cell->holder);

		bio_list_merge(inmates, &cell->bios);

	}



	mempool_free(cell, prison->cell_pool);

}



void dm_cell_release(struct dm_bio_prison_cell *cell, struct bio_list *bios)

{

	unsigned long flags;

	struct dm_bio_prison *prison = cell->prison;



	spin_lock_irqsave(&prison->lock, flags);

	__cell_release(cell, bios);

	spin_unlock_irqrestore(&prison->lock, flags);

}

EXPORT_SYMBOL_GPL(dm_cell_release);



/*

 * There are a couple of places where we put a bio into a cell briefly

 * before taking it out again.  In these situations we know that no other

 * bio may be in the cell.  This function releases the cell, and also does

 * a sanity check.

 */

static void __cell_release_singleton(struct dm_bio_prison_cell *cell, struct bio *bio)

{

	BUG_ON(cell->holder != bio);

	BUG_ON(!bio_list_empty(&cell->bios));



	__cell_release(cell, NULL);

}



void dm_cell_release_singleton(struct dm_bio_prison_cell *cell, struct bio *bio)

{

	unsigned long flags;

	struct dm_bio_prison *prison = cell->prison;



	spin_lock_irqsave(&prison->lock, flags);

	__cell_release_singleton(cell, bio);

	spin_unlock_irqrestore(&prison->lock, flags);

}

EXPORT_SYMBOL_GPL(dm_cell_release_singleton);



/*

 * Sometimes we don't want the holder, just the additional bios.

 */

static void __cell_release_no_holder(struct dm_bio_prison_cell *cell, struct bio_list *inmates)

{

	struct dm_bio_prison *prison = cell->prison;



	hlist_del(&cell->list);

	bio_list_merge(inmates, &cell->bios);



	mempool_free(cell, prison->cell_pool);

}



void dm_cell_release_no_holder(struct dm_bio_prison_cell *cell, struct bio_list *inmates)

{

	unsigned long flags;

	struct dm_bio_prison *prison = cell->prison;



	spin_lock_irqsave(&prison->lock, flags);

	__cell_release_no_holder(cell, inmates);

	spin_unlock_irqrestore(&prison->lock, flags);

}

EXPORT_SYMBOL_GPL(dm_cell_release_no_holder);



void dm_cell_error(struct dm_bio_prison_cell *cell)

{

	struct dm_bio_prison *prison = cell->prison;

	struct bio_list bios;

	struct bio *bio;

	unsigned long flags;



	bio_list_init(&bios);



	spin_lock_irqsave(&prison->lock, flags);

	__cell_release(cell, &bios);

	spin_unlock_irqrestore(&prison->lock, flags);



	while ((bio = bio_list_pop(&bios)))

		bio_io_error(bio);

}

EXPORT_SYMBOL_GPL(dm_cell_error);



/*----------------------------------------------------------------*/



#define DEFERRED_SET_SIZE 64



struct dm_deferred_entry {

	struct dm_deferred_set *ds;

	unsigned count;

	struct list_head work_items;

};



struct dm_deferred_set {

	spinlock_t lock;

	unsigned current_entry;

	unsigned sweeper;

	struct dm_deferred_entry entries[DEFERRED_SET_SIZE];

};



struct dm_deferred_set *dm_deferred_set_create(void)

{

	int i;

	struct dm_deferred_set *ds;



	ds = kmalloc(sizeof(*ds), GFP_KERNEL);

	if (!ds)

		return NULL;



	spin_lock_init(&ds->lock);

	ds->current_entry = 0;

	ds->sweeper = 0;

	for (i = 0; i < DEFERRED_SET_SIZE; i++) {

		ds->entries[i].ds = ds;

		ds->entries[i].count = 0;

		INIT_LIST_HEAD(&ds->entries[i].work_items);

	}



	return ds;

}

EXPORT_SYMBOL_GPL(dm_deferred_set_create);



void dm_deferred_set_destroy(struct dm_deferred_set *ds)

{

	kfree(ds);

}

EXPORT_SYMBOL_GPL(dm_deferred_set_destroy);



struct dm_deferred_entry *dm_deferred_entry_inc(struct dm_deferred_set *ds)

{

	unsigned long flags;

	struct dm_deferred_entry *entry;



	spin_lock_irqsave(&ds->lock, flags);

	entry = ds->entries + ds->current_entry;

	entry->count++;

	spin_unlock_irqrestore(&ds->lock, flags);



	return entry;

}

EXPORT_SYMBOL_GPL(dm_deferred_entry_inc);



static unsigned ds_next(unsigned index)

{

	return (index + 1) % DEFERRED_SET_SIZE;

}



static void __sweep(struct dm_deferred_set *ds, struct list_head *head)

{

	while ((ds->sweeper != ds->current_entry) &&

	       !ds->entries[ds->sweeper].count) {

		list_splice_init(&ds->entries[ds->sweeper].work_items, head);

		ds->sweeper = ds_next(ds->sweeper);

	}



	if ((ds->sweeper == ds->current_entry) && !ds->entries[ds->sweeper].count)

		list_splice_init(&ds->entries[ds->sweeper].work_items, head);

}



void dm_deferred_entry_dec(struct dm_deferred_entry *entry, struct list_head *head)

{

	unsigned long flags;



	spin_lock_irqsave(&entry->ds->lock, flags);

	BUG_ON(!entry->count);

	--entry->count;

	__sweep(entry->ds, head);

	spin_unlock_irqrestore(&entry->ds->lock, flags);

}

EXPORT_SYMBOL_GPL(dm_deferred_entry_dec);



/*

 * Returns 1 if deferred or 0 if no pending items to delay job.

 */

int dm_deferred_set_add_work(struct dm_deferred_set *ds, struct list_head *work)

{

	int r = 1;

	unsigned long flags;

	unsigned next_entry;



	spin_lock_irqsave(&ds->lock, flags);

	if ((ds->sweeper == ds->current_entry) &&

	    !ds->entries[ds->current_entry].count)

		r = 0;

	else {

		list_add(work, &ds->entries[ds->current_entry].work_items);

		next_entry = ds_next(ds->current_entry);

		if (!ds->entries[next_entry].count)

			ds->current_entry = next_entry;

	}

	spin_unlock_irqrestore(&ds->lock, flags);



	return r;

}

EXPORT_SYMBOL_GPL(dm_deferred_set_add_work);



/*----------------------------------------------------------------*/



static int __init dm_bio_prison_init(void)

{

	_cell_cache = KMEM_CACHE(dm_bio_prison_cell, 0);

	if (!_cell_cache)

		return -ENOMEM;



	return 0;

}



static void __exit dm_bio_prison_exit(void)

{

	kmem_cache_destroy(_cell_cache);

	_cell_cache = NULL;

}



/*

 * module hooks

 */

module_init(dm_bio_prison_init);

module_exit(dm_bio_prison_exit);



MODULE_DESCRIPTION(DM_NAME " bio prison");

MODULE_AUTHOR("Joe Thornber <dm-devel@redhat.com>");

MODULE_LICENSE("GPL");

drivers/md/dm-bio-prison.h

0 → 100644
+72 −0
Original line number Diff line number Diff line 
/*

 * Copyright (C) 2011-2012 Red Hat, Inc.

 *

 * This file is released under the GPL.

 */



#ifndef DM_BIO_PRISON_H

#define DM_BIO_PRISON_H



#include "persistent-data/dm-block-manager.h" /* FIXME: for dm_block_t */

#include "dm-thin-metadata.h" /* FIXME: for dm_thin_id */



#include <linux/list.h>

#include <linux/bio.h>



/*----------------------------------------------------------------*/



/*

 * Sometimes we can't deal with a bio straight away.  We put them in prison

 * where they can't cause any mischief.  Bios are put in a cell identified

 * by a key, multiple bios can be in the same cell.  When the cell is

 * subsequently unlocked the bios become available.

