aboutsummaryrefslogtreecommitdiff
path: root/drivers/md
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/md')
-rw-r--r--drivers/md/bcache/Kconfig1
-rw-r--r--drivers/md/bcache/bcache.h2
-rw-r--r--drivers/md/bcache/stats.c34
-rw-r--r--drivers/md/bcache/super.c185
-rw-r--r--drivers/md/bcache/writeback.c2
-rw-r--r--drivers/md/dm-bufio.c24
-rw-r--r--drivers/md/dm-cache-metadata.c4
-rw-r--r--drivers/md/dm-cache-policy.h4
-rw-r--r--drivers/md/dm-cache-target.c100
-rw-r--r--drivers/md/dm-mpath.c1
-rw-r--r--drivers/md/dm-snap.c1
-rw-r--r--drivers/md/dm-stripe.c11
-rw-r--r--drivers/md/dm-table.c2
-rw-r--r--drivers/md/dm-thin-metadata.c36
-rw-r--r--drivers/md/dm-thin-metadata.h7
-rw-r--r--drivers/md/dm-thin.c200
-rw-r--r--drivers/md/md.c2
-rw-r--r--drivers/md/persistent-data/dm-space-map-disk.c3
-rw-r--r--drivers/md/persistent-data/dm-space-map-metadata.c127
-rw-r--r--drivers/md/persistent-data/dm-space-map.h23
-rw-r--r--drivers/md/raid1.c38
-rw-r--r--drivers/md/raid10.c29
-rw-r--r--drivers/md/raid5.c6
23 files changed, 576 insertions, 266 deletions
diff --git a/drivers/md/bcache/Kconfig b/drivers/md/bcache/Kconfig
index 05c220d05e2..f950c9d29f3 100644
--- a/drivers/md/bcache/Kconfig
+++ b/drivers/md/bcache/Kconfig
@@ -1,7 +1,6 @@
config BCACHE
tristate "Block device as cache"
- select CLOSURES
---help---
Allows a block device to be used as cache for other devices; uses
a btree for indexing and the layout is optimized for SSDs.
diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index 340146d7c17..d3e15b42a4a 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -1241,7 +1241,7 @@ void bch_cache_set_stop(struct cache_set *);
struct cache_set *bch_cache_set_alloc(struct cache_sb *);
void bch_btree_cache_free(struct cache_set *);
int bch_btree_cache_alloc(struct cache_set *);
-void bch_writeback_init_cached_dev(struct cached_dev *);
+void bch_cached_dev_writeback_init(struct cached_dev *);
void bch_moving_init_cache_set(struct cache_set *);
void bch_cache_allocator_exit(struct cache *ca);
diff --git a/drivers/md/bcache/stats.c b/drivers/md/bcache/stats.c
index 64e679449c2..b8730e714d6 100644
--- a/drivers/md/bcache/stats.c
+++ b/drivers/md/bcache/stats.c
@@ -93,24 +93,6 @@ static struct attribute *bch_stats_files[] = {
};
static KTYPE(bch_stats);
-static void scale_accounting(unsigned long data);
-
-void bch_cache_accounting_init(struct cache_accounting *acc,
- struct closure *parent)
-{
- kobject_init(&acc->total.kobj, &bch_stats_ktype);
- kobject_init(&acc->five_minute.kobj, &bch_stats_ktype);
- kobject_init(&acc->hour.kobj, &bch_stats_ktype);
- kobject_init(&acc->day.kobj, &bch_stats_ktype);
-
- closure_init(&acc->cl, parent);
- init_timer(&acc->timer);
- acc->timer.expires = jiffies + accounting_delay;
- acc->timer.data = (unsigned long) acc;
- acc->timer.function = scale_accounting;
- add_timer(&acc->timer);
-}
-
int bch_cache_accounting_add_kobjs(struct cache_accounting *acc,
struct kobject *parent)
{
@@ -244,3 +226,19 @@ void bch_mark_sectors_bypassed(struct search *s, int sectors)
atomic_add(sectors, &dc->accounting.collector.sectors_bypassed);
atomic_add(sectors, &s->op.c->accounting.collector.sectors_bypassed);
}
+
+void bch_cache_accounting_init(struct cache_accounting *acc,
+ struct closure *parent)
+{
+ kobject_init(&acc->total.kobj, &bch_stats_ktype);
+ kobject_init(&acc->five_minute.kobj, &bch_stats_ktype);
+ kobject_init(&acc->hour.kobj, &bch_stats_ktype);
+ kobject_init(&acc->day.kobj, &bch_stats_ktype);
+
+ closure_init(&acc->cl, parent);
+ init_timer(&acc->timer);
+ acc->timer.expires = jiffies + accounting_delay;
+ acc->timer.data = (unsigned long) acc;
+ acc->timer.function = scale_accounting;
+ add_timer(&acc->timer);
+}
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index c8046bc4aa5..f88e2b653a3 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -634,11 +634,10 @@ static int open_dev(struct block_device *b, fmode_t mode)
return 0;
}
-static int release_dev(struct gendisk *b, fmode_t mode)
+static void release_dev(struct gendisk *b, fmode_t mode)
{
struct bcache_device *d = b->private_data;
closure_put(&d->cl);
- return 0;
}
static int ioctl_dev(struct block_device *b, fmode_t mode,
@@ -732,8 +731,7 @@ static void bcache_device_free(struct bcache_device *d)
if (d->c)
bcache_device_detach(d);
-
- if (d->disk)
+ if (d->disk && d->disk->flags & GENHD_FL_UP)
del_gendisk(d->disk);
if (d->disk && d->disk->queue)
blk_cleanup_queue(d->disk->queue);
@@ -756,12 +754,9 @@ static int bcache_device_init(struct bcache_device *d, unsigned block_size)
if (!(d->bio_split = bioset_create(4, offsetof(struct bbio, bio))) ||
!(d->unaligned_bvec = mempool_create_kmalloc_pool(1,
sizeof(struct bio_vec) * BIO_MAX_PAGES)) ||
- bio_split_pool_init(&d->bio_split_hook))
-
- return -ENOMEM;
-
- d->disk = alloc_disk(1);
- if (!d->disk)
+ bio_split_pool_init(&d->bio_split_hook) ||
+ !(d->disk = alloc_disk(1)) ||
+ !(q = blk_alloc_queue(GFP_KERNEL)))
return -ENOMEM;
snprintf(d->disk->disk_name, DISK_NAME_LEN, "bcache%i", bcache_minor);
@@ -771,10 +766,6 @@ static int bcache_device_init(struct bcache_device *d, unsigned block_size)
d->disk->fops = &bcache_ops;
d->disk->private_data = d;
- q = blk_alloc_queue(GFP_KERNEL);
- if (!q)
- return -ENOMEM;
-
blk_queue_make_request(q, NULL);
d->disk->queue = q;
q->queuedata = d;
@@ -999,14 +990,17 @@ static void cached_dev_free(struct closure *cl)
mutex_lock(&bch_register_lock);
- bd_unlink_disk_holder(dc->bdev, dc->disk.disk);
+ if (atomic_read(&dc->running))
+ bd_unlink_disk_holder(dc->bdev, dc->disk.disk);
bcache_device_free(&dc->disk);
list_del(&dc->list);
mutex_unlock(&bch_register_lock);
if (!IS_ERR_OR_NULL(dc->bdev)) {
- blk_sync_queue(bdev_get_queue(dc->bdev));
+ if (dc->bdev->bd_disk)
+ blk_sync_queue(bdev_get_queue(dc->bdev));
+
blkdev_put(dc->bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
}
@@ -1028,73 +1022,67 @@ static void cached_dev_flush(struct closure *cl)
static int cached_dev_init(struct cached_dev *dc, unsigned block_size)
{
- int err;
+ int ret;
struct io *io;
-
- closure_init(&dc->disk.cl, NULL);
- set_closure_fn(&dc->disk.cl, cached_dev_flush, system_wq);
+ struct request_queue *q = bdev_get_queue(dc->bdev);
__module_get(THIS_MODULE);
INIT_LIST_HEAD(&dc->list);
+ closure_init(&dc->disk.cl, NULL);
+ set_closure_fn(&dc->disk.cl, cached_dev_flush, system_wq);
kobject_init(&dc->disk.kobj, &bch_cached_dev_ktype);
-
- bch_cache_accounting_init(&dc->accounting, &dc->disk.cl);
-
- err = bcache_device_init(&dc->disk, block_size);
- if (err)
- goto err;
-
- spin_lock_init(&dc->io_lock);
- closure_init_unlocked(&dc->sb_write);
INIT_WORK(&dc->detach, cached_dev_detach_finish);
+ closure_init_unlocked(&dc->sb_write);
+ INIT_LIST_HEAD(&dc->io_lru);
