summaryrefslogtreecommitdiff
path: root/drivers/md
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/md')
-rw-r--r--drivers/md/bitmap.h2
-rw-r--r--drivers/md/dm-raid.c8
-rw-r--r--drivers/md/dm-region-hash.c2
-rw-r--r--drivers/md/dm-table.c114
-rw-r--r--drivers/md/faulty.c2
-rw-r--r--drivers/md/md.c90
-rw-r--r--drivers/md/md.h28
-rw-r--r--drivers/md/raid1.c29
-rw-r--r--drivers/md/raid10.c33
-rw-r--r--drivers/md/raid10.h4
-rw-r--r--drivers/md/raid5.c66
-rw-r--r--drivers/md/raid5.h2
12 files changed, 198 insertions, 182 deletions
diff --git a/drivers/md/bitmap.h b/drivers/md/bitmap.h
index 931a7a7c3796..d0aeaf46d932 100644
--- a/drivers/md/bitmap.h
+++ b/drivers/md/bitmap.h
@@ -45,7 +45,7 @@
*
* The counter counts pending write requests, plus the on-disk bit.
* When the counter is '1' and the resync bits are clear, the on-disk
- * bit can be cleared aswell, thus setting the counter to 0.
+ * bit can be cleared as well, thus setting the counter to 0.
* When we set a bit, or in the counter (to start a write), if the fields is
* 0, we first set the disk bit and set the counter to 1.
*
diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c
index 5ef136cdba91..e5d8904fc8f6 100644
--- a/drivers/md/dm-raid.c
+++ b/drivers/md/dm-raid.c
@@ -390,13 +390,6 @@ static int raid_is_congested(struct dm_target_callbacks *cb, int bits)
return md_raid5_congested(&rs->md, bits);
}
-static void raid_unplug(struct dm_target_callbacks *cb)
-{
- struct raid_set *rs = container_of(cb, struct raid_set, callbacks);
-
- md_raid5_kick_device(rs->md.private);
-}
-
/*
* Construct a RAID4/5/6 mapping:
* Args:
@@ -487,7 +480,6 @@ static int raid_ctr(struct dm_target *ti, unsigned argc, char **argv)
}
rs->callbacks.congested_fn = raid_is_congested;
- rs->callbacks.unplug_fn = raid_unplug;
dm_table_add_target_callbacks(ti->table, &rs->callbacks);
return 0;
diff --git a/drivers/md/dm-region-hash.c b/drivers/md/dm-region-hash.c
index dad011aed0c9..7771ed212182 100644
--- a/drivers/md/dm-region-hash.c
+++ b/drivers/md/dm-region-hash.c
@@ -419,7 +419,7 @@ void dm_rh_mark_nosync(struct dm_region_hash *rh, struct bio *bio)
/*
* Possible cases:
* 1) DM_RH_DIRTY
- * 2) DM_RH_NOSYNC: was dirty, other preceeding writes failed
+ * 2) DM_RH_NOSYNC: was dirty, other preceding writes failed
* 3) DM_RH_RECOVERING: flushing pending writes
* Either case, the region should have not been connected to list.
*/
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index 416d4e258df6..cb8380c9767f 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -927,20 +927,80 @@ static int dm_table_build_index(struct dm_table *t)
}
/*
+ * Get a disk whose integrity profile reflects the table's profile.
+ * If %match_all is true, all devices' profiles must match.
+ * If %match_all is false, all devices must at least have an
+ * allocated integrity profile; but uninitialized is ok.
+ * Returns NULL if integrity support was inconsistent or unavailable.
+ */
+static struct gendisk * dm_table_get_integrity_disk(struct dm_table *t,
+ bool match_all)
+{
+ struct list_head *devices = dm_table_get_devices(t);
+ struct dm_dev_internal *dd = NULL;
+ struct gendisk *prev_disk = NULL, *template_disk = NULL;
+
+ list_for_each_entry(dd, devices, list) {
+ template_disk = dd->dm_dev.bdev->bd_disk;
+ if (!blk_get_integrity(template_disk))
+ goto no_integrity;
+ if (!match_all && !blk_integrity_is_initialized(template_disk))
+ continue; /* skip uninitialized profiles */
+ else if (prev_disk &&
+ blk_integrity_compare(prev_disk, template_disk) < 0)
+ goto no_integrity;
+ prev_disk = template_disk;
+ }
+
+ return template_disk;
+
+no_integrity:
+ if (prev_disk)
+ DMWARN("%s: integrity not set: %s and %s profile mismatch",
+ dm_device_name(t->md),
+ prev_disk->disk_name,
+ template_disk->disk_name);
+ return NULL;
+}
+
+/*
* Register the mapped device for blk_integrity support if
- * the underlying devices support it.
+ * the underlying devices have an integrity profile. But all devices
+ * may not have matching profiles (checking all devices isn't reliable
+ * during table load because this table may use other DM device(s) which
+ * must be resumed before they will have an initialized integity profile).
+ * Stacked DM devices force a 2 stage integrity profile validation:
+ * 1 - during load, validate all initialized integrity profiles match
+ * 2 - during resume, validate all integrity profiles match
*/
static int dm_table_prealloc_integrity(struct dm_table *t, struct mapped_device *md)
{
- struct list_head *devices = dm_table_get_devices(t);
- struct dm_dev_internal *dd;
+ struct gendisk *template_disk = NULL;
- list_for_each_entry(dd, devices, list)
- if (bdev_get_integrity(dd->dm_dev.bdev)) {
- t->integrity_supported = 1;
- return blk_integrity_register(dm_disk(md), NULL);
- }
+ template_disk = dm_table_get_integrity_disk(t, false);
+ if (!template_disk)
+ return 0;
+ if (!blk_integrity_is_initialized(dm_disk(md))) {
+ t->integrity_supported = 1;
+ return blk_integrity_register(dm_disk(md), NULL);
+ }
+
+ /*
+ * If DM device already has an initalized integrity
+ * profile the new profile should not conflict.
+ */
+ if (blk_integrity_is_initialized(template_disk) &&
+ blk_integrity_compare(dm_disk(md), template_disk) < 0) {
+ DMWARN("%s: conflict with existing integrity profile: "
+ "%s profile mismatch",
+ dm_device_name(t->md),
+ template_disk->disk_name);
+ return 1;
+ }
+
+ /* Preserve existing initialized integrity profile */
+ t->integrity_supported = 1;
return 0;
}
@@ -1094,41 +1154,27 @@ combine_limits:
/*
* Set the integrity profile for this device if all devices used have
- * matching profiles.
+ * matching profiles. We're quite deep in the resume path but still
+ * don't know if all devices (particularly DM devices this device
+ * may be stacked on) have matching profiles. Even if the profiles
+ * don't match we have no way to fail (to resume) at this point.
*/
static void dm_table_set_integrity(struct dm_table *t)
{
- struct list_head *devices = dm_table_get_devices(t);
- struct dm_dev_internal *prev = NULL, *dd = NULL;
+ struct gendisk *template_disk = NULL;
if (!blk_get_integrity(dm_disk(t->md)))
return;
- list_for_each_entry(dd, devices, list) {
- if (prev &&
- blk_integrity_compare(prev->dm_dev.bdev->bd_disk,
- dd->dm_dev.bdev->bd_disk) < 0) {
- DMWARN("%s: integrity not set: %s and %s mismatch",
- dm_device_name(t->md),
- prev->dm_dev.bdev->bd_disk->disk_name,
- dd->dm_dev.bdev->bd_disk->disk_name);
- goto no_integrity;
- }
- prev = dd;
+ template_disk = dm_table_get_integrity_disk(t, true);
+ if (!template_disk &&
+ blk_integrity_is_initialized(dm_disk(t->md))) {
+ DMWARN("%s: device no longer has a valid integrity profile",
+ dm_device_name(t->md));
+ return;
}
-
- if (!prev || !bdev_get_integrity(prev->dm_dev.bdev))
- goto no_integrity;
-
blk_integrity_register(dm_disk(t->md),
- bdev_get_integrity(prev->dm_dev.bdev));
-
- return;
-
-no_integrity:
- blk_integrity_register(dm_disk(t->md), NULL);
-
- return;
+ blk_get_integrity(template_disk));
}
void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
diff --git a/drivers/md/faulty.c b/drivers/md/faulty.c
index 339fdc670751..23078dabb6df 100644
--- a/drivers/md/faulty.c
+++ b/drivers/md/faulty.c
@@ -30,7 +30,7 @@
*
* Different modes can be active at a time, but only
* one can be set at array creation. Others can be added later.
- * A mode can be one-shot or recurrent with the recurrance being
+ * A mode can be one-shot or recurrent with the recurrence being
* once in every N requests.
* The bottom 5 bits of the "layout" indicate the mode. The
* remainder indicate a period, or 0 for one-shot.
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 8b66e04c2ea6..7d6f7f18a920 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -447,48 +447,59 @@ EXPORT_SYMBOL(md_flush_request);
/* Support for plugging.
