GART error

[Saurabh Barve]

Hi,

This is a follow up to my previous message.

I have a dual opteron system with 8 GB of RAM. I am running RHEL 3.0 AS 
on it. The original kernel version was 2.4.21-4.ELsmp. The motherboard I 
am using is the Tyan Thunder K8S Pro - 2882 motherboard. I was getting 
the following error every two minutes or so:

GART error 11
Lost an northbridge error
NB error address some-hex-number
Error uncorrected

In the hope of resolving this problem, I recently upgraded my system to 
the 2.4.21-15.ELsmp #1 SMP kernel. However, I keep getting the following 
error:

CPU 0: Silent Northbridge MCE
Northbridge status a40000000005001b
   GART TLB error generic level generic
   extended error gart error
   link number 0
   error address valid
   error uncorrected
   previous error lost
   error address 00000000fafe2950

With the upgraded kernel though, the error message does not appear on the terminal. There is, however, a little catch. The default login screen for the system is text-based (as opposed to a graphic login screen). When the text login screen is displayed, the error message appears continuously on the screen, similar to the previous error. However, the machine does not beep continuously like it used to earlier. But once the X windows session is started, the error messages no longer appear on the screen like they used to earlier. So, if I changed the login screen to a graphical screen, the message would get totally suppressed. However, the system logs these messages, so I need to get rid of them.

I wanted to upgrade my kernel to 2.6.x version since I read that the new 
kernels have fixes for these kinds of GART errors. I read somewhere that 
this was a hardware error. However, I have checked my hardware, and I 
don't seem to have any bad RAM or anything.

Also, does anybody know whether upgrading to the 2.6.x kernel would 
invalidate the license for RHEL that I bought from Red Hat?

Thanks.

Regards,
Saurabh.

-- 
=============================================================================                                                                                
Saurabh Barve                                        Phone:
System Administrator/Data Specialist                 970-491-7714 (voice)
Montgomery Research Group,                           970-491-8449 (Fax)
Atmospheric Sciences Department,
Fort Collins, Colorado                             
Colorado State University                         

Mail : sa at atmos.colostate.edu
Web  : http://fjortoft.atmos.colostate.edu/~sa