+ spin_lock_init(&dc->io_lock);
+ bch_cache_accounting_init(&dc->accounting, &dc->disk.cl);
dc->sequential_merge = true;
dc->sequential_cutoff = 4 << 20;
- INIT_LIST_HEAD(&dc->io_lru);
- dc->sb_bio.bi_max_vecs = 1;
- dc->sb_bio.bi_io_vec = dc->sb_bio.bi_inline_vecs;
-
for (io = dc->io; io < dc->io + RECENT_IO; io++) {
list_add(&io->lru, &dc->io_lru);
hlist_add_head(&io->hash, dc->io_hash + RECENT_IO);
}
- bch_writeback_init_cached_dev(dc);
+ ret = bcache_device_init(&dc->disk, block_size);
+ if (ret)
+ return ret;
+
+ set_capacity(dc->disk.disk,
+ dc->bdev->bd_part->nr_sects - dc->sb.data_offset);
+
+ dc->disk.disk->queue->backing_dev_info.ra_pages =
+ max(dc->disk.disk->queue->backing_dev_info.ra_pages,
+ q->backing_dev_info.ra_pages);
+
+ bch_cached_dev_request_init(dc);
+ bch_cached_dev_writeback_init(dc);
return 0;
-err:
- bcache_device_stop(&dc->disk);
- return err;
}
/* Cached device - bcache superblock */
-static const char *register_bdev(struct cache_sb *sb, struct page *sb_page,
+static void register_bdev(struct cache_sb *sb, struct page *sb_page,
struct block_device *bdev,
struct cached_dev *dc)
{
char name[BDEVNAME_SIZE];
const char *err = "cannot allocate memory";
- struct gendisk *g;
struct cache_set *c;
- if (!dc || cached_dev_init(dc, sb->block_size << 9) != 0)
- return err;
-
memcpy(&dc->sb, sb, sizeof(struct cache_sb));
- dc->sb_bio.bi_io_vec[0].bv_page = sb_page;
dc->bdev = bdev;
dc->bdev->bd_holder = dc;
- g = dc->disk.disk;
-
- set_capacity(g, dc->bdev->bd_part->nr_sects - dc->sb.data_offset);
-
- g->queue->backing_dev_info.ra_pages =
- max(g->queue->backing_dev_info.ra_pages,
- bdev->bd_queue->backing_dev_info.ra_pages);
+ bio_init(&dc->sb_bio);
+ dc->sb_bio.bi_max_vecs = 1;
+ dc->sb_bio.bi_io_vec = dc->sb_bio.bi_inline_vecs;
+ dc->sb_bio.bi_io_vec[0].bv_page = sb_page;
+ get_page(sb_page);
- bch_cached_dev_request_init(dc);
+ if (cached_dev_init(dc, sb->block_size << 9))
+ goto err;
err = "error creating kobject";
if (kobject_add(&dc->disk.kobj, &part_to_dev(bdev->bd_part)->kobj,
@@ -1103,6 +1091,8 @@ static const char *register_bdev(struct cache_sb *sb, struct page *sb_page,
if (bch_cache_accounting_add_kobjs(&dc->accounting, &dc->disk.kobj))
goto err;
+ pr_info("registered backing device %s", bdevname(bdev, name));
+
list_add(&dc->list, &uncached_devices);
list_for_each_entry(c, &bch_cache_sets, list)
bch_cached_dev_attach(dc, c);
@@ -1111,15 +1101,10 @@ static const char *register_bdev(struct cache_sb *sb, struct page *sb_page,
BDEV_STATE(&dc->sb) == BDEV_STATE_STALE)
bch_cached_dev_run(dc);
- return NULL;
+ return;
err:
- kobject_put(&dc->disk.kobj);
pr_notice("error opening %s: %s", bdevname(bdev, name), err);
- /*
- * Return NULL instead of an error because kobject_put() cleans
- * everything up
- */
- return NULL;
+ bcache_device_stop(&dc->disk);
}
/* Flash only volumes */
@@ -1717,20 +1702,11 @@ static int cache_alloc(struct cache_sb *sb, struct cache *ca)
size_t free;
struct bucket *b;
- if (!ca)
- return -ENOMEM;
-
__module_get(THIS_MODULE);
kobject_init(&ca->kobj, &bch_cache_ktype);
- memcpy(&ca->sb, sb, sizeof(struct cache_sb));
-
INIT_LIST_HEAD(&ca->discards);
- bio_init(&ca->sb_bio);
- ca->sb_bio.bi_max_vecs = 1;
- ca->sb_bio.bi_io_vec = ca->sb_bio.bi_inline_vecs;
-
bio_init(&ca->journal.bio);
ca->journal.bio.bi_max_vecs = 8;
ca->journal.bio.bi_io_vec = ca->journal.bio.bi_inline_vecs;
@@ -1742,18 +1718,17 @@ static int cache_alloc(struct cache_sb *sb, struct cache *ca)
!init_fifo(&ca->free_inc, free << 2, GFP_KERNEL) ||
!init_fifo(&ca->unused, free << 2, GFP_KERNEL) ||
!init_heap(&ca->heap, free << 3, GFP_KERNEL) ||
- !(ca->buckets = vmalloc(sizeof(struct bucket) *
+ !(ca->buckets = vzalloc(sizeof(struct bucket) *
ca->sb.nbuckets)) ||
!(ca->prio_buckets = kzalloc(sizeof(uint64_t) * prio_buckets(ca) *
2, GFP_KERNEL)) ||
!(ca->disk_buckets = alloc_bucket_pages(GFP_KERNEL, ca)) ||
!(ca->alloc_workqueue = alloc_workqueue("bch_allocator", 0, 1)) ||
bio_split_pool_init(&ca->bio_split_hook))
- goto err;
+ return -ENOMEM;
ca->prio_last_buckets = ca->prio_buckets + prio_buckets(ca);
- memset(ca->buckets, 0, ca->sb.nbuckets * sizeof(struct bucket));
for_each_bucket(b, ca)
atomic_set(&b->pin, 0);
@@ -1766,22 +1741,28 @@ err:
return -ENOMEM;
}
-static const char *register_cache(struct cache_sb *sb, struct page *sb_page,
+static void register_cache(struct cache_sb *sb, struct page *sb_page,
struct block_device *bdev, struct cache *ca)
{
char name[BDEVNAME_SIZE];
const char *err = "cannot allocate memory";
- if (cache_alloc(sb, ca) != 0)
- return err;
-
- ca->sb_bio.bi_io_vec[0].bv_page = sb_page;
+ memcpy(&ca->sb, sb, sizeof(struct cache_sb));
ca->bdev = bdev;
ca->bdev->bd_holder = ca;
+ bio_init(&ca->sb_bio);
+ ca->sb_bio.bi_max_vecs = 1;
+ ca->sb_bio.bi_io_vec = ca->sb_bio.bi_inline_vecs;
+ ca->sb_bio.bi_io_vec[0].bv_page = sb_page;
+ get_page(sb_page);
+
if (blk_queue_discard(bdev_get_queue(ca->bdev)))
ca->discard = CACHE_DISCARD(&ca->sb);
+ if (cache_alloc(sb, ca) != 0)
+ goto err;
+
err = "error creating kobject";
if (kobject_add(&ca->kobj, &part_to_dev(bdev->bd_part)->kobj, "bcache"))
goto err;
@@ -1791,15 +1772,10 @@ static const char *register_cache(struct cache_sb *sb, struct page *sb_page,
goto err;
pr_info("registered cache device %s", bdevname(bdev, name));
-
- return NULL;
+ return;
err:
+ pr_notice("error opening %s: %s", bdevname(bdev, name), err);
kobject_put(&ca->kobj);
- pr_info("error opening %s: %s", bdevname(bdev, name), err);
- /* Return NULL instead of an error because kobject_put() cleans
- * everything up
- */
- return NULL;
}
/* Global interfaces/init */
@@ -1833,12 +1809,15 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
bdev = blkdev_get_by_path(strim(path),
FMODE_READ|FMODE_WRITE|FMODE_EXCL,
sb);
- if (bdev == ERR_PTR(-EBUSY))
- err = "device busy";
-
- if (IS_ERR(bdev) ||
- set_blocksize(bdev, 4096))
+ if (IS_ERR(bdev)) {
+ if (bdev == ERR_PTR(-EBUSY))
+ err = "device busy";
goto err;
+ }
+
+ err = "failed to set blocksize";
+ if (set_blocksize(bdev, 4096))
+ goto err_close;
err = read_super(sb, bdev, &sb_page);
if (err)
@@ -1846,33 +1825,33 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
if (SB_IS_BDEV(sb)) {
struct cached_dev *dc = kzalloc(sizeof(*dc), GFP_KERNEL);
+ if (!dc)
+ goto err_close;
- err = register_bdev(sb, sb_page, bdev, dc);
+ register_bdev(sb, sb_page, bdev, dc);
} else {
struct cache *ca = kzalloc(sizeof(*ca), GFP_KERNEL);
+ if (!ca)
+ goto err_close;
- err = register_cache(sb, sb_page, bdev, ca);
+ register_cache(sb, sb_page, bdev, ca);
}
-
- if (err) {
- /* register_(bdev|cache) will only return an error if they
- * didn't get far enough to create the kobject - if they did,
- * the kobject destructor will do this cleanup.
- */
+out:
+ if (sb_page)
put_page(sb_page);
-err_close:
- blkdev_put(bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
-err:
- if (attr != &ksysfs_register_quiet)
- pr_info("error opening %s: %s", path, err);
- ret = -EINVAL;
- }
-
kfree(sb);
kfree(path);
mutex_unlock(&bch_register_lock);
module_put(THIS_MODULE);
return ret;
+
+err_close:
+ blkdev_put(bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
+err:
+ if (attr != &ksysfs_register_quiet)
+ pr_info("error opening %s: %s", path, err);
+ ret = -EINVAL;
+ goto out;
}
static int bcache_reboot(struct notifier_block *n, unsigned long code, void *x)
diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c
index 93e7e31a4bd..2714ed3991d 100644
--- a/drivers/md/bcache/writeback.c
+++ b/drivers/md/bcache/writeback.c
@@ -375,7 +375,7 @@ err:
refill_dirty(cl);
}
-void bch_writeback_init_cached_dev(struct cached_dev *dc)
+void bch_cached_dev_writeback_init(struct cached_dev *dc)
{
closure_init_unlocked(&dc->writeback);
init_rwsem(&dc->writeback_lock);
diff --git a/drivers/md/dm-bufio.c b/drivers/md/dm-bufio.c
index c6083132c4b..0387e05cdb9 100644
--- a/drivers/md/dm-bufio.c
+++ b/drivers/md/dm-bufio.c
@@ -319,6 +319,9 @@ static void __cache_size_refresh(void)
static void *alloc_buffer_data(struct dm_bufio_client *c, gfp_t gfp_mask,
enum data_mode *data_mode)
{
+ unsigned noio_flag;
+ void *ptr;
+
if (c->block_size <= DM_BUFIO_BLOCK_SIZE_SLAB_LIMIT) {
*data_mode = DATA_MODE_SLAB;
return kmem_cache_alloc(DM_BUFIO_CACHE(c), gfp_mask);
@@ -332,7 +335,26 @@ static void *alloc_buffer_data(struct dm_bufio_client *c, gfp_t gfp_mask,
}
*data_mode = DATA_MODE_VMALLOC;
- return __vmalloc(c->block_size, gfp_mask, PAGE_KERNEL);
+
+ /*
+ * __vmalloc allocates the data pages and auxiliary structures with
+ * gfp_flags that were specified, but pagetables are always allocated
+ * with GFP_KERNEL, no matter what was specified as gfp_mask.
+ *
+ * Consequently, we must set per-process flag PF_MEMALLOC_NOIO so that
+ * all allocations done by this process (including pagetables) are done
+ * as if GFP_NOIO was specified.
+ */
+
+ if (gfp_mask & __GFP_NORETRY)
+ noio_flag = memalloc_noio_save();
+
+ ptr = __vmalloc(c->block_size, gfp_mask, PAGE_KERNEL);
+
+ if (gfp_mask & __GFP_NORETRY)
+ memalloc_noio_restore(noio_flag);
+
+ return ptr;
}
/*
diff --git a/drivers/md/dm-cache-metadata.c b/drivers/md/dm-cache-metadata.c
index 83e995fece8..1af7255bbff 100644
--- a/drivers/md/dm-cache-metadata.c
+++ b/drivers/md/dm-cache-metadata.c
@@ -1044,7 +1044,7 @@ void dm_cache_metadata_get_stats(struct dm_cache_metadata *cmd,
struct dm_cache_statistics *stats)
{
down_read(&cmd->root_lock);
- memcpy(stats, &cmd->stats, sizeof(*stats));
+ *stats = cmd->stats;
up_read(&cmd->root_lock);
}
@@ -1052,7 +1052,7 @@ void dm_cache_metadata_set_stats(struct dm_cache_metadata *cmd,
struct dm_cache_statistics *stats)
{
down_write(&cmd->root_lock);
- memcpy(&cmd->stats, stats, sizeof(*stats));
+ cmd->stats = *stats;
up_write(&cmd->root_lock);
}
diff --git a/drivers/md/dm-cache-policy.h b/drivers/md/dm-cache-policy.h
index 558bdfdabf5..33369ca9614 100644
--- a/drivers/md/dm-cache-policy.h
+++ b/drivers/md/dm-cache-policy.h
@@ -130,8 +130,8 @@ struct dm_cache_policy {
*
* Must not block.
*
- * Returns 1 iff in cache, 0 iff not, < 0 on error (-EWOULDBLOCK
- * would be typical).
+ * Returns 0 if in cache, -ENOENT if not, < 0 for other errors
+ * (-EWOULDBLOCK would be typical).
*/
int (*lookup)(struct dm_cache_policy *p, dm_oblock_t oblock, dm_cblock_t *cblock);
diff --git a/drivers/md/dm-cache-target.c b/drivers/md/dm-cache-target.c
index 10744091e6c..df44b60e66f 100644
--- a/drivers/md/dm-cache-target.c
+++ b/drivers/md/dm-cache-target.c
@@ -205,7 +205,7 @@ struct per_bio_data {
/*
* writethrough fields. These MUST remain at the end of this
* structure and the 'cache' member must be the first as it
- * is used to determine the offsetof the writethrough fields.
+ * is used to determine the offset of the writethrough fields.
*/
struct cache *cache;
dm_cblock_t cblock;
@@ -393,7 +393,7 @@ static int get_cell(struct cache *cache,
return r;
}
- /*----------------------------------------------------------------*/
+/*----------------------------------------------------------------*/
static bool is_dirty(struct cache *cache, dm_cblock_t b)
{
@@ -419,6 +419,7 @@ static void clear_dirty(struct cache *cache, dm_oblock_t oblock, dm_cblock_t cbl
}
/*----------------------------------------------------------------*/
+
static bool block_size_is_power_of_two(struct cache *cache)
{
return cache->sectors_per_block_shift >= 0;
@@ -667,7 +668,7 @@ static void writethrough_endio(struct bio *bio, int err)
/*
* We can't issue this bio directly, since we're in interrupt
- * context. So it get's put on a bio list for processing by the
+ * context. So it gets put on a bio list for processing by the
* worker thread.
