wait for pending after-state-change work and similar before attaching

Otherwise, if we do a very quick detach/attach, and the
after-state-change send_state() of the detach got delayed for some
reason, it may be processed after the new diskless->attaching state
change, probably confusing the peer.

Paranoia fix, not yet observed in real life.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>

drivers/block/drbd/drbd_nl.c

@@ -919,6 +919,8 @@ static int drbd_nl_disk_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp
 	drbd_suspend_io(mdev);
 	/* also wait for the last barrier ack. */
 	wait_event(mdev->misc_wait, !atomic_read(&mdev->ap_pending_cnt));
+	/* and for any other previously queued work */
+	drbd_flush_workqueue(mdev);
 
 	retcode = _drbd_request_state(mdev, NS(disk, D_ATTACHING), CS_VERBOSE);
 	drbd_resume_io(mdev);