please see PR 59688 (re: [solved] Re: kernel hangs during boot)

[Doug White]

On Mon, 8 Dec 2003 pseniura at techie.com wrote:

>
> I filed a PR several weeks ago on this 'device-md'-in-the-kernel bug.
> The PR has yet to be assigned.
> Now it's gone out to 5.2-Release, I bet everyone is going to see this bug.

Enabling MD_ROOT_SIZE specifies that a MDROOT has been compiled into the
kernel, and gives its size so the volume will be appropriately
initialized.  So specifying this option in the beginning was a mistake.
There is absolutely no need for MD_ROOT_SIZE in any production kernel,
unless you are booting diskless or some other special need.

I think we have to blame PEBKAC on this although a word of caution in
NOTES is probably warranted.

>
>
> ________
>
>
> On Sun, 7 Dec 2003, Doug White wrote:
>
> > On Sat, 6 Dec 2003, Wesley Morgan wrote:
> >
> > > > Eh, this is you hitting the gdb key.  how about 'ps'?
> > >
> > > Hooray for serial consoles!
> >
> > You didn't compile with BREAK_TO_DEBUGGER? you still hit the ddb key :)
> >
> > I'm still not seeing anything hanging. Have you tried booting with -v and
> > seeing if ata is trying anything, for instance?
>
> Ok, I've found the culprit:
>
> options        MD_ROOT_SIZE=10
>
> Removing that from my kernel config allows it to boot (this is
> some cruft left over from previous experiments). As for WHY it won't boot,
> I'll leave that to those with much more knowledge than me!
>
> --
> Hi! I'm a .signature virus! Copy me into your ~/.signature to help me spread!
>
> ________
>

-- 
Doug White                    |  FreeBSD: The Power to Serve
dwhite at gumbysoft.com          |  www.FreeBSD.org