Hroller used to first compute a coloring of the node graph and then
filter out the nodes that it had to work on. While the only filtering
was according to node groups this did not make a difference, as there
shouldn't be any instance with primary and secondary node on different
node groups. With more elaborate filtering, however, reducing the graph
first can lead to better reboot groups.

Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>