aboutsummaryrefslogtreecommitdiff
path: root/drivers/md
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/md')
-rw-r--r--drivers/md/bitmap.c9
-rw-r--r--drivers/md/dm-mpath.c68
-rw-r--r--drivers/md/dm-raid1.c1
-rw-r--r--drivers/md/dm-snap.c6
-rw-r--r--drivers/md/dm-table.c1
-rw-r--r--drivers/md/dm.c27
-rw-r--r--drivers/md/md.c1
-rw-r--r--drivers/md/raid0.c8
-rw-r--r--drivers/md/raid1.c37
9 files changed, 92 insertions, 66 deletions
diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c
index 95980ad6b27..0c2ed99a383 100644
--- a/drivers/md/bitmap.c
+++ b/drivers/md/bitmap.c
@@ -1345,7 +1345,8 @@ void bitmap_endwrite(struct bitmap *bitmap, sector_t offset, unsigned long secto
}
}
-int bitmap_start_sync(struct bitmap *bitmap, sector_t offset, int *blocks)
+int bitmap_start_sync(struct bitmap *bitmap, sector_t offset, int *blocks,
+ int degraded)
{
bitmap_counter_t *bmc;
int rv;
@@ -1362,8 +1363,10 @@ int bitmap_start_sync(struct bitmap *bitmap, sector_t offset, int *blocks)
rv = 1;
else if (NEEDED(*bmc)) {
rv = 1;
- *bmc |= RESYNC_MASK;
- *bmc &= ~NEEDED_MASK;
+ if (!degraded) { /* don't set/clear bits if degraded */
+ *bmc |= RESYNC_MASK;
+ *bmc &= ~NEEDED_MASK;
+ }
}
}
spin_unlock_irq(&bitmap->lock);
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index 0c1b8520ef8..785806bdb24 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -63,6 +63,7 @@ struct multipath {
unsigned nr_priority_groups;
struct list_head priority_groups;
unsigned pg_init_required; /* pg_init needs calling? */
+ unsigned pg_init_in_progress; /* Only one pg_init allowed at once */
unsigned nr_valid_paths; /* Total number of usable paths */
struct pgpath *current_pgpath;
@@ -72,7 +73,7 @@ struct multipath {
unsigned queue_io; /* Must we queue all I/O? */
unsigned queue_if_no_path; /* Queue I/O if last path fails? */
- unsigned suspended; /* Has dm core suspended our I/O? */
+ unsigned saved_queue_if_no_path;/* Saved state during suspension */
struct work_struct process_queued_ios;
struct bio_list queued_ios;
@@ -304,11 +305,12 @@ static int map_io(struct multipath *m, struct bio *bio, struct mpath_io *mpio,
m->queue_size--;
if ((pgpath && m->queue_io) ||
- (!pgpath && m->queue_if_no_path && !m->suspended)) {
+ (!pgpath && m->queue_if_no_path)) {
/* Queue for the daemon to resubmit */
bio_list_add(&m->queued_ios, bio);
m->queue_size++;
- if (m->pg_init_required || !m->queue_io)
+ if ((m->pg_init_required && !m->pg_init_in_progress) ||
+ !m->queue_io)
queue_work(kmultipathd, &m->process_queued_ios);
pgpath = NULL;
r = 0;
@@ -333,8 +335,9 @@ static int queue_if_no_path(struct multipath *m, unsigned queue_if_no_path)
spin_lock_irqsave(&m->lock, flags);
+ m->saved_queue_if_no_path = m->queue_if_no_path;
m->queue_if_no_path = queue_if_no_path;
- if (!m->queue_if_no_path)
+ if (!m->queue_if_no_path && m->queue_size)
queue_work(kmultipathd, &m->process_queued_ios);
spin_unlock_irqrestore(&m->lock, flags);
@@ -379,25 +382,31 @@ static void process_queued_ios(void *data)
{
struct multipath *m = (struct multipath *) data;
struct hw_handler *hwh = &m->hw_handler;
- struct pgpath *pgpath;
- unsigned init_required, must_queue = 0;