* This mirrors the plugging support in request_queue, but does not
- * require having a whole queue
+ * require having a whole queue or request structures.
+ * We allocate an md_plug_cb for each md device and each thread it gets
+ * plugged on. This links tot the private plug_handle structure in the
+ * personality data where we keep a count of the number of outstanding
+ * plugs so other code can see if a plug is active.
*/
-static void plugger_work(struct work_struct *work)
-{
- struct plug_handle *plug =
- container_of(work, struct plug_handle, unplug_work);
- plug->unplug_fn(plug);
-}
-static void plugger_timeout(unsigned long data)
-{
- struct plug_handle *plug = (void *)data;
- kblockd_schedule_work(NULL, &plug->unplug_work);
-}
-void plugger_init(struct plug_handle *plug,
- void (*unplug_fn)(struct plug_handle *))
-{
- plug->unplug_flag = 0;
- plug->unplug_fn = unplug_fn;
- init_timer(&plug->unplug_timer);
- plug->unplug_timer.function = plugger_timeout;
- plug->unplug_timer.data = (unsigned long)plug;
- INIT_WORK(&plug->unplug_work, plugger_work);
-}
-EXPORT_SYMBOL_GPL(plugger_init);
+struct md_plug_cb {
+ struct blk_plug_cb cb;
+ mddev_t *mddev;
+};
-void plugger_set_plug(struct plug_handle *plug)
+static void plugger_unplug(struct blk_plug_cb *cb)
{
- if (!test_and_set_bit(PLUGGED_FLAG, &plug->unplug_flag))
- mod_timer(&plug->unplug_timer, jiffies + msecs_to_jiffies(3)+1);
+ struct md_plug_cb *mdcb = container_of(cb, struct md_plug_cb, cb);
+ if (atomic_dec_and_test(&mdcb->mddev->plug_cnt))
+ md_wakeup_thread(mdcb->mddev->thread);
+ kfree(mdcb);
}
-EXPORT_SYMBOL_GPL(plugger_set_plug);
-int plugger_remove_plug(struct plug_handle *plug)
+/* Check that an unplug wakeup will come shortly.
+ * If not, wakeup the md thread immediately
+ */
+int mddev_check_plugged(mddev_t *mddev)
{
- if (test_and_clear_bit(PLUGGED_FLAG, &plug->unplug_flag)) {
- del_timer(&plug->unplug_timer);
- return 1;
- } else
+ struct blk_plug *plug = current->plug;
+ struct md_plug_cb *mdcb;
+
+ if (!plug)
return 0;
-}
-EXPORT_SYMBOL_GPL(plugger_remove_plug);
+ list_for_each_entry(mdcb, &plug->cb_list, cb.list) {
+ if (mdcb->cb.callback == plugger_unplug &&
+ mdcb->mddev == mddev) {
+ /* Already on the list, move to top */
+ if (mdcb != list_first_entry(&plug->cb_list,
+ struct md_plug_cb,
+ cb.list))
+ list_move(&mdcb->cb.list, &plug->cb_list);
+ return 1;
+ }
+ }
+ /* Not currently on the callback list */
+ mdcb = kmalloc(sizeof(*mdcb), GFP_ATOMIC);
+ if (!mdcb)
+ return 0;
+
+ mdcb->mddev = mddev;
+ mdcb->cb.callback = plugger_unplug;
+ atomic_inc(&mddev->plug_cnt);
+ list_add(&mdcb->cb.list, &plug->cb_list);
+ return 1;
+}
+EXPORT_SYMBOL_GPL(mddev_check_plugged);
static inline mddev_t *mddev_get(mddev_t *mddev)
{
@@ -538,6 +549,7 @@ void mddev_init(mddev_t *mddev)
atomic_set(&mddev->active, 1);
atomic_set(&mddev->openers, 0);
atomic_set(&mddev->active_io, 0);
+ atomic_set(&mddev->plug_cnt, 0);
spin_lock_init(&mddev->write_lock);
atomic_set(&mddev->flush_pending, 0);
init_waitqueue_head(&mddev->sb_wait);
@@ -3158,6 +3170,7 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
mddev->layout = mddev->new_layout;
mddev->chunk_sectors = mddev->new_chunk_sectors;
mddev->delta_disks = 0;
+ mddev->degraded = 0;
if (mddev->pers->sync_request == NULL) {
/* this is now an array without redundancy, so
* it must always be in_sync
@@ -4723,7 +4736,6 @@ static void md_clean(mddev_t *mddev)
mddev->bitmap_info.chunksize = 0;
mddev->bitmap_info.daemon_sleep = 0;
mddev->bitmap_info.max_write_behind = 0;
- mddev->plug = NULL;
}
static void __md_stop_writes(mddev_t *mddev)
@@ -6266,7 +6278,7 @@ static void status_resync(struct seq_file *seq, mddev_t * mddev)
* rt is a sector_t, so could be 32bit or 64bit.
* So we divide before multiply in case it is 32bit and close
* to the limit.
- * We scale the divisor (db) by 32 to avoid loosing precision
+ * We scale the divisor (db) by 32 to avoid losing precision
* near the end of resync when the number of remaining sectors
* is close to 'db'.
* We then divide rt by 32 after multiplying by db to compensate.
@@ -6688,12 +6700,6 @@ int md_allow_write(mddev_t *mddev)
}
EXPORT_SYMBOL_GPL(md_allow_write);
-void md_unplug(mddev_t *mddev)
-{
- if (mddev->plug)
- mddev->plug->unplug_fn(mddev->plug);
-}
-
#define SYNC_MARKS 10
#define SYNC_MARK_STEP (3*HZ)
void md_do_sync(mddev_t *mddev)
diff --git a/drivers/md/md.h b/drivers/md/md.h
index 12215d437fcc..0b1fd3f1d85b 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -29,26 +29,6 @@
typedef struct mddev_s mddev_t;
typedef struct mdk_rdev_s mdk_rdev_t;
-/* generic plugging support - like that provided with request_queue,
- * but does not require a request_queue
- */
-struct plug_handle {
- void (*unplug_fn)(struct plug_handle *);
- struct timer_list unplug_timer;
- struct work_struct unplug_work;
- unsigned long unplug_flag;
-};
-#define PLUGGED_FLAG 1
-void plugger_init(struct plug_handle *plug,
- void (*unplug_fn)(struct plug_handle *));
-void plugger_set_plug(struct plug_handle *plug);
-int plugger_remove_plug(struct plug_handle *plug);
-static inline void plugger_flush(struct plug_handle *plug)
-{
- del_timer_sync(&plug->unplug_timer);
- cancel_work_sync(&plug->unplug_work);
-}
-
/*
* MD's 'extended' device
*/
@@ -94,7 +74,7 @@ struct mdk_rdev_s
#define In_sync 2 /* device is in_sync with rest of array */
#define WriteMostly 4 /* Avoid reading if at all possible */
#define AutoDetected 7 /* added by auto-detect */
-#define Blocked 8 /* An error occured on an externally
+#define Blocked 8 /* An error occurred on an externally
* managed array, don't allow writes
* until it is cleared */
wait_queue_head_t blocked_wait;
@@ -199,6 +179,9 @@ struct mddev_s
int delta_disks, new_level, new_layout;
int new_chunk_sectors;
+ atomic_t plug_cnt; /* If device is expecting
+ * more bios soon.
+ */
struct mdk_thread_s *thread; /* management thread */
struct mdk_thread_s *sync_thread; /* doing resync or reconstruct */
sector_t curr_resync; /* last block scheduled */
@@ -336,7 +319,6 @@ struct mddev_s
struct list_head all_mddevs;
struct attribute_group *to_remove;
- struct plug_handle *plug; /* if used by personality */
struct bio_set *bio_set;
@@ -516,7 +498,6 @@ extern int md_integrity_register(mddev_t *mddev);
extern void md_integrity_add_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
extern int strict_strtoul_scaled(const char *cp, unsigned long *res, int scale);
extern void restore_bitmap_write_access(struct file *file);
-extern void md_unplug(mddev_t *mddev);
extern void mddev_init(mddev_t *mddev);
extern int md_run(mddev_t *mddev);
@@ -530,4 +511,5 @@ extern struct bio *bio_clone_mddev(struct bio *bio, gfp_t gfp_mask,
mddev_t *mddev);
extern struct bio *bio_alloc_mddev(gfp_t gfp_mask, int nr_iovecs,
mddev_t *mddev);
+extern int mddev_check_plugged(mddev_t *mddev);
#endif /* _MD_MD_H */
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index c2a21ae56d97..2b7a7ff401dc 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -565,12 +565,6 @@ static void flush_pending_writes(conf_t *conf)
spin_unlock_irq(&conf->device_lock);
}
-static void md_kick_device(mddev_t *mddev)
-{
- blk_flush_plug(current);
- md_wakeup_thread(mddev->thread);
-}
-
/* Barriers....
