Merge commit 'md/for-linus' into async-tx-next

Conflicts:
	drivers/md/raid5.c

18 files changed, 297 insertions, 183 deletions

diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index 9933eb861c7..ed103816401 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)
 		 * But don't wait if split was due to the io size restriction
 		 */
 		if (unlikely(out_of_pages))
-			congestion_wait(WRITE, HZ/100);
+			congestion_wait(BLK_RW_ASYNC, HZ/100);
 
 		/*
 		 * With async crypto it is unsafe to share the crypto context
@@ -1318,7 +1318,7 @@ static int crypt_iterate_devices(struct dm_target *ti,
 {
 	struct crypt_config *cc = ti->private;
 
-	return fn(ti, cc->dev, cc->start, data);
+	return fn(ti, cc->dev, cc->start, ti->len, data);
 }
 
 static struct target_type crypt_target = {
diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c
index 4e5b843cd4d..ebe7381f47c 100644
--- a/drivers/md/dm-delay.c
+++ b/drivers/md/dm-delay.c
@@ -324,12 +324,12 @@ static int delay_iterate_devices(struct dm_target *ti,
 	struct delay_c *dc = ti->private;
 	int ret = 0;
 
-	ret = fn(ti, dc->dev_read, dc->start_read, data);
+	ret = fn(ti, dc->dev_read, dc->start_read, ti->len, data);
 	if (ret)
 		goto out;
 
 	if (dc->dev_write)
-		ret = fn(ti, dc->dev_write, dc->start_write, data);
+		ret = fn(ti, dc->dev_write, dc->start_write, ti->len, data);
 
 out:
 	return ret;
diff --git a/drivers/md/dm-exception-store.c b/drivers/md/dm-exception-store.c
index c3ae51584b1..3710ff88fc1 100644
--- a/drivers/md/dm-exception-store.c
+++ b/drivers/md/dm-exception-store.c
@@ -195,7 +195,7 @@ int dm_exception_store_create(struct dm_target *ti, int argc, char **argv,
 			      struct dm_exception_store **store)
 {
 	int r = 0;
-	struct dm_exception_store_type *type;
+	struct dm_exception_store_type *type = NULL;
 	struct dm_exception_store *tmp_store;
 	char persistent;
 
@@ -211,12 +211,15 @@ int dm_exception_store_create(struct dm_target *ti, int argc, char **argv,
 	}
 
 	persistent = toupper(*argv[1]);
-	if (persistent != 'P' && persistent != 'N') {
+	if (persistent == 'P')
+		type = get_type("P");
+	else if (persistent == 'N')
+		type = get_type("N");
+	else {
 		ti->error = "Persistent flag is not P or N";
 		return -EINVAL;
 	}
 
-	type = get_type(&persistent);
 	if (!type) {
 		ti->error = "Exception store type not recognised";
 		r = -EINVAL;
diff --git a/drivers/md/dm-linear.c b/drivers/md/dm-linear.c
index 9184b6deb86..82f7d6e6b1e 100644
--- a/drivers/md/dm-linear.c
+++ b/drivers/md/dm-linear.c
@@ -139,7 +139,7 @@ static int linear_iterate_devices(struct dm_target *ti,
 {
 	struct linear_c *lc = ti->private;
 
-	return fn(ti, lc->dev, lc->start, data);
+	return fn(ti, lc->dev, lc->start, ti->len, data);
 }
 
 static struct target_type linear_target = {
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index c70604a2089..6f0d90d4a54 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -1453,7 +1453,7 @@ static int multipath_iterate_devices(struct dm_target *ti,
 
 	list_for_each_entry(pg, &m->priority_groups, list) {
 		list_for_each_entry(p, &pg->pgpaths, list) {
-			ret = fn(ti, p->path.dev, ti->begin, data);
+			ret = fn(ti, p->path.dev, ti->begin, ti->len, data);
 			if (ret)
 				goto out;
 		}
diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c
index ce8868c768c..9726577cde4 100644
--- a/drivers/md/dm-raid1.c
+++ b/drivers/md/dm-raid1.c
@@ -638,6 +638,7 @@ static void do_writes(struct mirror_set *ms, struct bio_list *writes)
 		spin_lock_irq(&ms->lock);
 		bio_list_merge(&ms->writes, &requeue);
 		spin_unlock_irq(&ms->lock);
+		delayed_wake(ms);
 	}
 
 	/*
@@ -1292,7 +1293,7 @@ static int mirror_iterate_devices(struct dm_target *ti,
 
 	for (i = 0; !ret && i < ms->nr_mirrors; i++)
 		ret = fn(ti, ms->mirror[i].dev,
-			 ms->mirror[i].offset, data);
+			 ms->mirror[i].offset, ti->len, data);
 
 	return ret;
 }
diff --git a/drivers/md/dm-stripe.c b/drivers/md/dm-stripe.c
index b240e85ae39..4e0e5937e42 100644
--- a/drivers/md/dm-stripe.c
+++ b/drivers/md/dm-stripe.c
@@ -320,10 +320,11 @@ static int stripe_iterate_devices(struct dm_target *ti,
 	int ret = 0;
 	unsigned i = 0;
 
-	do
+	do {
 		ret = fn(ti, sc->stripe[i].dev,
-			 sc->stripe[i].physical_start, data);
-	while (!ret && ++i < sc->stripes);
+			 sc->stripe[i].physical_start,
+			 sc->stripe_width, data);
+	} while (!ret && ++i < sc->stripes);
 
 	return ret;
 }
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index 4899ebe767c..d952b344191 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -346,7 +346,7 @@ static void close_dev(struct dm_dev_internal *d, struct mapped_device *md)
  * If possible, this checks an area of a destination device is valid.
  */
 static int device_area_is_valid(struct dm_target *ti, struct dm_dev *dev,
-				sector_t start, void *data)
+				sector_t start, sector_t len, void *data)
 {
 	struct queue_limits *limits = data;
 	struct block_device *bdev = dev->bdev;
@@ -359,7 +359,7 @@ static int device_area_is_valid(struct dm_target *ti, struct dm_dev *dev,
 	if (!dev_size)
 		return 1;
 
