md: Fix failed allocation of md_register_thread

commit e406f12 upstream.

mddev->sync_thread can be set to NULL on kzalloc failure downstream.
The patch checks for such a scenario and frees allocated resources.

Committer node:

Added similar fix to raid5.c, as suggested by Guoqing.

Cc: stable@vger.kernel.org # v3.16+
Acked-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Aditya Pakki <pakki001@umn.edu>
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers/md/raid10.c

@@ -3755,6 +3755,8 @@ static int run(struct mddev *mddev)
 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
 							"reshape");
+		if (!mddev->sync_thread)
+			goto out_free_conf;
 	}
 
 	return 0;

drivers/md/raid5.c

@@ -6973,6 +6973,8 @@ static int run(struct mddev *mddev)
 		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
 		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
 							"reshape");
+		if (!mddev->sync_thread)
+			goto abort;
 	}
 
 	/* Ok, everything is just fine now */