* Sometimes we need to suspend IO while we do something else,
* either some resync/recovery, or reconfigure the array.
@@ -600,7 +594,7 @@ static void raise_barrier(conf_t *conf)
/* Wait until no block IO is waiting */
wait_event_lock_irq(conf->wait_barrier, !conf->nr_waiting,
- conf->resync_lock, md_kick_device(conf->mddev));
+ conf->resync_lock, );
/* block any new IO from starting */
conf->barrier++;
@@ -608,7 +602,7 @@ static void raise_barrier(conf_t *conf)
/* Now wait for all pending IO to complete */
wait_event_lock_irq(conf->wait_barrier,
!conf->nr_pending && conf->barrier < RESYNC_DEPTH,
- conf->resync_lock, md_kick_device(conf->mddev));
+ conf->resync_lock, );
spin_unlock_irq(&conf->resync_lock);
}
@@ -630,7 +624,7 @@ static void wait_barrier(conf_t *conf)
conf->nr_waiting++;
wait_event_lock_irq(conf->wait_barrier, !conf->barrier,
conf->resync_lock,
- md_kick_device(conf->mddev));
+ );
conf->nr_waiting--;
}
conf->nr_pending++;
@@ -666,8 +660,7 @@ static void freeze_array(conf_t *conf)
wait_event_lock_irq(conf->wait_barrier,
conf->nr_pending == conf->nr_queued+1,
conf->resync_lock,
- ({ flush_pending_writes(conf);
- md_kick_device(conf->mddev); }));
+ flush_pending_writes(conf));
spin_unlock_irq(&conf->resync_lock);
}
static void unfreeze_array(conf_t *conf)
@@ -729,6 +722,7 @@ static int make_request(mddev_t *mddev, struct bio * bio)
const unsigned long do_sync = (bio->bi_rw & REQ_SYNC);
const unsigned long do_flush_fua = (bio->bi_rw & (REQ_FLUSH | REQ_FUA));
mdk_rdev_t *blocked_rdev;
+ int plugged;
/*
* Register the new request and wait if the reconstruction
@@ -820,6 +814,8 @@ static int make_request(mddev_t *mddev, struct bio * bio)
* inc refcount on their rdev. Record them by setting
* bios[x] to bio
*/
+ plugged = mddev_check_plugged(mddev);
+
disks = conf->raid_disks;
retry_write:
blocked_rdev = NULL;
@@ -925,7 +921,7 @@ static int make_request(mddev_t *mddev, struct bio * bio)
/* In case raid1d snuck in to freeze_array */
wake_up(&conf->wait_barrier);
- if (do_sync || !bitmap)
+ if (do_sync || !bitmap || !plugged)
md_wakeup_thread(mddev->thread);
return 0;
@@ -1516,13 +1512,16 @@ static void raid1d(mddev_t *mddev)
conf_t *conf = mddev->private;
struct list_head *head = &conf->retry_list;
mdk_rdev_t *rdev;
+ struct blk_plug plug;
md_check_recovery(mddev);
-
+
+ blk_start_plug(&plug);
for (;;) {
char b[BDEVNAME_SIZE];
- flush_pending_writes(conf);
+ if (atomic_read(&mddev->plug_cnt) == 0)
+ flush_pending_writes(conf);
spin_lock_irqsave(&conf->device_lock, flags);
if (list_empty(head)) {
@@ -1593,6 +1592,7 @@ static void raid1d(mddev_t *mddev)
}
cond_resched();
}
+ blk_finish_plug(&plug);
}
@@ -2039,7 +2039,6 @@ static int stop(mddev_t *mddev)
md_unregister_thread(mddev->thread);
mddev->thread = NULL;
- blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
if (conf->r1bio_pool)
mempool_destroy(conf->r1bio_pool);
kfree(conf->mirrors);
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index f7b62370b374..8e9462626ec5 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -5,7 +5,7 @@
*
* RAID-10 support for md.
*
- * Base on code in raid1.c. See raid1.c for futher copyright information.
+ * Base on code in raid1.c. See raid1.c for further copyright information.
*
*
* This program is free software; you can redistribute it and/or modify
@@ -340,14 +340,14 @@ static void raid10_end_write_request(struct bio *bio, int error)
/*
* RAID10 layout manager
- * Aswell as the chunksize and raid_disks count, there are two
+ * As well as the chunksize and raid_disks count, there are two
* parameters: near_copies and far_copies.
* near_copies * far_copies must be <= raid_disks.
* Normally one of these will be 1.
* If both are 1, we get raid0.
* If near_copies == raid_disks, we get raid1.
