I'm not sure that addresses the original poster's question.
There are two "union-y" thingies. The MNT_UNION option to the mount(2) system call (or "-o union" to mount(8)) allows a new filesystem to be grafted into the namespace and at the top level of that new mountpoint, files/directories in the underlying mountpoint can show through if not overriden. But once you "cd" to a directory in the lower level, you are fully in the lower filesystem again, and can't access the upper filesystem.
unionfs allows deep integration of two filesystem hierarchies to create a new and composite hierarchy. When you "cd" into a subdirectory, you continue to be in the virtual unionfs filesystem. When you try to make a new file, the upper filesystem is modified, no matter how deep in the composite hierarchy you are. That is the functionality that is not supported.
Do not post admin requests to the list. They will be ignored.
Filesystem-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
|