Lindsay Haisley posted on Sun, 20 Mar 2011 13:48:43 -0500 as excerpted:

> On Sun, 2011-03-20 at 11:50 -0500, Lindsay Haisley wrote:
>> This problem box has been running Gentoo since 2004, but my work and
>> priorities on my time have changed over the years and it's probably
>> time to move on.
> 
> To be a bit more precise here, I'm actively working on replacing this
> box, and will be running a Linux distribution on it other than Gentoo.
> But in the meantime, I do need to solve the problem about which I
> posted.

I appreciate that you /have/ decided to ultimately switch, and obviously 
believe it best or I'd have not recommend it.

I also appreciate that you need a solution for now.  That's what I suspect 
I may have provided (at least to a point) further down that post, which 
you may not have read given the order.

I was hoping to read all these replies and have you tell me whether I was 
on the right track or not, answering the question about initrd/initramfs 
so we could go from there if need be.  However, I don't see that answer, 
which suggests you didn't read that far down my post, for which I can't 
really blame you.

But you can always go back and do it now... =:^)

Briefly repeating so you can see if it's worth your trouble going back to 
get the details.

If the kernel could not load its initial root, be that initr* or the real-
root, it'd fail to load any userspace at all as it couldn't get to it, 
panicking instead when it couldn't get to a root at all.  (I know this 
from experience.)

Thus, that it's getting even a /limited/ userspace indicates it's getting 
to it's initial root.  That's a very significant point.

The question from there is whether that limited userspace is loading from 
the initrd/initramfs if you have one, indicating it's not successfully 
doing the initial real-root mount and pivot-root from the initr*, taking 
you down one path toward a solution (missing or incorrect drivers in the 
initr* or wrong initrd), or whether it's loading the real-root (as just 
getting to userspace at all indicates if you don't have an initr*), thus 
indicating that it has the necessary drivers (both pata and fs) to do so, 
and the problem is later, with the fsck or remount, taking you down a 
different path toward a solution (userspace issue).

If that sounds useful, check the earlier post for more.  If it doesn't, 
don't bother, as that's what I was detailing in the earlier post.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman