Coredns svc always using ipv4 even in dualStack

[manuelbuil]

Environmental Info:
RKE2 Version:

Node(s) CPU architecture, OS, and Version:

Cluster Configuration:

Describe the bug:

Coredns service is not changing to dualStack when rke2 is deployed in dualStack mode:

NAMESPACE       NAME                        TYPE        CLUSTER-IP             EXTERNAL-IP   PORT(S)          AGE
calico-system   calico-typha                ClusterIP   2001:cafe:42:1::e555   <none>        5473/TCP         2m21s
cattle-system   cattle-cluster-agent        ClusterIP   2001:cafe:42:1::b38e   <none>        80/TCP,443/TCP   2m55s
default         kubernetes                  ClusterIP   2001:cafe:42:1::1      <none>        443/TCP          3m39s
kube-system     rke2-coredns-rke2-coredns   ClusterIP   10.43.0.10             <none>        53/UDP,53/TCP    3m13s
twistlock       defender                    ClusterIP   2001:cafe:42:1::fef3   <none>        443/TCP          2m36s

Steps To Reproduce:

• Installed RKE2:
  Deploy rke2 in dualStack mode and check the coredns service

Expected behavior:

coredns service follows along and has ipv4 and ipv6 addresses

Actual behavior:

Additional context / logs:

[manuelbuil]

This issue should be divided into three:
1 - coredns not picking ipv6 address when ipv6 is preferred in dual-stack mode
2 - k3s/rke2 not taking IPFamilies order into account when setting dual-stack mode
3 - services in general not setting ipFamilyPolicy: PreferDualStack when dual-stack mode is set

[manuelbuil]

The three issues that were required are now closed, so we can close this one