On Thu, Jul 21, 2016 at 08:38:17PM +0200, Tomasz Flendrich wrote:
[…]
But the problem is that you are not checking any assignment. The code
is not executed at all. That's because you specify 'target dev=' from
which we're trying to guess the right address.
Because
There's a plan to rework the address handling, so testcases
that verify hotplugging ccw devices will help in avoiding
regression.
In this commit, some device files are duplicated because
of the way qemuhotplug.c calculates the expected xml filenames.
I plan on changing that to explicitly stating
>> […]
>
> But the problem is that you are not checking any assignment. The code
> is not executed at all. That's because you specify 'target dev=' from
> which we're trying to guess the right address.
>
> Because virDomainDeviceDefPostParseInternal() is ran for the device
> definition and it
On Mon, Jul 18, 2016 at 08:25:40PM +0200, Tomasz Flendrich wrote:
Thank you for your reply!
I know you weren't aiming for that, but this got me thinking, are we
checking that unplug works with not fully specified XML? I mean that
when unplugging, you don't need to specify everything.
Thank you for your reply!
> I know you weren't aiming for that, but this got me thinking, are we
> checking that unplug works with not fully specified XML? I mean that
> when unplugging, you don't need to specify everything. Whatever
> uniquely identifies the device should be enough. So for
On Tue, Jul 12, 2016 at 03:08:22PM +0200, Tomasz Flendrich wrote:
These are a few simple and one complex testcases.
The simple ones test attaching and detaching ccw devices
with both implicitly and explicitly stated addresses.
In the complex one, attaching and detaching a device
should make the
These are a few simple and one complex testcases.
The simple ones test attaching and detaching ccw devices
with both implicitly and explicitly stated addresses.
In the complex one, attaching and detaching a device
should make the address free to reuse.
I have plan to rework the address handling,