-	if ((start >= dev_size) || (start + ti->len > dev_size)) {
+	if ((start >= dev_size) || (start + len > dev_size)) {
 		DMWARN("%s: %s too small for target",
 		       dm_device_name(ti->table->md), bdevname(bdev, b));
 		return 0;
@@ -377,11 +377,11 @@ static int device_area_is_valid(struct dm_target *ti, struct dm_dev *dev,
 		return 0;
 	}
 
-	if (ti->len & (logical_block_size_sectors - 1)) {
+	if (len & (logical_block_size_sectors - 1)) {
 		DMWARN("%s: len=%llu not aligned to h/w "
 		       "logical block size %hu of %s",
 		       dm_device_name(ti->table->md),
-		       (unsigned long long)ti->len,
+		       (unsigned long long)len,
 		       limits->logical_block_size, bdevname(bdev, b));
 		return 0;
 	}
@@ -482,7 +482,7 @@ static int __table_get_device(struct dm_table *t, struct dm_target *ti,
 #define min_not_zero(l, r) (l == 0) ? r : ((r == 0) ? l : min(l, r))
 
 int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev,
-			 sector_t start, void *data)
+			 sector_t start, sector_t len, void *data)
 {
 	struct queue_limits *limits = data;
 	struct block_device *bdev = dev->bdev;
@@ -495,7 +495,7 @@ int dm_set_device_limits(struct dm_target *ti, struct dm_dev *dev,
 		return 0;
 	}
 
-	if (blk_stack_limits(limits, &q->limits, start) < 0)
+	if (blk_stack_limits(limits, &q->limits, start << 9) < 0)
 		DMWARN("%s: target device %s is misaligned",
 		       dm_device_name(ti->table->md), bdevname(bdev, b));
 
@@ -830,11 +830,6 @@ unsigned dm_table_get_type(struct dm_table *t)
 	return t->type;
 }
 
-bool dm_table_bio_based(struct dm_table *t)
-{
-	return dm_table_get_type(t) == DM_TYPE_BIO_BASED;
-}
-
 bool dm_table_request_based(struct dm_table *t)
 {
 	return dm_table_get_type(t) == DM_TYPE_REQUEST_BASED;
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 3c6d4ee8921..8a311ea0d44 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -1017,7 +1017,7 @@ static struct bio *split_bvec(struct bio *bio, sector_t sector,
 	clone->bi_flags |= 1 << BIO_CLONED;
 
 	if (bio_integrity(bio)) {
-		bio_integrity_clone(clone, bio, GFP_NOIO);
+		bio_integrity_clone(clone, bio, GFP_NOIO, bs);
 		bio_integrity_trim(clone,
 				   bio_sector_offset(bio, idx, offset), len);
 	}
@@ -1045,7 +1045,7 @@ static struct bio *clone_bio(struct bio *bio, sector_t sector,
 	clone->bi_flags &= ~(1 << BIO_SEG_VALID);
 
 	if (bio_integrity(bio)) {
-		bio_integrity_clone(clone, bio, GFP_NOIO);
+		bio_integrity_clone(clone, bio, GFP_NOIO, bs);
 
 		if (idx != bio->bi_idx || clone->bi_size < bio->bi_size)
 			bio_integrity_trim(clone,
@@ -2203,16 +2203,6 @@ int dm_swap_table(struct mapped_device *md, struct dm_table *table)
 		goto out;
 	}
 
-	/*
-	 * It is enought that blk_queue_ordered() is called only once when
-	 * the first bio-based table is bound.
-	 *
-	 * This setting should be moved to alloc_dev() when request-based dm
-	 * supports barrier.
-	 */
-	if (!md->map && dm_table_bio_based(table))
-		blk_queue_ordered(md->queue, QUEUE_ORDERED_DRAIN, NULL);
-
 	__unbind(md);
 	r = __bind(md, table, &limits);
 
diff --git a/drivers/md/dm.h b/drivers/md/dm.h
index 23278ae80f0..a7663eba17e 100644
--- a/drivers/md/dm.h
+++ b/drivers/md/dm.h
@@ -61,7 +61,6 @@ int dm_table_any_congested(struct dm_table *t, int bdi_bits);
 int dm_table_any_busy_target(struct dm_table *t);
 int dm_table_set_type(struct dm_table *t);
 unsigned dm_table_get_type(struct dm_table *t);
-bool dm_table_bio_based(struct dm_table *t);
 bool dm_table_request_based(struct dm_table *t);
 int dm_table_alloc_md_mempools(struct dm_table *t);
 void dm_table_free_md_mempools(struct dm_table *t);
diff --git a/drivers/md/linear.c b/drivers/md/linear.c
index 15c8b7b25a9..5fe39c2a3d2 100644
--- a/drivers/md/linear.c
+++ b/drivers/md/linear.c
@@ -166,8 +166,8 @@ static linear_conf_t *linear_conf(mddev_t *mddev, int raid_disks)
 			rdev->sectors = sectors * mddev->chunk_sectors;
 		}
 
-		blk_queue_stack_limits(mddev->queue,
-				       rdev->bdev->bd_disk->queue);
+		disk_stack_limits(mddev->gendisk, rdev->bdev,
+				  rdev->data_offset << 9);
 		/* as we don't honour merge_bvec_fn, we must never risk
 		 * violating it, so limit ->max_sector to one PAGE, as
 		 * a one page request is never in violation.
@@ -220,6 +220,7 @@ static int linear_run (mddev_t *mddev)
 	mddev->queue->unplug_fn = linear_unplug;
 	mddev->queue->backing_dev_info.congested_fn = linear_congested;
 	mddev->queue->backing_dev_info.congested_data = mddev;
+	md_integrity_register(mddev);
 	return 0;
 }
 
@@ -256,6 +257,7 @@ static int linear_add(mddev_t *mddev, mdk_rdev_t *rdev)
 	rcu_assign_pointer(mddev->private, newconf);
 	md_set_array_sectors(mddev, linear_size(mddev, 0, 0));
 	set_capacity(mddev->gendisk, mddev->array_sectors);
+	revalidate_disk(mddev->gendisk);
 	call_rcu(&oldconf->rcu, free_conf);
 	return 0;
 }
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 09be637d52c..9dd872000ce 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -359,6 +359,7 @@ static mddev_t * mddev_find(dev_t unit)
 	else
 		new->md_minor = MINOR(unit) >> MdpMinorShift;
 
+	mutex_init(&new->open_mutex);
 	mutex_init(&new->reconfig_mutex);
 	INIT_LIST_HEAD(&new->disks);
 	INIT_LIST_HEAD(&new->all_mddevs);
@@ -1308,7 +1309,12 @@ static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 	}
 	if (mddev->level != LEVEL_MULTIPATH) {
 		int role;
-		role = le16_to_cpu(sb->dev_roles[rdev->desc_nr]);
+		if (rdev->desc_nr < 0 ||
+		    rdev->desc_nr >= le32_to_cpu(sb->max_dev)) {
+			role = 0xffff;
+			rdev->desc_nr = -1;
+		} else
+			role = le16_to_cpu(sb->dev_roles[rdev->desc_nr]);
 		switch(role) {
 		case 0xffff: /* spare */
 			break;
@@ -1394,8 +1400,14 @@ static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		if (rdev2->desc_nr+1 > max_dev)
 			max_dev = rdev2->desc_nr+1;
 
-	if (max_dev > le32_to_cpu(sb->max_dev))
+	if (max_dev > le32_to_cpu(sb->max_dev)) {
+		int bmask;
 		sb->max_dev = cpu_to_le32(max_dev);
+		rdev->sb_size = max_dev * 2 + 256;
+		bmask = queue_logical_block_size(rdev->bdev->bd_disk->queue)-1;
+		if (rdev->sb_size & bmask)
+			rdev->sb_size = (rdev->sb_size | bmask) + 1;
+	}
 	for (i=0; i<max_dev;i++)
 		sb->dev_roles[i] = cpu_to_le16(0xfffe);
 	
@@ -1487,37 +1499,76 @@ static int match_mddev_units(mddev_t *mddev1, mddev_t *mddev2)
 
 static LIST_HEAD(pending_raid_disks);
 
-static void md_integrity_check(mdk_rdev_t *rdev, mddev_t *mddev)
+/*
+ * Try to register data integrity profile for an mddev
+ *
+ * This is called when an array is started and after a disk has been kicked
+ * from the array. It only succeeds if all working and active component devices
+ * are integrity capable with matching profiles.
+ */
+int md_integrity_register(mddev_t *mddev)
+{
+	mdk_rdev_t *rdev, *reference = NULL;
+
+	if (list_empty(&mddev->disks))
+		return 0; /* nothing to do */
+	if (blk_get_integrity(mddev->gendisk))
+		return 0; /* already registered */
+	list_for_each_entry(rdev, &mddev->disks, same_set) {
+		/* skip spares and non-functional disks */
+		if (test_bit(Faulty, &rdev->flags))
+			continue;
+		if (rdev->raid_disk < 0)
+			continue;
+		/*
+		 * If at least one rdev is not integrity capable, we can not
+		 * enable data integrity for the md device.
+		 */
+		if (!bdev_get_integrity(rdev->bdev))
+			return -EINVAL;
+		if (!reference) {
+			/* Use the first rdev as the reference */
+			reference = rdev;
+			continue;
+		}
+		/* does this rdev's profile match the reference profile? */
+		if (blk_integrity_compare(reference->bdev->bd_disk,
+				rdev->bdev->bd_disk) < 0)
+			return -EINVAL;
+	}
+	/*
+	 * All component devices are integrity capable and have matching
+	 * profiles, register the common profile for the md device.
+	 */
+	if (blk_integrity_register(mddev->gendisk,
+			bdev_get_integrity(reference->bdev)) != 0) {
+		printk(KERN_ERR "md: failed to register integrity for %s\n",
+			mdname(mddev));
+		return -EINVAL;
+	}
+	printk(KERN_NOTICE "md: data integrity on %s enabled\n",
+		mdname(mddev));
+	return 0;
+}
+EXPORT_SYMBOL(md_integrity_register);
+
+/* Disable data integrity if non-capable/non-matching disk is being added */
+void md_integrity_add_rdev(mdk_rdev_t *rdev, mddev_t *mddev)
 {
-	struct mdk_personality *pers = mddev->pers;
-	struct gendisk *disk = mddev->gendisk;
 	struct blk_integrity *bi_rdev = bdev_get_integrity(rdev->bdev);
-	struct blk_integrity *bi_mddev = blk_get_integrity(disk);
+	struct blk_integrity *bi_mddev = blk_get_integrity(mddev->gendisk);
 
-	/* Data integrity passthrough not supported on RAID 4, 5 and 6 */
-	if (pers && pers->level >= 4 && pers->level <= 6)
+	if (!bi_mddev) /* nothing to do */
 		return;
-
-	/* If rdev is integrity capable, register profile for mddev */
-	if (!bi_mddev && bi_rdev) {
-		if (blk_integrity_register(disk, bi_rdev))
-			printk(KERN_ERR "%s: %s Could not register integrity!\n",
-			       __func__, disk->disk_name);
-		else
-			printk(KERN_NOTICE "Enabling data integrity on %s\n",
-			       disk->disk_name);
+	if (rdev->raid_disk < 0) /* skip spares */
 		return;
-	}
-
-	/* Check that mddev and rdev have matching profiles */
-	if (blk_integrity_compare(disk, rdev->bdev->bd_disk) < 0) {
-		printk(KERN_ERR "%s: %s/%s integrity mismatch!\n", __func__,
-		       disk->disk_name, rdev->bdev->bd_disk->disk_name);
-		printk(KERN_NOTICE "Disabling data integrity on %s\n",
-		       disk->disk_name);
-		blk_integrity_unregister(disk);
-	}
+	if (bi_rdev && blk_integrity_compare(mddev->gendisk,
+					     rdev->bdev->bd_disk) >= 0)
+		return;
+	printk(KERN_NOTICE "disabling data integrity on %s\n", mdname(mddev));
+	blk_integrity_unregister(mddev->gendisk);
 }
+EXPORT_SYMBOL(md_integrity_add_rdev);
 
 static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 {
@@ -1591,7 +1642,6 @@ static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 	/* May as well allow recovery to be retried once */
 	mddev->recovery_disabled = 0;
 
