can't find loop device when create pod

[yanxuean]

Signed-off-by: yanxuean yan.xuean@zte.com.cn

It can't find loop device when we create loop device after deploying virtlet.

virtlet v1.1.2

pod spec:

[cloud@cloud-test1:/root/virtlet/yaml/yan]$ vi cirros-vm-raw-device.yaml
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: extraRuntime
            operator: In
            values:
            - virtlet
  containers:
  - name: cirros-vm
    imagePullPolicy: IfNotPresent
    #image: virtlet.cloud/myimage/centos-7-x86_64-genericcloud-1805.qcow2
    #image: virtlet.cloud/myimage/xenial-server-cloudimg-amd64-disk1.img
    image: virtlet.cloud/myimage/bionic-server-cloudimg-amd64.img
    #image: virtlet.cloud/cirros
    # tty and stdin required for `kubectl attach -t` to work
    tty: true
    stdin: true
    volumeMounts:
    - name: rawvol
      mountPath: /rawvol
    - name: somevol
      mountPath: /somevol
  volumes:
  - name: rawvol
    flexVolume:
      driver: "virtlet/flexvolume_driver"
      options:
        type: raw
        # this assumes that some file is associated with /dev/loop0 on
        # the virtlet node using losetup
        path: /dev/loop0
        part: "0"

kubectl describe:

Events:
  Type     Reason                 Age               From                  Message
  ----     ------                 ----              ----                  -------
  Normal   Scheduled              20s               default-scheduler     Successfully assigned cirros-vm-raw to cloud-test2
  Normal   SuccessfulMountVolume  20s               kubelet, cloud-test2  MountVolume.SetUp succeeded for volume "default-token-jfk6n"
  Normal   SuccessfulMountVolume  19s               kubelet, cloud-test2  MountVolume.SetUp succeeded for volume "somevol"
  Normal   SuccessfulMountVolume  19s               kubelet, cloud-test2  MountVolume.SetUp succeeded for volume "rawvol"
  Normal   Pulled                 5s (x3 over 19s)  kubelet, cloud-test2  Container image "virtlet.cloud/myimage/bionic-server-cloudimg-amd64.img" already present on machine
  Warning  Failed                 4s (x3 over 18s)  kubelet, cloud-test2  Error: "/run/virtlet.sock": rpc error: code = 5 desc = stat /dev/loop0: no such file or directory

in node:

[root@cloud-test2:~/virtlet/loopdev]$ losetup -l
NAME       SIZELIMIT OFFSET AUTOCLEAR RO BACK-FILE
/dev/loop0         0      0         0  0 /root/virtlet/loopdev/testloop0
[root@cloud-test2:~/virtlet/loopdev]$ ll /dev/loop*
brw-rw---- 1 root disk  7,   0 Jul 27 17:02 /dev/loop0
crw-rw---- 1 root disk 10, 237 Jul 27 17:02 /dev/loop-control

---

This change is