Xserver hosed after 2.6.29 kernel upgrade

derekr44 · 2009-04-20 20:33:38

I upgraded from 2.6.27 to 2.6.29 today via pacman -Syu. Since then, I have not been able to start X at all. I was running the nvidia driver, but my attempt to run vesa drivers had the same result.

hostname: Host name lookup failure
xauth: creating new authority file /root/.serverauth.4913

X.Org X Server 1.6.1
Release Date: 2009-4-14
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.29-ARCH i686
Current Operating System: Linux hubble 2.6.29-ARCH #1 SMP PREEMPT Fri Apr 17 12:46:01 UTC 2009 i686
Build Date: 15 April 2009 11:09:10AM

Before reporting problems, check http://wiki.x.org
to make sure that you have the latest version.
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: Mon Apr 20 13:14:55 2009
(==) Using config file: "/etc/X11/xorg.conf"
Parse error on line 9 of section Files in file /etc/X11/xorg.conf
Ignoring obsolete keyword "RgbPath".
(EE) Failed to load module "xtrap" (module does not exist, 0)
(EE) Failed to load module "freetype" (module does not exist, 0)
(EE) module ABI major version (2) doesn't match the server's version (5)
(EE) Failed to load module "vesa" (module requirement mismatch, 0)
(EE) No drivers available

Fatal server error:
no screens found

Please consult the The X.Org Foundation support
at http://wiki.x.org for help
Please also check the log file at "/var/log/Xorg.0.log" for additional information

giving up.
xinit: No such file or directory (errno 2): unable to connect to X server
xinit: No such process (errno 3): Server error.

I've never encountered this before... it won't even load vesa drivers...

graysky · 2009-04-20 21:05:41

When I ran Ubuntu and used the official nvidia driver package, I had to rebuild the kernel mod every time I rebuilt/upgraded my kernel. I dunno how arch handles it but I would think that you'd need a matching module for the 2.6.29 kernel, no? Does pacman handle this automatically? Have you tried to remove your current nvidia package and then reinstall it?

EDIT: can you post your /etc/X11/xorg.conf? The errors in your log you posted seem to indicate that you're missing several modules you're trying to load.

(EE) Failed to load module "xtrap" (module does not exist, 0)
(EE) Failed to load module "freetype" (module does not exist, 0)
(EE) module ABI major version (2) doesn't match the server's version (5)
(EE) Failed to load module "vesa" (module requirement mismatch, 0)
(EE) No drivers available

I'm no expert with Arch by any means (been running it exclusively for only a week now), but it looks like your kernel modules aren't right...? I'd wait to hear from the forum experts - this is just my naive guess.

Last edited by graysky (2009-04-20 21:07:23)

ataraxia · 2009-04-20 21:13:16

The ABI version "2" shows that you have an old VESA module. DId you do a full system upgrade? Do you have any ignored packages?

(Also, you appear to have a network problem. Fix your hostname in /etc/hosts and /etc/rc.conf)

derekr44 · 2009-04-20 21:22:28

My Vesa package upgrade was ignored because of a previous issue with the newest vesa driver.

I fixed the host issue (thanks). I made the change in rc.conf but not /etc/hosts.

Good thing is I've made a little progress. Instead of X not loading at all, I'm at a black screen. Because I have an old Toshiba laptop, I need to use my custom EDID file for the screen resolution. I need to shut down X in order to get back to the console, but ctrl+alt+backspace is not working.

Last edited by derekr44 (2009-04-20 21:25:11)

derekr44 · 2009-04-20 22:21:32

Ok, I managed to get back into xorg.conf in order to reference my custom EDID file. It seems that the nvidia driver is recognized as well as my screen resolution because there are no errors when the driver is loaded (according to the log file). However, I'm getting the following:

Backtrace:
0: /usr/bin/X(xorg_backtrace+0x3b) [0x813278b]
1: /usr/bin/X(xf86SigHandler+0x51) [0x80dda61]
2: [0xb80c6400]

Fatal server error:
Caught signal 11. Server aborting

Here is how my xorg.conf file is setup (the device section)

Section "Device"
Identifier "Card0"
Driver "nvidia"
Option "UseDisplayDevice" "DFP"
Option "CustomEDID" "DFP-0:/home/derek/edid/edid.bin"
Option "AddARGBVisuals" "True"
Option "AddARGBLXVisuals" "True"
Option "NoLogo" "False"
Option "nvagp" "0"
Option "ConnectedMonitor" "DFP"
EndSection

derekr44 · 2009-04-20 22:47:41

After doing some more research, it appears this "Caught Signal 11" error is caused by kernel memory accessing. Mods might want to move this one over to the Kernel discussion forum instead.

