From: Mike Snitzer <snitzer@redhat.com>
To: dm-devel@redhat.com
Cc: ejt@redhat.com
Subject: [PATCH 2/2 v2] dm thin: elevate the bio_prison and deferred_set code to allow for re-use
Date: Fri, 2 Mar 2012 22:13:21 -0500 [thread overview]
Message-ID: <20120303031320.GA20103@redhat.com> (raw)
In-Reply-To: <1330732770-18621-2-git-send-email-snitzer@redhat.com>
Allow other DM targets to make use of the bio_prison and deferred_set
infrastructure. Code elevated to drivers/md/dm-bio-prison.[hc]
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
---
drivers/md/Makefile | 2
drivers/md/dm-bio-prison.c | 336 ++++++++++++++++++++++++++++++++
drivers/md/dm-bio-prison.h | 109 ++++++++++
drivers/md/dm-thin-metadata.h | 6
drivers/md/dm-thin.c | 438 +++---------------------------------------
5 files changed, 483 insertions(+), 408 deletions(-)
create mode 100644 drivers/md/dm-bio-prison.c
create mode 100644 drivers/md/dm-bio-prison.h
Index: linux-2.6/drivers/md/Makefile
===================================================================
--- linux-2.6.orig/drivers/md/Makefile
+++ linux-2.6/drivers/md/Makefile
@@ -3,7 +3,7 @@
#
dm-mod-y += dm.o dm-table.o dm-target.o dm-linear.o dm-stripe.o \
- dm-ioctl.o dm-io.o dm-kcopyd.o dm-sysfs.o
+ dm-ioctl.o dm-io.o dm-kcopyd.o dm-sysfs.o dm-bio-prison.o
dm-multipath-y += dm-path-selector.o dm-mpath.o
dm-snapshot-y += dm-snap.o dm-exception-store.o dm-snap-transient.o \
dm-snap-persistent.o
Index: linux-2.6/drivers/md/dm-bio-prison.c
===================================================================
--- /dev/null
+++ linux-2.6/drivers/md/dm-bio-prison.c
@@ -0,0 +1,336 @@
+/*
+ * Copyright (C) 2012 Red Hat.
+ *
+ * This file is released under the GPL.
+ */
+
+#include "dm-bio-prison.h"
+
+#include <linux/module.h>
+
+/*----------------------------------------------------------------*/
+
+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;
+}
+
+/*
+ * @nr_cells should be the number of cells you want in use _concurrently_.
+ * Don't confuse it with the number of distinct keys.
+ */
+struct bio_prison *dm_prison_create(unsigned nr_cells)
+{
+ unsigned i;
+ uint32_t nr_buckets = calc_nr_buckets(nr_cells);
+ size_t len = sizeof(struct bio_prison) +
+ (sizeof(struct hlist_head) * nr_buckets);
+ struct bio_prison *prison = kmalloc(len, GFP_KERNEL);
+
+ if (!prison)
+ return NULL;
+
+ spin_lock_init(&prison->lock);
+ prison->cell_pool = mempool_create_kmalloc_pool(nr_cells,
+ sizeof(struct cell));
+ 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_prison_create);
+
+void dm_prison_destroy(struct bio_prison *prison)
+{
+ mempool_destroy(prison->cell_pool);
+ kfree(prison);
+}
+EXPORT_SYMBOL_GPL(dm_prison_destroy);
+
+static uint32_t hash_key(struct bio_prison *prison, struct 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 cell_key *lhs, struct cell_key *rhs)
+{
+ return (lhs->virtual == rhs->virtual) &&
+ (lhs->dev == rhs->dev) &&
+ (lhs->block == rhs->block);
+}
+
+static struct cell *__search_bucket(struct hlist_head *bucket,
+ struct cell_key *key)
+{
+ struct 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 bio_prison *prison, struct cell_key *key,
+ struct bio *inmate, struct cell **ref)
+{
+ int r;
+ unsigned long flags;
+ uint32_t hash = hash_key(prison, key);
+ struct cell *uninitialized_var(cell), *cell2 = NULL;
+
+ BUG_ON(hash > prison->nr_buckets);
+
+ spin_lock_irqsave(&prison->lock, flags);
+ cell = __search_bucket(prison->cells + hash, key);
+
