RDMA/mlx5: Order num_pending_prefetch properly with synchronize_srcu

During destroy setting live = 0 and then synchronize_srcu() prevents
num_pending_prefetch from incrementing, and also, ensures that all work
holding that count is queued on the WQ. Testing before causes races of the
form:

    CPU0                                         CPU1
  dereg_mr()
                                          mlx5_ib_advise_mr_prefetch()
            				   srcu_read_lock()
                                            num_pending_prefetch_inc()
					      if (!live)
   live = 0
   atomic_read() == 0
     // skip flush_workqueue()
                                              atomic_inc()
 					      queue_work();
            				   srcu_read_unlock()
   WARN_ON(atomic_read())  // Fails

Swap the order so that the synchronize_srcu() prevents this.

Fixes: a6bc387 ("IB/mlx5: Protect against prefetch of invalid MR")
Link: https://lore.kernel.org/r/20191001153821.23621-5-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>

drivers/infiniband/hw/mlx5/mr.c

@@ -1609,13 +1609,14 @@ static void dereg_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
 		 */
 		mr->live = 0;
 
+		/* Wait for all running page-fault handlers to finish. */
+		synchronize_srcu(&dev->mr_srcu);
+
 		/* dequeue pending prefetch requests for the mr */
 		if (atomic_read(&mr->num_pending_prefetch))
 			flush_workqueue(system_unbound_wq);
 		WARN_ON(atomic_read(&mr->num_pending_prefetch));
 
-		/* Wait for all running page-fault handlers to finish. */
-		synchronize_srcu(&dev->mr_srcu);
 		/* Destroy all page mappings */
 		if (!umem_odp->is_implicit_odp)
 			mlx5_ib_invalidate_range(umem_odp,