problem loading KDE after updating portage

[mesman00]

ok, i just updated my linux system by doing "emerge -u system". After it was done compiling and installing all the new ebuilds, i updated my config files as i was directed. then, after restarting my computer linux wouldn't start kde. when i typed "startx" i get the following error:
failed to initialized NVDriver module in kernel

is it possible that i have to change something in my XF86Config file because updating the system intalled new drivers? right now in my XF86Config file my driver is "nvidia". It has worked liked this fine up until i updated all my ebuilds. Any input would be great. Thanks.

[Slougi]

Try running the NVmakedevices script. Not quite sure of the name, it is either in /sbin or /usr/sbin somewhere. If that does not work you need to add the nvidia module to /etc/modules.autoload.

[Slougi]

OK, just checked bu SSH'ing into my PC ;)

The script is called NVmakedevices.sh and is in /sbin.

[radTube]

Mesman: Wouldn't you be better off searching the gentoo forums? Just a thought...

BTW Slougi, shouldn't XF86 automatically load the driver module if it's correctly configured to do so in XF86Config? Why should it be in modules.autoload?

[Slougi]

Yes it should always do that *theoretically*, and it used to work fine. Some update to baselayout and/or devfs however caused some breakage, so that one of the two things I mentioned above needs to be done for it to work correctly. It's a widespread problem, just search the gentoo forums.

[Malcolm]

do an lsmod to see if the modules loaded up
do an opengl-update nvidia
make sure you have your XF86Config file setup properly (from the error it sounds like you do)

try to run it again

[mesman00]

Originally posted by Slougi
Try running the NVmakedevices script. Not quite sure of the name, it is either in /sbin or /usr/sbin somewhere. If that does not work you need to add the nvidia module to /etc/modules.autoload.

i ran the script, tried to start x, and got the same errors. How should it appear in the modules file?

[mesman00]

Originally posted by Malcolm
do an lsmod to see if the modules loaded up
do an opengl-update nvidia
make sure you have your XF86Config file setup properly (from the error it sounds like you do)

try to run it again

what exactly is "Ismod". anyways, i tried opengl-update nvidia and it still doesn't work. everthing as far as i am concerned is set up correctly in the XF86Config file. oh yah, i did post this on the gentoo forums, but they were of no help. i only got one response. that was it. so i came here

[Malcolm]

its not Ismod it lsmod (with an L)

It lists all loaded modules, if NVdriver isn't loaded then you'll always get an error on loading XFree86

[mesman00]

the nvidia module isnt loaded...there are no modules loaded as a matter of fact, becuase my modules.autoload file is empty (except for a few instructions at the top). so, how do i add nvidia to my modules.autuload file? what does it have to be entered as?

[Slougi]

What happens when you try to load the module? (modprobe nvidia or modprobe NVdriver, depending on driver version)

If it works you should be able to start X. And modules.autoload is just a text file where you can add modules (drivers) to be loaded on boot. Just add the driver name to it, in this case either nvidia or NVdriver. Here is some info from my computer:



Gondolin root # cat /etc/modules.autoload
# /etc/modules.autoload.d/kernel-2.4: kernel modules to load when system boots.
# $Header: /home/cvsroot/gentoo-src/rc-scripts/etc/modules.autoload.d/kernel-2.4,v 1.1 2003/03/09 09:06:23 azarah Exp $
#
# Note that this file is for 2.4 kernels. If you need different modules
# for a 2.5 kernel, you can create /etc/modules.autoload.d/kernel-2.5
#
# Add the names of modules that you'd like to load when the system
# starts into this file, one per line. Comments begin with # and
# are ignored. Read man modules.autoload for additional details.

# For example:
# 3c59x
joydev
sis900
floppy
ide-scsi




Section "Device"
Identifier "NVIDIA GeForce"
Driver "nvidia"
#VideoRam 65536
# Insert Clocks lines here if appropriate
EndSection




* media-video/nvidia-glx
Latest version available: 1.0.4363
Latest version installed: 1.0.4363
Size of downloaded files: 4,099 kB
Homepage: http://www.nvidia.com/
Description: XFree86 GLX libraries for the NVIDIA's X driver

* media-video/nvidia-kernel
Latest version available: 1.0.4363-r1
Latest version installed: 1.0.4363-r1
Size of downloaded files: 645 kB
Homepage: http://www.nvidia.com/
Description: Linux kernel module for the NVIDIA's X driver




Gondolin root # lsmod
Module Size Used by Tainted: P
udf 90176 0 (autoclean)
sr_mod 18648 0 (autoclean)
agpgart 14416 3 (autoclean)
nvidia 1542240 10 (autoclean)
nfs 53236 1 (autoclean)
nfsd 49256 8 (autoclean)
lockd 39216 1 (autoclean) [nfs nfsd]
sunrpc 69596 1 (autoclean) [nfs nfsd lockd]
hid 19684 0 (unused)
usb-ohci 20008 0 (unused)
snd-pcm-oss 40196 1
snd-mixer-oss 13880 0 [snd-pcm-oss]
snd-intel8x0 18404 1
snd-ac97-codec 39848 0 [snd-intel8x0]
snd-pcm 64448 0 [snd-pcm-oss snd-intel8x0]
snd-timer 15880 0 [snd-pcm]
snd-page-alloc 5308 0 [snd-intel8x0 snd-pcm]
snd-mpu401-uart 3824 0 [snd-intel8x0]
snd-rawmidi 14624 0 [snd-mpu401-uart]
snd-seq-device 4432 0 [snd-rawmidi]
snd 32100 0 [snd-pcm-oss snd-mixer-oss snd-intel8x0 snd-ac97-codec snd-pcm snd-timer snd-mpu401-uart snd-rawmidi snd-seq-device]
soundcore 4164 3 [snd]
rtc 7484 0 (autoclean)
usbcore 65664 1 [hid usb-ohci]
supermount 13920 1 (autoclean)
ide-scsi 8944 0
scsi_mod 59284 2 [sr_mod ide-scsi]
floppy 51708 0
sis900 14636 1
joydev 5984 0 (unused)
input 3712 0 [hid joydev]

[mesman00]

ok after adding the module and editing the make.conf and the rc.conf file i was able to bring kde back up. however, i no longer have internet connection, even though when the kernel is loading it says that eth0 was brought up sucessfully. also, when i log into my user account using kde and open a konsole, it won't let me switch to the root user. instead it says permission denied. i had this problem before, and i know there's something i have to change in the user settings, but i don't exactly remember what. thanks for the help so far.

[Malcolm]

find your net.eth0 and make sure that dhcpcd is assigning everything, I can't remember how to do this off hand and I'm at school but I think its somewhere in the Gentoo install docs.

About the root stuff, no idea
/me looks at Slougi

[AlphaWolf]

btw, when you want to make sure that all of the packages on your system are up to date, do "emerge -U --deep world" rather than system. System refers to the core components, where world refers to all major packages. Also, --deep checks all dependencies, rather than just the immediate ones that are flagged for updating when any given package is updated.

[Slougi]

Baselayout 1.8.6.8 is broken, I had the same problem. Downgrade to 1.8.6.7, reboot, emerge sync, emerge baselayout 1.8.6.8-r1. If needed boot from gentoo livecd and chroot.

http://bugs.gentoo.org/show_bug.cgi?id=21376

< edit >

From your description (now that I read it properly) It seems like you might have replaced /etc/groups, or something similar. Possibly you need to readd your user to the wheel group.


usermod -g users -G wheel [user]

[mesman00]

hey slougi...how do i downgrade from the livecd?

[Slougi]

Boot from CD, set up networking, mount partitions, chroot (same as when you installed), use emerge as always :)

[mesman00]

heres the deal, i messed with something i wasn't supposed to, becuase now after emerge baselayout, when i try to load linux i get the following error after going through some of the loading process:

kernel panic: no kernel init!!

insteading of emerging after setting up my network card and mounting my drives, i did.

tar -xvjpf /mnt/cdrom/gentoo/stag1-*.tbz2
mount -o bind /proc /mnt/gentoo/proc

and unpacked the stage1 tarbell. it wasn't until after this that i did:

chroot /mnt/gentoo /bin/bash
evn-update
source /etc/profile

and emerged baselayout

could this be the reason that is causing my newfound problems? any help would be great, and thanks for all the help so far. i guess the only way to learn things is by fuxoring them up.

[Slougi]

Originally posted by mesman00
heres the deal, i messed with something i wasn't supposed to, becuase now after emerge baselayout, when i try to load linux i get the following error after going through some of the loading process:

kernel panic: no kernel init!!

insteading of emerging after setting up my network card and mounting my drives, i did.

tar -xvjpf /mnt/cdrom/gentoo/stag1-*.tbz2
Ack :|
Basically you will need to bootstrap again at this point, maybe emerge system. emerge sync before this though, so you get the new fixed baselayout.


could this be the reason that is causing my newfound problems? any help would be great, and thanks for all the help so far. i guess the only way to learn things is by fuxoring them up.
True. Or having a system where you do not have super user rights, and a broken tcl and gcc, and you need to compile an eggdrop :doh:
That's how I learned my stuff mostly :P

[mesman00]

this bootstrap you speak of, can i get some more details.

[Malcolm]

http://www.gentoo.org/doc/en/gentoo-x86-install.xml#doc_chap11

[mesman00]

after bootstrapping an emerging system, am i gonna have to re-setup my kernel, or will that all be saved?

[Malcolm]

you're going to have to recompile your kernel but your .conf file _shouldn't_ be overridden; I'd make a back-up anyways.

[Slougi]

Umm, no, you don't need to recompile the kernel. Also alternatively to bootstrapping you could just get a stage 3 tarball and extract that.

[mesman00]

still having some problems. let me try to explain it as best as possible: this is exactly what i'm doing

first, set up network:

modprobe 3189too (my ethernet drivers)
net-setup eth0
ifconfig (test to make sure it works)

next, mount drives:

mkdir /mnt/gentoo
mount /dev/hda7 /mnt/gentoo (root)
mount /dev/hda5 /mnt/gentoo/boot (boot)

setting up from stage 3 (as slougi suggested for ease, atleast for right now):

cd /mnt/gentoo
tar -xvjpf /mnt/cdrom/gentoo/stage3-*.tbz2
mount -o /proc /mnt/gentoo/proc
cp /etc/resolv.conf /mnt/gentoo/etc/resolv.conf
chroot /mnt/gentoo /bin/bash
env-update
source /etc/profile

ok, now i do emerge baselayout to downgrade:

emerge baselayout

at the end of the merge i get the following error:

error scanning /etc
error scanning /etc/kde3.1/... (not sure the rest of the path)

if i try to go to /etc by doing cd /etc i get the error:

the specified path does not exist

if i try to boot up gentoo, i still get the same kernel error:

kernel panic: no kernel init

what do you guys got for me?

[Slougi]

Erm, check that you *have* an /etc directory at all, because it seems that you for whatever reason do not. Maybe you need to run fsck on the partition? Remember to do that only if it is mounted read only (boot from livecd -> mount -o ro -t [type of partition] /dev/hdx[1-9] /mnt/whatever

[mesman00]

what exactly does running fsck do? is that the same as efsck?

[Slougi]

Fsck is the *nix equivalent of scandisk, afaik it stands for file system check.

[mesman00]

ok, i ran e2fsck i the drive and repaired it. then i did unpacked stage 3 tarbell, re-emerge baselayout. but, i still get the error scanning /etc and error scaning /usr/kde/.... (not sure the entire path). i can acess the etc directory, and i can get into /usr. However, in /usr there is no kde directory.

*edit* so i tried recompiling my kernel, but to no avail. i get the same error. is there anything i can do after recompiling the kernel, or before. obviously i am not doing something right.

[Slougi]

In that case I really do not know. The last thing I can think of is a permission problem, but that is a very very long shot.

[mesman00]

eeek, the last thing i want to do is scratch the whole root and start over.

[Slougi]

Can you still access /etc and so on if you just change to that dir in a terminal?

[mesman00]

i can acess /etc fine, but it doesn't seem like the /usr/kde directory exists