On Mon, Feb 20, 2023 at 7:17 PM Andrea Bolognani <abologna@redhat.com> wrote:
On Mon, Feb 20, 2023 at 05:12:53PM +0100, Kristina Hanicova wrote:
> v1 here:
> https://listman.redhat.com/archives/libvir-list/2023-February/237622.html
>
> diff to v1:
> * reduced test files (thanks Andrea)
> * removed redundant check for address type (noticed by Peter)
> * plugging the device directly into pcie.0 if its address was not
>   specified

Just for completeness' sake, note that the use of
VIR_PCI_CONNECT_INTEGRATED results in libvirt outright rejecting
attempts to put the device anywhere but on pcie.0, even when the
address comes directly from the user. This is technically a
limitation compared to what QEMU allows, but in practice the other
configurations are untested and overall it feels like a fair
trade-off. We can also decide to lift this limitation in the future,
if it ever comes to that.

> Kristina Hanicova (6):
>   qemu: introduce QEMU_CAPS_DEVICE_PANIC_PCI
>   conf: add panic model 'pvpanic'
>   tests: add test cases for device pvpanic-pci
>   qemu: assign PCI address to device pvpanic-pci
>   tests: add case for pvpanic-pci without address
>   docs: document panic device 'pvpanic-pci'

Everything looks great, so

  Reviewed-by: Andrea Bolognani <abologna@redhat.com>

The first patch is missing the Signed-off-by tag. Can you please
confirm that you're okay with me adding it before pushing?

Yes, thank you.
 

I'm also not seeing any updates to the NEWS file. Please post that as
a follow-up patch before release :)


Thanks for the reminder, the patch updating NEWS is on the list now.


Kristina