-	md_integrity_check(rdev, mddev);
 	return 0;
 
  fail:
@@ -1756,9 +1806,10 @@ static void print_sb_1(struct mdp_superblock_1 *sb)
 	__u8 *uuid;
 
 	uuid = sb->set_uuid;
-	printk(KERN_INFO "md:  SB: (V:%u) (F:0x%08x) Array-ID:<%02x%02x%02x%02x"
-			":%02x%02x:%02x%02x:%02x%02x:%02x%02x%02x%02x%02x%02x>\n"
-	       KERN_INFO "md:    Name: \"%s\" CT:%llu\n",
+	printk(KERN_INFO
+	       "md:  SB: (V:%u) (F:0x%08x) Array-ID:<%02x%02x%02x%02x"
+	       ":%02x%02x:%02x%02x:%02x%02x:%02x%02x%02x%02x%02x%02x>\n"
+	       "md:    Name: \"%s\" CT:%llu\n",
 		le32_to_cpu(sb->major_version),
 		le32_to_cpu(sb->feature_map),
 		uuid[0], uuid[1], uuid[2], uuid[3],
@@ -1770,12 +1821,13 @@ static void print_sb_1(struct mdp_superblock_1 *sb)
 		       & MD_SUPERBLOCK_1_TIME_SEC_MASK);
 
 	uuid = sb->device_uuid;
-	printk(KERN_INFO "md:       L%u SZ%llu RD:%u LO:%u CS:%u DO:%llu DS:%llu SO:%llu"
+	printk(KERN_INFO
+	       "md:       L%u SZ%llu RD:%u LO:%u CS:%u DO:%llu DS:%llu SO:%llu"
 			" RO:%llu\n"
-	       KERN_INFO "md:     Dev:%08x UUID: %02x%02x%02x%02x:%02x%02x:%02x%02x:%02x%02x"
-			":%02x%02x%02x%02x%02x%02x\n"
-	       KERN_INFO "md:       (F:0x%08x) UT:%llu Events:%llu ResyncOffset:%llu CSUM:0x%08x\n"
-	       KERN_INFO "md:         (MaxDev:%u) \n",
+	       "md:     Dev:%08x UUID: %02x%02x%02x%02x:%02x%02x:%02x%02x:%02x%02x"
+	                ":%02x%02x%02x%02x%02x%02x\n"
+	       "md:       (F:0x%08x) UT:%llu Events:%llu ResyncOffset:%llu CSUM:0x%08x\n"
+	       "md:         (MaxDev:%u) \n",
 		le32_to_cpu(sb->level),
 		(unsigned long long)le64_to_cpu(sb->size),
 		le32_to_cpu(sb->raid_disks),
@@ -1923,17 +1975,14 @@ repeat:
 		/* otherwise we have to go forward and ... */
 		mddev->events ++;
 		if (!mddev->in_sync || mddev->recovery_cp != MaxSector) { /* not clean */
-			/* .. if the array isn't clean, insist on an odd 'events' */
-			if ((mddev->events&1)==0) {
-				mddev->events++;
+			/* .. if the array isn't clean, an 'even' event must also go
+			 * to spares. */
+			if ((mddev->events&1)==0)
 				nospares = 0;
-			}
 		} else {
-			/* otherwise insist on an even 'events' (for clean states) */
-			if ((mddev->events&1)) {
-				mddev->events++;
+			/* otherwise an 'odd' event must go to spares */
+			if ((mddev->events&1))
 				nospares = 0;
-			}
 		}
 	}
 
@@ -2655,6 +2704,7 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
 	ssize_t rv = len;
 	struct mdk_personality *pers;
 	void *priv;
+	mdk_rdev_t *rdev;
 
 	if (mddev->pers == NULL) {
 		if (len == 0)
@@ -2734,6 +2784,12 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
 	mddev_suspend(mddev);
 	mddev->pers->stop(mddev);
 	module_put(mddev->pers->owner);
+	/* Invalidate devices that are now superfluous */
+	list_for_each_entry(rdev, &mddev->disks, same_set)
+		if (rdev->raid_disk >= mddev->raid_disks) {
+			rdev->raid_disk = -1;
+			clear_bit(In_sync, &rdev->flags);
+		}
 	mddev->pers = pers;
 	mddev->private = priv;
 	strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
@@ -3543,6 +3599,7 @@ max_sync_store(mddev_t *mddev, const char *buf, size_t len)
 		if (max < mddev->resync_min)
 			return -EINVAL;
 		if (max < mddev->resync_max &&
+		    mddev->ro == 0 &&
 		    test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
 			return -EBUSY;
 
@@ -3573,7 +3630,8 @@ suspend_lo_store(mddev_t *mddev, const char *buf, size_t len)
 	char *e;
 	unsigned long long new = simple_strtoull(buf, &e, 10);
 
-	if (mddev->pers->quiesce == NULL)
+	if (mddev->pers == NULL || 
+	    mddev->pers->quiesce == NULL)
 		return -EINVAL;
 	if (buf == e || (*e && *e != '\n'))
 		return -EINVAL;
@@ -3601,7 +3659,8 @@ suspend_hi_store(mddev_t *mddev, const char *buf, size_t len)
 	char *e;
 	unsigned long long new = simple_strtoull(buf, &e, 10);
 
-	if (mddev->pers->quiesce == NULL)
+	if (mddev->pers == NULL ||
+	    mddev->pers->quiesce == NULL)
 		return -EINVAL;
 	if (buf == e || (*e && *e != '\n'))
 		return -EINVAL;
@@ -3681,17 +3740,8 @@ array_size_store(mddev_t *mddev, const char *buf, size_t len)
 
 	mddev->array_sectors = sectors;
 	set_capacity(mddev->gendisk, mddev->array_sectors);
-	if (mddev->pers) {
-		struct block_device *bdev = bdget_disk(mddev->gendisk, 0);
-
-		if (bdev) {
-			mutex_lock(&bdev->bd_inode->i_mutex);
-			i_size_write(bdev->bd_inode,
-				     (loff_t)mddev->array_sectors << 9);
-			mutex_unlock(&bdev->bd_inode->i_mutex);
-			bdput(bdev);
-		}
-	}
+	if (mddev->pers)
+		revalidate_disk(mddev->gendisk);
 
 	return len;
 }
@@ -3844,11 +3894,9 @@ static int md_alloc(dev_t dev, char *name)
 	flush_scheduled_work();
 
 	mutex_lock(&disks_mutex);
-	if (mddev->gendisk) {
-		mutex_unlock(&disks_mutex);
-		mddev_put(mddev);
-		return -EEXIST;
-	}
+	error = -EEXIST;
+	if (mddev->gendisk)
+		goto abort;
 
 	if (name) {
 		/* Need to ensure that 'name' is not a duplicate.
@@ -3860,17 +3908,15 @@ static int md_alloc(dev_t dev, char *name)
 			if (mddev2->gendisk &&
 			    strcmp(mddev2->gendisk->disk_name, name) == 0) {
 				spin_unlock(&all_mddevs_lock);
-				return -EEXIST;
+				goto abort;
 			}
 		spin_unlock(&all_mddevs_lock);
 	}
 
+	error = -ENOMEM;
 	mddev->queue = blk_alloc_queue(GFP_KERNEL);
-	if (!mddev->queue) {
-		mutex_unlock(&disks_mutex);
-		mddev_put(mddev);
-		return -ENOMEM;
-	}
+	if (!mddev->queue)
+		goto abort;
 	mddev->queue->queuedata = mddev;
 
 	/* Can be unlocked because the queue is new: no concurrency */
@@ -3880,11 +3926,9 @@ static int md_alloc(dev_t dev, char *name)
 
 	disk = alloc_disk(1 << shift);
 	if (!disk) {
-		mutex_unlock(&disks_mutex);
 		blk_cleanup_queue(mddev->queue);
 		mddev->queue = NULL;
-		mddev_put(mddev);
-		return -ENOMEM;
+		goto abort;
 	}
 	disk->major = MAJOR(mddev->unit);
 	disk->first_minor = unit << shift;
@@ -3906,16 +3950,22 @@ static int md_alloc(dev_t dev, char *name)
 	mddev->gendisk = disk;
 	error = kobject_init_and_add(&mddev->kobj, &md_ktype,
 				     &disk_to_dev(disk)->kobj, "%s", "md");
-	mutex_unlock(&disks_mutex);
-	if (error)
+	if (error) {
+		/* This isn't possible, but as kobject_init_and_add is marked
+		 * __must_check, we must do something with the result
+		 */
 		printk(KERN_WARNING "md: cannot register %s/md - name in use\n",
 		       disk->disk_name);
-	else {
+		error = 0;
+	}
+ abort:
+	mutex_unlock(&disks_mutex);
+	if (!error) {
 		kobject_uevent(&mddev->kobj, KOBJ_ADD);
 		mddev->sysfs_state = sysfs_get_dirent(mddev->kobj.sd, "array_state");
 	}
 	mddev_put(mddev);
-	return 0;
+	return error;
 }
 
 static struct kobject *md_probe(dev_t dev, int *part, void *data)
@@ -4044,10 +4094,6 @@ static int do_md_run(mddev_t * mddev)
 	}
 	strlcpy(mddev->clevel, pers->name, sizeof(mddev->clevel));
 
-	if (pers->level >= 4 && pers->level <= 6)
-		/* Cannot support integrity (yet) */
-		blk_integrity_unregister(mddev->gendisk);
-
 	if (mddev->reshape_position != MaxSector &&
 	    pers->start_reshape == NULL) {
 		/* This personality cannot handle reshaping... */
@@ -4185,6 +4231,7 @@ static int do_md_run(mddev_t * mddev)
 	md_wakeup_thread(mddev->thread);
 	md_wakeup_thread(mddev->sync_thread); /* possibly kick off a reshape */
 
+	revalidate_disk(mddev->gendisk);
 	mddev->changed = 1;
 	md_new_event(mddev);
 	sysfs_notify_dirent(mddev->sysfs_state);
@@ -4256,12 +4303,11 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 	struct gendisk *disk = mddev->gendisk;
 	mdk_rdev_t *rdev;
 
+	mutex_lock(&mddev->open_mutex);
 	if (atomic_read(&mddev->openers) > is_open) {
 		printk("md: %s still in use.\n",mdname(mddev));
-		return -EBUSY;
-	}
-
-	if (mddev->pers) {
+		err = -EBUSY;
+	} else if (mddev->pers) {
 
 		if (mddev->sync_thread) {
 			set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
@@ -4318,8 +4364,12 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 		if (mode == 1)
 			set_disk_ro(disk, 1);
 		clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
+		err = 0;
 	}
-
+out:
+	mutex_unlock(&mddev->open_mutex);
+	if (err)
+		return err;
 	/*
 	 * Free resources if final stop
 	 */
@@ -4385,7 +4435,6 @@ static int do_md_stop(mddev_t * mddev, int mode, int is_open)
 	blk_integrity_unregister(disk);
 	md_new_event(mddev);
 	sysfs_notify_dirent(mddev->sysfs_state);
-out:
 	return err;
 }
 
@@ -5083,18 +5132,8 @@ static int update_size(mddev_t *mddev, sector_t num_sectors)
 			return -ENOSPC;
 	}
 	rv = mddev->pers->resize(mddev, num_sectors);
-	if (!rv) {
-		struct block_device *bdev;
-
-		bdev = bdget_disk(mddev->gendisk, 0);
-		if (bdev) {
-			mutex_lock(&bdev->bd_inode->i_mutex);
-			i_size_write(bdev->bd_inode,
-				     (loff_t)mddev->array_sectors << 9);
-			mutex_unlock(&bdev->bd_inode->i_mutex);
-			bdput(bdev);
-		}
-	}
+	if (!rv)
+		revalidate_disk(mddev->gendisk);
 	return rv;
 }
 
@@ -5480,12 +5519,12 @@ static int md_open(struct block_device *bdev, fmode_t mode)
 	}
 	BUG_ON(mddev != bdev->bd_disk->private_data);
 
-	if ((err = mutex_lock_interruptible_nested(&mddev->reconfig_mutex, 1)))
+	if ((err = mutex_lock_interruptible(&mddev->open_mutex)))
 		goto out;
 
 	err = 0;
 	atomic_inc(&mddev->openers);
-	mddev_unlock(mddev);
+	mutex_unlock(&mddev->open_mutex);
 
 	check_disk_change(bdev);
  out:
@@ -6334,10 +6373,16 @@ void md_do_sync(mddev_t *mddev)
 			sysfs_notify(&mddev->kobj, NULL, "sync_completed");
 		}
 
-		if (j >= mddev->resync_max)
-			wait_event(mddev->recovery_wait,
-				   mddev->resync_max > j
-				   || kthread_should_stop());
+		while (j >= mddev->resync_max && !kthread_should_stop()) {
+			/* As this condition is controlled by user-space,
+			 * we can block indefinitely, so use '_interruptible'
+			 * to avoid triggering warnings.
+			 */
+			flush_signals(current); /* just in case */
+			wait_event_interruptible(mddev->recovery_wait,
+						 mddev->resync_max > j
+						 || kthread_should_stop());
+		}
 
 		if (kthread_should_stop())
 			goto interrupted;
diff --git a/drivers/md/md.h b/drivers/md/md.h
index 9430a110db9..f8fc188bc76 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -223,6 +223,16 @@ struct mddev_s
 							    * so we don't loop trying */
 
 	int				in_sync;	/* know to not need resync */
+	/* 'open_mutex' avoids races between 'md_open' and 'do_md_stop', so
+	 * that we are never stopping an array while it is open.
+	 * 'reconfig_mutex' protects all other reconfiguration.
+	 * These locks are separate due to conflicting interactions
+	 * with bdev->bd_mutex.
+	 * Lock ordering is:
+	 *  reconfig_mutex -> bd_mutex : e.g. do_md_run -> revalidate_disk
+	 *  bd_mutex -> open_mutex:  e.g. __blkdev_get -> md_open
+	 */
+	struct mutex			open_mutex;
 	struct mutex			reconfig_mutex;
 	atomic_t			active;		/* general refcount */
 	atomic_t			openers;	/* number of active opens */
@@ -431,5 +441,7 @@ extern int md_allow_write(mddev_t *mddev);
 extern void md_wait_for_blocked_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
 extern void md_set_array_sectors(mddev_t *mddev, sector_t array_sectors);
 extern int md_check_no_bitmap(mddev_t *mddev);
+extern int md_integrity_register(mddev_t *mddev);
+void md_integrity_add_rdev(mdk_rdev_t *rdev, mddev_t *mddev);
 
 #endif /* _MD_MD_H */
diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c
index cbe368fa659..7140909f666 100644
--- a/drivers/md/multipath.c
+++ b/drivers/md/multipath.c
@@ -294,7 +294,8 @@ static int multipath_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 	for (path = first; path <= last; path++)
 		if ((p=conf->multipaths+path)->rdev == NULL) {
 			q = rdev->bdev->bd_disk->queue;
-			blk_queue_stack_limits(mddev->queue, q);
+			disk_stack_limits(mddev->gendisk, rdev->bdev,
+					  rdev->data_offset << 9);
 
 		/* as we don't honour merge_bvec_fn, we must never risk
 		 * violating it, so limit ->max_sector to one PAGE, as
@@ -312,6 +313,7 @@ static int multipath_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 			set_bit(In_sync, &rdev->flags);
 			rcu_assign_pointer(p->rdev, rdev);
 			err = 0;
+			md_integrity_add_rdev(rdev, mddev);
 			break;
 		}
 
@@ -344,7 +346,9 @@ static int multipath_remove_disk(mddev_t *mddev, int number)
 			/* lost the race, try later */
 			err = -EBUSY;
 			p->rdev = rdev;
+			goto abort;
 		}
+		md_integrity_register(mddev);
 	}
 abort:
 
@@ -463,9 +467,9 @@ static int multipath_run (mddev_t *mddev)
 
 		disk = conf->multipaths + disk_idx;
 		disk->rdev = rdev;
+		disk_stack_limits(mddev->gendisk, rdev->bdev,
+				  rdev->data_offset << 9);
 
-		blk_queue_stack_limits(mddev->queue,
-				       rdev->bdev->bd_disk->queue);
 		/* as we don't honour merge_bvec_fn, we must never risk
 		 * violating it, not that we ever expect a device with
 		 * a merge_bvec_fn to be involved in multipath */
@@ -518,7 +522,7 @@ static int multipath_run (mddev_t *mddev)
 	mddev->queue->unplug_fn = multipath_unplug;
 	mddev->queue->backing_dev_info.congested_fn = multipath_congested;
 	mddev->queue->backing_dev_info.congested_data = mddev;
-
+	md_integrity_register(mddev);
 	return 0;
 
 out_free_conf:
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index ab4a489d869..898e2bdfee4 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -170,8 +170,8 @@ static int create_strip_zones(mddev_t *mddev)
 		}
 		dev[j] = rdev1;
 
-		blk_queue_stack_limits(mddev->queue,
-				       rdev1->bdev->bd_disk->queue);
+		disk_stack_limits(mddev->gendisk, rdev1->bdev,
+				  rdev1->data_offset << 9);
 		/* as we don't honour merge_bvec_fn, we must never risk
 		 * violating it, so limit ->max_sector to one PAGE, as
 		 * a one page request is never in violation.
@@ -250,6 +250,11 @@ static int create_strip_zones(mddev_t *mddev)
 		       mddev->chunk_sectors << 9);
 		goto abort;
 	}
+
+	blk_queue_io_min(mddev->queue, mddev->chunk_sectors << 9);
+	blk_queue_io_opt(mddev->queue,
+			 (mddev->chunk_sectors << 9) * mddev->raid_disks);
+
 	printk(KERN_INFO "raid0: done.\n");
 	mddev->private = conf;
 	return 0;
@@ -346,6 +351,7 @@ static int raid0_run(mddev_t *mddev)
 
 	blk_queue_merge_bvec(mddev->queue, raid0_mergeable_bvec);
 	dump_zones(mddev);
+	md_integrity_register(mddev);
 	return 0;
 }
 
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 89939a7aef5..8726fd7ebce 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1123,8 +1123,8 @@ static int raid1_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 	for (mirror = first; mirror <= last; mirror++)
 		if ( !(p=conf->mirrors+mirror)->rdev) {
 
-			blk_queue_stack_limits(mddev->queue,
-					       rdev->bdev->bd_disk->queue);
+			disk_stack_limits(mddev->gendisk, rdev->bdev,
+					  rdev->data_offset << 9);
 			/* as we don't honour merge_bvec_fn, we must never risk
 			 * violating it, so limit ->max_sector to one PAGE, as
 			 * a one page request is never in violation.
@@ -1144,7 +1144,7 @@ static int raid1_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 			rcu_assign_pointer(p->rdev, rdev);
 			break;
 		}
-
+	md_integrity_add_rdev(rdev, mddev);
 	print_conf(conf);
 	return err;
 }
@@ -1178,7 +1178,9 @@ static int raid1_remove_disk(mddev_t *mddev, int number)
 			/* lost the race, try later */
 			err = -EBUSY;
 			p->rdev = rdev;
+			goto abort;
 		}
+		md_integrity_register(mddev);
 	}
 abort:
 