*/
defer_writethrough_bio(pb->cache, bio);
@@ -1445,6 +1446,7 @@ static void do_worker(struct work_struct *ws)
static void do_waker(struct work_struct *ws)
{
struct cache *cache = container_of(to_delayed_work(ws), struct cache, waker);
+ policy_tick(cache->policy);
wake_worker(cache);
queue_delayed_work(cache->wq, &cache->waker, COMMIT_PERIOD);
}
@@ -1809,7 +1811,37 @@ static int parse_cache_args(struct cache_args *ca, int argc, char **argv,
static struct kmem_cache *migration_cache;
-static int set_config_values(struct dm_cache_policy *p, int argc, const char **argv)
+#define NOT_CORE_OPTION 1
+
+static int process_config_option(struct cache *cache, const char *key, const char *value)
+{
+ unsigned long tmp;
+
+ if (!strcasecmp(key, "migration_threshold")) {
+ if (kstrtoul(value, 10, &tmp))
+ return -EINVAL;
+
+ cache->migration_threshold = tmp;
+ return 0;
+ }
+
+ return NOT_CORE_OPTION;
+}
+
+static int set_config_value(struct cache *cache, const char *key, const char *value)
+{
+ int r = process_config_option(cache, key, value);
+
+ if (r == NOT_CORE_OPTION)
+ r = policy_set_config_value(cache->policy, key, value);
+
+ if (r)
+ DMWARN("bad config value for %s: %s", key, value);
+
+ return r;
+}
+
+static int set_config_values(struct cache *cache, int argc, const char **argv)
{
int r = 0;
@@ -1819,12 +1851,9 @@ static int set_config_values(struct dm_cache_policy *p, int argc, const char **a
}
while (argc) {
- r = policy_set_config_value(p, argv[0], argv[1]);
- if (r) {
- DMWARN("policy_set_config_value failed: key = '%s', value = '%s'",
- argv[0], argv[1]);
- return r;
- }
+ r = set_config_value(cache, argv[0], argv[1]);
+ if (r)
+ break;
argc -= 2;
argv += 2;
@@ -1836,8 +1865,6 @@ static int set_config_values(struct dm_cache_policy *p, int argc, const char **a
static int create_cache_policy(struct cache *cache, struct cache_args *ca,
char **error)
{
- int r;
-
cache->policy = dm_cache_policy_create(ca->policy_name,
cache->cache_size,
cache->origin_sectors,
@@ -1847,14 +1874,7 @@ static int create_cache_policy(struct cache *cache, struct cache_args *ca,
return -ENOMEM;
}
- r = set_config_values(cache->policy, ca->policy_argc, ca->policy_argv);
- if (r) {
- *error = "Error setting cache policy's config values";
- dm_cache_policy_destroy(cache->policy);
- cache->policy = NULL;
- }
-
- return r;
+ return 0;
}
/*
@@ -1886,7 +1906,7 @@ static sector_t calculate_discard_block_size(sector_t cache_block_size,
return discard_block_size;
}
-#define DEFAULT_MIGRATION_THRESHOLD (2048 * 100)
+#define DEFAULT_MIGRATION_THRESHOLD 2048
static int cache_create(struct cache_args *ca, struct cache **result)
{
@@ -1911,7 +1931,7 @@ static int cache_create(struct cache_args *ca, struct cache **result)
ti->discards_supported = true;
ti->discard_zeroes_data_unsupported = true;
- memcpy(&cache->features, &ca->features, sizeof(cache->features));
+ cache->features = ca->features;
ti->per_bio_data_size = get_per_bio_data_size(cache);
cache->callbacks.congested_fn = cache_is_congested;
@@ -1948,7 +1968,15 @@ static int cache_create(struct cache_args *ca, struct cache **result)
r = create_cache_policy(cache, ca, error);
if (r)
goto bad;
+
cache->policy_nr_args = ca->policy_argc;
+ cache->migration_threshold = DEFAULT_MIGRATION_THRESHOLD;
+
+ r = set_config_values(cache, ca->policy_argc, ca->policy_argv);
+ if (r) {
+ *error = "Error setting cache policy's config values";
+ goto bad;
+ }
cmd = dm_cache_metadata_open(cache->metadata_dev->bdev,
ca->block_size, may_format,
@@ -1967,10 +1995,10 @@ static int cache_create(struct cache_args *ca, struct cache **result)
INIT_LIST_HEAD(&cache->quiesced_migrations);
INIT_LIST_HEAD(&cache->completed_migrations);
INIT_LIST_HEAD(&cache->need_commit_migrations);
- cache->migration_threshold = DEFAULT_MIGRATION_THRESHOLD;
atomic_set(&cache->nr_migrations, 0);
init_waitqueue_head(&cache->migration_wait);
+ r = -ENOMEM;
cache->nr_dirty = 0;
cache->dirty_bitset = alloc_bitset(from_cblock(cache->cache_size));
if (!cache->dirty_bitset) {
@@ -2517,23 +2545,6 @@ err:
DMEMIT("Error");
}
-#define NOT_CORE_OPTION 1
-
-static int process_config_option(struct cache *cache, char **argv)
-{
- unsigned long tmp;
-
- if (!strcasecmp(argv[0], "migration_threshold")) {
- if (kstrtoul(argv[1], 10, &tmp))
- return -EINVAL;
-
- cache->migration_threshold = tmp;
- return 0;
- }
-
- return NOT_CORE_OPTION;
-}
-
/*
* Supports <key> <value>.
*
@@ -2541,17 +2552,12 @@ static int process_config_option(struct cache *cache, char **argv)
*/
static int cache_message(struct dm_target *ti, unsigned argc, char **argv)
{
- int r;
struct cache *cache = ti->private;
if (argc != 2)
return -EINVAL;
- r = process_config_option(cache, argv);
- if (r == NOT_CORE_OPTION)
- return policy_set_config_value(cache->policy, argv[0], argv[1]);
-
- return r;
+ return set_config_value(cache, argv[0], argv[1]);
}
static int cache_iterate_devices(struct dm_target *ti,
@@ -2609,7 +2615,7 @@ static void cache_io_hints(struct dm_target *ti, struct queue_limits *limits)
static struct target_type cache_target = {
.name = "cache",
- .version = {1, 1, 0},
+ .version = {1, 1, 1},
.module = THIS_MODULE,
.ctr = cache_ctr,
.dtr = cache_dtr,
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index 51bb81676be..bdf26f5bd32 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -907,6 +907,7 @@ static int multipath_ctr(struct dm_target *ti, unsigned int argc,
ti->num_flush_bios = 1;
ti->num_discard_bios = 1;
+ ti->num_write_same_bios = 1;
return 0;
diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index c0e07026a8d..c434e5aab2d 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -1121,6 +1121,7 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv)
s->pending_pool = mempool_create_slab_pool(MIN_IOS, pending_cache);
if (!s->pending_pool) {
ti->error = "Could not allocate mempool for pending exceptions";
+ r = -ENOMEM;
goto bad_pending_pool;
}
diff --git a/drivers/md/dm-stripe.c b/drivers/md/dm-stripe.c
index ea5e878a30b..d907ca6227c 100644
--- a/drivers/md/dm-stripe.c
+++ b/drivers/md/dm-stripe.c
@@ -94,7 +94,7 @@ static int get_stripe(struct dm_target *ti, struct stripe_c *sc,
static int stripe_ctr(struct dm_target *ti, unsigned int argc, char **argv)
{
struct stripe_c *sc;
- sector_t width;
+ sector_t width, tmp_len;
uint32_t stripes;
uint32_t chunk_size;
int r;
@@ -116,15 +116,16 @@ static int stripe_ctr(struct dm_target *ti, unsigned int argc, char **argv)
}
width = ti->len;
- if (sector_div(width, chunk_size)) {
+ if (sector_div(width, stripes)) {
ti->error = "Target length not divisible by "
- "chunk size";
+ "number of stripes";
return -EINVAL;
}
- if (sector_div(width, stripes)) {
+ tmp_len = width;
+ if (sector_div(tmp_len, chunk_size)) {
ti->error = "Target length not divisible by "
- "number of stripes";
+ "chunk size";
return -EINVAL;
}
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index e50dad0c65f..1ff252ab7d4 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -1442,7 +1442,7 @@ static bool dm_table_supports_write_same(struct dm_table *t)
return false;
if (!ti->type->iterate_devices ||
- !ti->type->iterate_devices(ti, device_not_write_same_capable, NULL))
+ ti->type->iterate_devices(ti, device_not_write_same_capable, NULL))
return false;
}
diff --git a/drivers/md/dm-thin-metadata.c b/drivers/md/dm-thin-metadata.c
index 00cee02f8fc..60bce435f4f 100644
--- a/drivers/md/dm-thin-metadata.c
+++ b/drivers/md/dm-thin-metadata.c
@@ -1645,12 +1645,12 @@ int dm_thin_get_highest_mapped_block(struct dm_thin_device *td,
return r;
}
-static int __resize_data_dev(struct dm_pool_metadata *pmd, dm_block_t new_count)
+static int __resize_space_map(struct dm_space_map *sm, dm_block_t new_count)
{
int r;
dm_block_t old_count;
- r = dm_sm_get_nr_blocks(pmd->data_sm, &old_count);
+ r = dm_sm_get_nr_blocks(sm, &old_count);
if (r)
return r;
@@ -1658,11 +1658,11 @@ static int __resize_data_dev(struct dm_pool_metadata *pmd, dm_block_t new_count)
return 0;
if (new_count < old_count) {
- DMERR("cannot reduce size of data device");
+ DMERR("cannot reduce size of space map");
return -EINVAL;
}
- return dm_sm_extend(pmd->data_sm, new_count - old_count);
+ return dm_sm_extend(sm, new_count - old_count);
}
int dm_pool_resize_data_dev(struct dm_pool_metadata *pmd, dm_block_t new_count)
@@ -1671,7 +1671,19 @@ int dm_pool_resize_data_dev(struct dm_pool_metadata *pmd, dm_block_t new_count)
down_write(&pmd->root_lock);
if (!pmd->fail_io)
- r = __resize_data_dev(pmd, new_count);
+ r = __resize_space_map(pmd->data_sm, new_count);
+ up_write(&pmd->root_lock);
+
+ return r;
+}
+
+int dm_pool_resize_metadata_dev(struct dm_pool_metadata *pmd, dm_block_t new_count)
+{
+ int r = -EINVAL;
+
+ down_write(&pmd->root_lock);
+ if (!pmd->fail_io)
+ r = __resize_space_map(pmd->metadata_sm, new_count);
up_write(&pmd->root_lock);
return r;
@@ -1684,3 +1696,17 @@ void dm_pool_metadata_read_only(struct dm_pool_metadata *pmd)
dm_bm_set_read_only(pmd->bm);
up_write(&pmd->root_lock);
}
+
+int dm_pool_register_metadata_threshold(struct dm_pool_metadata *pmd,
+ dm_block_t threshold,
+ dm_sm_threshold_fn fn,
+ void *context)
+{
+ int r;
+
+ down_write(&pmd->root_lock);
+ r = dm_sm_register_threshold_callback(pmd->metadata_sm, threshold, fn, context);
+ up_write(&pmd->root_lock);
+
+ return r;
+}
diff --git a/drivers/md/dm-thin-metadata.h b/drivers/md/dm-thin-metadata.h
index 0cecc370288..845ebbe589a 100644
--- a/drivers/md/dm-thin-metadata.h
+++ b/drivers/md/dm-thin-metadata.h
@@ -8,6 +8,7 @@
#define DM_THIN_METADATA_H
#include "persistent-data/dm-block-manager.h"
+#include "persistent-data/dm-space-map.h"
#define THIN_METADATA_BLOCK_SIZE 4096
@@ -185,6 +186,7 @@ int dm_pool_get_data_dev_size(struct dm_pool_metadata *pmd, dm_block_t *result);
* blocks would be lost.
