torusblk aoe hangs while stopping the process

[nak3]

version:

• 100add2

steps to reproduce

1. run torusblk aoe test lo 1 1 and stop the process by Ctrl+C

[root@torus1 torus]# ./bin/torusblk aoe test eth0 1 1
2016-11-30 13:37:42.185671 I | torus: creating metadata service: etcd
2016-11-30 13:37:42.188524 N | ring: Using ring that requests replication level 2, but has only 0 peers. Add nodes to match replication.
2016-11-30 13:37:42.188765 I | torus: creating blockstore: temp
2016-11-30 13:37:42.189618 N | ring: Using ring that requests replication level 2, but has only 0 peers. Add nodes to match replication.
Attached to AoE device (eth0, Major: 1, Minor: 1). Server loop begins ...
^C
Received an interrupt, stopping services...

additional info

• loopback device can work well. So, I thought that torus suppoorts loopback device only, but according to aoe: tiny: update command help to use loopback device #387, this be fixed to work well.

[root@torus1 torus]# ./bin/torusblk aoe test lo 1 1
2016-11-30 13:37:35.953565 I | torus: creating metadata service: etcd
2016-11-30 13:37:35.955117 N | ring: Using ring that requests replication level 2, but has only 0 peers. Add nodes to match replication.
2016-11-30 13:37:35.955290 I | torus: creating blockstore: temp
2016-11-30 13:37:35.955759 N | ring: Using ring that requests replication level 2, but has only 0 peers. Add nodes to match replication.
Attached to AoE device (lo, Major: 1, Minor: 1). Server loop begins ...
^C
Received an interrupt, stopping services...
2016-11-30 13:37:38.341962 E | aoe: failed to handle frame: bad file descriptor
2016-11-30 13:37:38.342128 E | aoe: ReadFrom failed: bad file descriptor