Merge branch 'linux-3.4.35' into rel-17

Bug 1243631

Change-Id: I915826047b2e20f0ad0a7d75df295c6cbf6e5b0a

13 files changed, 214 insertions, 140 deletions

diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c
index a1a3e6df17b8..f011d4b3c139 100644
--- a/drivers/md/dm-ioctl.c
+++ b/drivers/md/dm-ioctl.c
@@ -1563,6 +1563,14 @@ static int copy_params(struct dm_ioctl __user *user, struct dm_ioctl **param)
 	if (copy_from_user(dmi, user, tmp.data_size))
 		goto bad;
 
+	/*
+	 * Abort if something changed the ioctl data while it was being copied.
+	 */
+	if (dmi->data_size != tmp.data_size) {
+		DMERR("rejecting ioctl: data size modified while processing parameters");
+		goto bad;
+	}
+
 	/* Wipe the user buffer so we do not return it to userspace */
 	if (secure_data && clear_user(user, tmp.data_size))
 		goto bad;
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index 2e227fbf1622..f220a695a4b1 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -1351,17 +1351,25 @@ static int device_is_nonrot(struct dm_target *ti, struct dm_dev *dev,
 	return q && blk_queue_nonrot(q);
 }
 
-static bool dm_table_is_nonrot(struct dm_table *t)
+static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,
+			     sector_t start, sector_t len, void *data)
+{
+	struct request_queue *q = bdev_get_queue(dev->bdev);
+
+	return q && !blk_queue_add_random(q);
+}
+
+static bool dm_table_all_devices_attribute(struct dm_table *t,
+					   iterate_devices_callout_fn func)
 {
 	struct dm_target *ti;
 	unsigned i = 0;
 
-	/* Ensure that all underlying device are non-rotational. */
 	while (i < dm_table_get_num_targets(t)) {
 		ti = dm_table_get_target(t, i++);
 
 		if (!ti->type->iterate_devices ||
-		    !ti->type->iterate_devices(ti, device_is_nonrot, NULL))
+		    !ti->type->iterate_devices(ti, func, NULL))
 			return 0;
 	}
 
@@ -1393,7 +1401,8 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	if (!dm_table_discard_zeroes_data(t))
 		q->limits.discard_zeroes_data = 0;
 
-	if (dm_table_is_nonrot(t))
+	/* Ensure that all underlying devices are non-rotational. */
+	if (dm_table_all_devices_attribute(t, device_is_nonrot))
 		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, q);
 	else
 		queue_flag_clear_unlocked(QUEUE_FLAG_NONROT, q);
@@ -1401,6 +1410,15 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	dm_table_set_integrity(t);
 
 	/*
+	 * Determine whether or not this queue's I/O timings contribute
+	 * to the entropy pool, Only request-based targets use this.
+	 * Clear QUEUE_FLAG_ADD_RANDOM if any underlying device does not
+	 * have it set.
+	 */
+	if (blk_queue_add_random(q) && dm_table_all_devices_attribute(t, device_is_not_random))
+		queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, q);
+
+	/*
 	 * QUEUE_FLAG_STACKABLE must be set after all queue settings are
 	 * visible to other CPUs because, once the flag is set, incoming bios
 	 * are processed by request-based dm, which refers to the queue
diff --git a/drivers/md/dm-verity.c b/drivers/md/dm-verity.c
index fa365d39b612..68bf5c37c3fe 100644
--- a/drivers/md/dm-verity.c
+++ b/drivers/md/dm-verity.c
@@ -718,8 +718,8 @@ static int verity_ctr(struct dm_target *ti, unsigned argc, char **argv)
 	v->hash_dev_block_bits = ffs(num) - 1;
 
 	if (sscanf(argv[5], "%llu%c", &num_ll, &dummy) != 1 ||
-	    num_ll << (v->data_dev_block_bits - SECTOR_SHIFT) !=
-	    (sector_t)num_ll << (v->data_dev_block_bits - SECTOR_SHIFT)) {
+	    (sector_t)(num_ll << (v->data_dev_block_bits - SECTOR_SHIFT))
+	    >> (v->data_dev_block_bits - SECTOR_SHIFT) != num_ll) {
 		ti->error = "Invalid data blocks";
 		r = -EINVAL;
 		goto bad;
@@ -733,8 +733,8 @@ static int verity_ctr(struct dm_target *ti, unsigned argc, char **argv)
 	}
 
 	if (sscanf(argv[6], "%llu%c", &num_ll, &dummy) != 1 ||
-	    num_ll << (v->hash_dev_block_bits - SECTOR_SHIFT) !=
-	    (sector_t)num_ll << (v->hash_dev_block_bits - SECTOR_SHIFT)) {
+	    (sector_t)(num_ll << (v->hash_dev_block_bits - SECTOR_SHIFT))
+	    >> (v->hash_dev_block_bits - SECTOR_SHIFT) != num_ll) {
 		ti->error = "Invalid hash start";
 		r = -EINVAL;
 		goto bad;
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index e24143cc2040..32370ea32e22 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -754,8 +754,14 @@ static void rq_completed(struct mapped_device *md, int rw, int run_queue)
 	if (!md_in_flight(md))
 		wake_up(&md->wait);
 
+	/*
+	 * Run this off this callpath, as drivers could invoke end_io while
+	 * inside their request_fn (and holding the queue lock). Calling
+	 * back into ->request_fn() could deadlock attempting to grab the
+	 * queue lock again.
+	 */
 	if (run_queue)
-		blk_run_queue(md->queue);
+		blk_run_queue_async(md->queue);
 
 	/*
 	 * dm_put() must be at the end of this function. See the comment above
@@ -865,10 +871,14 @@ static void dm_done(struct request *clone, int error, bool mapped)
 {
 	int r = error;
 	struct dm_rq_target_io *tio = clone->end_io_data;
-	dm_request_endio_fn rq_end_io = tio->ti->type->rq_end_io;
+	dm_request_endio_fn rq_end_io = NULL;
 
-	if (mapped && rq_end_io)
-		r = rq_end_io(tio->ti, clone, error, &tio->info);
+	if (tio->ti) {
+		rq_end_io = tio->ti->type->rq_end_io;
+
+		if (mapped && rq_end_io)
+			r = rq_end_io(tio->ti, clone, error, &tio->info);
+	}
 
 	if (r <= 0)
 		/* The target wants to complete the I/O */
@@ -1566,15 +1576,6 @@ static int map_request(struct dm_target *ti, struct request *clone,
 	int r, requeued = 0;
 	struct dm_rq_target_io *tio = clone->end_io_data;
 
-	/*
-	 * Hold the md reference here for the in-flight I/O.
-	 * We can't rely on the reference count by device opener,
-	 * because the device may be closed during the request completion
-	 * when all bios are completed.
-	 * See the comment in rq_completed() too.
-	 */
-	dm_get(md);
-
 	tio->ti = ti;
 	r = ti->type->map_rq(ti, clone, &tio->info);
 	switch (r) {
@@ -1606,6 +1607,26 @@ static int map_request(struct dm_target *ti, struct request *clone,
 	return requeued;
 }
 
+static struct request *dm_start_request(struct mapped_device *md, struct request *orig)
+{
+	struct request *clone;
+
+	blk_start_request(orig);
+	clone = orig->special;
+	atomic_inc(&md->pending[rq_data_dir(clone)]);
+
+	/*
+	 * Hold the md reference here for the in-flight I/O.
+	 * We can't rely on the reference count by device opener,
+	 * because the device may be closed during the request completion
+	 * when all bios are completed.
+	 * See the comment in rq_completed() too.
+	 */
+	dm_get(md);
+
+	return clone;
+}
+
 /*
  * q->request_fn for request-based dm.
  * Called with the queue lock held.
@@ -1635,14 +1656,21 @@ static void dm_request_fn(struct request_queue *q)
 			pos = blk_rq_pos(rq);
 
 		ti = dm_table_find_target(map, pos);
-		BUG_ON(!dm_target_is_valid(ti));
+		if (!dm_target_is_valid(ti)) {
+			/*
+			 * Must perform setup, that dm_done() requires,
+			 * before calling dm_kill_unmapped_request
+			 */
+			DMERR_LIMIT("request attempted access beyond the end of device");
+			clone = dm_start_request(md, rq);
+			dm_kill_unmapped_request(clone, -EIO);
+			continue;
+		}
 
 		if (ti->type->busy && ti->type->busy(ti))
 			goto delay_and_out;
 
-		blk_start_request(rq);
-		clone = rq->special;
-		atomic_inc(&md->pending[rq_data_dir(clone)]);
+		clone = dm_start_request(md, rq);
 
 		spin_unlock(q->queue_lock);
 		if (map_request(ti, clone, md))
@@ -1662,8 +1690,6 @@ delay_and_out:
 	blk_delay_queue(q, HZ / 10);
 out:
 	dm_table_put(map);
-
-	return;
 }
 
 int dm_underlying_device_busy(struct request_queue *q)
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 9ee8ce3e9650..0a447a1be2ca 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -1143,8 +1143,11 @@ static int super_90_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor
 			ret = 0;
 	}
 	rdev->sectors = rdev->sb_start;
-	/* Limit to 4TB as metadata cannot record more than that */
-	if (rdev->sectors >= (2ULL << 32))
+	/* Limit to 4TB as metadata cannot record more than that.
+	 * (not needed for Linear and RAID0 as metadata doesn't
+	 * record this size)
+	 */
+	if (rdev->sectors >= (2ULL << 32) && sb->level >= 1)
 		rdev->sectors = (2ULL << 32) - 2;
 
