Lukáš Doktor

Yes, @pevogam, the problem is that aexpect was strictly specialized tool with no magic (almost). The `set_output_func` and `set_output_params` functions were IMO a bit redundant as the caller could just...

Officially we don't support it, but we do want to test it to uncover bugs. Also note that src and dst doesn't necessarily mean older and newer qemu. In my...

@luckyh I talked to Dave and as a matter of fact we do support migrating there and back as well. So this actually adds coverage that is currently missing in...

Changes: * fixed the check for `None` (both `None` and `False` should fail) Not changed: * the timeout as param/variable. I am open to turning it into a class variable...

Hello @skycastlelily2 was it necessary to create a new pypi package? I thought the existing `passfd` would do, otherwise I don't think it's worth supporting it at all as it's...

@clebergnu @wainersm would anyone of you please review this? It's been a long time and I believe it's ready to go...

Hi @cevich, yes, `qemu_devices` is fully pickable and I'd like to see all related code inside of it. The problem is that I still have a lot to cover in...

@clebergnu would you please take a look at this?

Hello @clebergnu one think I don't understand is the x86 variant. This is a power specific test focusing on existing BZ. That is why I created this test I don't...

As for the number of migration, it doesn't matter how many. It's as many as possible but at least one (before and after).