Hi,

I upgraded my Debian 3 via apt-get dist-upgrade. It is a server, and I am using a remote X-terminal to work on it. I usually install gdm and XFree86 on my machines, do not start a local server, but simply let gdm start the X, then I connect to it.

After my dist-upgrade, X kept crashing. So I reinstalled xserver-xorg, and then gdm - both to no avail. I then checked the gdm log, to find, to my astonishment, that the mouse could not be mounted anymore - /dev/psaux was gone.

I went to /dev, and indeed, no psaux, no mouse or mice anywhere to be seen. The server indeed does not have a mouse connected to it, but had during initial installation.

In order to get my X running again, I need to somehow tell him that there is a mouse, although there isn't. Does anyone have a smart idea on how I can do so? I am not a total beginner, but not (yet) a Linux user.

Thanks for any help offered!

Oh, one more thing - would it be a good idea to replace the normal xorg-server with Xvfb? Would this work? My X-Client needs to work graphically, the server assuredly not! :-)

Try Option AllowMouseOpenFail "true" in the server flags section.

You do know you don't need to have X running on the server in order to run X apps on the server and have the display be sent back to your client. If you ssh to the server (ssh -X server) and you have xforwarding enabled in sshd on the server, if you run an X app you will see it on your client. You might be able to trick your server into thinking it has a mouse, but there isn't much reason for that.

HTH

Forrest

If you are actually running Sarge (ver3), I suspect your problems may be related to an incorrect sources.list file. Post yours.

The output of $ uname -r ... may also be useful.

@Happy Tux - I tried that, i.e. I added

Section "ServerFlags"
Option "AllowMouseOpenFail"
EndSection

to my xorg.conf, and indeed, the gdm logfile now tells me that there were no fatal errors - nevertheless, there is no X running.

I also tried using the void driver, by replacing "Driver = mouse" with "Driver = void". It had the same effect: gdm was satisfied, but X crashes all the same :-(

If I remove the void or the server flag, then the message:

Fatal server error:
failed to initialize core devices

appears once more. Stupid, since now I do not know where to look for the error anymore. /var/log/Xorg.o.log gives me no reason, either. *sniff*

@forestt - My reason for havin an XServer running on all my machines so far is simple: it worked fine :-)) I am running Windows 2000 on my X-Client, and am using NetSarang to connect to all my Linux boxes. NetSarang needs an XServer broadcasting that it is there.

@rikh - uname -r gives me 2.6.8-2-686-smp. My sources.list says:

#deb file:///cdrom/ sarge main

#deb ftp://ftp.tiscali.de/pub/debian/debian/ stable main
#deb-src ftp://ftp.tiscali.de/pub/debian/debian/ stable main

#deb http://security.debian.org/ stable/updates main

deb ftp://debian.uni-essen.de/debian/ stable main
deb-src ftp://debian.uni-essen.de/debian/ stable main

deb ftp://ftp.tiscali.de/pub/debian/debian/ stable main
deb-src ftp://ftp.tiscali.de/pub/debian/debian/ stable main

deb ftp://ftp.de.debian.org/debian/ stable main
deb-src ftp://ftp.de.debian.org/debian/ stable main

deb ftp://sunsite.informatik.rwth-aachen.../Linux/debian/ stable main
deb-src ftp://sunsite.informatik.rwth-aachen.../Linux/debian/ stable main

deb ftp://ftp2.de.debian.org/debian/ stable main
deb-src ftp://ftp2.de.debian.org/debian/ stable main

deb http://security.debian.org/ stable/updates main

Do I really need to tear the machine out of my rack and re-install from scratch? (Since there would be a real mouse attached then, I suspect that would actually work)

You have a serious problem. You are running Sarge, and you did a dist-upgrade to Etch. The Release notes on the Debian webpage have a lot of instructions about successfully making that conversion. The upgrade from Xfree to Xorg is particularly tricky.

I would make sure you have any really critical data backed up just in case.

First, upgrade to the current Etch kernel (2.6.18-5, I think). Then run dist-upgrade again to see if you pick up anything else. Finally, make sure your xorg.conf file is using the best xorg driver for your video card. You can try switching to a proprietary driver later. I would back up your xorg.conf file with another name, then run "dpkg-reconfigure xserver-xorg" forcing it to write a new one.

From where you are, it's too late to try returning to Sarge ... your best bet is to complete the upgrade to Etch.

@rikh - damn. I read for hours before doing it, but obviously, I did something wrong :-(

Just to educate me - how could you tell that I am still running sarge, despite the dist-upgrade? The kernel? How do I update a kernel (never did this before - I usually never touch running systems, it was the first time I had to, because something I wanted to do did not work, and dist-upgrade was recommended)? Should I use apt-get? People usually tell me that this is not the best way... Can I use the normal kernel? This machine here has 2 CPUs...

I will also ask google, and will do what you told me... lets see if we can get this machine back up... thanks for all the help. I AM getting better, but it is a painful process :-)

Make sure you know the manufacturer of your video card.

