I would like to get an answer for a question about kernel related
filesystems mounting, before and after initramfs's treatment. Before
calling udev's "stuff" within my initramfs, I have to mount kernel
filesystems on my temporary system : /proc, /sys and /dev. Quite normal.
Then I call udev, it makes its job and we can go further. It's at this
point I have a doubt.
When I proceed with the 'pivot_root' with run-init to mount the
'real' root filesystem, current /proc, /sys and /dev mounts are
theorically "lost". During the 'real' boot process, I have to remount
these filesystems on the new root. I don't think it's a problem for
/proc, but what about /sys and /dev ? Particulary, /dev is of tmpfs's
type, it only remains in memory and is lost at unmounting time. So
devices nodes created by udev from my initramfs may be lost too.
Maybe run-init makes a kind of special job about this, but I would
greatly appreciate some precisions. If not, how should I proceed to have
the desired behavior ?
Thanks for help
Mickaël
Dan Nicholson a écrit :
> On 3/20/07, MickB <[EMAIL PROTECTED]> wrote:
>
>> Clear and precise : thank you very much ! I'll try this as soon as
>> possible, I'll give you my feedback soon :)
>>
>
> One other thing I forgot. The generated init script for the initramfs
> will use sh and ps, but the ones from klibc are called sh.shared and
> minips. So, I move them to the names I want since I don't install any
> other utilities of the same name:
>
> # use klibc sh and ps as defaults
> mv -v ${PKG_DEST}/usr/lib/klibc/bin/sh{.shared,}
> mv -v ${PKG_DEST}/usr/lib/klibc/bin/{mini,}ps
>
> --
> Dan
>
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
