OpenShift Cluster Upgrade breaks the cluster when upgrading ≤4.12 to 4.13+ with FIPS enabled or custom MachineConfigs are used.

Issue

• The OpenShift Upgrade fails with nodein NotReady,SchedulingDisabled status while doing upgrade from ≤ 4.12 to 4.13+ while FIPS enabled or custom MachineConfig are used for services like OpenSSL or OpenSSH.

# oc get nodes
NAME                                          STATUS              ROLES          AGE    VERSION
master0.ocp.example.com    NotReady,SchedulingDisabled            master         574d   v1.26.14+03ee898
master1.ocp.example.com    Ready                                  master         574d   v1.26.14+03ee898
master2.ocp.example.com    Ready                                  master         574d   v1.26.14+03ee898

• The kubelet service fails to start on NotReady node with error FIPS mode is enabled, but the required OpenSSL backend is unavailable as,

Apr 15 07:20:15 master0.ocp.example.com kubenswrapper[2692]: FIPS mode is enabled, but the required OpenSSL backend is unavailable
Apr 15 07:20:15 master0.ocp.example.com systemd[1]: kubelet.service: Main process exited, code=exited, status=1/FAILURE
Apr 15 07:20:15 master0.ocp.example.com systemd[1]: kubelet.service: Failed with result 'exit-code'

• The sshd service fails to start with error Bad configuration option: CRYPTO_POLICY as,

master0.ocp.example.com systemd[1]: Stopped OpenSSH server daemon.
master0.ocp.example.com systemd[1]: Starting OpenSSH server daemon...
master0.ocp.example.com sshd[397714]: 
master0.ocp.example.com sshd[401448]: /etc/crypto-policies/back-ends/opensshserver.config: line 1: Bad configuration option: CRYPTO_POLICY
master0.ocp.example.com sshd[401448]: /etc/crypto-policies/back-ends/opensshserver.config: terminating, 1 bad configuration options
master0.ocp.example.com systemd[1]: sshd.service: Main proces