*/
int dm_pool_resize_data_dev(struct dm_pool_metadata *pmd, dm_block_t new_size);
+int dm_pool_resize_metadata_dev(struct dm_pool_metadata *pmd, dm_block_t new_size);
/*
* Flicks the underlying block manager into read only mode, so you know
@@ -192,6 +194,11 @@ int dm_pool_resize_data_dev(struct dm_pool_metadata *pmd, dm_block_t new_size);
*/
void dm_pool_metadata_read_only(struct dm_pool_metadata *pmd);
+int dm_pool_register_metadata_threshold(struct dm_pool_metadata *pmd,
+ dm_block_t threshold,
+ dm_sm_threshold_fn fn,
+ void *context);
+
/*----------------------------------------------------------------*/
#endif
diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c
index 004ad1652b7..88f2f802d52 100644
--- a/drivers/md/dm-thin.c
+++ b/drivers/md/dm-thin.c
@@ -922,7 +922,7 @@ static int alloc_data_block(struct thin_c *tc, dm_block_t *result)
return r;
if (free_blocks <= pool->low_water_blocks && !pool->low_water_triggered) {
- DMWARN("%s: reached low water mark, sending event.",
+ DMWARN("%s: reached low water mark for data device: sending event.",
dm_device_name(pool->pool_md));
spin_lock_irqsave(&pool->lock, flags);
pool->low_water_triggered = 1;
@@ -1281,6 +1281,10 @@ static void process_bio_fail(struct thin_c *tc, struct bio *bio)
bio_io_error(bio);
}
+/*
+ * FIXME: should we also commit due to size of transaction, measured in
+ * metadata blocks?
+ */
static int need_commit_due_to_time(struct pool *pool)
{
return jiffies < pool->last_commit_jiffies ||
@@ -1909,6 +1913,56 @@ static int parse_pool_features(struct dm_arg_set *as, struct pool_features *pf,
return r;
}
+static void metadata_low_callback(void *context)
+{
+ struct pool *pool = context;
+
+ DMWARN("%s: reached low water mark for metadata device: sending event.",
+ dm_device_name(pool->pool_md));
+
+ dm_table_event(pool->ti->table);
+}
+
+static sector_t get_metadata_dev_size(struct block_device *bdev)
+{
+ sector_t metadata_dev_size = i_size_read(bdev->bd_inode) >> SECTOR_SHIFT;
+ char buffer[BDEVNAME_SIZE];
+
+ if (metadata_dev_size > THIN_METADATA_MAX_SECTORS_WARNING) {
+ DMWARN("Metadata device %s is larger than %u sectors: excess space will not be used.",
+ bdevname(bdev, buffer), THIN_METADATA_MAX_SECTORS);
+ metadata_dev_size = THIN_METADATA_MAX_SECTORS_WARNING;
+ }
+
+ return metadata_dev_size;
+}
+
+static dm_block_t get_metadata_dev_size_in_blocks(struct block_device *bdev)
+{
+ sector_t metadata_dev_size = get_metadata_dev_size(bdev);
+
+ sector_div(metadata_dev_size, THIN_METADATA_BLOCK_SIZE >> SECTOR_SHIFT);
+
+ return metadata_dev_size;
+}
+
+/*
+ * When a metadata threshold is crossed a dm event is triggered, and
+ * userland should respond by growing the metadata device. We could let
+ * userland set the threshold, like we do with the data threshold, but I'm
+ * not sure they know enough to do this well.
+ */
+static dm_block_t calc_metadata_threshold(struct pool_c *pt)
+{
+ /*
+ * 4M is ample for all ops with the possible exception of thin
+ * device deletion which is harmless if it fails (just retry the
+ * delete after you've grown the device).
+ */
+ dm_block_t quarter = get_metadata_dev_size_in_blocks(pt->metadata_dev->bdev) / 4;
+ return min((dm_block_t)1024ULL /* 4M */, quarter);
+}
+
/*
* thin-pool <metadata dev> <data dev>
* <data block size (sectors)>
@@ -1931,8 +1985,7 @@ static int pool_ctr(struct dm_target *ti, unsigned argc, char **argv)
unsigned long block_size;
dm_block_t low_water_blocks;
struct dm_dev *metadata_dev;
- sector_t metadata_dev_size;
- char b[BDEVNAME_SIZE];
+ fmode_t metadata_mode;
/*
* FIXME Remove validation from scope of lock.
@@ -1944,19 +1997,32 @@ static int pool_ctr(struct dm_target *ti, unsigned argc, char **argv)
r = -EINVAL;
goto out_unlock;
}
+
as.argc = argc;
as.argv = argv;
- r = dm_get_device(ti, argv[0], FMODE_READ | FMODE_WRITE, &metadata_dev);
+ /*
+ * Set default pool features.
+ */
+ pool_features_init(&pf);
+
+ dm_consume_args(&as, 4);
+ r = parse_pool_features(&as, &pf, ti);
+ if (r)
+ goto out_unlock;
+
+ metadata_mode = FMODE_READ | ((pf.mode == PM_READ_ONLY) ? 0 : FMODE_WRITE);
+ r = dm_get_device(ti, argv[0], metadata_mode, &metadata_dev);
if (r) {
ti->error = "Error opening metadata block device";
goto out_unlock;
}
- metadata_dev_size = i_size_read(metadata_dev->bdev->bd_inode) >> SECTOR_SHIFT;
- if (metadata_dev_size > THIN_METADATA_MAX_SECTORS_WARNING)
- DMWARN("Metadata device %s is larger than %u sectors: excess space will not be used.",
- bdevname(metadata_dev->bdev, b), THIN_METADATA_MAX_SECTORS);
+ /*
+ * Run for the side-effect of possibly issuing a warning if the
+ * device is too big.
+ */
+ (void) get_metadata_dev_size(metadata_dev->bdev);
r = dm_get_device(ti, argv[1], FMODE_READ | FMODE_WRITE, &data_dev);
if (r) {
@@ -1979,16 +2045,6 @@ static int pool_ctr(struct dm_target *ti, unsigned argc, char **argv)
goto out;
}
- /*
- * Set default pool features.
