This cache is used to avoid mapping and unmapping buffer objects
unnecessarily. Mappings are cached per client and stay hot until
the buffer object is destroyed.
Signed-off-by: NThierry Reding <treding@nvidia.com>