Re: radeon KMS causes GART Table Walk Errors (was: K8 ECC error with linux-2.6.32)

[Borislav Petkov <borislav.petkov@amd.com>]

On Thu, Dec 17, 2009 at 08:03:29PM +0100, Johannes Hirte wrote:
> GART Error Reporting was disabled. Here is the output after enabling it:
>
> datengrab ~ # lsmsr MC4 -V3
> MC4_CTL              = 0x0000000000003bff
>   CorrEccEn=0x1
>   UnCorrEccEn=0x1
>   CrcErr0En=0x1
>   CrcErr1En=0x1
>   CrcErr2En=0x1
>   SyncPkt0En=0x1
>   SyncPkt1En=0x1
>   SyncPkt2En=0x1
>   MstrAbrtEn=0x1
>   TgtAbrtEn=0x1
>   GartTblWkEn=0
>   AtomicRMWEn=0x1
>   WchDogTmrEn=0x1
>   DramParEn=0

[.. ]

> MC4_CTL_MASK         = 0x0000000000000000
>   CorrEccEn=0
>   UnCorrEccEn=0
>   CrcErr0En=0
>   CrcErr1En=0
>   CrcErr2En=0
>   SyncPkt0En=0
>   SyncPkt1En=0
>   SyncPkt2En=0
>   MstrAbrtEn=0
>   TgtAbrtEn=0
>   GartTblWkEn=0

Ok, thanks for testing. It looks like your BIOS is applying the wrong
workaround when the option is enabled. It clears MC4_CTL[GartTblWkEn],
which means, it disables reporting of GART table walk errors while
they're still being logged by the hw. What it should do is to set
MC4_CTL_MASK[GartTblWkEn] to 1 so that logging gets disabled, as it is
recommended in the K8 BKDG.

-- 
Regards/Gruss,
Boris.

Operating | Advanced Micro Devices GmbH
  System  | Karl-Hammerschmidt-Str. 34, 85609 Dornach b. München, Germany
 Research | Geschäftsführer: Andrew Bowd, Thomas M. McCoy, Giuliano Meroni
  Center  | Sitz: Dornach, Gemeinde Aschheim, Landkreis München
  (OSRC)  | Registergericht München, HRB Nr. 43632