You are not logged in.
Hi,
Yet another user with a broken system after kernel and mkinitrd upgrade.
The system drops into the busybox shell no matter what I do. I first tried initrd26.img using mkintrd auto, then initrd26-full.img using mkinitrd full.
Either way I am dropped into the busybox shell with
/bin/sh: can't access tty; job control turned off
When there, using the full initrd, cat /proc/partitions is correct and shows my partitions sda1 ... sda9. sda6 is the Arch root partition. The hdd is sata, running on an nForce 4 chipset. I think sata_nv is the correct module.
Anyway, using the full initrd, cat /proc/modules lists hundreds of modules so not loading a module doesn't appear to be the problem.
But I still can't get anywhere past this shell. Any suggestions as to how I might fix it?
Thanks
Francis
Offline
Yesterday I have pacman -Syu and thus installed the new 'initrd kernel'. When booting just few hours ago system halts when trying to uncompress the kernel.
It says:
invalid compressed format (err=2).
I don't now wetter this has anything todo with initrd but I never had this problem before.
Some people say this could have to do with corrupt packages (but pacman verifies after downloading - doesn't it?).
Can anybody help me with this?
PS: before anybody asks: yes I've put the initrd line in my lilo conf and ran lilo afterwards. (initrd line points to initrd26.img not the full one.) I use the stock kernel.
Offline
Ok, more info. I used the 0.7 boot CD and chrooted into my Arch system. I tried enabling everything in an attempt to get it to work and then mkinitrd'd again.
When doing a mkinitrd auto I get on the screen
/sys/devices is not present -- mount /sys first
3 timesbefore it creates the initrd.
When doing mkinitrd full no such error, but the system still refuses to boot.
I've wasted about 3 evenings so far working on this - have read through this entire forum thread and the wiki pages and nothing has worked. I really hope it's something simple I've just missed...
Cheers
Francis
Offline
come on irc i'll try to help you
Offline
Thanks to a quick chat with tpowa my problem is solved.
As usual, the error was obvious with a bit of insight:
My GRUB menu.lst entry for Arch was incorrect.. my system has one SATA hard drive and no IDE.
The correct kernel line for Arch on my system is
kernel /boot/vmlinuz26 root=/dev/sda6
I incorrectly had /dev/hda6 in there.
I think this must be because I simply copied the new parts of menu.lst.pacnew into my menu.lst and checked all the numbers etc. but not the letter!
So - completely my fault, not Arch's. A massive thank you to tpowa for his assistance.
Cheers
Francis
Offline
Thanks to a quick chat with tpowa my problem is solved.
As usual, the error was obvious with a bit of insight:
My GRUB menu.lst entry for Arch was incorrect.. my system has one SATA hard drive and no IDE.
The correct kernel line for Arch on my system is
kernel /boot/vmlinuz26 root=/dev/sda6
I incorrectly had /dev/hda6 in there.
Well that's surprising me,
main system has a single Hitachi SATAII HDD and since it upgraded to kernel >=2.6.14 it always booted fine with
title Arch Linux 2.6.15 [/boot/vmlinuz26]
root (hd0,5)
kernel /boot/vmlinuz26 root=/dev/hda6 ro devfs=nomount vga=792
initrd /boot/initrd26.img
Well I just upgraded to 2.6.15 and will reboot just to be sure.
[EDIT:ON]
yeap! had to edit /boot/grub/menu.lst from
kernel /boot/vmlinuz26 root=/dev/hda6
to
kernel /boot/vmlinuz26 root=/dev/sda6
[EDIT:OFF]
Also have a question about hwdetect.
On 2 of my machines, hwdetect do not print anything. I tried with:
hwdetect --show-modules, ~ --show-all, ~ --show-all, ~ --sata (etc.) both as user and root.
Which makes me unsure on what module to put into /etc/mkinitrd.conf on my old gateway, that i just updated from kernel 2.6.13 to 2.6.15.
This box has no screen/keyboard, it's administred by ssh only. So a single error on initrd configuration will be be uneasilly corrected.
Seeded last month: Arch 50 gig, derivatives 1 gig
Desktop @3.3GHz 8 gig RAM, linux-ck
laptop #1 Atom 2 gig RAM, Arch linux stock i686 (6H w/ 6yrs old battery ) #2: ARM Tegra K1, 4 gig RAM, ChrOS
Atom Z520 2 gig RAM, OMV (Debian 7) kernel 3.16 bpo on SDHC | PGP Key: 0xFF0157D9
Offline
i just installed a fresh arch 0.7.1 , then did the initrd custom config as it says on the wiki , and my system is working flawlesly as usual on arch.