+ struct pgpath *pgpath = NULL;
+ unsigned init_required = 0, must_queue = 1;
unsigned long flags;
spin_lock_irqsave(&m->lock, flags);
+ if (!m->queue_size)
+ goto out;
+
if (!m->current_pgpath)
__choose_pgpath(m);
pgpath = m->current_pgpath;
- if ((pgpath && m->queue_io) ||
- (!pgpath && m->queue_if_no_path && !m->suspended))
- must_queue = 1;
+ if ((pgpath && !m->queue_io) ||
+ (!pgpath && !m->queue_if_no_path))
+ must_queue = 0;
- init_required = m->pg_init_required;
- if (init_required)
+ if (m->pg_init_required && !m->pg_init_in_progress) {
m->pg_init_required = 0;
+ m->pg_init_in_progress = 1;
+ init_required = 1;
+ }
+out:
spin_unlock_irqrestore(&m->lock, flags);
if (init_required)
@@ -752,6 +761,8 @@ static int multipath_ctr(struct dm_target *ti, unsigned int argc,
static void multipath_dtr(struct dm_target *ti)
{
struct multipath *m = (struct multipath *) ti->private;
+
+ flush_workqueue(kmultipathd);
free_multipath(m);
}
@@ -765,6 +776,9 @@ static int multipath_map(struct dm_target *ti, struct bio *bio,
struct mpath_io *mpio;
struct multipath *m = (struct multipath *) ti->private;
+ if (bio_barrier(bio))
+ return -EOPNOTSUPP;
+
mpio = mempool_alloc(m->mpio_pool, GFP_NOIO);
dm_bio_record(&mpio->details, bio);
@@ -837,7 +851,7 @@ static int reinstate_path(struct pgpath *pgpath)
pgpath->path.is_active = 1;
m->current_pgpath = NULL;
- if (!m->nr_valid_paths++)
+ if (!m->nr_valid_paths++ && m->queue_size)
queue_work(kmultipathd, &m->process_queued_ios);
queue_work(kmultipathd, &m->trigger_event);
@@ -963,12 +977,13 @@ void dm_pg_init_complete(struct path *path, unsigned err_flags)
bypass_pg(m, pg, 1);
spin_lock_irqsave(&m->lock, flags);
- if (!err_flags)
- m->queue_io = 0;
- else {
+ if (err_flags) {
m->current_pgpath = NULL;
m->current_pg = NULL;
- }
+ } else if (!m->pg_init_required)
+ m->queue_io = 0;
+
+ m->pg_init_in_progress = 0;
queue_work(kmultipathd, &m->process_queued_ios);
spin_unlock_irqrestore(&m->lock, flags);
}
@@ -988,9 +1003,12 @@ static int do_end_io(struct multipath *m, struct bio *bio,
if ((error == -EWOULDBLOCK) && bio_rw_ahead(bio))
return error;
+ if (error == -EOPNOTSUPP)
+ return error;
+
spin_lock(&m->lock);
if (!m->nr_valid_paths) {
- if (!m->queue_if_no_path || m->suspended) {
+ if (!m->queue_if_no_path) {
spin_unlock(&m->lock);
return -EIO;
} else {
@@ -1051,27 +1069,27 @@ static int multipath_end_io(struct dm_target *ti, struct bio *bio,
/*
* Suspend can't complete until all the I/O is processed so if
- * the last path failed we will now error any queued I/O.
+ * the last path fails we must error any remaining I/O.
+ * Note that if the freeze_bdev fails while suspending, the
+ * queue_if_no_path state is lost - userspace should reset it.
*/
static void multipath_presuspend(struct dm_target *ti)
{
struct multipath *m = (struct multipath *) ti->private;
- unsigned long flags;
- spin_lock_irqsave(&m->lock, flags);
- m->suspended = 1;
- if (m->queue_if_no_path)
- queue_work(kmultipathd, &m->process_queued_ios);
- spin_unlock_irqrestore(&m->lock, flags);
+ queue_if_no_path(m, 0);
}
+/*
+ * Restore the queue_if_no_path setting.