 	if (rdev->sectors < ((sector_t)sb->size) * 2 && sb->level >= 1)
@@ -1426,7 +1429,7 @@ super_90_rdev_size_change(struct md_rdev *rdev, sector_t num_sectors)
 	/* Limit to 4TB as metadata cannot record more than that.
 	 * 4TB == 2^32 KB, or 2*2^32 sectors.
 	 */
-	if (num_sectors >= (2ULL << 32))
+	if (num_sectors >= (2ULL << 32) && rdev->mddev->level >= 1)
 		num_sectors = (2ULL << 32) - 2;
 	md_super_write(rdev->mddev, rdev, rdev->sb_start, rdev->sb_size,
 		       rdev->sb_page);
@@ -1802,10 +1805,10 @@ retry:
 			memset(bbp, 0xff, PAGE_SIZE);
 
 			for (i = 0 ; i < bb->count ; i++) {
-				u64 internal_bb = *p++;
+				u64 internal_bb = p[i];
 				u64 store_bb = ((BB_OFFSET(internal_bb) << 10)
 						| BB_LEN(internal_bb));
-				*bbp++ = cpu_to_le64(store_bb);
+				bbp[i] = cpu_to_le64(store_bb);
 			}
 			bb->changed = 0;
 			if (read_seqretry(&bb->lock, seq))
@@ -7417,6 +7420,8 @@ static int remove_and_add_spares(struct mddev *mddev)
 			}
 		}
 	}
+	if (removed)
+		set_bit(MD_CHANGE_DEVS, &mddev->flags);
 	return spares;
 }
 
@@ -7430,9 +7435,11 @@ static void reap_sync_thread(struct mddev *mddev)
 	    !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
 		/* success...*/
 		/* activate any spares */
-		if (mddev->pers->spare_active(mddev))
+		if (mddev->pers->spare_active(mddev)) {
 			sysfs_notify(&mddev->kobj, NULL,
 				     "degraded");
+			set_bit(MD_CHANGE_DEVS, &mddev->flags);
+		}
 	}
 	if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery) &&
 	    mddev->pers->finish_reshape)
@@ -7687,9 +7694,9 @@ int md_is_badblock(struct badblocks *bb, sector_t s, int sectors,
 		   sector_t *first_bad, int *bad_sectors)
 {
 	int hi;
-	int lo = 0;
+	int lo;
 	u64 *p = bb->page;
-	int rv = 0;
+	int rv;
 	sector_t target = s + sectors;
 	unsigned seq;
 
@@ -7704,7 +7711,8 @@ int md_is_badblock(struct badblocks *bb, sector_t s, int sectors,
 
 retry:
 	seq = read_seqbegin(&bb->lock);
-
+	lo = 0;
+	rv = 0;
 	hi = bb->count;
 
 	/* Binary search between lo and hi for 'target'
diff --git a/drivers/md/persistent-data/dm-btree-internal.h b/drivers/md/persistent-data/dm-btree-internal.h
index 5709bfeab1e8..accbb05f17b6 100644
--- a/drivers/md/persistent-data/dm-btree-internal.h
+++ b/drivers/md/persistent-data/dm-btree-internal.h
@@ -36,13 +36,13 @@ struct node_header {
 	__le32 padding;
 } __packed;
 
-struct node {
+struct btree_node {
 	struct node_header header;
 	__le64 keys[0];
 } __packed;
 
 
-void inc_children(struct dm_transaction_manager *tm, struct node *n,
+void inc_children(struct dm_transaction_manager *tm, struct btree_node *n,
 		  struct dm_btree_value_type *vt);
 
 int new_block(struct dm_btree_info *info, struct dm_block **result);
@@ -64,7 +64,7 @@ struct ro_spine {
 void init_ro_spine(struct ro_spine *s, struct dm_btree_info *info);
 int exit_ro_spine(struct ro_spine *s);
 int ro_step(struct ro_spine *s, dm_block_t new_child);
-struct node *ro_node(struct ro_spine *s);
+struct btree_node *ro_node(struct ro_spine *s);
 
 struct shadow_spine {
 	struct dm_btree_info *info;
@@ -98,17 +98,17 @@ int shadow_root(struct shadow_spine *s);
 /*
  * Some inlines.
  */
-static inline __le64 *key_ptr(struct node *n, uint32_t index)
+static inline __le64 *key_ptr(struct btree_node *n, uint32_t index)
 {
 	return n->keys + index;
 }
 
-static inline void *value_base(struct node *n)
+static inline void *value_base(struct btree_node *n)
 {
 	return &n->keys[le32_to_cpu(n->header.max_entries)];
 }
 
-static inline void *value_ptr(struct node *n, uint32_t index)
+static inline void *value_ptr(struct btree_node *n, uint32_t index)
 {
 	uint32_t value_size = le32_to_cpu(n->header.value_size);
 	return value_base(n) + (value_size * index);
@@ -117,7 +117,7 @@ static inline void *value_ptr(struct node *n, uint32_t index)
 /*
  * Assumes the values are suitably-aligned and converts to core format.
  */
-static inline uint64_t value64(struct node *n, uint32_t index)
+static inline uint64_t value64(struct btree_node *n, uint32_t index)
 {
 	__le64 *values_le = value_base(n);
 
@@ -127,7 +127,7 @@ static inline uint64_t value64(struct node *n, uint32_t index)
 /*
  * Searching for a key within a single node.
  */
-int lower_bound(struct node *n, uint64_t key);
+int lower_bound(struct btree_node *n, uint64_t key);
 
 extern struct dm_block_validator btree_node_validator;
 
diff --git a/drivers/md/persistent-data/dm-btree-remove.c b/drivers/md/persistent-data/dm-btree-remove.c
index aa71e2359a07..c4f28133ef82 100644
--- a/drivers/md/persistent-data/dm-btree-remove.c
+++ b/drivers/md/persistent-data/dm-btree-remove.c
@@ -53,7 +53,7 @@
 /*
  * Some little utilities for moving node data around.
  */
-static void node_shift(struct node *n, int shift)
+static void node_shift(struct btree_node *n, int shift)
 {
 	uint32_t nr_entries = le32_to_cpu(n->header.nr_entries);
 	uint32_t value_size = le32_to_cpu(n->header.value_size);
@@ -79,7 +79,7 @@ static void node_shift(struct node *n, int shift)
 	}
 }
 
-static void node_copy(struct node *left, struct node *right, int shift)
+static void node_copy(struct btree_node *left, struct btree_node *right, int shift)
 {
 	uint32_t nr_left = le32_to_cpu(left->header.nr_entries);
 	uint32_t value_size = le32_to_cpu(left->header.value_size);
@@ -108,7 +108,7 @@ static void node_copy(struct node *left, struct node *right, int shift)
 /*
  * Delete a specific entry from a leaf node.
  */
-static void delete_at(struct node *n, unsigned index)
+static void delete_at(struct btree_node *n, unsigned index)
 {
 	unsigned nr_entries = le32_to_cpu(n->header.nr_entries);
 	unsigned nr_to_copy = nr_entries - (index + 1);
@@ -128,7 +128,7 @@ static void delete_at(struct node *n, unsigned index)
 	n->header.nr_entries = cpu_to_le32(nr_entries - 1);
 }
 
-static unsigned merge_threshold(struct node *n)
+static unsigned merge_threshold(struct btree_node *n)
 {
 	return le32_to_cpu(n->header.max_entries) / 3;
 }
@@ -136,7 +136,7 @@ static unsigned merge_threshold(struct node *n)
 struct child {
 	unsigned index;
 	struct dm_block *block;
-	struct node *n;
+	struct btree_node *n;
 };
 
 static struct dm_btree_value_type le64_type = {
@@ -147,7 +147,7 @@ static struct dm_btree_value_type le64_type = {
 	.equal = NULL
 };
 
-static int init_child(struct dm_btree_info *info, struct node *parent,
+static int init_child(struct dm_btree_info *info, struct btree_node *parent,
 		      unsigned index, struct child *result)
 {
 	int r, inc;
@@ -177,7 +177,7 @@ static int exit_child(struct dm_btree_info *info, struct child *c)
 	return dm_tm_unlock(info->tm, c->block);
 }
 
-static void shift(struct node *left, struct node *right, int count)
+static void shift(struct btree_node *left, struct btree_node *right, int count)
 {
 	uint32_t nr_left = le32_to_cpu(left->header.nr_entries);
 	uint32_t nr_right = le32_to_cpu(right->header.nr_entries);
@@ -203,11 +203,11 @@ static void shift(struct node *left, struct node *right, int count)
 	right->header.nr_entries = cpu_to_le32(nr_right + count);
 }
 
-static void __rebalance2(struct dm_btree_info *info, struct node *parent,
+static void __rebalance2(struct dm_btree_info *info, struct btree_node *parent,
 			 struct child *l, struct child *r)
 {
-	struct node *left = l->n;
-	struct node *right = r->n;
+	struct btree_node *left = l->n;
+	struct btree_node *right = r->n;
 	uint32_t nr_left = le32_to_cpu(left->header.nr_entries);
 	uint32_t nr_right = le32_to_cpu(right->header.nr_entries);
 	unsigned threshold = 2 * merge_threshold(left) + 1;
@@ -239,7 +239,7 @@ static int rebalance2(struct shadow_spine *s, struct dm_btree_info *info,
 		      unsigned left_index)
 {
 	int r;
-	struct node *parent;
+	struct btree_node *parent;
 	struct child left, right;
 
 	parent = dm_block_data(shadow_current(s));
@@ -270,9 +270,9 @@ static int rebalance2(struct shadow_spine *s, struct dm_btree_info *info,
  * in right, then rebalance2.  This wastes some cpu, but I want something
  * simple atm.
  */
-static void delete_center_node(struct dm_btree_info *info, struct node *parent,
+static void delete_center_node(struct dm_btree_info *info, struct btree_node *parent,
 			       struct child *l, struct child *c, struct child *r,
-			       struct node *left, struct node *center, struct node *right,
+			       struct btree_node *left, struct btree_node *center, struct btree_node *right,
 			       uint32_t nr_left, uint32_t nr_center, uint32_t nr_right)
 {
 	uint32_t max_entries = le32_to_cpu(left->header.max_entries);
@@ -301,9 +301,9 @@ static void delete_center_node(struct dm_btree_info *info, struct node *parent,
 /*
  * Redistributes entries among 3 sibling nodes.
  */
-static void redistribute3(struct dm_btree_info *info, struct node *parent,
+static void redistribute3(struct dm_btree_info *info, struct btree_node *parent,
 			  struct child *l, struct child *c, struct child *r,
-			  struct node *left, struct node *center, struct node *right,
+			  struct btree_node *left, struct btree_node *center, struct btree_node *right,
 			  uint32_t nr_left, uint32_t nr_center, uint32_t nr_right)
 {
 	int s;
@@ -343,12 +343,12 @@ static void redistribute3(struct dm_btree_info *info, struct node *parent,
 	*key_ptr(parent, r->index) = right->keys[0];
 }
 
-static void __rebalance3(struct dm_btree_info *info, struct node *parent,
+static void __rebalance3(struct dm_btree_info *info, struct btree_node *parent,
 			 struct child *l, struct child *c, struct child *r)
 {
-	struct node *left = l->n;
-	struct node *center = c->n;
-	struct node *right = r->n;
+	struct btree_node *left = l->n;
+	struct btree_node *center = c->n;
+	struct btree_node *right = r->n;
 
 	uint32_t nr_left = le32_to_cpu(left->header.nr_entries);
 	uint32_t nr_center = le32_to_cpu(center->header.nr_entries);
@@ -371,7 +371,7 @@ static int rebalance3(struct shadow_spine *s, struct dm_btree_info *info,
 		      unsigned left_index)
 {
 	int r;
-	struct node *parent = dm_block_data(shadow_current(s));
+	struct btree_node *parent = dm_block_data(shadow_current(s));
 	struct child left, center, right;
 
 	/*
@@ -421,7 +421,7 @@ static int get_nr_entries(struct dm_transaction_manager *tm,
 {
 	int r;
 	struct dm_block *block;
-	struct node *n;
+	struct btree_node *n;
 
 	r = dm_tm_read_lock(tm, b, &btree_node_validator, &block);
 	if (r)
@@ -438,7 +438,7 @@ static int rebalance_children(struct shadow_spine *s,
 {
 	int i, r, has_left_sibling, has_right_sibling;
 	uint32_t child_entries;
-	struct node *n;
+	struct btree_node *n;
 
 	n = dm_block_data(shadow_current(s));
 
@@ -483,7 +483,7 @@ static int rebalance_children(struct shadow_spine *s,
 	return r;
 }
 
-static int do_leaf(struct node *n, uint64_t key, unsigned *index)
+static int do_leaf(struct btree_node *n, uint64_t key, unsigned *index)
 {
 	int i = lower_bound(n, key);
 
@@ -506,7 +506,7 @@ static int remove_raw(struct shadow_spine *s, struct dm_btree_info *info,
 		      uint64_t key, unsigned *index)
 {
 	int i = *index, r;
-	struct node *n;
+	struct btree_node *n;
 
 	for (;;) {
 		r = shadow_step(s, root, vt);
@@ -556,7 +556,7 @@ int dm_btree_remove(struct dm_btree_info *info, dm_block_t root,
 	unsigned level, last_level = info->levels - 1;
 	int index = 0, r = 0;
 	struct shadow_spine spine;
-	struct node *n;
+	struct btree_node *n;
 
 	init_shadow_spine(&spine, info);
 	for (level = 0; level < info->levels; level++) {
diff --git a/drivers/md/persistent-data/dm-btree-spine.c b/drivers/md/persistent-data/dm-btree-spine.c
index d9a7912ee8ee..2f0805c3263e 100644
--- a/drivers/md/persistent-data/dm-btree-spine.c
+++ b/drivers/md/persistent-data/dm-btree-spine.c
@@ -23,7 +23,7 @@ static void node_prepare_for_write(struct dm_block_validator *v,
 				   struct dm_block *b,
 				   size_t block_size)
 {
-	struct node *n = dm_block_data(b);
+	struct btree_node *n = dm_block_data(b);
 	struct node_header *h = &n->header;
 
 	h->blocknr = cpu_to_le64(dm_block_location(b));
@@ -38,7 +38,7 @@ static int node_check(struct dm_block_validator *v,
 		      struct dm_block *b,
 		      size_t block_size)
 {
-	struct node *n = dm_block_data(b);
+	struct btree_node *n = dm_block_data(b);
 	struct node_header *h = &n->header;
 	size_t value_size;
 	__le32 csum_disk;
@@ -164,7 +164,7 @@ int ro_step(struct ro_spine *s, dm_block_t new_child)
 	return r;
 }
 
-struct node *ro_node(struct ro_spine *s)
+struct btree_node *ro_node(struct ro_spine *s)
 {
 	struct dm_block *block;
 
diff --git a/drivers/md/persistent-data/dm-btree.c b/drivers/md/persistent-data/dm-btree.c
index d12b2cc51f1a..371f3d49d18e 100644
--- a/drivers/md/persistent-data/dm-btree.c
+++ b/drivers/md/persistent-data/dm-btree.c
@@ -38,7 +38,7 @@ static void array_insert(void *base, size_t elt_size, unsigned nr_elts,
 /*----------------------------------------------------------------*/
 
 /* makes the assumption that no two keys are the same. */
-static int bsearch(struct node *n, uint64_t key, int want_hi)
+static int bsearch(struct btree_node *n, uint64_t key, int want_hi)
 {
 	int lo = -1, hi = le32_to_cpu(n->header.nr_entries);
 
@@ -58,12 +58,12 @@ static int bsearch(struct node *n, uint64_t key, int want_hi)
 	return want_hi ? hi : lo;
 }
 
-int lower_bound(struct node *n, uint64_t key)
+int lower_bound(struct btree_node *n, uint64_t key)
 {
 	return bsearch(n, key, 0);
 }
 
-void inc_children(struct dm_transaction_manager *tm, struct node *n,
+void inc_children(struct dm_transaction_manager *tm, struct btree_node *n,
 		  struct dm_btree_value_type *vt)
 {
 	unsigned i;
@@ -77,7 +77,7 @@ void inc_children(struct dm_transaction_manager *tm, struct node *n,
 			vt->inc(vt->context, value_ptr(n, i));
 }
 
-static int insert_at(size_t value_size, struct node *node, unsigned index,
+static int insert_at(size_t value_size, struct btree_node *node, unsigned index,
 		      uint64_t key, void *value)
 		      __dm_written_to_disk(value)
 {
@@ -122,7 +122,7 @@ int dm_btree_empty(struct dm_btree_info *info, dm_block_t *root)
 {
 	int r;
 	struct dm_block *b;
-	struct node *n;
+	struct btree_node *n;
 	size_t block_size;
 	uint32_t max_entries;
 
@@ -154,7 +154,7 @@ EXPORT_SYMBOL_GPL(dm_btree_empty);
 #define MAX_SPINE_DEPTH 64
 struct frame {
 	struct dm_block *b;
-	struct node *n;
+	struct btree_node *n;
 	unsigned level;
 	unsigned nr_children;
 	unsigned current_child;
@@ -295,7 +295,7 @@ EXPORT_SYMBOL_GPL(dm_btree_del);
 /*----------------------------------------------------------------*/
 
 static int btree_lookup_raw(struct ro_spine *s, dm_block_t block, uint64_t key,
-			    int (*search_fn)(struct node *, uint64_t),
+			    int (*search_fn)(struct btree_node *, uint64_t),
 			    uint64_t *result_key, void *v, size_t value_size)
 {
 	int i, r;
@@ -406,7 +406,7 @@ static int btree_split_sibling(struct shadow_spine *s, dm_block_t root,
 	size_t size;
 	unsigned nr_left, nr_right;
 	struct dm_block *left, *right, *parent;
-	struct node *ln, *rn, *pn;
+	struct btree_node *ln, *rn, *pn;
 	__le64 location;
 
 	left = shadow_current(s);
@@ -491,7 +491,7 @@ static int btree_split_beneath(struct shadow_spine *s, uint64_t key)
 	size_t size;
 	unsigned nr_left, nr_right;
 	struct dm_block *left, *right, *new_parent;
-	struct node *pn, *ln, *rn;
+	struct btree_node *pn, *ln, *rn;
 	__le64 val;
 
 	new_parent = shadow_current(s);
@@ -576,7 +576,7 @@ static int btree_insert_raw(struct shadow_spine *s, dm_block_t root,
 			    uint64_t key, unsigned *index)
 {
 	int r, i = *index, top = 1;
-	struct node *node;
+	struct btree_node *node;
 
 	for (;;) {
 		r = shadow_step(s, root, vt);
@@ -643,7 +643,7 @@ static int insert(struct dm_btree_info *info, dm_block_t root,
 	unsigned level, index = -1, last_level = info->levels - 1;
 	dm_block_t block = root;
 	struct shadow_spine spine;
-	struct node *n;
+	struct btree_node *n;
 	struct dm_btree_value_type le64_type;
 
 	le64_type.context = NULL;
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 23904d233735..df445091384a 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -2564,7 +2564,7 @@ static struct r1conf *setup_conf(struct mddev *mddev)
 		    || disk_idx < 0)
 			continue;
 		if (test_bit(Replacement, &rdev->flags))
-			disk = conf->mirrors + conf->raid_disks + disk_idx;
+			disk = conf->mirrors + mddev->raid_disks + disk_idx;
 		else
 			disk = conf->mirrors + disk_idx;
 
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index a954c95d7c92..6137d0041d4d 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -476,7 +476,7 @@ static void raid10_end_write_request(struct bio *bio, int error)
 	 */
 	one_write_done(r10_bio);
 	if (dec_rdev)
-		rdev_dec_pending(conf->mirrors[dev].rdev, conf->mddev);
+		rdev_dec_pending(rdev, conf->mddev);
 }
 
