md-cluster: call md_kick_rdev_from_array once ack failed

GuoqingJiang · shligit · commit e566aef12a16 · 2016-09-21T09:09:44.000-07:00
The new_disk_ack could return failure if WAITING_FOR_NEWDISK
is not set, so we need to kick the dev from array in case
failure happened.

And we missed to check err before call new_disk_ack othwise
we could kick a rdev which isn't in array, thanks for the
reminder from Shaohua.

Reviewed-by: NeilBrown &lt;neilb@suse.com&gt;
Signed-off-by: Guoqing Jiang &lt;gqjiang@suse.com&gt;
Signed-off-by: Shaohua Li &lt;shli@fb.com&gt;
diff --git a/drivers/md/md.c b/drivers/md/md.c
@@ -6101,9 +6101,14 @@ static int add_new_disk(struct mddev *mddev, mdu_disk_info_t *info)
 			export_rdev(rdev);
 
 		if (mddev_is_clustered(mddev)) {
-			if (info->state & (1 << MD_DISK_CANDIDATE))
-				md_cluster_ops->new_disk_ack(mddev, (err == 0));
-			else {
+			if (info->state & (1 << MD_DISK_CANDIDATE)) {
+				if (!err) {
+					err = md_cluster_ops->new_disk_ack(mddev,
+						err == 0);
+					if (err)
+						md_kick_rdev_from_array(rdev);
+				}
+			} else {
 				if (err)
 					md_cluster_ops->add_new_disk_cancel(mddev);
 				else
