Cover _interface_name_to_common_name uncovered branch by writing a test that specifies a non-existant block device

[mulkieran]

I believe that this should be quite straightforward.

[mulkieran]

It turns out that we can not do this, because the CLI currently has no action that specifies a block device once it is added to a pool. So, we'll have to justify the no-cover pragma, instead, by explaining this fact.

Eventually, we may have a stratis-cli command that does specify a blockdev, like one to set the user info, see issue #922, and then we could take the pragma off. But now, we have nothing.