 /*
@@ -612,20 +612,24 @@ static int raid10_mergeable_bvec(struct request_queue *q,
 		max = biovec->bv_len;
 
 	if (mddev->merge_check_needed) {
-		struct r10bio r10_bio;
+		struct {
+			struct r10bio r10_bio;
+			struct r10dev devs[conf->copies];
+		} on_stack;
+		struct r10bio *r10_bio = &on_stack.r10_bio;
 		int s;
-		r10_bio.sector = sector;
-		raid10_find_phys(conf, &r10_bio);
+		r10_bio->sector = sector;
+		raid10_find_phys(conf, r10_bio);
 		rcu_read_lock();
 		for (s = 0; s < conf->copies; s++) {
-			int disk = r10_bio.devs[s].devnum;
+			int disk = r10_bio->devs[s].devnum;
 			struct md_rdev *rdev = rcu_dereference(
 				conf->mirrors[disk].rdev);
 			if (rdev && !test_bit(Faulty, &rdev->flags)) {
 				struct request_queue *q =
 					bdev_get_queue(rdev->bdev);
 				if (q->merge_bvec_fn) {
-					bvm->bi_sector = r10_bio.devs[s].addr
+					bvm->bi_sector = r10_bio->devs[s].addr
 						+ rdev->data_offset;
 					bvm->bi_bdev = rdev->bdev;
 					max = min(max, q->merge_bvec_fn(
@@ -637,7 +641,7 @@ static int raid10_mergeable_bvec(struct request_queue *q,
 				struct request_queue *q =
 					bdev_get_queue(rdev->bdev);
 				if (q->merge_bvec_fn) {
-					bvm->bi_sector = r10_bio.devs[s].addr
+					bvm->bi_sector = r10_bio->devs[s].addr
 						+ rdev->data_offset;
 					bvm->bi_bdev = rdev->bdev;
 					max = min(max, q->merge_bvec_fn(
@@ -1178,18 +1182,21 @@ retry_write:
 			blocked_rdev = rrdev;
 			break;
 		}
+		if (rdev && (test_bit(Faulty, &rdev->flags)
+			     || test_bit(Unmerged, &rdev->flags)))
+			rdev = NULL;
 		if (rrdev && (test_bit(Faulty, &rrdev->flags)
 			      || test_bit(Unmerged, &rrdev->flags)))
 			rrdev = NULL;
 
 		r10_bio->devs[i].bio = NULL;
 		r10_bio->devs[i].repl_bio = NULL;
-		if (!rdev || test_bit(Faulty, &rdev->flags) ||
-		    test_bit(Unmerged, &rdev->flags)) {
+
+		if (!rdev && !rrdev) {
 			set_bit(R10BIO_Degraded, &r10_bio->state);
 			continue;
 		}
-		if (test_bit(WriteErrorSeen, &rdev->flags)) {
+		if (rdev && test_bit(WriteErrorSeen, &rdev->flags)) {
 			sector_t first_bad;
 			sector_t dev_sector = r10_bio->devs[i].addr;
 			int bad_sectors;
@@ -1231,8 +1238,10 @@ retry_write:
 					max_sectors = good_sectors;
 			}
 		}
-		r10_bio->devs[i].bio = bio;
-		atomic_inc(&rdev->nr_pending);
+		if (rdev) {
+			r10_bio->devs[i].bio = bio;
+			atomic_inc(&rdev->nr_pending);
+		}
 		if (rrdev) {
 			r10_bio->devs[i].repl_bio = bio;
 			atomic_inc(&rrdev->nr_pending);
@@ -1288,51 +1297,52 @@ retry_write:
 	for (i = 0; i < conf->copies; i++) {
 		struct bio *mbio;
 		int d = r10_bio->devs[i].devnum;
-		if (!r10_bio->devs[i].bio)
-			continue;
+		if (r10_bio->devs[i].bio) {
+			struct md_rdev *rdev = conf->mirrors[d].rdev;
+			mbio = bio_clone_mddev(bio, GFP_NOIO, mddev);
+			md_trim_bio(mbio, r10_bio->sector - bio->bi_sector,
+				    max_sectors);
+			r10_bio->devs[i].bio = mbio;
+
+			mbio->bi_sector	= (r10_bio->devs[i].addr+
+					   rdev->data_offset);
+			mbio->bi_bdev = rdev->bdev;
+			mbio->bi_end_io	= raid10_end_write_request;
+			mbio->bi_rw = WRITE | do_sync | do_fua;
+			mbio->bi_private = r10_bio;
 