@@ -1988,9 +1990,8 @@ static int run(mddev_t *mddev)
 		disk = conf->mirrors + disk_idx;
 
 		disk->rdev = rdev;
-
-		blk_queue_stack_limits(mddev->queue,
-				       rdev->bdev->bd_disk->queue);
+		disk_stack_limits(mddev->gendisk, rdev->bdev,
+				  rdev->data_offset << 9);
 		/* as we don't honour merge_bvec_fn, we must never risk
 		 * violating it, so limit ->max_sector to one PAGE, as
 		 * a one page request is never in violation.
@@ -2068,7 +2069,7 @@ static int run(mddev_t *mddev)
 	mddev->queue->unplug_fn = raid1_unplug;
 	mddev->queue->backing_dev_info.congested_fn = raid1_congested;
 	mddev->queue->backing_dev_info.congested_data = mddev;
-
+	md_integrity_register(mddev);
 	return 0;
 
 out_no_mem:
@@ -2133,6 +2134,7 @@ static int raid1_resize(mddev_t *mddev, sector_t sectors)
 		return -EINVAL;
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	mddev->changed = 1;
+	revalidate_disk(mddev->gendisk);
 	if (sectors > mddev->dev_sectors &&
 	    mddev->recovery_cp == MaxSector) {
 		mddev->recovery_cp = mddev->dev_sectors;
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index ae12ceafe10..3d9020cf6f6 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -1151,8 +1151,8 @@ static int raid10_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 	for ( ; mirror <= last ; mirror++)
 		if ( !(p=conf->mirrors+mirror)->rdev) {
 
-			blk_queue_stack_limits(mddev->queue,
-					       rdev->bdev->bd_disk->queue);
+			disk_stack_limits(mddev->gendisk, rdev->bdev,
+					  rdev->data_offset << 9);
 			/* as we don't honour merge_bvec_fn, we must never risk
 			 * violating it, so limit ->max_sector to one PAGE, as
 			 * a one page request is never in violation.
@@ -1170,6 +1170,7 @@ static int raid10_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
 			break;
 		}
 
+	md_integrity_add_rdev(rdev, mddev);
 	print_conf(conf);
 	return err;
 }
@@ -1203,7 +1204,9 @@ static int raid10_remove_disk(mddev_t *mddev, int number)
 			/* lost the race, try later */
 			err = -EBUSY;
 			p->rdev = rdev;
+			goto abort;
 		}
+		md_integrity_register(mddev);
 	}
 abort:
 
@@ -2044,7 +2047,7 @@ raid10_size(mddev_t *mddev, sector_t sectors, int raid_disks)
 static int run(mddev_t *mddev)
 {
 	conf_t *conf;
-	int i, disk_idx;
+	int i, disk_idx, chunk_size;
 	mirror_info_t *disk;
 	mdk_rdev_t *rdev;
 	int nc, fc, fo;
@@ -2130,6 +2133,14 @@ static int run(mddev_t *mddev)
 	spin_lock_init(&conf->device_lock);
 	mddev->queue->queue_lock = &conf->device_lock;
 
+	chunk_size = mddev->chunk_sectors << 9;
+	blk_queue_io_min(mddev->queue, chunk_size);
+	if (conf->raid_disks % conf->near_copies)
+		blk_queue_io_opt(mddev->queue, chunk_size * conf->raid_disks);
+	else
+		blk_queue_io_opt(mddev->queue, chunk_size *
+				 (conf->raid_disks / conf->near_copies));
+
 	list_for_each_entry(rdev, &mddev->disks, same_set) {
 		disk_idx = rdev->raid_disk;
 		if (disk_idx >= mddev->raid_disks
@@ -2138,9 +2149,8 @@ static int run(mddev_t *mddev)
 		disk = conf->mirrors + disk_idx;
 
 		disk->rdev = rdev;
-
-		blk_queue_stack_limits(mddev->queue,
-				       rdev->bdev->bd_disk->queue);
+		disk_stack_limits(mddev->gendisk, rdev->bdev,
+				  rdev->data_offset << 9);
 		/* as we don't honour merge_bvec_fn, we must never risk
 		 * violating it, so limit ->max_sector to one PAGE, as
 		 * a one page request is never in violation.
@@ -2218,6 +2228,7 @@ static int run(mddev_t *mddev)
 
 	if (conf->near_copies < mddev->raid_disks)
 		blk_queue_merge_bvec(mddev->queue, raid10_mergeable_bvec);
+	md_integrity_register(mddev);
 	return 0;
 
 out_free_conf:
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index cac6f4d3a14..9b00a229015 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -3911,13 +3911,21 @@ static int make_request(struct request_queue *q, struct bio * bi)
 					goto retry;
 				}
 			}
-			/* FIXME what if we get a false positive because these
-			 * are being updated.
-			 */
-			if (logical_sector >= mddev->suspend_lo &&
+
+			if (bio_data_dir(bi) == WRITE &&
+			    logical_sector >= mddev->suspend_lo &&
 			    logical_sector < mddev->suspend_hi) {
 				release_stripe(sh);
-				schedule();
+				/* As the suspend_* range is controlled by
+				 * userspace, we want an interruptible
+				 * wait.
+				 */
+				flush_signals(current);
+				prepare_to_wait(&conf->wait_for_overlap,
+						&w, TASK_INTERRUPTIBLE);
+				if (logical_sector >= mddev->suspend_lo &&
+				    logical_sector < mddev->suspend_hi)
+					schedule();
 				goto retry;
 			}
 
@@ -3989,7 +3997,7 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped
 		    conf->reshape_progress < raid5_size(mddev, 0, 0)) {
 			sector_nr = raid5_size(mddev, 0, 0)
 				- conf->reshape_progress;
-		} else if (mddev->delta_disks > 0 &&
+		} else if (mddev->delta_disks >= 0 &&
 			   conf->reshape_progress > 0)
 			sector_nr = conf->reshape_progress;
 		sector_div(sector_nr, new_data_disks);
@@ -4203,6 +4211,9 @@ static inline sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *ski
 		return 0;
 	}
 