- */
- pool_features_init(&pf);
-
- dm_consume_args(&as, 4);
- r = parse_pool_features(&as, &pf, ti);
- if (r)
- goto out;
-
pt = kzalloc(sizeof(*pt), GFP_KERNEL);
if (!pt) {
r = -ENOMEM;
@@ -2040,6 +2096,13 @@ static int pool_ctr(struct dm_target *ti, unsigned argc, char **argv)
}
ti->private = pt;
+ r = dm_pool_register_metadata_threshold(pt->pool->pmd,
+ calc_metadata_threshold(pt),
+ metadata_low_callback,
+ pool);
+ if (r)
+ goto out_free_pt;
+
pt->callbacks.congested_fn = pool_is_congested;
dm_table_add_target_callbacks(ti->table, &pt->callbacks);
@@ -2079,18 +2142,7 @@ static int pool_map(struct dm_target *ti, struct bio *bio)
return r;
}
-/*
- * Retrieves the number of blocks of the data device from
- * the superblock and compares it to the actual device size,
- * thus resizing the data device in case it has grown.
- *
- * This both copes with opening preallocated data devices in the ctr
- * being followed by a resume
- * -and-
- * calling the resume method individually after userspace has
- * grown the data device in reaction to a table event.
- */
-static int pool_preresume(struct dm_target *ti)
+static int maybe_resize_data_dev(struct dm_target *ti, bool *need_commit)
{
int r;
struct pool_c *pt = ti->private;
@@ -2098,12 +2150,7 @@ static int pool_preresume(struct dm_target *ti)
sector_t data_size = ti->len;
dm_block_t sb_data_size;
- /*
- * Take control of the pool object.
- */
- r = bind_control_target(pool, ti);
- if (r)
- return r;
+ *need_commit = false;
(void) sector_div(data_size, pool->sectors_per_block);
@@ -2114,7 +2161,7 @@ static int pool_preresume(struct dm_target *ti)
}
if (data_size < sb_data_size) {
- DMERR("pool target too small, is %llu blocks (expected %llu)",
+ DMERR("pool target (%llu blocks) too small: expected %llu",
(unsigned long long)data_size, sb_data_size);
return -EINVAL;
@@ -2122,17 +2169,90 @@ static int pool_preresume(struct dm_target *ti)
r = dm_pool_resize_data_dev(pool->pmd, data_size);
if (r) {
DMERR("failed to resize data device");
- /* FIXME Stricter than necessary: Rollback transaction instead here */
set_pool_mode(pool, PM_READ_ONLY);
return r;
}
- (void) commit_or_fallback(pool);
+ *need_commit = true;
}
return 0;
}
+static int maybe_resize_metadata_dev(struct dm_target *ti, bool *need_commit)
+{
+ int r;
+ struct pool_c *pt = ti->private;
+ struct pool *pool = pt->pool;
+ dm_block_t metadata_dev_size, sb_metadata_dev_size;
+
+ *need_commit = false;
+
+ metadata_dev_size = get_metadata_dev_size_in_blocks(pool->md_dev);
+
+ r = dm_pool_get_metadata_dev_size(pool->pmd, &sb_metadata_dev_size);
+ if (r) {
+ DMERR("failed to retrieve data device size");
+ return r;
+ }
+
+ if (metadata_dev_size < sb_metadata_dev_size) {
+ DMERR("metadata device (%llu blocks) too small: expected %llu",
+ metadata_dev_size, sb_metadata_dev_size);
+ return -EINVAL;
+
+ } else if (metadata_dev_size > sb_metadata_dev_size) {
+ r = dm_pool_resize_metadata_dev(pool->pmd, metadata_dev_size);
+ if (r) {
+ DMERR("failed to resize metadata device");
+ return r;
+ }
+
+ *need_commit = true;
+ }
+
+ return 0;
+}
+
+/*
+ * Retrieves the number of blocks of the data device from
+ * the superblock and compares it to the actual device size,
+ * thus resizing the data device in case it has grown.
+ *
+ * This both copes with opening preallocated data devices in the ctr
+ * being followed by a resume
+ * -and-
+ * calling the resume method individually after userspace has
+ * grown the data device in reaction to a table event.
+ */
+static int pool_preresume(struct dm_target *ti)
+{
+ int r;
+ bool need_commit1, need_commit2;
+ struct pool_c *pt = ti->private;
+ struct pool *pool = pt->pool;
+
+ /*
+ * Take control of the pool object.
+ */
+ r = bind_control_target(pool, ti);
+ if (r)
+ return r;
+
+ r = maybe_resize_data_dev(ti, &need_commit1);
+ if (r)
+ return r;
+
+ r = maybe_resize_metadata_dev(ti, &need_commit2);
+ if (r)
+ return r;
+
+ if (need_commit1 || need_commit2)
+ (void) commit_or_fallback(pool);
+
+ return 0;
+}
+
static void pool_resume(struct dm_target *ti)
{
struct pool_c *pt = ti->private;
@@ -2549,7 +2669,7 @@ static struct target_type pool_target = {
.name = "thin-pool",
.features = DM_TARGET_SINGLETON | DM_TARGET_ALWAYS_WRITEABLE |
DM_TARGET_IMMUTABLE,
- .version = {1, 7, 0},
+ .version = {1, 8, 0},
.module = THIS_MODULE,
.ctr = pool_ctr,
.dtr = pool_dtr,
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 681d1099a2d..9b82377a833 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -5268,8 +5268,8 @@ static void md_clean(struct mddev *mddev)
static void __md_stop_writes(struct mddev *mddev)
{
+ set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
if (mddev->sync_thread) {
- set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
set_bit(MD_RECOVERY_INTR, &mddev->recovery);
md_reap_sync_thread(mddev);
}
diff --git a/drivers/md/persistent-data/dm-space-map-disk.c b/drivers/md/persistent-data/dm-space-map-disk.c
index f6d29e614ab..e735a6d5a79 100644
--- a/drivers/md/persistent-data/dm-space-map-disk.c
+++ b/drivers/md/persistent-data/dm-space-map-disk.c
@@ -248,7 +248,8 @@ static struct dm_space_map ops = {
.new_block = sm_disk_new_block,
.commit = sm_disk_commit,
.root_size = sm_disk_root_size,
- .copy_root = sm_disk_copy_root
+ .copy_root = sm_disk_copy_root,
+ .register_threshold_callback = NULL
};
struct dm_space_map *dm_sm_disk_create(struct dm_transaction_manager *tm,
diff --git a/drivers/md/persistent-data/dm-space-map-metadata.c b/drivers/md/persistent-data/dm-space-map-metadata.c
index 906cf3df71a..1c959684cae 100644
--- a/drivers/md/persistent-data/dm-space-map-metadata.c
+++ b/drivers/md/persistent-data/dm-space-map-metadata.c
@@ -17,6 +17,55 @@
/*----------------------------------------------------------------*/
/*
+ * An edge triggered threshold.
+ */
+struct threshold {
+ bool threshold_set;
+ bool value_set;
+ dm_block_t threshold;
+ dm_block_t current_value;
+ dm_sm_threshold_fn fn;
+ void *context;
+};
+
+static void threshold_init(struct threshold *t)
+{
+ t->threshold_set = false;
+ t->value_set = false;
+}
+
+static void set_threshold(struct threshold *t, dm_block_t value,
+ dm_sm_threshold_fn fn, void *context)
+{
+ t->threshold_set = true;
+ t->threshold = value;
+ t->fn = fn;
+ t->context = context;
+}
+
+static bool below_threshold(struct threshold *t, dm_block_t value)
+{
+ return t->threshold_set && value <= t->threshold;
+}
+
+static bool threshold_already_triggered(struct threshold *t)
+{
+ return t->value_set && below_threshold(t, t->current_value);
+}
+
+static void check_threshold(struct threshold *t, dm_block_t value)
+{
+ if (below_threshold(t, value) &&
+ !threshold_already_triggered(t))
+ t->fn(t->context);
+
+ t->value_set = true;
+ t->current_value = value;
+}
+
+/*----------------------------------------------------------------*/
+
+/*
* Space map interface.
*
* The low level disk format is written using the standard btree and
@@ -54,6 +103,8 @@ struct sm_metadata {
unsigned allocated_this_transaction;
unsigned nr_uncommitted;
struct block_op uncommitted[MAX_RECURSIVE_ALLOCATIONS];
+
+ struct threshold threshold;
};
static int add_bop(struct sm_metadata *smm, enum block_op_type type, dm_block_t b)
@@ -144,12 +195,6 @@ static void sm_metadata_destroy(struct dm_space_map *sm)
kfree(smm);
}
-static int sm_metadata_extend(struct dm_space_map *sm, dm_block_t extra_blocks)
-{
- DMERR("doesn't support extend");
- return -EINVAL;
-}
-
static int sm_metadata_get_nr_blocks(struct dm_space_map *sm, dm_block_t *count)
{
struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm);
@@ -335,9 +380,19 @@ static int sm_metadata_new_block_(struct dm_space_map *sm, dm_block_t *b)
static int sm_metadata_new_block(struct dm_space_map *sm, dm_block_t *b)
{
+ dm_block_t count;
+ struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm);
+
int r = sm_metadata_new_block_(sm, b);
if (r)
DMERR("unable to allocate new metadata block");
+
+ r = sm_metadata_get_nr_free(sm, &count);
+ if (r)
+ DMERR("couldn't get free block count");
+
+ check_threshold(&smm->threshold, count);
+
return r;
}
@@ -357,6 +412,18 @@ static int sm_metadata_commit(struct dm_space_map *sm)
return 0;
}
+static int sm_metadata_register_threshold_callback(struct dm_space_map *sm,
+ dm_block_t threshold,
+ dm_sm_threshold_fn fn,
+ void *context)
+{
+ struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm);
+
+ set_threshold(&smm->threshold, threshold, fn, context);
+
+ return 0;
+}
+
static int sm_metadata_root_size(struct dm_space_map *sm, size_t *result)
{
*result = sizeof(struct disk_sm_root);
@@ -382,6 +449,8 @@ static int sm_metadata_copy_root(struct dm_space_map *sm, void *where_le, size_t
return 0;
}
+static int sm_metadata_extend(struct dm_space_map *sm, dm_block_t extra_blocks);
+
static struct dm_space_map ops = {
.destroy = sm_metadata_destroy,
.extend = sm_metadata_extend,
@@ -395,7 +464,8 @@ static struct dm_space_map ops = {
.new_block = sm_metadata_new_block,
.commit = sm_metadata_commit,
.root_size = sm_metadata_root_size,
- .copy_root = sm_metadata_copy_root
+ .copy_root = sm_metadata_copy_root,
+ .register_threshold_callback = sm_metadata_register_threshold_callback
};
/*----------------------------------------------------------------*/
@@ -410,7 +480,7 @@ static void sm_bootstrap_destroy(struct dm_space_map *sm)
static int sm_bootstrap_extend(struct dm_space_map *sm, dm_block_t extra_blocks)
{
- DMERR("boostrap doesn't support extend");
+ DMERR("bootstrap doesn't support extend");
return -EINVAL;
}
@@ -450,7 +520,7 @@ static int sm_bootstrap_count_is_more_than_one(struct dm_space_map *sm,
static int sm_bootstrap_set_count(struct dm_space_map *sm, dm_block_t b,
uint32_t count)
{
- DMERR("boostrap doesn't support set_count");
+ DMERR("bootstrap doesn't support set_count");
return -EINVAL;
}
@@ -491,7 +561,7 @@ static int sm_bootstrap_commit(struct dm_space_map *sm)
static int sm_bootstrap_root_size(struct dm_space_map *sm, size_t *result)
{
- DMERR("boostrap doesn't support root_size");
+ DMERR("bootstrap doesn't support root_size");
return -EINVAL;
}
@@ -499,7 +569,7 @@ static int sm_bootstrap_root_size(struct dm_space_map *sm, size_t *result)
static int sm_bootstrap_copy_root(struct dm_space_map *sm, void *where,
size_t max)
{
- DMERR("boostrap doesn't support copy_root");
+ DMERR("bootstrap doesn't support copy_root");
return -EINVAL;
}
@@ -517,11 +587,42 @@ static struct dm_space_map bootstrap_ops = {
.new_block = sm_bootstrap_new_block,
.commit = sm_bootstrap_commit,
.root_size = sm_bootstrap_root_size,
- .copy_root = sm_bootstrap_copy_root
+ .copy_root = sm_bootstrap_copy_root,
+ .register_threshold_callback = NULL
};
/*----------------------------------------------------------------*/
+static int sm_metadata_extend(struct dm_space_map *sm, dm_block_t extra_blocks)
+{
+ int r, i;
+ enum allocation_event ev;
+ struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm);
+ dm_block_t old_len = smm->ll.nr_blocks;
+
+ /*
+ * Flick into a mode where all blocks get allocated in the new area.
+ */
+ smm->begin = old_len;
+ memcpy(&smm->sm, &bootstrap_ops, sizeof(smm->sm));
+
+ /*
+ * Extend.
+ */
+ r = sm_ll_extend(&smm->ll, extra_blocks);
+
+ /*
+ * Switch back to normal behaviour.
+ */
+ memcpy(&smm->sm, &ops, sizeof(smm->sm));
+ for (i = old_len; !r && i < smm->begin; i++)
+ r = sm_ll_inc(&smm->ll, i, &ev);
+
+ return r;
+}
+
+/*----------------------------------------------------------------*/
+
struct dm_space_map *dm_sm_metadata_init(void)
{
struct sm_metadata *smm;
@@ -549,6 +650,7 @@ int dm_sm_metadata_create(struct dm_space_map *sm,
smm->recursion_count = 0;
smm->allocated_this_transaction = 0;
smm->nr_uncommitted = 0;
+ threshold_init(&smm->threshold);
memcpy(&smm->sm, &bootstrap_ops, sizeof(smm->sm));
@@ -590,6 +692,7 @@ int dm_sm_metadata_open(struct dm_space_map *sm,
smm->recursion_count = 0;
smm->allocated_this_transaction = 0;
smm->nr_uncommitted = 0;
+ threshold_init(&smm->threshold);
memcpy(&smm->old_ll, &smm->ll, sizeof(smm->old_ll));
return 0;
diff --git a/drivers/md/persistent-data/dm-space-map.h b/drivers/md/persistent-data/dm-space-map.h
index 1cbfc6b1638..3e6d1153b7c 100644
--- a/drivers/md/persistent-data/dm-space-map.h
+++ b/drivers/md/persistent-data/dm-space-map.h
@@ -9,6 +9,8 @@
#include "dm-block-manager.h"
+typedef void (*dm_sm_threshold_fn)(void *context);
+
/*
* struct dm_space_map keeps a record of how many times each block in a device
* is referenced. It needs to be fixed on disk as part of the transaction.
@@ -59,6 +61,15 @@ struct dm_space_map {
*/
int (*root_size)(struct dm_space_map *sm, size_t *result);
int (*copy_root)(struct dm_space_map *sm, void *copy_to_here_le, size_t len);
+
+ /*
+ * You can register one threshold callback which is edge-triggered
+ * when the free space in the space map drops below the threshold.
+ */
+ int (*register_threshold_callback)(struct dm_space_map *sm,
+ dm_block_t threshold,
+ dm_sm_threshold_fn fn,
+ void *context);
};
/*----------------------------------------------------------------*/
@@ -131,4 +142,16 @@ static inline int dm_sm_copy_root(struct dm_space_map *sm, void *copy_to_here_le
return sm->copy_root(sm, copy_to_here_le, len);
}
+static inline int dm_sm_register_threshold_callback(struct dm_space_map *sm,
+ dm_block_t threshold,
+ dm_sm_threshold_fn fn,
+ void *context)
+{
+ if (sm->register_threshold_callback)
+ return sm->register_threshold_callback(sm, threshold, fn, context);
+
+ return -EINVAL;
+}
+
+
#endif /* _LINUX_DM_SPACE_MAP_H */
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 55951182af7..6e17f8181c4 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -417,7 +417,17 @@ static void raid1_end_write_request(struct bio *bio, int error)
r1_bio->bios[mirror] = NULL;
to_put = bio;
- set_bit(R1BIO_Uptodate, &r1_bio->state);
+ /*
+ * Do not set R1BIO_Uptodate if the current device is
+ * rebuilding or Faulty. This is because we cannot use
+ * such device for properly reading the data back (we could
+ * potentially use it, if the current write would have felt
+ * before rdev->recovery_offset, but for simplicity we don't
+ * check this here.
+ */
+ if (test_bit(In_sync, &conf->mirrors[mirror].rdev->flags) &&
+ !test_bit(Faulty, &conf->mirrors[mirror].rdev->flags))
+ set_bit(R1BIO_Uptodate, &r1_bio->state);
/* Maybe we can clear some bad blocks. */
if (is_badblock(conf->mirrors[mirror].rdev,
@@ -870,17 +880,17 @@ static void allow_barrier(struct r1conf *conf)
wake_up(&conf->wait_barrier);
}
-static void freeze_array(struct r1conf *conf)
+static void freeze_array(struct r1conf *conf, int extra)
{
/* stop syncio and normal IO and wait for everything to
* go quite.
