Add L2 cache and make it the next level of cache for each of the CPUs.
Signed-off-by: Joseph Lo <josephl@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>