*
- * Chunks are layed out in raid0 style with near_copies copies of the
+ * Chunks are laid out in raid0 style with near_copies copies of the
* first chunk, followed by near_copies copies of the next chunk and
* so on.
* If far_copies > 1, then after 1/far_copies of the array has been assigned
@@ -634,12 +634,6 @@ static void flush_pending_writes(conf_t *conf)
spin_unlock_irq(&conf->device_lock);
}
-static void md_kick_device(mddev_t *mddev)
-{
- blk_flush_plug(current);
- md_wakeup_thread(mddev->thread);
-}
-
/* Barriers....
* Sometimes we need to suspend IO while we do something else,
* either some resync/recovery, or reconfigure the array.
@@ -669,15 +663,15 @@ static void raise_barrier(conf_t *conf, int force)
/* Wait until no block IO is waiting (unless 'force') */
wait_event_lock_irq(conf->wait_barrier, force || !conf->nr_waiting,
- conf->resync_lock, md_kick_device(conf->mddev));
+ conf->resync_lock, );
/* block any new IO from starting */
conf->barrier++;
- /* No wait for all pending IO to complete */
+ /* Now wait for all pending IO to complete */
wait_event_lock_irq(conf->wait_barrier,
!conf->nr_pending && conf->barrier < RESYNC_DEPTH,
- conf->resync_lock, md_kick_device(conf->mddev));
+ conf->resync_lock, );
spin_unlock_irq(&conf->resync_lock);
}
@@ -698,7 +692,7 @@ static void wait_barrier(conf_t *conf)
conf->nr_waiting++;
wait_event_lock_irq(conf->wait_barrier, !conf->barrier,
conf->resync_lock,
- md_kick_device(conf->mddev));
+ );
conf->nr_waiting--;
}
conf->nr_pending++;
@@ -734,8 +728,8 @@ static void freeze_array(conf_t *conf)
wait_event_lock_irq(conf->wait_barrier,
conf->nr_pending == conf->nr_queued+1,
conf->resync_lock,
- ({ flush_pending_writes(conf);
- md_kick_device(conf->mddev); }));
+ flush_pending_writes(conf));
+
spin_unlock_irq(&conf->resync_lock);
}
@@ -762,6 +756,7 @@ static int make_request(mddev_t *mddev, struct bio * bio)
const unsigned long do_fua = (bio->bi_rw & REQ_FUA);
unsigned long flags;
mdk_rdev_t *blocked_rdev;
+ int plugged;
if (unlikely(bio->bi_rw & REQ_FLUSH)) {
md_flush_request(mddev, bio);
@@ -870,6 +865,8 @@ static int make_request(mddev_t *mddev, struct bio * bio)
* inc refcount on their rdev. Record them by setting
* bios[x] to bio
*/
+ plugged = mddev_check_plugged(mddev);
+
raid10_find_phys(conf, r10_bio);
retry_write:
blocked_rdev = NULL;
@@ -946,9 +943,8 @@ static int make_request(mddev_t *mddev, struct bio * bio)
/* In case raid10d snuck in to freeze_array */
wake_up(&conf->wait_barrier);
- if (do_sync || !mddev->bitmap)
+ if (do_sync || !mddev->bitmap || !plugged)
md_wakeup_thread(mddev->thread);
-
return 0;
}
@@ -1640,9 +1636,11 @@ static void raid10d(mddev_t *mddev)
conf_t *conf = mddev->private;
struct list_head *head = &conf->retry_list;
mdk_rdev_t *rdev;
+ struct blk_plug plug;
md_check_recovery(mddev);
+ blk_start_plug(&plug);
for (;;) {
char b[BDEVNAME_SIZE];
@@ -1716,6 +1714,7 @@ static void raid10d(mddev_t *mddev)
}
cond_resched();
}
+ blk_finish_plug(&plug);
}
diff --git a/drivers/md/raid10.h b/drivers/md/raid10.h
index 2316ac2e8e21..944b1104d3b4 100644
--- a/drivers/md/raid10.h
+++ b/drivers/md/raid10.h
@@ -17,8 +17,8 @@ struct r10_private_data_s {
spinlock_t device_lock;
/* geometry */
- int near_copies; /* number of copies layed out raid0 style */
- int far_copies; /* number of copies layed out
+ int near_copies; /* number of copies laid out raid0 style */
+ int far_copies; /* number of copies laid out
* at large strides across drives
*/
int far_offset; /* far_copies are offset by 1 stripe
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index e867ee42b152..49bf5f891435 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -27,12 +27,12 @@
*
* We group bitmap updates into batches. Each batch has a number.
* We may write out several batches at once, but that isn't very important.
- * conf->bm_write is the number of the last batch successfully written.
- * conf->bm_flush is the number of the last batch that was closed to
+ * conf->seq_write is the number of the last batch successfully written.
+ * conf->seq_flush is the number of the last batch that was closed to
* new additions.
* When we discover that we will need to write to any block in a stripe
* (in add_stripe_bio) we update the in-memory bitmap and record in sh->bm_seq
- * the number of the batch it will be in. This is bm_flush+1.
+ * the number of the batch it will be in. This is seq_flush+1.
* When we are ready to do a write, if that batch hasn't been written yet,
* we plug the array and queue the stripe for later.
* When an unplug happens, we increment bm_flush, thus closing the current
@@ -199,14 +199,12 @@ static void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh)
BUG_ON(!list_empty(&sh->lru));
BUG_ON(atomic_read(&conf->active_stripes)==0);
if (test_bit(STRIPE_HANDLE, &sh->state)) {
- if (test_bit(STRIPE_DELAYED, &sh->state)) {
+ if (test_bit(STRIPE_DELAYED, &sh->state))
list_add_tail(&sh->lru, &conf->delayed_list);
- plugger_set_plug(&conf->plug);
- } else if (test_bit(STRIPE_BIT_DELAY, &sh->state) &&
- sh->bm_seq - conf->seq_write > 0) {
+ else if (test_bit(STRIPE_BIT_DELAY, &sh->state) &&
+ sh->bm_seq - conf->seq_write > 0)
list_add_tail(&sh->lru, &conf->bitmap_list);
- plugger_set_plug(&conf->plug);
- } else {
+ else {
clear_bit(STRIPE_BIT_DELAY, &sh->state);
list_add_tail(&sh->lru, &conf->handle_list);
}
@@ -461,7 +459,7 @@ get_active_stripe(raid5_conf_t *conf, sector_t sector,
< (conf->max_nr_stripes *3/4)
|| !conf->inactive_blocked),
conf->device_lock,
- md_raid5_kick_device(conf));
+ );
conf->inactive_blocked = 0;
} else
init_stripe(sh, sector, previous);
@@ -1470,7 +1468,7 @@ static int resize_stripes(raid5_conf_t *conf, int newsize)
wait_event_lock_irq(conf->wait_for_stripe,
!list_empty(&conf->inactive_list),
conf->device_lock,
- blk_flush_plug(current));
+ );
osh = get_free_stripe(conf);
spin_unlock_irq(&conf->device_lock);
atomic_set(&nsh->count, 1);
@@ -3623,8 +3621,7 @@ static void raid5_activate_delayed(raid5_conf_t *conf)
atomic_inc(&conf->preread_active_stripes);
list_add_tail(&sh->lru, &conf->hold_list);
}
- } else
- plugger_set_plug(&conf->plug);
+ }
}
static void activate_bit_delay(raid5_conf_t *conf)
@@ -3641,21 +3638,6 @@ static void activate_bit_delay(raid5_conf_t *conf)
}
}
-void md_raid5_kick_device(raid5_conf_t *conf)
-{
- blk_flush_plug(current);
- raid5_activate_delayed(conf);
- md_wakeup_thread(conf->mddev->thread);
-}
-EXPORT_SYMBOL_GPL(md_raid5_kick_device);
-
-static void raid5_unplug(struct plug_handle *plug)
-{
- raid5_conf_t *conf = container_of(plug, raid5_conf_t, plug);
-
- md_raid5_kick_device(conf);
-}
-
int md_raid5_congested(mddev_t *mddev, int bits)
{
raid5_conf_t *conf = mddev->private;
@@ -3945,6 +3927,7 @@ static int make_request(mddev_t *mddev, struct bio * bi)
struct stripe_head *sh;
const int rw = bio_data_dir(bi);
int remaining;
+ int plugged;
if (unlikely(bi->bi_rw & REQ_FLUSH)) {
md_flush_request(mddev, bi);
@@ -3963,6 +3946,7 @@ static int make_request(mddev_t *mddev, struct bio * bi)
bi->bi_next = NULL;
bi->bi_phys_segments = 1; /* over-loaded to count active stripes */
+ plugged = mddev_check_plugged(mddev);
for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
DEFINE_WAIT(w);
int disks, data_disks;