* We increment barrier and nr_waiting, and then
- * wait until nr_pending match nr_queued+1
+ * wait until nr_pending match nr_queued+extra
* This is called in the context of one normal IO request
* that has failed. Thus any sync request that might be pending
* will be blocked by nr_pending, and we need to wait for
* pending IO requests to complete or be queued for re-try.
- * Thus the number queued (nr_queued) plus this request (1)
+ * Thus the number queued (nr_queued) plus this request (extra)
* must match the number of pending IOs (nr_pending) before
* we continue.
*/
@@ -888,7 +898,7 @@ static void freeze_array(struct r1conf *conf)
conf->barrier++;
conf->nr_waiting++;
wait_event_lock_irq_cmd(conf->wait_barrier,
- conf->nr_pending == conf->nr_queued+1,
+ conf->nr_pending == conf->nr_queued+extra,
conf->resync_lock,
flush_pending_writes(conf));
spin_unlock_irq(&conf->resync_lock);
@@ -1544,8 +1554,8 @@ static int raid1_add_disk(struct mddev *mddev, struct md_rdev *rdev)
* we wait for all outstanding requests to complete.
*/
synchronize_sched();
- raise_barrier(conf);
- lower_barrier(conf);
+ freeze_array(conf, 0);
+ unfreeze_array(conf);
clear_bit(Unmerged, &rdev->flags);
}
md_integrity_add_rdev(rdev, mddev);
@@ -1595,11 +1605,11 @@ static int raid1_remove_disk(struct mddev *mddev, struct md_rdev *rdev)
*/
struct md_rdev *repl =
conf->mirrors[conf->raid_disks + number].rdev;
- raise_barrier(conf);
+ freeze_array(conf, 0);
clear_bit(Replacement, &repl->flags);
p->rdev = repl;
conf->mirrors[conf->raid_disks + number].rdev = NULL;
- lower_barrier(conf);
+ unfreeze_array(conf);
clear_bit(WantReplacement, &rdev->flags);
} else
clear_bit(WantReplacement, &rdev->flags);
@@ -2195,7 +2205,7 @@ static void handle_read_error(struct r1conf *conf, struct r1bio *r1_bio)
* frozen
*/
if (mddev->ro == 0) {
- freeze_array(conf);
+ freeze_array(conf, 1);
fix_read_error(conf, r1_bio->read_disk,
r1_bio->sector, r1_bio->sectors);
unfreeze_array(conf);
@@ -2780,8 +2790,8 @@ static int run(struct mddev *mddev)
return PTR_ERR(conf);
if (mddev->queue)
- blk_queue_max_write_same_sectors(mddev->queue,
- mddev->chunk_sectors);
+ blk_queue_max_write_same_sectors(mddev->queue, 0);
+
rdev_for_each(rdev, mddev) {
if (!mddev->gendisk)
continue;
@@ -2963,7 +2973,7 @@ static int raid1_reshape(struct mddev *mddev)
return -ENOMEM;
}
- raise_barrier(conf);
+ freeze_array(conf, 0);
/* ok, everything is stopped */
oldpool = conf->r1bio_pool;
@@ -2994,7 +3004,7 @@ static int raid1_reshape(struct mddev *mddev)
conf->raid_disks = mddev->raid_disks = raid_disks;
mddev->delta_disks = 0;
- lower_barrier(conf);
+ unfreeze_array(conf);
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
md_wakeup_thread(mddev->thread);
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 59d4daa5f4c..6ddae2501b9 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -490,7 +490,17 @@ static void raid10_end_write_request(struct bio *bio, int error)
sector_t first_bad;
int bad_sectors;
- set_bit(R10BIO_Uptodate, &r10_bio->state);
+ /*
+ * Do not set R10BIO_Uptodate if the current device is
+ * rebuilding or Faulty. This is because we cannot use
+ * such device for properly reading the data back (we could
+ * potentially use it, if the current write would have felt
+ * before rdev->recovery_offset, but for simplicity we don't
+ * check this here.
+ */
+ if (test_bit(In_sync, &rdev->flags) &&
+ !test_bit(Faulty, &rdev->flags))
+ set_bit(R10BIO_Uptodate, &r10_bio->state);
/* Maybe we can clear some bad blocks. */
if (is_badblock(rdev,
@@ -1055,17 +1065,17 @@ static void allow_barrier(struct r10conf *conf)
wake_up(&conf->wait_barrier);
}
-static void freeze_array(struct r10conf *conf)
+static void freeze_array(struct r10conf *conf, int extra)
{
/* stop syncio and normal IO and wait for everything to
* go quiet.
* We increment barrier and nr_waiting, and then
- * wait until nr_pending match nr_queued+1
+ * wait until nr_pending match nr_queued+extra
* This is called in the context of one normal IO request
* that has failed. Thus any sync request that might be pending
* will be blocked by nr_pending, and we need to wait for
* pending IO requests to complete or be queued for re-try.
- * Thus the number queued (nr_queued) plus this request (1)
+ * Thus the number queued (nr_queued) plus this request (extra)
* must match the number of pending IOs (nr_pending) before
* we continue.
*/
@@ -1073,7 +1083,7 @@ static void freeze_array(struct r10conf *conf)
conf->barrier++;
conf->nr_waiting++;
wait_event_lock_irq_cmd(conf->wait_barrier,
- conf->nr_pending == conf->nr_queued+1,
+ conf->nr_pending == conf->nr_queued+extra,
conf->resync_lock,
flush_pending_writes(conf));
@@ -1837,8 +1847,8 @@ static int raid10_add_disk(struct mddev *mddev, struct md_rdev *rdev)
* we wait for all outstanding requests to complete.
*/
synchronize_sched();
- raise_barrier(conf, 0);
- lower_barrier(conf);
+ freeze_array(conf, 0);
+ unfreeze_array(conf);
clear_bit(Unmerged, &rdev->flags);
}
md_integrity_add_rdev(rdev, mddev);
@@ -2612,7 +2622,7 @@ static void handle_read_error(struct mddev *mddev, struct r10bio *r10_bio)
r10_bio->devs[slot].bio = NULL;
if (mddev->ro == 0) {
- freeze_array(conf);
+ freeze_array(conf, 1);
fix_read_error(conf, mddev, r10_bio);
unfreeze_array(conf);
} else
@@ -3609,8 +3619,7 @@ static int run(struct mddev *mddev)
if (mddev->queue) {
blk_queue_max_discard_sectors(mddev->queue,
mddev->chunk_sectors);
- blk_queue_max_write_same_sectors(mddev->queue,
- mddev->chunk_sectors);
+ blk_queue_max_write_same_sectors(mddev->queue, 0);
blk_queue_io_min(mddev->queue, chunk_size);
if (conf->geo.raid_disks % conf->geo.near_copies)
blk_queue_io_opt(mddev->queue, chunk_size * conf->geo.raid_disks);
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 9359828ffe2..05e4a105b9c 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -664,6 +664,7 @@ static void ops_run_io(struct stripe_head *sh, struct stripe_head_state *s)
if (test_bit(R5_ReadNoMerge, &sh->dev[i].flags))
bi->bi_rw |= REQ_FLUSH;
+ bi->bi_vcnt = 1;
bi->bi_io_vec[0].bv_len = STRIPE_SIZE;
bi->bi_io_vec[0].bv_offset = 0;
bi->bi_size = STRIPE_SIZE;
@@ -701,6 +702,7 @@ static void ops_run_io(struct stripe_head *sh, struct stripe_head_state *s)
else
rbi->bi_sector = (sh->sector
+ rrdev->data_offset);
+ rbi->bi_vcnt = 1;
rbi->bi_io_vec[0].bv_len = STRIPE_SIZE;
rbi->bi_io_vec[0].bv_offset = 0;
rbi->bi_size = STRIPE_SIZE;
@@ -5464,7 +5466,7 @@ static int run(struct mddev *mddev)
if (mddev->major_version == 0 &&
mddev->minor_version > 90)
rdev->recovery_offset = reshape_offset;
-
+
if (rdev->recovery_offset < reshape_offset) {
/* We need to check old and new layout */
if (!only_parity(rdev->raid_disk,
@@ -5587,6 +5589,8 @@ static int run(struct mddev *mddev)
*/
mddev->queue->limits.discard_zeroes_data = 0;
+ blk_queue_max_write_same_sectors(mddev->queue, 0);
+
rdev_for_each(rdev, mddev) {
disk_stack_limits(mddev->gendisk, rdev->bdev,
rdev->data_offset << 9);
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-18 00:44:33 +0000