Jules Dubois wrote:
On Thursday 18 August 2005 09:49, Hugo Vanwoerkom <hvw59601@care2.com> (<[🔎] de2amb$85e$1@sea.gmane.org>) wrote:This is for info in case anybody googles for this motherboard, by chance.It's not very good information, except that you have what appear to be hardware problems. What you said was insufficient to diagnose the defect but you assigned blame anyway.Who knows whether it is the CPU, the mobo or the memory, not a clue.Raise the voltage on the memory by 0.05V. It fixed those problems for me.
Thanks. I'll try that. There is very little information to provide! It is an Epox EP-8VTAI mobo and an XP 2700+ Thoroughbred CPU.Install is painless except for the lm_sensors support for the Winbond w83687thf but the lm_sensors people provided a patch and I am monitoring now.
The failure is either a segfault in a compiler, usually gcc (Everything runs Sarge. But it will happen in Sid also)or a kernel panic someplace that may or may not show up on your display. No trace anywhere else.
I have been running 2.6.11 for 2 days now without a hitch: run compiles etc.It is definitely related to these 2 pieces of equipment because with the same software (2.4.31 - 2.6.13 + Sarge or Sid) the old KT7A + Athlon Thunderbird was rocksolid.
But where, there is the question. The segfaults provide no info and go away again when you try it the next day.
The kernel panics point to a failure, but lead to a dead end because you cannot repeat it.
A hardware info database for Linux would be nice, that is up to date. Thanks again for the reply. H.