rgmanager segfaults with central_processing enabled in RHEL 6

Issue

• Managing a service with clusvcadm while central_processing is enabled leads to an rgmanager segfault
• When building a VM resource on a different node than the current RG-Master, the final clusvcadm -d and subsequent clusvcadm -e of the new ressource segfaults the rgmanager, leading to an immediate fencing of the node. When running the same procedure on the current RG-Manager, everything works as expected.
• A cluster node crashes and gets fenced while enabling, disabling, relocating, restarting, or migrating a service withi clusvcadm and central_processing enabled.