 */

struct dm_bio_prison;

struct dm_bio_prison_cell;



/* FIXME: this needs to be more abstract */

struct dm_cell_key {

	int virtual;

	dm_thin_id dev;

	dm_block_t block;

};



struct dm_bio_prison *dm_bio_prison_create(unsigned nr_cells);

void dm_bio_prison_destroy(struct dm_bio_prison *prison);



/*

 * This may block if a new cell needs allocating.  You must ensure that

 * cells will be unlocked even if the calling thread is blocked.

 *

 * Returns 1 if the cell was already held, 0 if @inmate is the new holder.

 */

int dm_bio_detain(struct dm_bio_prison *prison, struct dm_cell_key *key,

		  struct bio *inmate, struct dm_bio_prison_cell **ref);



void dm_cell_release(struct dm_bio_prison_cell *cell, struct bio_list *bios);

void dm_cell_release_singleton(struct dm_bio_prison_cell *cell, struct bio *bio); // FIXME: bio arg not needed

void dm_cell_release_no_holder(struct dm_bio_prison_cell *cell, struct bio_list *inmates);

void dm_cell_error(struct dm_bio_prison_cell *cell);



/*----------------------------------------------------------------*/



/*

 * We use the deferred set to keep track of pending reads to shared blocks.

 * We do this to ensure the new mapping caused by a write isn't performed

 * until these prior reads have completed.  Otherwise the insertion of the

 * new mapping could free the old block that the read bios are mapped to.

 */



struct dm_deferred_set;

struct dm_deferred_entry;



struct dm_deferred_set *dm_deferred_set_create(void);

void dm_deferred_set_destroy(struct dm_deferred_set *ds);



struct dm_deferred_entry *dm_deferred_entry_inc(struct dm_deferred_set *ds);

void dm_deferred_entry_dec(struct dm_deferred_entry *entry, struct list_head *head);

int dm_deferred_set_add_work(struct dm_deferred_set *ds, struct list_head *work);



/*----------------------------------------------------------------*/



#endif

drivers/md/dm-bufio.c

+4 −9
Original line number Diff line number Diff line
@@ -280,9 +280,7 @@ static void __cache_size_refresh(void) 
	BUG_ON(!mutex_is_locked(&dm_bufio_clients_lock));

	BUG_ON(dm_bufio_client_count < 0);



	dm_bufio_cache_size_latch = dm_bufio_cache_size;



	barrier();

	dm_bufio_cache_size_latch = ACCESS_ONCE(dm_bufio_cache_size);



	/*

	 * Use default if set to 0 and report the actual cache size used.
@@ -441,8 +439,7 @@ static void __relink_lru(struct dm_buffer *b, int dirty) 
	c->n_buffers[b->list_mode]--;

	c->n_buffers[dirty]++;

	b->list_mode = dirty;

	list_del(&b->lru_list);

	list_add(&b->lru_list, &c->lru[dirty]);

	list_move(&b->lru_list, &c->lru[dirty]);

}



/*----------------------------------------------------------------
@@ -813,7 +810,7 @@ static void __get_memory_limit(struct dm_bufio_client *c, 
{

	unsigned long buffers;



	if (dm_bufio_cache_size != dm_bufio_cache_size_latch) {

	if (ACCESS_ONCE(dm_bufio_cache_size) != dm_bufio_cache_size_latch) {

		mutex_lock(&dm_bufio_clients_lock);

		__cache_size_refresh();

		mutex_unlock(&dm_bufio_clients_lock);
@@ -1591,11 +1588,9 @@ EXPORT_SYMBOL_GPL(dm_bufio_client_destroy); 


static void cleanup_old_buffers(void)

{

	unsigned long max_age = dm_bufio_max_age;

	unsigned long max_age = ACCESS_ONCE(dm_bufio_max_age);

	struct dm_bufio_client *c;



	barrier();



	if (max_age > ULONG_MAX / HZ)

		max_age = ULONG_MAX / HZ;

CRA Git | Maintained and supported by SUSTech CRA and CCSE