neonvm-controller: drop generated default value

We use kubebuilder:default annotation for scaling mode fields,
which effiectively means we never allow it to be non set and that
bypasses the logic with --default-cpu-scaling-mode controller argument

neonvm-controller/deployment.yaml

@@ -55,6 +55,7 @@ spec:
         - "--concurrency-limit=128"
         - "--skip-update-validation-for="
         - "--disable-runner-cgroup"
+        - "--default-cpu-scaling-mode=QmpScaling"
         # See #775 and its links.
         # * cache.writeback=on - don't set O_DSYNC (don't flush every write)
         # * cache.direct=on    - use O_DIRECT (don't abuse host's page cache!)

neonvm/apis/neonvm/v1/virtualmachine_types.go

@@ -169,7 +169,6 @@ type VirtualMachineSpec struct {
 	TargetRevision *RevisionWithTime `json:"targetRevision,omitempty"`
 
 	// Controls how CPU scaling is performed, either hotplug new CPUs with QMP, or enable them in sysfs.
-	// +kubebuilder:default:=QmpScaling
 	// +optional
 	CpuScalingMode *CpuScalingMode `json:"cpuScalingMode,omitempty"`
 

neonvm/config/crd/bases/vm.neon.tech_virtualmachines.yaml

@@ -998,7 +998,6 @@ spec:
                     type: object
                 type: object
               cpuScalingMode:
-                default: QmpScaling
                 description: Controls how CPU scaling is performed, either hotplug
                   new CPUs with QMP, or enable them in sysfs.
                 enum: