A new API for mounting filesystems

> It might look something like:
>
> SETDEV\0/dev/sda1
> SETOPTS\0seclabel\0data=ordered
> ...

I"m not the world's biggest fan of configfs, but this appears to be duplicating exactly that functionality.

> One attendee pointed out that would mean that some other mechanism would need to be used to mount /proc or /sys.

Clearly a non-issue given:

> Szeredi said that the existing mount() would not be going away.

Both md and dm already have mechanisms for "build some kernel object using these storage devices". While I don't claim either of them represent perfection, but it would seem sensible to use existing practice as the starting point for a discussion, rather than creating something totally different.