+	/* Allow raid5_quiesce to complete */
+	wait_event(conf->wait_for_overlap, conf->quiesce != 2);
+
 	if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
 		return reshape_request(mddev, sector_nr, skipped);
 
@@ -4803,7 +4814,7 @@ static raid5_conf_t *setup_conf(mddev_t *mddev)
 static int run(mddev_t *mddev)
 {
 	raid5_conf_t *conf;
-	int working_disks = 0;
+	int working_disks = 0, chunk_size;
 	mdk_rdev_t *rdev;
 
 	if (mddev->recovery_cp != MaxSector)
@@ -4844,7 +4855,26 @@ static int run(mddev_t *mddev)
 			   (old_disks-max_degraded));
 		/* here_old is the first stripe that we might need to read
 		 * from */
-		if (here_new >= here_old) {
+		if (mddev->delta_disks == 0) {
+			/* We cannot be sure it is safe to start an in-place
+			 * reshape.  It is only safe if user-space if monitoring
+			 * and taking constant backups.
+			 * mdadm always starts a situation like this in
+			 * readonly mode so it can take control before
+			 * allowing any writes.  So just check for that.
+			 */
+			if ((here_new * mddev->new_chunk_sectors != 
+			     here_old * mddev->chunk_sectors) ||
+			    mddev->ro == 0) {
+				printk(KERN_ERR "raid5: in-place reshape must be started"
+				       " in read-only mode - aborting\n");
+				return -EINVAL;
+			}
+		} else if (mddev->delta_disks < 0
+		    ? (here_new * mddev->new_chunk_sectors <=
+		       here_old * mddev->chunk_sectors)
+		    : (here_new * mddev->new_chunk_sectors >=
+		       here_old * mddev->chunk_sectors)) {
 			/* Reading from the same stripe as writing to - bad */
 			printk(KERN_ERR "raid5: reshape_position too early for "
 			       "auto-recovery - aborting.\n");
@@ -4958,6 +4988,14 @@ static int run(mddev_t *mddev)
 	md_set_array_sectors(mddev, raid5_size(mddev, 0, 0));
 
 	blk_queue_merge_bvec(mddev->queue, raid5_mergeable_bvec);
+	chunk_size = mddev->chunk_sectors << 9;
+	blk_queue_io_min(mddev->queue, chunk_size);
+	blk_queue_io_opt(mddev->queue, chunk_size *
+			 (conf->raid_disks - conf->max_degraded));
+
+	list_for_each_entry(rdev, &mddev->disks, same_set)
+		disk_stack_limits(mddev->gendisk, rdev->bdev,
+				  rdev->data_offset << 9);
 
 	return 0;
 abort:
@@ -5185,6 +5223,7 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors)
 		return -EINVAL;
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	mddev->changed = 1;
+	revalidate_disk(mddev->gendisk);
 	if (sectors > mddev->dev_sectors && mddev->recovery_cp == MaxSector) {
 		mddev->recovery_cp = mddev->dev_sectors;
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
@@ -5330,7 +5369,7 @@ static int raid5_start_reshape(mddev_t *mddev)
 		spin_unlock_irqrestore(&conf->device_lock, flags);
 	}
 	mddev->raid_disks = conf->raid_disks;
-	mddev->reshape_position = 0;
+	mddev->reshape_position = conf->reshape_progress;
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);
 
 	clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
@@ -5385,7 +5424,6 @@ static void end_reshape(raid5_conf_t *conf)
  */
 static void raid5_finish_reshape(mddev_t *mddev)
 {
-	struct block_device *bdev;
 	raid5_conf_t *conf = mddev->private;
 
 	if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
@@ -5394,15 +5432,7 @@ static void raid5_finish_reshape(mddev_t *mddev)
 			md_set_array_sectors(mddev, raid5_size(mddev, 0, 0));
 			set_capacity(mddev->gendisk, mddev->array_sectors);
 			mddev->changed = 1;
-
-			bdev = bdget_disk(mddev->gendisk, 0);
-			if (bdev) {
-				mutex_lock(&bdev->bd_inode->i_mutex);
-				i_size_write(bdev->bd_inode,
-					     (loff_t)mddev->array_sectors << 9);
-				mutex_unlock(&bdev->bd_inode->i_mutex);
-				bdput(bdev);
-			}
+			revalidate_disk(mddev->gendisk);
 		} else {
 			int d;
 			mddev->degraded = conf->raid_disks;
@@ -5413,8 +5443,15 @@ static void raid5_finish_reshape(mddev_t *mddev)
 					mddev->degraded--;
 			for (d = conf->raid_disks ;
 			     d < conf->raid_disks - mddev->delta_disks;
-			     d++)
-				raid5_remove_disk(mddev, d);
+			     d++) {
+				mdk_rdev_t *rdev = conf->disks[d].rdev;
+				if (rdev && raid5_remove_disk(mddev, d) == 0) {
+					char nm[20];
+					sprintf(nm, "rd%d", rdev->raid_disk);
+					sysfs_remove_link(&mddev->kobj, nm);
+					rdev->raid_disk = -1;
+				}
+			}
 		}
 		mddev->layout = conf->algorithm;
 		mddev->chunk_sectors = conf->chunk_sectors;
@@ -5434,12 +5471,18 @@ static void raid5_quiesce(mddev_t *mddev, int state)
 
 	case 1: /* stop all writes */
 		spin_lock_irq(&conf->device_lock);
-		conf->quiesce = 1;
+		/* '2' tells resync/reshape to pause so that all
+		 * active stripes can drain
+		 */
+		conf->quiesce = 2;
 		wait_event_lock_irq(conf->wait_for_stripe,
 				    atomic_read(&conf->active_stripes) == 0 &&
 				    atomic_read(&conf->active_aligned_reads) == 0,
 				    conf->device_lock, /* nothing */);
+		conf->quiesce = 1;
 		spin_unlock_irq(&conf->device_lock);
+		/* allow reshape to continue */
+		wake_up(&conf->wait_for_overlap);
 		break;
 
 	case 0: /* re-enable writes */