podman machine rm for QEMU provider removed connections before confirmation

[arixmkii]

Issue Description

After calling podman machine rm and pressing n to abort, the machine is not functional as connections are erased. Confirmed for QEMU on Windows and macOS.

The code part causing this is https://github.com/containers/podman/blob/main/pkg/machine/qemu/machine.go#L960 which is immediate, but should go inside callback https://github.com/containers/podman/blob/main/pkg/machine/qemu/machine.go#L992 (like in WSL implementation). WSL implementation is not affected.

Steps to reproduce the issue

Steps to reproduce the issue

1. podman machine init
2. podman machine rm
3. press n
4. podman system connection list (no connections)
5. podman machine list (machine still exists)

Describe the results you received

Connections are erased.

Describe the results you expected

Connections persisted like in WSL case on Widnows

podman info output

Intentionally skipped as machine was already ruined.

Podman in a container

No

Privileged Or Rootless

None

Upstream Latest Release

Yes

Additional environment details

On Windows was tested with built from latest main + QEMU machine supporting changes on top.

Additional information

No response

[baude]

this is a legit bug ... if anyone on the team can get to this, it'd be great ... the bug is also in the hyperv implementation.

[Luap99]

@baude I can take it, I see that it is broken in qemu but it looks correct in hyperV.