@Saucerhead Tharpe
As I remember it, it was the JFS code that IBM wrote for AIX 3.1 that was the main issue. SCO were arguing that this was a derived wok from the AT&T source, although it was more like an evolution of the BSD4.4 filesystem, complete with distributed Inodes and block bitmap for the free space. The whole concept of derivative works currently appears to be sparking controversy with GPL2, which just shows how convoluted US Copyright law is. See http://www.theregister.co.uk/2009/10/15/black_duck_gpl_web_conference_copenhaver_radcliffe/
All the LVM and hooks to extend filesystems were as far as I am aware IBM innovations that were actually new code unique to AIX.
Subsequently, IBM contributed the original LVM and JFS code to the Open Software Foundation, where it became used by several different vendors who implemented OSF/1 (although I don't believe that DEC used it in their implementation, they preferred their own Advanced File System [not to be confused with the Andrew File System]). This did not cause a problem, because all OSF members were already UNIX source licensees.
It is obviously now available for Linux, which caused the whole furore.
If anybody sees my copy of the Lyons annotated V6 UNIX code, I would be grateful to hear, as it appears to have fallen out of my coat pocket whilst in the cloakroom.