Re: [coLinux-devel] Initrd support?

[<ch...@to...>]

did you create a new kernel with initrd support?

did you read /usr/src/Documentation/initrd.txt
I think root=/dev/cobd0 init=/linuxrc
needs the explicit change root commands at the end of linuxrc
where as intrd=  can use a linuxrc that will automaticly pivot root at the
end.

if you checked out all this you could privatly e-mail me your linuxrc and
.config for your kernel and I can look it over. I could also do some
further experiments with what I have tomorow. I think I stopped at the
same place you are due to not having the commands needed on my initrd and
realizing that I had stripped my linuxrc down to almost nothing and I came
up with another way to do it.

chris

> initrd= doesn't work for coLinux because it can't find the file. This is
> fine, as I should be able to simulate the same result by using
> root=/dev/cobd0 init=/linuxrc, which works. The problem comes when the
> pivot_root call is made..
>
> EXT3-fs: mounted filesystem with ordered data mode.
> Kernel panic: Attempted to kill init!
>  daemon: monitor terminated, reason 3
> switch_message: freed message a047010 (2 to 4)
> daemon: module disconnected: console
> colinux: shutting down
> daemon: daemon cleanup
> Removing kernel driver
> Stopping driver service
> Removing driver service
>
> Unfortunately, this also makes the Console go completely blank on me, so
> it's hard to see if there were any other problems.
>
> If anyone could give some assistance, that would be great.
>
> --
> Eric Andresen
> Junior Systems Administrator
> Mars Space Flight Facility
> Arizona State University
> ea...@ma...
>
> On Wed, 2004-03-03 at 05:49, ch...@to... wrote:
>> see initrd.txt in a 2.4.24 kernel source documentation.
>> the newer way of specifying initrd with the explicit change_root may
>> work.
>> you will need to compile your own kernel at this time as the coLinux
>> binary kernel does not support initrd. Just apply the coLinux patch and
>> make shure you have all the initrd related options selected in
>> .configure
>> .
>>
>> I was experimenting with an initrd based setup that I was using but
>> after
>> eleminating the stuff I didn't need for coLinux mode from linuxrc I
>> realized that I could just skip the initrd and boot directly to the root
>> image.
>>
>> initrd supplying hardware modules should not be necessary because the
>> "hardware enviroment" is fixed and taken care of by the coLinux patch.
>> the initrd mounting loops may also be unnecessary because you could just
>> set that all up in the coLinux config and fstab.
>>
>> so It may be worth considering do you really need the initrd.
>>
>> chris
>>
>> > Hi,
>> >
>> > I'm trying to essentially port my current UML configuration over to a
>> > similar coLinux setup, but the first obvious problem that comes to
>> mind
>> > is that there does not appear to be a way to specify initrds.
>> >
>> > Am I mistaken, or is this so? This is a requirement for my setup,
>> > because the initrd creates a number of tmpfs filesystems for /etc,
>> /var,
>> > and /tmp (for specific reasons outside of the realm of this topic).
>> >
>> > Also, regarding the memory restrictions: is there no other possible
>> way
>> > to allocate physical ram? I find it hard to think that no Windows
>> > application surpasses this memory usage (VMware? Most games?); is
>> there
>> > a chance that one of these alternate memory allocation sources could
>> be
>> > used, thus allowing more realistic physical memory usages?
>> >
>> > Thanks,
>> > -- Eric Andresen
>> >    Junior Systems Administrator
>> >    Mars Space Flight Facility
>> >    Arizona State University
>> >    ea...@ma...
>> >
>
>