+ if (!cell) {
+ /*
+ * 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) {
+ cell = cell2;
+ cell2 = NULL;
+
+ 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;
+
+ } else {
+ mempool_free(cell2, prison->cell_pool);
+ cell2 = NULL;
+ r = 1;
+ bio_list_add(&cell->bios, inmate);
+ }
+
+ } else {
+ r = 1;
+ bio_list_add(&cell->bios, inmate);
+ }
+ 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 cell *cell, struct bio_list *inmates)
+{
+ struct 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 cell *cell, struct bio_list *bios)
+{
+ unsigned long flags;
+ struct 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 cell *cell, struct bio *bio)
+{
+ hlist_del(&cell->list);
+ BUG_ON(cell->holder != bio);
+ BUG_ON(!bio_list_empty(&cell->bios));
+}
+
+void dm_cell_release_singleton(struct cell *cell, struct bio *bio)
+{
+ unsigned long flags;
+ struct 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 cell *cell, struct bio_list *inmates)
+{
+ struct bio_prison *prison = cell->prison;
+
+ hlist_del(&cell->list);
+ if (inmates)
+ bio_list_merge(inmates, &cell->bios);
+
+ mempool_free(cell, prison->cell_pool);
+}
+
+void dm_cell_release_no_holder(struct cell *cell, struct bio_list *inmates)
+{
+ unsigned long flags;
+ struct 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 cell *cell)
+{
+ struct 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);
+
+/*----------------------------------------------------------------*/
+
+void dm_ds_init(struct deferred_set *ds)
+{
+ int i;
+
+ 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);
+ }
+}
+EXPORT_SYMBOL_GPL(dm_ds_init);
+
+struct deferred_entry *dm_ds_inc(struct deferred_set *ds)
+{
+ unsigned long flags;
+ struct 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_ds_inc);
+
+static unsigned ds_next(unsigned index)
+{
+ return (index + 1) % DEFERRED_SET_SIZE;
+}
+
+static void __sweep(struct 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_ds_dec(struct 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_ds_dec);
+
+/*
+ * Returns 1 if deferred or 0 if no pending items to delay job.
+ */
+int dm_ds_add_work(struct 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_ds_add_work);
+
+/*----------------------------------------------------------------*/
Index: linux-2.6/drivers/md/dm-bio-prison.h
===================================================================
--- /dev/null
+++ linux-2.6/drivers/md/dm-bio-prison.h
@@ -0,0 +1,109 @@
+/*
+ * Copyright (C) 2012 Red Hat.
+ *
+ * This file is released under the GPL.
+ */
+
+#ifndef DM_BIO_PRISON_H
+#define DM_BIO_PRISON_H
+
+#include "persistent-data/dm-block-manager.h"
+
+#include <linux/list.h>
+
+/*----------------------------------------------------------------*/
+
+/*
+ * Device identifier
+ */
+typedef uint64_t dm_dev_id;
+
+/*----------------------------------------------------------------*/
+
+/*
+ * 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 bio_prison;
+
+struct cell_key {
+ int virtual;
+ dm_dev_id dev;
+ dm_block_t block;
+};
+
+struct cell {
+ struct hlist_node list;
+ struct bio_prison *prison;
+ struct cell_key key;
+ struct bio *holder;
+ struct bio_list bios;
+};
+
+struct bio_prison {
+ spinlock_t lock;
+ mempool_t *cell_pool;
+
+ unsigned nr_buckets;
+ unsigned hash_mask;
+ struct hlist_head *cells;
+};
+
+/*
+ * @nr_cells should be the number of cells you want in use _concurrently_.
+ * Don't confuse it with the number of distinct keys.
+ */
+struct bio_prison *dm_prison_create(unsigned nr_cells);
+void dm_prison_destroy(struct 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 bio_prison *prison, struct cell_key *key,
+ struct bio *inmate, struct cell **ref);
+void dm_cell_release(struct cell *cell, struct bio_list *bios);
+void dm_cell_release_singleton(struct cell *cell, struct bio *bio);
+void dm_cell_release_no_holder(struct cell *cell, struct bio_list *inmates);
+void dm_cell_error(struct 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.
+ */
+#define DEFERRED_SET_SIZE 64
+
+struct deferred_set;
+struct deferred_entry {
+ struct deferred_set *ds;
+ unsigned count;
+ struct list_head work_items;
+};
+
+struct deferred_set {
+ spinlock_t lock;
+ unsigned current_entry;
+ unsigned sweeper;
+ struct deferred_entry entries[DEFERRED_SET_SIZE];
+};
+
+void dm_ds_init(struct deferred_set *ds);
+struct deferred_entry *dm_ds_inc(struct deferred_set *ds);
+void dm_ds_dec(struct deferred_entry *entry, struct list_head *head);
+
+/*
+ * Returns 1 if deferred or 0 if no pending items to delay job.
+ */
+int dm_ds_add_work(struct deferred_set *ds, struct list_head *work);
+
+/*----------------------------------------------------------------*/
+
+#endif
Index: linux-2.6/drivers/md/dm-thin-metadata.h
===================================================================
--- linux-2.6.orig/drivers/md/dm-thin-metadata.h
+++ linux-2.6/drivers/md/dm-thin-metadata.h
@@ -7,6 +7,7 @@
#ifndef DM_THIN_METADATA_H
#define DM_THIN_METADATA_H
+#include "dm-bio-prison.h"
#include "persistent-data/dm-block-manager.h"
#define THIN_METADATA_BLOCK_SIZE 4096
@@ -32,11 +33,6 @@ struct dm_pool_metadata;
struct dm_thin_device;
/*
- * Device identifier
- */
-typedef uint64_t dm_dev_id;
-
-/*
* Reopens or creates a new, empty metadata volume.
*/
struct dm_pool_metadata *dm_pool_metadata_open(struct block_device *bdev,
Index: linux-2.6/drivers/md/dm-thin.c
===================================================================
--- linux-2.6.orig/drivers/md/dm-thin.c
+++ linux-2.6/drivers/md/dm-thin.c
@@ -4,6 +4,7 @@
* This file is released under the GPL.
*/
+#include "dm-bio-prison.h"
#include "dm-thin-metadata.h"
#include <linux/device-mapper.h>
@@ -20,7 +21,6 @@
* Tunable constants
*/
#define ENDIO_HOOK_POOL_SIZE 10240
-#define DEFERRED_SET_SIZE 64
#define MAPPING_POOL_SIZE 1024
#define PRISON_CELLS 1024
#define COMMIT_PERIOD HZ
@@ -98,372 +98,6 @@
/*----------------------------------------------------------------*/
/*
- * 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 bio_prison;
-
-struct cell_key {
- int virtual;
- dm_dev_id dev;
- dm_block_t block;
-};
-
-struct cell {
- struct hlist_node list;
- struct bio_prison *prison;
- struct cell_key key;
- struct bio *holder;
- struct bio_list bios;
-};
-
-struct 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;
-}
-
-/*
- * @nr_cells should be the number of cells you want in use _concurrently_.
- * Don't confuse it with the number of distinct keys.
- */
-static struct bio_prison *prison_create(unsigned nr_cells)
-{
- unsigned i;
- uint32_t nr_buckets = calc_nr_buckets(nr_cells);
- size_t len = sizeof(struct bio_prison) +
- (sizeof(struct hlist_head) * nr_buckets);
- struct bio_prison *prison = kmalloc(len, GFP_KERNEL);
-
- if (!prison)
- return NULL;
-
- spin_lock_init(&prison->lock);
- prison->cell_pool = mempool_create_kmalloc_pool(nr_cells,
- sizeof(struct cell));
- 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;
-}
-
-static void prison_destroy(struct bio_prison *prison)
-{
- mempool_destroy(prison->cell_pool);
- kfree(prison);
-}
-
-static uint32_t hash_key(struct bio_prison *prison, struct 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 cell_key *lhs, struct cell_key *rhs)
-{
- return (lhs->virtual == rhs->virtual) &&
- (lhs->dev == rhs->dev) &&
- (lhs->block == rhs->block);
-}
-
-static struct cell *__search_bucket(struct hlist_head *bucket,
- struct cell_key *key)
-{
- struct 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.
- */
-static int bio_detain(struct bio_prison *prison, struct cell_key *key,
- struct bio *inmate, struct cell **ref)
-{
- int r;
- unsigned long flags;
- uint32_t hash = hash_key(prison, key);
- struct cell *uninitialized_var(cell), *cell2 = NULL;
-
- BUG_ON(hash > prison->nr_buckets);
-
- spin_lock_irqsave(&prison->lock, flags);
- cell = __search_bucket(prison->cells + hash, key);
-
- if (!cell) {
- /*
- * 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) {
- cell = cell2;
- cell2 = NULL;
-
- 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;
-
- } else {
- mempool_free(cell2, prison->cell_pool);
- cell2 = NULL;
- r = 1;
- bio_list_add(&cell->bios, inmate);
- }
-
- } else {
- r = 1;
- bio_list_add(&cell->bios, inmate);
- }
- spin_unlock_irqrestore(&prison->lock, flags);
-
- *ref = cell;
- return r;
-}
-
-/*
- * @inmates must have been initialised prior to this call
- */
-static void __cell_release(struct cell *cell, struct bio_list *inmates)
-{
- struct 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);
-}
-
-static void cell_release(struct cell *cell, struct bio_list *bios)
-{
- unsigned long flags;
- struct bio_prison *prison = cell->prison;
-
- spin_lock_irqsave(&prison->lock, flags);
- __cell_release(cell, bios);
- spin_unlock_irqrestore(&prison->lock, flags);
-}
-
-/*
- * 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 cell *cell, struct bio *bio)
-{
- hlist_del(&cell->list);
- BUG_ON(cell->holder != bio);
- BUG_ON(!bio_list_empty(&cell->bios));
-}
-
-static void cell_release_singleton(struct cell *cell, struct bio *bio)
-{
- unsigned long flags;
- struct bio_prison *prison = cell->prison;
-
- spin_lock_irqsave(&prison->lock, flags);
- __cell_release_singleton(cell, bio);
- spin_unlock_irqrestore(&prison->lock, flags);
-}
-
-/*
- * Sometimes we don't want the holder, just the additional bios.
- */
-static void __cell_release_no_holder(struct cell *cell, struct bio_list *inmates)
-{
- struct bio_prison *prison = cell->prison;
-
- hlist_del(&cell->list);
- if (inmates)
- bio_list_merge(inmates, &cell->bios);
-
- mempool_free(cell, prison->cell_pool);
-}
-
-static void cell_release_no_holder(struct cell *cell, struct bio_list *inmates)
-{
- unsigned long flags;
- struct bio_prison *prison = cell->prison;
-
- spin_lock_irqsave(&prison->lock, flags);
- __cell_release_no_holder(cell, inmates);
- spin_unlock_irqrestore(&prison->lock, flags);
-}
-
-static void cell_error(struct cell *cell)
-{
- struct 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);
-}
-
-/*----------------------------------------------------------------*/
-
-/*
- * 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 deferred_set;
-struct deferred_entry {
- struct deferred_set *ds;
- unsigned count;
- struct list_head work_items;
-};
-
-struct deferred_set {
- spinlock_t lock;
- unsigned current_entry;
- unsigned sweeper;
- struct deferred_entry entries[DEFERRED_SET_SIZE];
-};
-
-static void ds_init(struct deferred_set *ds)
-{
- int i;
-
- 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);
- }
-}
-
-static struct deferred_entry *ds_inc(struct deferred_set *ds)
-{
- unsigned long flags;
- struct 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;
-}
-
-static unsigned ds_next(unsigned index)
-{
- return (index + 1) % DEFERRED_SET_SIZE;
-}
-
-static void __sweep(struct 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);
-}
-
-static void ds_dec(struct 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);
-}
-
-/*
- * Returns 1 if deferred or 0 if no pending items to delay job.