+ */
static void multipath_resume(struct dm_target *ti)
{
struct multipath *m = (struct multipath *) ti->private;
unsigned long flags;
spin_lock_irqsave(&m->lock, flags);
- m->suspended = 0;
+ m->queue_if_no_path = m->saved_queue_if_no_path;
spin_unlock_irqrestore(&m->lock, flags);
}
diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c
index 6e3cf7e1345..12031c9d3f1 100644
--- a/drivers/md/dm-raid1.c
+++ b/drivers/md/dm-raid1.c
@@ -1060,6 +1060,7 @@ static int mirror_ctr(struct dm_target *ti, unsigned int argc, char **argv)
}
ti->private = ms;
+ ti->split_io = ms->rh.region_size;
r = kcopyd_client_create(DM_IO_PAGES, &ms->kcopyd_client);
if (r) {
diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index 7e691ab9a74..ab54f99b7c3 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -777,7 +777,7 @@ static int snapshot_map(struct dm_target *ti, struct bio *bio,
/* Full snapshots are not usable */
if (!s->valid)
- return -1;
+ return -EIO;
/*
* Write to snapshot - higher level takes care of RW/RO
@@ -931,6 +931,10 @@ static int __origin_write(struct list_head *snapshots, struct bio *bio)
if (!snap->valid)
continue;
+ /* Nothing to do if writing beyond end of snapshot */
+ if (bio->bi_sector >= dm_table_get_size(snap->table))
+ continue;
+
down_write(&snap->lock);
/*
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index 18e9b9953fc..a5a4c0ed8a1 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -943,6 +943,7 @@ EXPORT_SYMBOL(dm_vcalloc);
EXPORT_SYMBOL(dm_get_device);
EXPORT_SYMBOL(dm_put_device);
EXPORT_SYMBOL(dm_table_event);
+EXPORT_SYMBOL(dm_table_get_size);
EXPORT_SYMBOL(dm_table_get_mode);
EXPORT_SYMBOL(dm_table_put);
EXPORT_SYMBOL(dm_table_get);
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index f6b03957efc..54fabbf0667 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -384,7 +384,7 @@ static void __map_bio(struct dm_target *ti, struct bio *clone,
/* error the io and bail out */
struct dm_io *io = tio->io;
free_tio(tio->io->md, tio);
- dec_pending(io, -EIO);
+ dec_pending(io, r);
bio_put(clone);
}
}
@@ -966,23 +966,20 @@ static void __flush_deferred_io(struct mapped_device *md, struct bio *c)
*/
int dm_swap_table(struct mapped_device *md, struct dm_table *table)
{
- int r;
+ int r = -EINVAL;
down_write(&md->lock);
/* device must be suspended */
- if (!test_bit(DMF_SUSPENDED, &md->flags)) {
- up_write(&md->lock);
- return -EPERM;
- }
+ if (!test_bit(DMF_SUSPENDED, &md->flags))
+ goto out;
__unbind(md);
r = __bind(md, table);
- if (r)
- return r;
+out:
up_write(&md->lock);
- return 0;
+ return r;
}
/*
@@ -1055,14 +1052,17 @@ int dm_suspend(struct mapped_device *md)
if (test_bit(DMF_BLOCK_IO, &md->flags))
goto out_read_unlock;
- error = __lock_fs(md);
- if (error)
- goto out_read_unlock;
-
map = dm_get_table(md);
if (map)
+ /* This does not get reverted if there's an error later. */
dm_table_presuspend_targets(map);
+ error = __lock_fs(md);
+ if (error) {
+ dm_table_put(map);
+ goto out_read_unlock;
+ }
+
up_read(&md->lock);
/*
@@ -1121,7 +1121,6 @@ int dm_suspend(struct mapped_device *md)
return 0;
out_unfreeze:
- /* FIXME Undo dm_table_presuspend_targets */
__unlock_fs(md);
clear_bit(DMF_BLOCK_IO, &md->flags);
out_write_unlock:
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 3802f7a17f1..4a0c57db2b6 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -338,6 +338,7 @@ static int super_written(struct bio *bio, unsigned int bytes_done, int error)
if (atomic_dec_and_test(&rdev->mddev->pending_writes))
wake_up(&rdev->mddev->sb_wait);
+ bio_put(bio);
return 0;
}
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index e11dd14d0b4..2120710172c 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -314,16 +314,16 @@ static int raid0_run (mddev_t *mddev)
sector_t space = conf->hash_spacing;
int round;
conf->preshift = 0;
- if (sizeof(sector_t) > sizeof(unsigned long)) {
+ if (sizeof(sector_t) > sizeof(u32)) {
/*shift down space and s so that sector_div will work */
- while (space > (sector_t) (~(unsigned long)0)) {
+ while (space > (sector_t) (~(u32)0)) {
s >>= 1;
space >>= 1;
s += 1; /* force round-up */
conf->preshift++;
}
}
- round = sector_div(s, (unsigned long)space) ? 1 : 0;
+ round = sector_div(s, (u32)space) ? 1 : 0;
nb_zone = s + round;
}
printk("raid0 : nb_zone is %d.\n", nb_zone);
@@ -443,7 +443,7 @@ static int raid0_make_request (request_queue_t *q, struct bio *bio)
volatile
#endif
sector_t x = block >> conf->preshift;
- sector_div(x, (unsigned long)conf->hash_spacing);
+ sector_div(x, (u32)conf->hash_spacing);
zone = conf->hash_table[x];
}
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index ff1dbec864a..5f253ee536b 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1126,21 +1126,19 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
* only be one in raid1 resync.
* We can find the current addess in mddev->curr_resync
*/
- if (!conf->fullsync) {
- if (mddev->curr_resync < max_sector)
- bitmap_end_sync(mddev->bitmap,
- mddev->curr_resync,
+ if (mddev->curr_resync < max_sector) /* aborted */
+ bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
&sync_blocks, 1);
- bitmap_close_sync(mddev->bitmap);
- }
- if (mddev->curr_resync >= max_sector)
+ else /* completed sync */
conf->fullsync = 0;
+
+ bitmap_close_sync(mddev->bitmap);
close_sync(conf);
return 0;
}
- if (!conf->fullsync &&
- !bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks)) {
+ if (!bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, mddev->degraded) &&
+ !conf->fullsync) {
/* We can skip this block, and probably several more */
*skipped = 1;
return sync_blocks;
@@ -1243,15 +1241,15 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
len = (max_sector - sector_nr) << 9;
if (len == 0)
break;
- if (!conf->fullsync) {
- if (sync_blocks == 0) {
- if (!bitmap_start_sync(mddev->bitmap,
- sector_nr, &sync_blocks))
- break;
- if (sync_blocks < (PAGE_SIZE>>9))
- BUG();
- if (len > (sync_blocks<<9)) len = sync_blocks<<9;
- }
+ if (sync_blocks == 0) {
+ if (!bitmap_start_sync(mddev->bitmap, sector_nr,
+ &sync_blocks, mddev->degraded) &&
+ !conf->fullsync)
+ break;
+ if (sync_blocks < (PAGE_SIZE>>9))
+ BUG();
+ if (len > (sync_blocks<<9))
+ len = sync_blocks<<9;
}
for (i=0 ; i < conf->raid_disks; i++) {
@@ -1264,7 +1262,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
while (i > 0) {
i--;
bio = r1_bio->bios[i];
- if (bio->bi_end_io==NULL) continue;
+ if (bio->bi_end_io==NULL)
+ continue;
/* remove last page from this bio */
bio->bi_vcnt--;
bio->bi_size -= len;