One change was the addition of a new file system, the OVERLAY file system. The OVERLAY file system is functionally similar to the NULLFS[4] file system except that it does not export the directory hierarchy into another part of the file name space. It places itself between the file name space and the overlayed file system4. It is intended for certain specific circumstances, such as that of DMFS, where the layered file system has a strong need to block access to the underlying file system. DMFS does so to prevent metadata inconsistencies. One other variant of the OVERLAY file system might be a file system which attempts to validate executables with a checksum before permitting their execution. This file system would need to block access as that is very key to its security model. Unless there is a strong need for interposition, all future layered file systems should be based on the NULLFS layer rather than the OVERLAY layer as it simplifies the fcntl(2) interactions described in section 3.8.