- */
-static int ds_add_work(struct 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;
-}
-
-/*----------------------------------------------------------------*/
-
-/*
* Key building.
*/
static void build_data_key(struct dm_thin_device *td,
@@ -816,7 +450,7 @@ static void cell_defer(struct thin_c *tc
unsigned long flags;
spin_lock_irqsave(&pool->lock, flags);
- cell_release(cell, &pool->deferred_bios);
+ dm_cell_release(cell, &pool->deferred_bios);
spin_unlock_irqrestore(&tc->pool->lock, flags);
wake_worker(pool);
@@ -835,7 +469,7 @@ static void cell_defer_except(struct thi
bio_list_init(&bios);
spin_lock_irqsave(&pool->lock, flags);
- cell_release_no_holder(cell, &pool->deferred_bios);
+ dm_cell_release_no_holder(cell, &pool->deferred_bios);
spin_unlock_irqrestore(&pool->lock, flags);
wake_worker(pool);
@@ -852,7 +486,7 @@ static void process_prepared_mapping(str
bio->bi_end_io = m->saved_bi_end_io;
if (m->err) {
- cell_error(m->cell);
+ dm_cell_error(m->cell);
return;
}
@@ -864,7 +498,7 @@ static void process_prepared_mapping(str
r = dm_thin_insert_block(tc->td, m->virt_block, m->data_block);
if (r) {
DMERR("dm_thin_insert_block() failed");
- cell_error(m->cell);
+ dm_cell_error(m->cell);
return;
}
@@ -983,7 +617,7 @@ static void schedule_copy(struct thin_c
m->err = 0;
m->bio = NULL;
- if (!ds_add_work(&pool->shared_read_ds, &m->list))
+ if (!dm_ds_add_work(&pool->shared_read_ds, &m->list))
m->quiesced = 1;
/*
@@ -1014,7 +648,7 @@ static void schedule_copy(struct thin_c
if (r < 0) {
mempool_free(m, pool->mapping_pool);
DMERR("dm_kcopyd_copy() failed");
- cell_error(cell);
+ dm_cell_error(cell);
}
}
}
@@ -1079,7 +713,7 @@ static void schedule_zero(struct thin_c
if (r < 0) {
mempool_free(m, pool->mapping_pool);
DMERR("dm_kcopyd_zero() failed");
- cell_error(cell);
+ dm_cell_error(cell);
}
}
}
@@ -1167,7 +801,7 @@ static void no_space(struct cell *cell)
struct bio_list bios;
bio_list_init(&bios);
- cell_release(cell, &bios);
+ dm_cell_release(cell, &bios);
while ((bio = bio_list_pop(&bios)))
retry_on_resume(bio);
@@ -1184,7 +818,7 @@ static void process_discard(struct thin_
struct new_mapping *m;
build_virtual_key(tc->td, block, &key);
- if (bio_detain(tc->pool->prison, &key, bio, &cell))
+ if (dm_bio_detain(tc->pool->prison, &key, bio, &cell))
return;
r = dm_thin_find_block(tc->td, block, 1, &lookup_result);
@@ -1196,8 +830,8 @@ static void process_discard(struct thin_
* on this block.
*/
build_data_key(tc->td, lookup_result.block, &key2);
- if (bio_detain(tc->pool->prison, &key2, bio, &cell2)) {
- cell_release_singleton(cell, bio);
+ if (dm_bio_detain(tc->pool->prison, &key2, bio, &cell2)) {
+ dm_cell_release_singleton(cell, bio);
break;
}
@@ -1216,7 +850,7 @@ static void process_discard(struct thin_
m->err = 0;
m->bio = bio;
- if (!ds_add_work(&pool->all_io_ds, &m->list)) {
+ if (!dm_ds_add_work(&pool->all_io_ds, &m->list)) {
list_add(&m->list, &pool->prepared_discards);
wake_worker(pool);
}
@@ -1231,8 +865,8 @@ static void process_discard(struct thin_
unsigned remaining = (pool->sectors_per_block - offset) << 9;
bio->bi_size = min(bio->bi_size, remaining);
- cell_release_singleton(cell, bio);
- cell_release_singleton(cell2, bio);
+ dm_cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell2, bio);
remap_and_issue(tc, bio, lookup_result.block);
}
break;
@@ -1241,13 +875,13 @@ static void process_discard(struct thin_
/*
* It isn't provisioned, just forget it.
*/
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
bio_endio(bio, 0);
break;
default:
DMERR("discard: find block unexpectedly returned %d\n", r);
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
bio_io_error(bio);
break;
}
@@ -1274,7 +908,7 @@ static void break_sharing(struct thin_c
default:
DMERR("%s: alloc_data_block() failed, error = %d", __func__, r);
- cell_error(cell);
+ dm_cell_error(cell);
break;
}
}
@@ -1292,7 +926,7 @@ static void process_shared_bio(struct th
* of being broken so we have nothing further to do here.
*/
build_data_key(tc->td, lookup_result->block, &key);
- if (bio_detain(pool->prison, &key, bio, &cell))
+ if (dm_bio_detain(pool->prison, &key, bio, &cell))
return;
if (bio_data_dir(bio) == WRITE)
@@ -1300,9 +934,9 @@ static void process_shared_bio(struct th
else {
struct endio_hook *h = dm_get_mapinfo(bio)->ptr;
- h->shared_read_entry = ds_inc(&pool->shared_read_ds);
+ h->shared_read_entry = dm_ds_inc(&pool->shared_read_ds);
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
remap_and_issue(tc, bio, lookup_result->block);
}
}
@@ -1317,7 +951,7 @@ static void provision_block(struct thin_
* Remap empty bios (flushes) immediately, without provisioning.
*/
if (!bio->bi_size) {
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
remap_and_issue(tc, bio, 0);
return;
}
@@ -1327,7 +961,7 @@ static void provision_block(struct thin_
*/
if (bio_data_dir(bio) == READ) {
zero_fill_bio(bio);
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
bio_endio(bio, 0);
return;
}
@@ -1347,7 +981,7 @@ static void provision_block(struct thin_
default:
DMERR("%s: alloc_data_block() failed, error = %d", __func__, r);
- cell_error(cell);
+ dm_cell_error(cell);
break;
}
}
@@ -1365,7 +999,7 @@ static void process_bio(struct thin_c *t
* being provisioned so we have nothing further to do here.
*/
build_virtual_key(tc->td, block, &key);
- if (bio_detain(tc->pool->prison, &key, bio, &cell))
+ if (dm_bio_detain(tc->pool->prison, &key, bio, &cell))
return;
r = dm_thin_find_block(tc->td, block, 1, &lookup_result);
@@ -1380,7 +1014,7 @@ static void process_bio(struct thin_c *t
* TODO: this will probably have to change when discard goes
* back in.
