Re: "Warning: unable to open an initial console."
On Thu, 14 Aug 2008 11:20:02 +0200 (CEST) "C.J.Wagenius" <cjw@slackware.se> wrote:
_THIS IS JUST A GUESS_ but I think this is why I don't receive $SERVICES[] loading status messages at boot time [...]
That's right...
When your rootfs isn't live (up and running), do you got the console device in /dev? If not... See to it that the rootfs isn't live and create it with "mknod console c 5 1 && chmod og-r console".
I'm afraid I don't follow you. But I think I've figured out what happened. When I moved the partitions I didn't touch /dev /proc /sys (and directories like those ones) because I thought it wasn't neded. But, for the system to boot up properly, there have to be some dev files in /dev before devfs is mounted (console and stuff like that). Is that correct? Thanks for the pointer. Just a couple of question: * What do you mean when you say "When your rootfs isn't live (up and running)"? * Is there a way to check if those dev files exist without booting from another partition? Thanks in advance. Regards.
ncrfgs@tin.it wrote:
On Thu, 14 Aug 2008 11:20:02 +0200 (CEST) "C.J.Wagenius" <cjw@slackware.se> wrote:
_THIS IS JUST A GUESS_ but I think this is why I don't receive $SERVICES[] loading status messages at boot time [...] That's right...
When your rootfs isn't live (up and running), do you got the console device in /dev? If not... See to it that the rootfs isn't live and create it with "mknod console c 5 1 && chmod og-r console".
When I moved the partitions I didn't touch /dev /proc /sys (and directories like those ones) because I thought it wasn't neded.
only some files in /dev are needed (i guess null and console), they're needed before udev starts and /dev is handled by it... you can watch in the filesystem port to see how to create them with mknod... http://crux.nu/gitweb/?p=ports/core.git;a=blob;f=filesystem/Pkgfile;h=87eb59...
But, for the system to boot up properly, there have to be some dev files in /dev before devfs is mounted (console and stuff like that). Is that correct?
yep :) but devfs is ancient stuff (devfs was kernel space and so it was up as soon as the kernel was up, so you didn't need files in dev)... now we have udev in userspace...
Just a couple of question:
* What do you mean when you say "When your rootfs isn't live (up and running)"?
the system is not running from that rootfs (i.e. when you booted from a live cd or with the system on another partition), aka udev is not handling it.
* Is there a way to check if those dev files exist without booting from another partition?
hmm dont'know, it's up to your fantasy... if the system gets up and running you could put the check and subsequent mknod in rc and/or rc.single before udev starts... bye, have fun, giorgio -- NullPointer || GnuPG/PGP Key-Id: 0x343B22E6 http://cruxppc.sunsite.dk
On Fri, 15 Aug 2008 13:16:52 +0200 Giorgio Agrelli <giorgio_a@libero.it> wrote:
But, for the system to boot up properly, there have to be some dev files in /dev before devfs is mounted (console and stuff like that). Is that correct?
yep :) but devfs is ancient stuff (devfs was kernel space and so it was up as soon as the kernel was up, so you didn't need files in dev)... now we have udev in userspace...
Sorry I meant devpts which is just for virtual terminals as far as I've understood.
* Is there a way to check if those dev files exist without booting from another partition?
hmm dont'know, it's up to your fantasy... if the system gets up and running you could put the check and subsequent mknod in rc and/or rc.single before udev starts...
It looks like mounting / a second time in /mnt/mnt/ worked.
* What do you mean when you say "When your rootfs isn't live (up and running)"?
the system is not running from that rootfs (i.e. when you booted from a live cd or with the system on another partition), aka udev is not handling it.
Mmm I haven't understood. Thanks for help anyway. :) -- Value your freedom, or you will lose it, teaches history. ``Don't bother us with politics,'' respond those who don't want to learn. -- Richard M. Stallman http://www.gnu.org/philosophy/linux-gnu-freedom.html
participants (2)
-
Giorgio Agrelli -
ncrfgs＠tin.it