EDIT:
Ok, so apparently nopat needs to be added to my GRUB menu.lst. Any ideas why pat would need to be disabled after a kernel upgrade?

Last edited by derekr44 (2009-04-20 22:57:07)

nemrood · 2009-04-20 23:18:06

I was gonna post a thread like this, but I'm glad I've found this one . Seems like I have the same issue for about 5 days now and can't resolve it in a simple way (without changing the distro for egzample).

Here is my xorg.conf: http://pastebin.com/f34a761e
Here is my Xorg.0.log: http://pastebin.com/fc443b0f

The most important fragment of log (as of my point of view):

Backtrace:
0: /usr/bin/X(xorg_backtrace+0x3b) [0x813278b]
1: /usr/bin/X(xf86SigHandler+0x51) [0x80dda61]
2: [0xb7fb2400]
 
Fatal server error:
Caught signal 11.  Server aborting

The problem doesn't exist when I use nv driver, so it may be nvidia drivers issue. My nvidia driver: nvidia-driver-96xx

PS. And please don't tell me about Type1 and Freetype modules - the crash like here takes place even when I comment loading those modules in xorg.conf.

Last edited by nemrood (2009-04-20 23:20:02)

ataraxia · 2009-04-20 23:34:54

The xtrap, freetype, and type1 modules should come out of everyone's config. They don't exist any longer. (They're also not causing your problems, but they make noise in your logs.)

It looks to me that the NVidia driver is causing the crash. I don't use it, so I can't go any further.

nemrood · 2009-04-23 21:46:07

The only solution to this I've found is downgrade to 2.6.28 kernel and earlier version of xorg-server and nvidia-96xx (to be sure). It works for me now, but I have broken dependencies.

I'm not sure if it's nvidia driver issue (it doesn't work when I use older nvidia driver too). It can be something in the kernel or xorg-server 1.6.0/1 (I use xorg-server 1.5.3 again).

Hope this will be fixed soon.

derekr44 · 2009-04-24 16:18:16

The only way I got mine to work was to adjust my /boot/grub/menu.lst file and add nopat to my Arch option:

kernel /vmlinuz26 root=/dev/sda3 ro vga=773 nopat

Last edited by derekr44 (2009-04-24 16:18:33)

.:B:. · 2009-04-24 19:25:32

hostname: Host name lookup failure

You might wanna fix that too, it's minor and no showstopper at all, but your network doesn't look configured the way it should be .

guitarMan666 · 2009-04-26 01:00:19

I am having the same problem on my Intel i845g and it also makes some noise about the BO front buffer. Specifically it says:

Fatal Server Error:
Failed to bind memory for BO front buffer.

I entered

sudo pacman -Syu

to do an upgrade this evening (2009, April 25) and when I exited Fluxbox to reboot, SLiM was missing and I was greeted with the text login prompt. I proceeded to reboot and when it came back up SLiM was still gone so I logged in and naively tried to run startx which resulted in the errors in my xorg.0.log, http://pastebin.com/m432cf161. I have also tried the nopat and enable_mtrr_cleanup boot flags both separately and combined.

My XOrg.conf (which is based on one I found on this forum from another user with my card and worked flawlessly until now) is included here http://pastebin.com/d7f238d9a for completeness.

.:B:. · 2009-04-26 05:57:30

It looks to me like this is a different problem (with the same result). You might wanna open a separate topic for this.

Arch Linux

#1 2009-04-20 20:33:38

Xserver hosed after 2.6.29 kernel upgrade

#2 2009-04-20 21:05:41

Re: Xserver hosed after 2.6.29 kernel upgrade

#3 2009-04-20 21:13:16

Re: Xserver hosed after 2.6.29 kernel upgrade

#4 2009-04-20 21:22:28

Re: Xserver hosed after 2.6.29 kernel upgrade

#5 2009-04-20 22:21:32

Re: Xserver hosed after 2.6.29 kernel upgrade

#6 2009-04-20 22:47:41

Re: Xserver hosed after 2.6.29 kernel upgrade

#7 2009-04-20 23:18:06

Re: Xserver hosed after 2.6.29 kernel upgrade

#8 2009-04-20 23:34:54

Re: Xserver hosed after 2.6.29 kernel upgrade

#9 2009-04-23 21:46:07

Re: Xserver hosed after 2.6.29 kernel upgrade

#10 2009-04-24 16:18:16

Re: Xserver hosed after 2.6.29 kernel upgrade

#11 2009-04-24 19:25:32

Re: Xserver hosed after 2.6.29 kernel upgrade

#12 2009-04-26 01:00:19

Re: Xserver hosed after 2.6.29 kernel upgrade

#13 2009-04-26 05:57:30

Re: Xserver hosed after 2.6.29 kernel upgrade

Board footer