I cant belive my machine boots up to the KDM on 20 seconds !!
Offline
Folks I may have been asleep at the wheel here, but I just rolled up my Arch, and happened to see the initrd message scroll by.
Mmmm - wonder what that was about, so I checked pacman.log. Nothing of interest. Are these messages saved elsewhere ???
Whilst a new menu.lst is saved, it might be nice to be able to get all the messages, even if stored elsewhere, aka Gentoo with portage.log and LOG_PORTDIR.
Is of concern to me as I use this Arch system as kernel patch testing, so it has a lot of kernels, and I chainload in from another (stable) system.
And (obviously) I'm not up-to-date on the announcements.
Offline
This is usually caused because a module needed to access your hd or fs is not loaded by initrd.
try this for grub
initrd /boot/initrd26-full.img
and this for lilo
initrd=/boot/initrd26-full.img
This will boot with all available modules loaded. If it works, you can then try to figure out which modules are needed to boot your system.
Check the wiki-entry for initrd for details.
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
clean chroot building not flexible enough ?
Try clean chroot manager by graysky
Offline
i keep getting the
ERROR: All of your loopback devices are in use!
I'm just finally moving ahead to udev and initrd, and its a nightmare. I have the /dev/zero node, whats the problem here? i'm lost, ive been screwing with this all day. i've swapped all manner of kernels and uninstalled and installed every combination i can imagine. HELP
"Ignorance is bliss, for stupid people."
"open-source is [...] programming Darwinism."
Vaughan-Nichols
Offline
try using root user, normal users cannot build initrd
Offline
i am logged in as root
"Ignorance is bliss, for stupid people."
"open-source is [...] programming Darwinism."
Vaughan-Nichols
Offline
Support for loopback block devices is available?
bash-3.00$ grep loop /proc/devices
7 loop
Also check your device files:
bash-3.00$ ls -l /dev/loop
total 0
brw-rw---- 1 root disk 7, 0 2006-01-19 18:46 0
brw-rw---- 1 root disk 7, 1 2006-01-19 18:46 1
brw-rw---- 1 root disk 7, 2 2006-01-19 18:46 2
brw-rw---- 1 root disk 7, 3 2006-01-19 18:46 3
brw-rw---- 1 root disk 7, 4 2006-01-19 18:46 4
brw-rw---- 1 root disk 7, 5 2006-01-19 18:46 5
brw-rw---- 1 root disk 7, 6 2006-01-19 18:46 6
brw-rw---- 1 root disk 7, 7 2006-01-19 18:46 7
Offline
Support for loopback block devices is available?
bash-3.00$ grep loop /proc/devices 7 loop
Also check your device files:
bash-3.00$ ls -l /dev/loop total 0 brw-rw---- 1 root disk 7, 0 2006-01-19 18:46 0 brw-rw---- 1 root disk 7, 1 2006-01-19 18:46 1 brw-rw---- 1 root disk 7, 2 2006-01-19 18:46 2 brw-rw---- 1 root disk 7, 3 2006-01-19 18:46 3 brw-rw---- 1 root disk 7, 4 2006-01-19 18:46 4 brw-rw---- 1 root disk 7, 5 2006-01-19 18:46 5 brw-rw---- 1 root disk 7, 6 2006-01-19 18:46 6 brw-rw---- 1 root disk 7, 7 2006-01-19 18:46 7
its all there, still no dice
"Ignorance is bliss, for stupid people."
"open-source is [...] programming Darwinism."
Vaughan-Nichols
Offline
Hmm... Unfortunately losetup's stderr gets lost in /dev/null in mkinitrd, so try to setup a loopback device manually:
# touch /tmp/test.img
# losetup /dev/loop/0 /tmp/test.img
Jürgen
Offline
[root@thebeast linux-2.6.3]# touch /tmp/test.img
[root@thebeast linux-2.6.3]# losetup /dev/loop/0 /tmp/test.img
memlock: Cannot allocate memory
Couldn't lock into memory, exiting.
aaaaack. WTF? why won't this ever work for me? this is going to be the 3rd time ive crashed a box because of this. :oops:
BTW, i appreciate the useful replies everyone
"Ignorance is bliss, for stupid people."
"open-source is [...] programming Darwinism."
Vaughan-Nichols
Offline
Your memory lock resource limit is exeeded? losetup works fine here with defaults:
# ulimit -l
40000
Try to increase:
# ulimit -l YOURLIMIT
Offline
Your memory lock resource limit is exeeded? losetup works fine here with defaults:
# ulimit -l 40000
Try to increase:
# ulimit -l YOURLIMIT
none of that is doing anything...no response, its like i didnt even type anything...? GAWD MY BOX IS BROKEN :oops:
"Ignorance is bliss, for stupid people."