*/
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
if (lookup_result.shared)
process_shared_bio(tc, bio, block, &lookup_result);
@@ -1390,7 +1024,7 @@ static void process_bio(struct thin_c *t
case -ENODATA:
if (bio_data_dir(bio) == READ && tc->origin_dev) {
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
remap_to_origin_and_issue(tc, bio);
} else
provision_block(tc, bio, block, cell);
@@ -1398,7 +1032,7 @@ static void process_bio(struct thin_c *t
default:
DMERR("dm_thin_find_block() failed, error = %d", r);
- cell_release_singleton(cell, bio);
+ dm_cell_release_singleton(cell, bio);
bio_io_error(bio);
break;
}
@@ -1522,7 +1156,7 @@ static struct endio_hook *thin_hook_bio(
h->tc = tc;
h->shared_read_entry = NULL;
- h->all_io_entry = bio->bi_rw & REQ_DISCARD ? NULL : ds_inc(&pool->all_io_ds);
+ h->all_io_entry = bio->bi_rw & REQ_DISCARD ? NULL : dm_ds_inc(&pool->all_io_ds);
h->overwrite_mapping = NULL;
return h;
@@ -1652,7 +1286,7 @@ static void __pool_destroy(struct pool *
if (dm_pool_metadata_close(pool->pmd) < 0)
DMWARN("%s: dm_pool_metadata_close() failed.", __func__);
- prison_destroy(pool->prison);
+ dm_prison_destroy(pool->prison);
dm_kcopyd_client_destroy(pool->copier);
if (pool->wq)
@@ -1693,7 +1327,7 @@ static struct pool *pool_create(struct m
pool->offset_mask = block_size - 1;
pool->low_water_blocks = 0;
pool_features_init(&pool->pf);
- pool->prison = prison_create(PRISON_CELLS);
+ pool->prison = dm_prison_create(PRISON_CELLS);
if (!pool->prison) {
*error = "Error creating pool's bio prison";
err_p = ERR_PTR(-ENOMEM);
@@ -1729,8 +1363,8 @@ static struct pool *pool_create(struct m
pool->low_water_triggered = 0;
pool->no_free_space = 0;
bio_list_init(&pool->retry_on_resume_list);
- ds_init(&pool->shared_read_ds);
- ds_init(&pool->all_io_ds);
+ dm_ds_init(&pool->shared_read_ds);
+ dm_ds_init(&pool->all_io_ds);
pool->next_mapping = NULL;
pool->mapping_pool =
@@ -1763,7 +1397,7 @@ bad_mapping_pool:
bad_wq:
dm_kcopyd_client_destroy(pool->copier);
bad_kcopyd_client:
- prison_destroy(pool->prison);
+ dm_prison_destroy(pool->prison);
bad_prison:
kfree(pool);
bad_pool:
@@ -2614,7 +2248,7 @@ static int thin_endio(struct dm_target *
if (h->shared_read_entry) {
INIT_LIST_HEAD(&work);
- ds_dec(h->shared_read_entry, &work);
+ dm_ds_dec(h->shared_read_entry, &work);
spin_lock_irqsave(&pool->lock, flags);
list_for_each_entry_safe(m, tmp, &work, list) {
@@ -2627,7 +2261,7 @@ static int thin_endio(struct dm_target *
if (h->all_io_entry) {
INIT_LIST_HEAD(&work);
- ds_dec(h->all_io_entry, &work);
+ dm_ds_dec(h->all_io_entry, &work);
list_for_each_entry_safe(m, tmp, &work, list)
list_add(&m->list, &pool->prepared_discards);
}
next prev parent reply other threads:[~2012-03-03 3:13 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-03-02 23:59 [PATCH 1/2] dm thin: rename dm_thin_id to more generic dm_dev_id Mike Snitzer
2012-03-02 23:59 ` [PATCH 2/2] dm thin: elevate the bio_prison and deferred_set code to allow for re-use Mike Snitzer
2012-03-03 0:26 ` Alasdair G Kergon
2012-03-03 3:13 ` Mike Snitzer [this message]
2012-03-05 10:29 ` [PATCH 1/2] dm thin: rename dm_thin_id to more generic dm_dev_id Joe Thornber
2012-03-05 14:08 ` Mike Snitzer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20120303031320.GA20103@redhat.com \
--to=snitzer@redhat.com \
--cc=dm-devel@redhat.com \
--cc=ejt@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.