# aptitude keep-all
# aptitude install linux-image-2.6.18-5-686
# aptitude dist-upgrade

Reboot to Single-User mode. Presumably you will have a network connection.

# aptitude install linux-headers-$(uname -r)
# aptitude install build-essential
# mv /etc/X11/xorg.conf /etc/X11/xorg_old.conf
# dpkg-reconfigure xserver-xorg

Reboot to regular user mode, praying fervently. If you generate serious error messages anywhere along the way, you may have big problems.

@rikh - unfortunately, video is onboard. But since the machine itself does not ever need to go visual, I will go forward anyway. I always backup my data, so no worries there. If things fail, and the machine does not boot anymore afterwards, I will just get it out of the rack and install from scratch. So, let's start praying fervently :-))

Well, it worked - and yet it did not :-)

I did exactly as you said, and the machine performed every command neatly, but, although my gdm.log tells me that X is up and running:

X Window System Version 7.1.1
Release Date: 12 May 2006
X Protocol Version 11, Revision 0, Release 7.1.1
Build Operating System: UNKNOWN
Current Operating System: Linux uriel 2.6.18-5-686 #1 SMP Wed Oct 3 00:12:50 UTC 2007 i686
Build Date: 04 September 2007
Before reporting problems, check http://wiki.x.org
to make sure that you have the latest version.
Module Loader present
Markers: (--) probed, (**) from config file, (==) default setting,
(++) from command line, (!!) notice, (II) informational,
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(==) Log file: "/var/log/Xorg.0.log", Time: Thu Dec 13 21:15:54 2007
(==) Using config file: "/etc/X11/xorg.conf"
(EE) AIGLX: Screen 0 is not DRI capable
xkb_keycodes { include "xfree86+aliases(qwertz)" };
xkb_types { include "complete" };
xkb_compatibility { include "complete" };
xkb_symbols { include "pc(pc105)+de(nodeadkeys)" };
xkb_geometry { include "pc(pc101)" };
The XKEYBOARD keymap compiler (xkbcomp) reports:
> Warning: Type "ONE_LEVEL" has 1 levels, but <RALT> has 2 symbols
> Ignoring extra symbols
Errors from xkbcomp are not fatal to the X server
FreeFontPath: FPE "/usr/share/fonts/X11/misc" refcount is 2, should be 1; fixing.

Nevertheless, I get no X. :-(

uname - r now tells me 2.6.18-5-686, joe /etc/debian_version tells me 4.0, and joe /proc/version tells me Linux version 2.6.18-5-686 (Debian 2.6.18.dfsg.1-13etch4) (dannf@debian.org) (gcc version 4.1.2 20061115 (prerelease) (Debian4.1.1-21)) #1 SMP Wed Oct 3 00:12:50 UTC 2007.

Well, at least psaux is back in /dev. Unfortunately, I still don't know why my X does not stay up. However, I have one hint. I did ps -efw|grep gdm, and he told me:

00:00:00 /usr/lib/gdm/gdmopen -l /bin/sh -c /usr/bin/whiptail --msgbox 'Der Anzeige-Server wurde in den letzten 90 Sekunden ca. 6 Mal heruntergefahren. Vermutlich läuft irgendetwas schief. Vor dem nächsten Anmeldeversuch auf Anzeige :0 wird eine Pause von 2 Minuten eingelegt.' 16 70

Which basically means gdm was, during the last 90 seconds, shutdown more than 6 times.

WHAT CAN THAT BE? WHY CAN IT BE? *sniff*

Post output of $ lspci | grep VGA

... and this section of /etc/X11/xorg.conf

Section "Device"
# Identifier "ATI Technologies Inc Rage XL AGP 2X"
# Driver "ati"
# BusID "PCI:1:0:0"
Identifier "Standardgrafikkarte"
Driver "vga"
EndSection

I tried both - the one currently active is the one working in my other two machines :-/


lspci | grep VGA =
01:00.0 VGA compatible controller: ATI Technologies Inc Rage XL AGP 2X (rev 27)

If this is an error which makes gdm crash again and again, how come it is not in the logs???

The first priority is to get X to work. For that I'd recommend you stick to the VESA driver for now. Once everything works a change to the ATI driver can be an option. Btw, you are aware that after installing a new kernel any propriety driver (like the videodrivers of ATI and nVidia) needs to be re-installed! Those drivers (nothing more than kernel modules) are build against a certain kernel, a new kernel needs a new kernel module

The ati driver is xorg's not from ATI so that would not be a problem here. To the OP I would say since you could care less about running a desktop that installing xdm and when the choice is provided say yes/select it as the default just to get rid of gdm in case it is this itself that is causing the problem then restart the X server. Plus the entire xorg.conf and /var/log/Xorg.0.log would be nice to see as well, the log would need to be put in a couple of posts as it is rather big and over the post size limit.