video memory test

Hi!

I need to test my graphic card memory, because I suspect video memory errors.
My LCD shows unclear image (like a distortion effect!) and vertical lines. It also happens connecting my laptop to an external CRT monitor.
No problem only working on a Terminal (X server not running). I suppose for the low memory needed.

What linux software could I run?

Hi,

Any changes to the system to produce the problem? Or was this a working system then the graphics problem occurred?

Hi!
The issue appeared suddenly two days ago on a working system.
(My laptop has been fully working for six years!! )
I noticed a screen corruption/distortion. Initially, I supposed it could be a driver issue so I tried to restart X, but no success. I guess the LCD is ok since the same issue appears connecting the laptop to an the external CRT.
I'm looking for a tool like memetest86 but related and specific for video cards.

Is your laptop video card a Geforce 8000 series?
There were a lot of faulty ones some years ago...

Hi,

Originally Posted by 4Strings Hi! The issue appeared suddenly two days ago on a working system. (My laptop has been fully working for six years!! ) I noticed a screen corruption/distortion. Initially, I supposed it could be a driver issue so I tried to restart X, but no success. I guess the LCD is ok since the same issue appears connecting the laptop to an the external CRT. I'm looking for a tool like memetest86 but related and specific for video cards.

Look at your logs in '/var/log', especially 'Xorg.0.log' & dmesg for errors or warnings. For diagnostics I would look at the manufacture sets.

Originally Posted by 414N Is your laptop video card a Geforce 8000 series? There were a lot of faulty ones some years ago...

No, it isn't. It's an ATI X1400.

Originally Posted by onebuck Hi, Look at your logs in '/var/log', especially 'Xorg.0.log' & dmesg for errors or warnings. For diagnostics I would look at the manufacture sets.

Hi,

I don't know if the following messages/warnings can be related to this hardware issue, but I've noticed the following:
dmesg:

[9.770769] [Firmware Bug]: Duplicate ACPI video bus devices for the same VGA controller, please try module parameter "video.allow_duplicates=1"if the current driver doesn't work.

syslog:

Aug 20 10:50:41 darkstar kernel: [   91.151741] NMI: PCI system error (SERR) for reason a1 on CPU 0.
Aug 20 10:50:41 darkstar kernel: [   91.151749] Dazed and confused, but trying to continue
...
Aug 20 15:18:29 darkstar kernel: [  851.465482] NMI: PCI system error (SERR) for reason b1 on CPU 0.
Aug 20 15:18:29 darkstar kernel: [  851.465488] Dazed and confused, but trying to continue

/var/log/Xorg.0.log doesn't report anything related to it.




Whenever I echoed the "/sys/class/drm/card0/device/power_profile" by entering one of the [auto], [default], [mid] or [low] profiles, I've always got the NMI error message:

echo low > /sys/class/drm/card0/device/power_profile
Uhhuh. NMI received for unknown reason a1 on CPU 0. You have some hardware problem, likely on the PCI bus.
Dazed and confused, but trying to continue

Today I've tried again to echo the "power_profile" but I've got a blank screen display and a system freeze (no MagicSysRq, no Ctrl+Alt+Fn, ....). I had to press the power-off button.



I've found out "VMT"-Video Memory Test: http://vmtce.sourceforge.net/.
It's a an opensource project under the GPL License.
I will try to test my video memory.

If the same effect happens with an external monitor then it probably is the graphics card.

I guess you could try a live CD and see if it happens there to exclude broken drivers on your current distro.

Originally Posted by H_TeXMeX_H If the same effect happens with an external monitor then it probably is the graphics card.

Hi H_TeXMeX_H, your guess is right.
The VMT test results report several video memory errors.

I burned the VMT iso image on a boot cd. Powering on the laptop, VMT prints on screen several of these warnings:

I/O card parity interrupt at 2775:3E88
Shut off NMI, Reboot or press other keys to continue

Then, the running tests report hundred of errors:

...
Error at [00098133] must be FF, but found 7F (bits: 100000000)
Error at [0009813B] must be FF, but found 7F (bits: 100000000)
Error at [0009814B] must be FF, but found 7F (bits: 100000000)
Error at [00098153] must be FF, but found 7F (bits: 100000000)
...

Now I can definitely say that my video card is dying and it's generating the last images to my lcd!

Thanks for the VMT utility, I didn't know about it and will probably need it for my main machine as it has been having some issues.