@@ -4057,7 +4041,7 @@ static int make_request(mddev_t *mddev, struct bio * bi)
* add failed due to overlap. Flush everything
* and wait a while
*/
- md_raid5_kick_device(conf);
+ md_wakeup_thread(mddev->thread);
release_stripe(sh);
schedule();
goto retry;
@@ -4077,6 +4061,9 @@ static int make_request(mddev_t *mddev, struct bio * bi)
}
}
+ if (!plugged)
+ md_wakeup_thread(mddev->thread);
+
spin_lock_irq(&conf->device_lock);
remaining = raid5_dec_bi_phys_segments(bi);
spin_unlock_irq(&conf->device_lock);
@@ -4478,24 +4465,30 @@ static void raid5d(mddev_t *mddev)
struct stripe_head *sh;
raid5_conf_t *conf = mddev->private;
int handled;
+ struct blk_plug plug;
pr_debug("+++ raid5d active\n");
md_check_recovery(mddev);
+ blk_start_plug(&plug);
handled = 0;
spin_lock_irq(&conf->device_lock);
while (1) {
struct bio *bio;
- if (conf->seq_flush != conf->seq_write) {
- int seq = conf->seq_flush;
+ if (atomic_read(&mddev->plug_cnt) == 0 &&
+ !list_empty(&conf->bitmap_list)) {
+ /* Now is a good time to flush some bitmap updates */
+ conf->seq_flush++;
spin_unlock_irq(&conf->device_lock);
bitmap_unplug(mddev->bitmap);
spin_lock_irq(&conf->device_lock);
- conf->seq_write = seq;
+ conf->seq_write = conf->seq_flush;
activate_bit_delay(conf);
}
+ if (atomic_read(&mddev->plug_cnt) == 0)
+ raid5_activate_delayed(conf);
while ((bio = remove_bio_from_retry(conf))) {
int ok;
@@ -4525,6 +4518,7 @@ static void raid5d(mddev_t *mddev)
spin_unlock_irq(&conf->device_lock);
async_tx_issue_pending_all();
+ blk_finish_plug(&plug);
pr_debug("--- raid5d inactive\n");
}
@@ -5141,8 +5135,6 @@ static int run(mddev_t *mddev)
mdname(mddev));
md_set_array_sectors(mddev, raid5_size(mddev, 0, 0));
- plugger_init(&conf->plug, raid5_unplug);
- mddev->plug = &conf->plug;
if (mddev->queue) {
int chunk_size;
/* read-ahead size must cover two whole stripes, which
@@ -5159,7 +5151,6 @@ static int run(mddev_t *mddev)
mddev->queue->backing_dev_info.congested_data = mddev;
mddev->queue->backing_dev_info.congested_fn = raid5_congested;
- mddev->queue->queue_lock = &conf->device_lock;
chunk_size = mddev->chunk_sectors << 9;
blk_queue_io_min(mddev->queue, chunk_size);
@@ -5192,7 +5183,6 @@ static int stop(mddev_t *mddev)
mddev->thread = NULL;
if (mddev->queue)
mddev->queue->backing_dev_info.congested_fn = NULL;
- plugger_flush(&conf->plug); /* the unplug fn references 'conf'*/
free_conf(conf);
mddev->private = NULL;
mddev->to_remove = &raid5_attrs_group;
@@ -5688,6 +5678,7 @@ static void raid5_quiesce(mddev_t *mddev, int state)
static void *raid45_takeover_raid0(mddev_t *mddev, int level)
{
struct raid0_private_data *raid0_priv = mddev->private;
+ sector_t sectors;
/* for raid0 takeover only one zone is supported */
if (raid0_priv->nr_strip_zones > 1) {
@@ -5696,6 +5687,9 @@ static void *raid45_takeover_raid0(mddev_t *mddev, int level)
return ERR_PTR(-EINVAL);
}
+ sectors = raid0_priv->strip_zone[0].zone_end;
+ sector_div(sectors, raid0_priv->strip_zone[0].nb_dev);
+ mddev->dev_sectors = sectors;
mddev->new_level = level;
mddev->new_layout = ALGORITHM_PARITY_N;
mddev->new_chunk_sectors = mddev->chunk_sectors;
diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
index 8d563a4f022a..3ca77a2613ba 100644
--- a/drivers/md/raid5.h
+++ b/drivers/md/raid5.h
@@ -400,8 +400,6 @@ struct raid5_private_data {
* Cleared when a sync completes.
*/
- struct plug_handle plug;
-
/* per cpu variables */
struct raid5_percpu {
struct page *spare_page; /* Used when checking P/Q in raid6 */
generated by cgit v1.2.3 (git 2.39.0) at 2024-11-19 20:29:25 +0300