-		mbio = bio_clone_mddev(bio, GFP_NOIO, mddev);
-		md_trim_bio(mbio, r10_bio->sector - bio->bi_sector,
-			    max_sectors);
-		r10_bio->devs[i].bio = mbio;
-
-		mbio->bi_sector	= (r10_bio->devs[i].addr+
-				   conf->mirrors[d].rdev->data_offset);
-		mbio->bi_bdev = conf->mirrors[d].rdev->bdev;
-		mbio->bi_end_io	= raid10_end_write_request;
-		mbio->bi_rw = WRITE | do_sync | do_fua;
-		mbio->bi_private = r10_bio;
-
-		atomic_inc(&r10_bio->remaining);
-		spin_lock_irqsave(&conf->device_lock, flags);
-		bio_list_add(&conf->pending_bio_list, mbio);
-		conf->pending_count++;
-		spin_unlock_irqrestore(&conf->device_lock, flags);
-
-		if (!r10_bio->devs[i].repl_bio)
-			continue;
-
-		mbio = bio_clone_mddev(bio, GFP_NOIO, mddev);
-		md_trim_bio(mbio, r10_bio->sector - bio->bi_sector,
-			    max_sectors);
-		r10_bio->devs[i].repl_bio = mbio;
+			atomic_inc(&r10_bio->remaining);
+			spin_lock_irqsave(&conf->device_lock, flags);
+			bio_list_add(&conf->pending_bio_list, mbio);
+			conf->pending_count++;
+			spin_unlock_irqrestore(&conf->device_lock, flags);
+		}
 
-		/* We are actively writing to the original device
-		 * so it cannot disappear, so the replacement cannot
-		 * become NULL here
-		 */
-		mbio->bi_sector	= (r10_bio->devs[i].addr+
-				   conf->mirrors[d].replacement->data_offset);
-		mbio->bi_bdev = conf->mirrors[d].replacement->bdev;
-		mbio->bi_end_io	= raid10_end_write_request;
-		mbio->bi_rw = WRITE | do_sync | do_fua;
-		mbio->bi_private = r10_bio;
+		if (r10_bio->devs[i].repl_bio) {
+			struct md_rdev *rdev = conf->mirrors[d].replacement;
+			if (rdev == NULL) {
+				/* Replacement just got moved to main 'rdev' */
+				smp_mb();
+				rdev = conf->mirrors[d].rdev;
+			}
+			mbio = bio_clone_mddev(bio, GFP_NOIO, mddev);
+			md_trim_bio(mbio, r10_bio->sector - bio->bi_sector,
+				    max_sectors);
+			r10_bio->devs[i].repl_bio = mbio;
+
+			mbio->bi_sector	= (r10_bio->devs[i].addr+
+					   rdev->data_offset);
+			mbio->bi_bdev = rdev->bdev;
+			mbio->bi_end_io	= raid10_end_write_request;
+			mbio->bi_rw = WRITE | do_sync | do_fua;
+			mbio->bi_private = r10_bio;
 
-		atomic_inc(&r10_bio->remaining);
-		spin_lock_irqsave(&conf->device_lock, flags);
-		bio_list_add(&conf->pending_bio_list, mbio);
-		conf->pending_count++;
-		spin_unlock_irqrestore(&conf->device_lock, flags);
+			atomic_inc(&r10_bio->remaining);
+			spin_lock_irqsave(&conf->device_lock, flags);
+			bio_list_add(&conf->pending_bio_list, mbio);
+			conf->pending_count++;
+			spin_unlock_irqrestore(&conf->device_lock, flags);
+		}
 	}
 
 	/* Don't remove the bias on 'remaining' (one_write_done) until
@@ -3015,7 +3025,7 @@ static sector_t sync_request(struct mddev *mddev, sector_t sector_nr,
 				else {
 					bad_sectors -= (sector - first_bad);
 					if (max_sync > bad_sectors)
-						max_sync = max_sync;
+						max_sync = bad_sectors;
 					continue;
 				}
 			}
diff --git a/drivers/md/raid10.h b/drivers/md/raid10.h
index 7c615613c381..24d45b8af5c9 100644
--- a/drivers/md/raid10.h
+++ b/drivers/md/raid10.h
@@ -104,7 +104,7 @@ struct r10bio {
 	 * We choose the number when they are allocated.
 	 * We sometimes need an extra bio to write to the replacement.
 	 */
-	struct {
+	struct r10dev {
 		struct bio	*bio;
 		union {
 			struct bio	*repl_bio; /* used for resync and
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 73a58007eb99..0240576564dc 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -380,6 +380,8 @@ static int calc_degraded(struct r5conf *conf)
 	degraded = 0;
 	for (i = 0; i < conf->previous_raid_disks; i++) {
 		struct md_rdev *rdev = rcu_dereference(conf->disks[i].rdev);
+		if (rdev && test_bit(Faulty, &rdev->flags))
+			rdev = rcu_dereference(conf->disks[i].replacement);
 		if (!rdev || test_bit(Faulty, &rdev->flags))
 			degraded++;
 		else if (test_bit(In_sync, &rdev->flags))
@@ -404,6 +406,8 @@ static int calc_degraded(struct r5conf *conf)
 	degraded2 = 0;
 	for (i = 0; i < conf->raid_disks; i++) {
 		struct md_rdev *rdev = rcu_dereference(conf->disks[i].rdev);
+		if (rdev && test_bit(Faulty, &rdev->flags))
+			rdev = rcu_dereference(conf->disks[i].replacement);
 		if (!rdev || test_bit(Faulty, &rdev->flags))
 			degraded2++;
 		else if (test_bit(In_sync, &rdev->flags))