"open-source is [...] programming Darwinism."
Vaughan-Nichols
Offline
Possibly a stupid question, but...do we have to add <code>etc/mkinitrd.conf</code> to the <code>NoUpgrade</code> option of <code>pacman.conf</code> in order to not have to re-customize it everytime there's a new kernel?
--==EDIT==--
DOH! Wrong file name...
Offline
Possibly a stupid question, but...do we have to add <code>etc/initrd.conf</code> to the <code>NoUpgrade</code> option of <code>pacman.conf</code> in order to not have to re-customize it everytime there's a new kernel?
Yes, but the filename is etc/mkinitrd.conf
Offline
I have:
michiel@vanessa ~$ grep FS /etc/mkinitrd.conf
# REMOVE_FS: removes the complete Filesystems
REMOVE_FS="xfs jfs reiserfs"
But this doesn't seem to help.. A cut from `lsmod`:
xfs 615552 0
reiserfs 269424 0
jfs 192636 0
All you got to do, is tell the story right...
Offline
argh wrong syntax, just specify your fs
eg:
FILESYSTEMS="reiserfs"
Offline
I reinstalled Arch and it got a horrible error...
/bin/sh: can't access tty; job control turned off
I get dropped into busybox.
I tried booting the kernel with initrd26-full.img, still no go. I tried mounting my drive on my main machine chrooted into it, and tried mkinitrd auto. It builds it, but when I reboot same error.
Initrds internal..
[root@RoXoR media]# ls -R /mnt/initrd
/mnt/initrd:
bin dev etc lib linuxrc new_root proc sbin sys/mnt/initrd/bin:
awk chroot fsck mkdir mount rm sleep
busybox cp insmod mknod pivot_root sed test
cat echo ls mkrootdev reboot sh umount/mnt/initrd/dev:
console tty/mnt/initrd/etc:
fstab/mnt/initrd/lib:
aec62xx.ko cy82c693.ko ide-disk.ko pdc202xx_old.ko sis5513.ko
alim15x3.ko ext3.ko ide-floppy.ko piix.ko slc90e66.ko
amd74xx.ko generic.ko it821x.ko reiserfs.ko triflex.ko
atiixp.ko hpt34x.ko jbd.ko rz1000.ko trm290.ko
cdrom.ko hpt366.ko jfs.ko sc1200.ko via82cxxx.ko
cmd64x.ko ide-cd.ko ns87415.ko serverworks.ko xfs.ko
cs5530.ko ide-core.ko pdc202xx_new.ko siimage.ko/mnt/initrd/new_root:
/mnt/initrd/proc:
/mnt/initrd/sbin:
fsck.ext2 fsck.ext3 fsck.jfs fsck.reiserfs fsck.xfs/mnt/initrd/sys:
I did a ftp install as I always do, I used auto partition as I always do. I used the official arch mirror so it's fully up to date.
The first harddrive is /dev/hda1 (ext2 for boot)
the second is /dev/hda2 (swap)
the third is the root partition /dev/hda3 (ext3 for /)
this is my menu.lst
[root@RoXoR ~]# cat /mnt/arch/boot/grub/menu.lst
# DEVICE NAME CONVERSIONS
#
# Linux Grub
# -------------------------
# /dev/fd0 (fd0)
# /dev/hda (hd0)
# /dev/hdb2 (hd1,1)
# /dev/hda3 (hd0,2)
## general configuration:
timeout 5
default 0
color light-blue/black light-cyan/blue# (0) Arch Linux
title Arch Linux
root (hd0,0)
kernel (hd0,0)/vmlinuz26 root=/dev/hda3 ro vga=773
initrd /initrd26.img
I'm now on my primary system that never seems to have problems with arch. :shock:
Offline
I just updated the wiki's initrd page a bit, and had a couple of things to note...
<code>hwdetect</code> doesn't show the proper format for the mkinitrd.conf file when using the hostcontroller flags. For example:
$ hwdetect --ide
HOSTCONTROLLER_IDE: generic piix
...is what you get, but it should show:
HOSTCONTROLLER_IDE="generic piix"
Also, what is the downside of disabling cdrom support? I noticed that the one person who had done it said his cdrom was no longer was accessible even after bootup, but shouldn't it be okay as long as they're not trying to boot from a kernel on a cd?
Offline
One last thing...I too have not been able to remove <code>ide-floppy</code> from loading.
Offline