Improvements to gvinum and it's future

Hi,

I've been working this summer to improve gvinum, and add features from old vinum
into the geom version of vinum. Lukas Ertl have done a fantastic amount of work
on it (and Chris Jones) since it started, but it's not complete. So far, I've
completed many of the missing commands such as attach/detach, setstate stop, and 
other commands (The manpage lists most of them). I've also implemented basic
functionality of drivegroups, which provides grouping of drives with the same
configuration, and sort of acts as a drivepool that abstract drives further, and
it can simplify volume management for larger systems.

However, although many of these features maybe needs more
testing, some of them also belongs in HEAD, but since I am not a committer, I
need someone to help me get these changes into the tree. The reason for this is
largely that le_at_ has not been visible for a couple of months, so the one person
actually maintaining gvinum is not available. So therefore I ask for someone
else to help me with this.

Now, one could ask why I bother doing work on gvinum now, since we have gmirror,
gconcat, graid3 and all that. The reason is that gvinum is important as a volume
manager. Ivan Voras' work on gvirstor seems very promising as a foundation of a
new volume manager, and I've been planning to start working on utilities (fvm,
freebsd volume manager) that utilize these new geom classes instead of having to
maintain a separate RAID implementations, but that will take time, and meanwhile
we need gvinum up to speed. Perhaps can a lot of the concepts of gvinum be
reused in a better volume manager for FreeBSD. At least this is how I see the
future.

My first set of patches are available in PR 99183, 99186 and more is coming from
the perforce repository as soon as I'm able to split them up. All suggestions
to me are welcome! I posted this on -current and -arch, and I'm sorry if this
bothers some people, but it's not just a post of my patches to current and to
get attention of some committers. It is also an appeal to further discussion on
this subject.

-- 
Ulf Lilleengen