Fedora 11: X Windows configuration gets messed up after reboot (X server crashing)

maverick280857 · #1 12th August 2009, 09:03 AM

Hi everyone,

I recently installed Fedora 11 on my machine, the configuration of which is: Intel Core i7 920 CPU, 3 GB DDR3 SDRAM, nVidia GTX295 Graphics Card.

I initially used the default kmod-nvidia driver, but with a yum update, somehow a non-PAE kernel got installed and the kmod-nvidia driver did not get installed with it during the update for some reason. So I uninstalled the PAE kernel and associated kmod drivers, and downloaded the nVidia driver from the nVidia website. It ran just fine for a few days, and then all of a sudden, after a reboot, the X Server crashed and I had to log into the text mode, run

Code:

setup

and do the default X configuration. Then, I re-installed kmod-nvidia, with no luck.

Sorry for being so vague about this, but I finally managed to get the CUDA driver (I need to use CUDA for my project work) to work (downloaded from http://www.nvidia.com/object/cuda_get.html). But after one reboot, the same problem resurfaced.

Now, I am in X Windows with a default 800x600 configuration using a generic visa driver, on a 23" monitor

I would appreciate help and advice to sort out this matter. I have been rebooting my computer all morning and trying various things to get the driver to work again.

I would be happy to furnish any details about the installed packages I have. For now, here is some information.

Thanks in advance, and someone please do reply! (SOS)

Code:

[root@farpoint vivek]# rpm -qa |grep kernel
kernel-PAEdebug-2.6.29.6-217.2.3.fc11.i686
kernel-doc-2.6.29.6-217.2.3.fc11.noarch
kernel-headers-2.6.29.6-217.2.3.fc11.i586
kernel-PAEdebug-devel-2.6.29.6-217.2.3.fc11.i686
kernel-2.6.29.6-217.2.3.fc11.i586
kernel-PAE-devel-2.6.29.6-217.2.3.fc11.i686
kerneloops-0.12-5.fc11.i586
kernel-devel-2.6.29.6-217.2.3.fc11.i586
kernel-firmware-2.6.29.6-217.2.3.fc11.noarch
kernel-PAE-2.6.29.6-217.2.3.fc11.i686
[root@farpoint vivek]#

maverick280857 · #2 12th August 2009, 04:53 PM

Ok, so I reinstalled Fedora 11.

Now, I am faced with a dilemma.

If I install kmod-nvidia, xorg-x11-nvidia-* and akmod-nvidia, the nVidia CUDA system does not work.

If I install the drivers obtained from nVidia directly, then, there is a chance of getting the error that "the kernel module nvidia.ko" could not be loaded. In this case, I reach a dead end as before.

I want to
(a) install the correct nVidia drivers on my system
(b) run CUDA

Question: What do I do? How do I install the correct nVidia driver so as to be able to use CUDA, and also not run into the situation of having a crashed X server every time my system reboots?

Code:

[root@farpoint vivek]# uname -a
Linux farpoint 2.6.29.6-217.2.3.fc11.i686.PAE #1 SMP Wed Jul 29 16:05:22 EDT 2009 i686 i686 i386 GNU/Linux
[root@farpoint vivek]#

***Hlingler*** · #3 12th August 2009, 05:06 PM

I would suggest removing one or the other of the "vanilla" .i586 (for lack of a better term) or PAE .i686 kernel[-devel]s. Since you seem to not need the PAE stuff, this option might be best, and you can "exclude" these packages from auto-updates (or disable auto-updates). Note that removing kernel-PAE* will make it much harder if you later add RAM. You might therefore want to remove the "vanilla" .i586 kernel stuff instead. Pick one or the other. The (effectively) duplicate/redundant, and possibly conflicting, kernel-devel packages may be interfering with NVidia kmods.

V

maverick280857 · #4 12th August 2009, 05:29 PM

Hi Hlinger, thanks for replying. I now have only the default PAE kernel that comes with FC11, as I just did a complete reinstall. I do not have the vanilla kernel anymore.

Now, if I do

Code:

yum install kmod-nvidia

I see options to download kmod-nvidia not for the PAE kernel, but for the vanilla kernel. This then forces yum to download -- for dependency resolution -- the vanilla kernel. I think that is when problems begin to crop up.

So to be sure, right now, I have a raw FC11 system with default drivers. What course of action do you suggest I take?

(Also, the kmod-nvidia drivers are outdated compared to the drivers nVidia is supplying on their drivers page, and also on the CUDA page.)

maverick280857 · #5 12th August 2009, 06:21 PM

Okay, I did

Code:

yum install kmod-nvidia-PAE

and restarted the system. The X server has crashed again

and when I log in (through the text console), and run

Code:

startx

the X server does not start of course, displaying the error message "nvidia kernel module could not be loaded".

What next?

joliveira · #6 12th August 2009, 07:16 PM

Same here!

It's always the same, every release as at least one extremely dramatic update!

maverick280857 · #7 12th August 2009, 07:29 PM

Any ideas?

***Hlingler*** · #8 12th August 2009, 05:31 PM

Quote:

yum install kmod-nvidia

No:
yum install kmod-nvidia-PAE

V

***Hlingler*** · #9 12th August 2009, 08:17 PM

Please post results:
uname -r
rpm -qa|grep -i kernel|sort
rpm -qa|grep -i nvid|sort

And attach (do NOT post) files: /etc/X11/xorg.conf and /var/log/Xorg.0.log

V

maverick280857 · #10 13th August 2009, 02:40 AM

Quote:

Originally Posted by Hlingler

Please post results:
uname -r
rpm -qa|grep -i kernel|sort
rpm -qa|grep -i nvid|sort

And attach (do NOT post) files: /etc/X11/xorg.conf and /var/log/Xorg.0.log

V

Ok, here goes

Code:

[vivek@farpoint ~]$ uname -r
2.6.29.6-217.2.3.fc11.i686.PAE

Code:

[vivek@farpoint ~]$ rpm -qa |grep -i kernel|sort
arm-gp2x-linux-kernel-headers-2.6.12.0-3.fc11.noarch
kernel-firmware-2.6.29.6-217.2.3.fc11.noarch
kernel-headers-2.6.29.6-217.2.3.fc11.i586
kerneloops-0.12-5.fc11.i586
kernel-PAE-2.6.29.6-217.2.3.fc11.i686
kernel-PAE-devel-2.6.29.6-217.2.3.fc11.i686
[vivek@farpoint ~]$

Code:

[vivek@farpoint ~]$ rpm -qa|grep -i nvid|sort
akmod-nvidia-185.18.14-1.fc11.i686
kmod-nvidia-2.6.29.6-217.2.3.fc11.i686.PAE-185.18.14-1.fc11.3.i686
kmod-nvidia-PAE-185.18.14-1.fc11.3.i686
xorg-x11-drv-nvidia-185.18.14-3.fc11.i586
xorg-x11-drv-nvidia-libs-185.18.14-3.fc11.i586
[vivek@farpoint ~]$

I have attached the files just as you instructed, but the Fedoraforum uploader had some issues with the .log extension so I have renamed the Xorg.0.log file to Xorg-0.txt

Please let me know what you think.

***Hlingler*** · #11 13th August 2009, 02:55 AM

This may be a problem specific to this version of the NVidia driver RPM, but try:

Code:

Section "Files"
    ModulePath      "/usr/lib/xorg/modules/extensions/nvidia"
    ModulePath      "/usr/lib/xorg/modules/extensions"
    ModulePath      "/usr/lib/xorg/modules"
EndSection

Re-boot.

If that fails: follow these instructions: http://forums.fedoraforum.org/showpo...91&postcount=4
Re-boot.

V

maverick280857 · #12 13th August 2009, 03:20 AM

Doing this

Code:

Section "Files"
    ModulePath      "/usr/lib/xorg/modules/extensions/nvidia"
    ModulePath      "/usr/lib/xorg/modules/extensions"
    ModulePath      "/usr/lib/xorg/modules"
EndSection

and rebooting, did not help. Then, I tried the commands given on the link you posted. The last command gave me this output:

Code:

[root@farpoint Download]# yum --nogpgcheck install nvidia-x11-drv-185.18.31-1.fc11.i586.rpm
Loaded plugins: fastestmirror, presto, refresh-packagekit
Loading mirror speeds from cached hostfile
 * fedora: mirror.cse.iitk.ac.in
 * livna: ftp-stud.fht-esslingen.de
 * rpmfusion-free: mirror.cse.iitk.ac.in
 * rpmfusion-free-updates: mirror.cse.iitk.ac.in
 * rpmfusion-nonfree: mirror.cse.iitk.ac.in
 * rpmfusion-nonfree-updates: mirror.cse.iitk.ac.in
 * updates: ftp.jaist.ac.jp
Setting up Install Process
Examining nvidia-x11-drv-185.18.31-1.fc11.i586.rpm: nvidia-x11-drv-185.18.31-1.fc11.i586
Marking nvidia-x11-drv-185.18.31-1.fc11.i586.rpm to be installed
Resolving Dependencies
--> Running transaction check
---> Package nvidia-x11-drv.i586 0:185.18.31-1.fc11 set to be updated
--> Finished Dependency Resolution

Dependencies Resolved

================================================================================
 Package
     Arch   Version               Repository                               Size
================================================================================
Installing:
 nvidia-x11-drv
     i586   185.18.31-1.fc11      /nvidia-x11-drv-185.18.31-1.fc11.i586    40 M

Transaction Summary
================================================================================
Install      1 Package(s)         
Update       0 Package(s)         
Remove       0 Package(s)         

Total size: 40 M
Is this ok [y/N]: y
Downloading Packages:
Running rpm_check_debug
Running Transaction Test
Finished Transaction Test
Transaction Test Succeeded
Running Transaction
  Installing     : nvidia-x11-drv-185.18.31-1.fc11.i586                     1/1 
*** glibc detected *** /usr/bin/python: free(): invalid pointer: 0x0014624a ***
======= Backtrace: =========
/lib/libc.so.6[0x2292a1]
/usr/lib/python2.6/site-packages/ixf86configmodule.so[0x140a57]
/usr/lib/python2.6/site-packages/ixf86configmodule.so[0x13d1ab]
/usr/lib/python2.6/site-packages/ixf86configmodule.so[0x1390a5]
/usr/lib/libpython2.6.so.1.0[0x7fc391]
/usr/lib/libpython2.6.so.1.0(PyDict_SetItem+0x81)[0x7ff0a1]
/usr/lib/libpython2.6.so.1.0(_PyModule_Clear+0x16e)[0x800f3e]
/usr/lib/libpython2.6.so.1.0(PyImport_Cleanup+0x409)[0x870c49]
/usr/lib/libpython2.6.so.1.0(Py_Finalize+0xb5)[0x879c85]
/usr/lib/libpython2.6.so.1.0(Py_Main+0x587)[0x886f97]
/usr/bin/python(main+0x27)[0x80485c7]
/lib/libc.so.6(__libc_start_main+0xe6)[0x1cfa66]
/usr/bin/python[0x8048501]
======= Memory map: ========
00133000-0014a000 r-xp 00000000 fd:00 217948     /usr/lib/python2.6/site-packages/ixf86configmodule.so
0014a000-0014e000 rw-p 00016000 fd:00 217948     /usr/lib/python2.6/site-packages/ixf86configmodule.so
0014e000-0018e000 r-xp 00000000 fd:00 89526      /lib/libgobject-2.0.so.0.2000.4
0018e000-0018f000 rw-p 00040000 fd:00 89526      /lib/libgobject-2.0.so.0.2000.4
0019e000-001b4000 r-xp 00000000 fd:00 22111      /lib/libpthread-2.10.1.so
001b4000-001b5000 ---p 00016000 fd:00 22111      /lib/libpthread-2.10.1.so
001b5000-001b6000 r--p 00016000 fd:00 22111      /lib/libpthread-2.10.1.so
001b6000-001b7000 rw-p 00017000 fd:00 22111      /lib/libpthread-2.10.1.so
001b7000-001b9000 rw-p 001b7000 00:00 0 
001b9000-00324000 r-xp 00000000 fd:00 22087      /lib/libc-2.10.1.so
00324000-00326000 r--p 0016b000 fd:00 22087      /lib/libc-2.10.1.so
00326000-00327000 rw-p 0016d000 fd:00 22087      /lib/libc-2.10.1.so
00327000-0032a000 rw-p 00327000 00:00 0 
00468000-0046c000 r-xp 00000000 fd:00 198360     /usr/lib/python2.6/lib-dynload/stropmodule.so
0046c000-0046e000 rw-p 00004000 fd:00 198360     /usr/lib/python2.6/lib-dynload/stropmodule.so
005dc000-005dd000 r-xp 005dc000 00:00 0          [vdso]
006d0000-006fa000 r-xp 00000000 fd:00 22056      /lib/libgcc_s-4.4.0-20090506.so.1
006fa000-006fb000 rw-p 00029000 fd:00 22056      /lib/libgcc_s-4.4.0-20090506.so.1
0070c000-0070f000 r-xp 00000000 fd:00 89524      /lib/libgmodule-2.0.so.0.2000.4
0070f000-00710000 rw-p 00002000 fd:00 89524      /lib/libgmodule-2.0.so.0.2000.4
00799000-008e5000 r-xp 00000000 fd:00 182855     /usr/lib/libpython2.6.so.1.0
008e5000-00910000 rw-p 0014c000 fd:00 182855     /usr/lib/libpython2.6.so.1.0
00910000-00919000 rw-p 00910000 00:00 0 
00c2d000-00c53000 r-xp 00000000 fd:00 22095      /lib/libm-2.10.1.so
00c53000-00c54000 r--p 00025000 fd:00 22095      /lib/libm-2.10.1.so
00c54000-00c55000 rw-p 00026000 fd:00 22095      /lib/libm-2.10.1.so
00d48000-00d4a000 r-xp 00000000 fd:00 22119      /lib/libutil-2.10.1.so
00d4a000-00d4b000 r--p 00001000 fd:00 22119      /lib/libutil-2.10.1.so
00d4b000-00d4c000 rw-p 00002000 fd:00 22119      /lib/libutil-2.10.1.so
00d61000-00d64000 r-xp 00000000 fd:00 22093      /lib/libdl-2.10.1.so
00d64000-00d65000 r--p 00002000 fd:00 22093      /lib/libdl-2.10.1.so
00d65000-00d66000 rw-p 00003000 fd:00 22093      /lib/libdl-2.10.1.so
00eab000-00ecb000 r-xp 00000000 fd:00 22080      /lib/ld-2.10.1.so
00ecb000-00ecc000 r--p 0001f000 fd:00 22080      /lib/ld-2.10.1.so
00ecc000-00ecd000 rw-p 00020000 fd:00 22080      /lib/ld-2.10.1.so
00eee000-00fcb000 r-xp 00000000 fd:00 89522      /lib/libglib-2.0.so.0.2000.4
00fcb000-00fcc000 rw-p 000dd000 fd:00 89522      /lib/libglib-2.0.so.0.2000.4
08048000-08049000 r-xp 00000000 fd:00 199091     /usr/bin/python
08049000-0804a000 rw-p 00000000 fd:00 199091     /usr/bin/python
09365000-09440000 rw-p 09365000 00:00 0          [heap]
b7cc8000-b7ec8000 r--p 00000000 fd:00 25351      /usr/lib/locale/locale-archive
b7ec8000-b7f4c000 rw-p b7ec8000 00:00 0 
b7f78000-b7f79000 rw-p b7f78000 00:00 0 
bf897000-bf8af000 rw-p bffe8000 00:00 0          [stack]
/var/tmp/rpm-tmp.yLyjWX: line 7:  4913 Aborted                 /usr/sbin/nvidia-config-display enable

Installed:
  nvidia-x11-drv.i586 0:185.18.31-1.fc11                                           

Complete!
[root@farpoint Download]#

Should I be concerned about these messages? Also what about "/var/tmp/rpm-tmp.yLyjWX: line 7: 4913 Aborted /usr/sbin/nvidia-config-display enable"

Anyway, I'm going to reboot to see if it works.

***Hlingler*** · #13 13th August 2009, 03:29 AM

Those "error" messages are indeed odd, and should not be - but I would not panic (yet). See what happens....

V

maverick280857 · #14 13th August 2009, 03:33 AM

It did not work. Here is the new log file (attached).

What next?

maverick280857 · #15 13th August 2009, 05:51 AM

Is it advisable to remove the PAE kernel altogether, installing something else, and then re-installing the NVIDIA drivers?

Why can't I use the drivers (in the pkg.run file) given by nVidia themselves? They don't seem to be able to compile. I think the problem is that they are not designed for PAE kernels, and so do not detect the kernel sources for the PAE kernel.

I think I will need to use nVidia's proprietary drivers to use CUDA. A side question: does anyone here have experience of using an nVidia driver on a PAE kernel to run CUDA?