(solved) Issues with a PC with NVDIA GForce FX5200 (NV34); which Xorg.conf ?

[floppy]

That was the headers..

When I created the rawdata, an error came with the nvidia...run
ERROR: The runtime configuration check failed for the library 'libnvidia-tls.so.173.14.31' (expected '/usr/lib/tls/libnvidia-tls.so.1', found '/usr/lib/libnvidia-tls.so.1'). The most likely reason for this is that conflicting OpenGL libraries are installed in a location not inspected by 'nvidia-installer'. Please be sure you have uninstalled any third-party OpenGL and/or thirsd-party graphics driver packages.

Few errors are coming now.

>>> at boot (with boot code showapps), just after the loading of the newly created extension nvidia... following message come to the screen:
ldconfig=/usr/lib/ next: libstdc++.so.6.0.16-gdb.py is not a shared library, skipping

>>> in the last lines of dmesg
NVRM: The NVIDIA GeForce FX 5200 GPU installed in this system is
NVRM:  supported through the NVIDIA 173.14.xx Legacy drivers. Please
NVRM:  visit http://www.nvidia.com/object/unix.html for more
NVRM:  information.  The 280.13 NVIDIA driver will ignore
NVRM:  this GPU.  Continuing probe...
NVRM: No NVIDIA graphics adapter found!

>>> the xorglog in the attachment

Perhaps I should delete the nvidia kernel drivers, or something else?

[Rich]

Hi floppy

>>> at boot just after trying to create the extension nvidia/... i made
ldconfig=/usr/lib/ next: libstdc++.so.6.0.16-gdb.py is not a shared library, skipping

I did not need to run anything except the scripts.

>>> in the last lines of dmesg
NVRM: The NVIDIA GeForce FX 5200 GPU installed in this system is
NVRM:  supported through the NVIDIA 173.14.xx Legacy drivers. Please
NVRM:  visit http://www.nvidia.com/object/unix.html for more
NVRM:  information.  The 280.13 NVIDIA driver will ignore
NVRM:  this GPU.  Continuing probe...
NVRM: No NVIDIA graphics adapter found!

You still have nvidia-module-3.0.x-tinycore.tcz in your onboot.lst, which is the 280.13 version. The message just says
that the nvidia-module-3.0.x-tinycore.tcz extension won't load anything for your board because it doesn't contain a
suitable driver. That's expected. Remove that extension from your onboot.lst, it doesn't support your board.
I'm not completely clear on how it's going, so:
1. Does the nvidia run script complete without errors? A few warnings are normal.
2. Does the create-nvidia-raw script complete without errors?
3. Does the create-nvidia-package script complete without errors?
If you answer yes to all three, you are almost done.
Attached is my xorg.conf which you can use as starting point if you wish. Also a dep file which the script does not
create. Replace the word KERNEL with the result of the command   uname -r

[Rich]

Hi floppy
I see you added something to your previous post:

When I created the rawdata, an error came with the nvidia...run
ERROR: The runtime configuration check failed for the library 'libnvidia-tls.so.173.14.31' (expected '/usr/lib/tls/libnvidia-tls.so.1', found '/usr/lib/libnvidia-tls.so.1'). The most likely reason for this is that conflicting OpenGL libraries are installed in a location not inspected by 'nvidia-installer'. Please be sure you have uninstalled any third-party OpenGL and/or thirsd-party graphics driver packages.

That's probably because you have other nvidia extensions installed.
If the scripts look like they all ran correctly, copy the tcz and dep files to tce/optional. make sure you have an
/etc/X11/xorg.conf file. Then enter  sudo tce-load -i /path to tce/optional/nvidia-173.14.....tcz. Then startx.

  @ALL: Rather than adding information to an existing post, add a new post so that it shows up on the unread topics page.

[floppy]

Hi floppy

>>> at boot just after trying to create the extension nvidia/... i made
ldconfig=/usr/lib/ next: libstdc++.so.6.0.16-gdb.py is not a shared library, skipping

I did not need to run anything except the scripts.

this is my mistake (very late post last night): the correct explanation would have been:

>>> at boot (with boot code showapps), just after the loading of the newly created extension nvidia... following message come to the screen:
"ldconfig=/usr/lib/ next: libstdc++.so.6.0.16-gdb.py is not a shared library, skipping"

so,
1) error during the rawdata creation with the "nvidia run" programm (see before)
2) raw script completed without error
3) package creation script completed without error
only errors during extension looading (see above) and error in dmesg.

I will delete the nvidia-module-3.0.x-tinycore.tcz extension as you suggested and post any progress here.
Thanks rich for your undestructible patience.

[Juanito]

libstdc++.so.6.0.16-gdb.py is not a shared library, skipping

You can safely ignore this message

[floppy]

Thanks all.
So, now the pc boots without error when no xorg.conf is created (the extension I made yesterday is so far ok; it has still to be fully tested with a running conf file; the scripts from rich are so far ok).
Im now facing a setup issue with my xorg.conf
The screen gets black or freeze as soon I start the pc with an xorg.conf file (conf file from rich or from the attachment) .. and cannot access any xorg.0.log file.
Im still searching the reason and a method how to debug the conf file.
Any comment  is welcome.

