DEVELOPER QUESTIONS:

Q. Why maintain both the VFS' inodes for the pseudo fs in addition to the inode
   information maintained in the store for the filesystem?

A. A normal filesystem has its own device where it maintains inode information
   "persistently".  The overlay filesystem does not make use of such a device
   (at least not at this time).  Therefore, the inode information is maintained
   in memory until the super block is written out.  So, why not just store the
   VFS' inodes?  Because they belong to the VFS who decides when and where to
   put those inodes, and it would not be a good idea to hold on to the VFS'
   inodes the entire time the overlay filesystem is mounted.

   Holding on to the VFS' inodes would cause the VFS to slow down every time it
   needs to read through its inode cache to find an inode or to obtain an un-
   used inode.  In addition, there may be changes to the way in which the VFS
   maintains its inodes in the future, and it is desireable to keep this
   filesystem's implementation independent of the VFS as much as possible so
   that it may be used with those future implementations without problem.