[Rich]

Hi floppy
In the dep file I supplied, did you remember to replace the word KERNEL with the result of the command   uname -r
This is to insure that the correct  graphics-x.x.x-tinycore.tcz  gets loaded.
Notes:
1. From the black screen, see if  Ctrl-Alt-Backspace  kicks you back into the console.
2. I used Xorg7.5. Xorg7.6 would freeze up on me.
3. Creating xorg.conf with
      A. sudo Xorg -configure  or
      B. sudo Xorg -configure  followed by  sudo nvidia-xconfig
   both resulted in a black screen. I had to use  sudo nvidia-xconfig --virtual=1920x1080 --mode=1920x1080 --force-generate
   The  --force-generate  option makes it delete the existing  xorg.conf and create a new one.
   I needed the  --virtual  and  --mode  switches to get my screen sized correctly.

[floppy]

progress made with the "rich" instructions.
xorg file and its log.
Im still searching the setup.
(screen is still hanging up)

[floppy]

wrong log file in the previous post  because it was overwriten by a startx with a vesa conf file.
Correct one is here.

[Rich]

Hi floppy
Try replacing the  Screen  section in your  xorg.conf  with:

Code: [Select]

Section "Screen"
    Identifier     "Screen0"
    Device         "Device0"
    Monitor        "Monitor0"
    DefaultDepth    24
    SubSection     "Display"
        Virtual     1024 768
        Depth       24
        Modes      "1024x768" "800x600"
    EndSubSection
EndSection
You can also try commenting out the  Modules  section in case one of the modules is causing a conflict.

[floppy]

a) xorg.conf.new file (see attachment) still gives a black screen
after sudo X -config /etc/X11/xorg.conf.new
I go back to the prompt with strg-alt-backspace
End of dmesg
...
ACPI: resource 0000:00:02.1 [io  0x0c00-0x0c1f] conflicts with ACPI region SMRG [io 0xc00-0xc1f]
ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver
agpgart-sis 0000:00:00.0: SiS chipset [1039/0746]
agpgart-sis 0000:00:00.0: AGP aperture is 64M @ 0xd0000000
nvidia 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
vgaarb: device changed decodes: PCI:0000:01:00.0,olddecodes=io+mem,decodes=none:owns=io+mem
NVRM: loading NVIDIA UNIX x86 Kernel Module  173.14.31  Sun Jul 17 22:11:28 PDT 2011
eth0: Media Link On 100mbps full-duplex
EXT4-fs (sda1): error count: 23
EXT4-fs (sda1): initial error at 1333886435: ext4_lookup:1044: inode 333932
EXT4-fs (sda1): last error at 1333889797: ext4_mb_generate_buddy:736
agpgart-sis 0000:00:00.0: AGP 3.0 bridge
agpgart-sis 0000:00:00.0: putting AGP V3 device at 0000:00:00.0 into 8x mode
agpgart-sis 0000:00:00.0: SiS delay workaround: giving bridge time to recover
agpgart-sis 0000:00:00.0: putting AGP V3 device at 0000:01:00.0 into 8x mode

b) acpi=off made an hangup.

c) When I then included hsync and Vrefresh params, a Nividia pictures appears on the screen (nomore black) after sudo X -config /etc/X11/xorg.conf.new
But it hangs up after startx.

[Rich]

Hi floppy
I'm running out of ideas here. The differences I see between what you have and what I have is that I don't have
a  Files  or  Modules  section in my xorg.conf and I'm running a 3.0.3 kernel. I could not run Xorg7.6 because it
would freeze up on me, but maybe it will work better for you.

[floppy]

Hi floppy
I'm running out of ideas here. The differences I see between what you have and what I have is that I don't have
a  Files  or  Modules  section in my xorg.conf and I'm running a 3.0.3 kernel. I could not run Xorg7.6 because it
would freeze up on me, but maybe it will work better for you.

thanks for the help!
This topic is probably like a good bottle of a bordeaux-wine: dont touch it, think about it, look at it.. it need time before it become mature. So, no panic at all.. the solution is there.. here.. I just have to find out where exactly.
My impression is the nvidia drivers are extremely sensitive with the xorg.conf format. I will delete the modules and files.
The Xorg7.5.tcz posted is from 2.6.33. Perhaps a new compilation with 3.0.3 or 21 could be good (Im running kernel 3.0.21).

[Rich]

Hi floppy
In the  Device  section of your xorg.conf, try adding
Option   "NvAGP" "0"
and see if that fixes the freezing problem.

[floppy]

according your recommendation and http://us.download.nvidia.com/XFree86/Linux-x86/173.14.31/README/chapter-12.html

Tried 0 and 1. Freeze with black screen (no PC freeze).
2 gives a pc freeze (nvidia picture comes up).
3 gives a screen freeze with black screen (no pc freeze).

Seems that somebody
- putted "NvAGP" "1"
- blacklisted the agpgart and sis_agp in ubuntu: http://forum.ubuntuusers.de/topic/dunkler-bildschirm-mit-nvidia-geforce-fx-5200/?highlight=x11#post-164618

or here, too.  http://www.mail-archive.com/linuxkernelnewbies@googlegroups.com/msg00437.html

Next try was:  (my chip is a sis746)
- according the first link in that post, the FX 5200 should use the AGPGART. -> "NvAGP"  "3".
- sis_agp blacklisted at boot
-> PC freeze

I will try "NvAGP"  "1" and blacklist AGPGART + SIS_AGP.