Re: System freeze after random program crashes associated with kernel oops

[Kristoffer Erlandsson <krier115@student.liu.se>]

On Wed, Jun 11, 2003 at 05:12:48PM -0400, Derrick 'dman' Hudson wrote:
> On Tue, Jun 10, 2003 at 10:43:55AM +0200, Kristoffer Erlandsson wrote:
> [...]
> 
> | Western Digital IDE ATA-100 60GB 7200RPM 2MB cache (WD600BB) (Running
> | one ntfs partition and one ext2 with the operating system in question
>       ^^^^
> | on)
> 
> Are you, by any chance, trying to read from the NTFS partition when
> the crashes occur?  NTFS support in linux is quite experimental (at
> least the last time I checked) largely due to MS not providing
> documentation.  It's possible that the ntfs module is causing the
> problems.  OTOH, since you recall heavy disk activity at the time of
> the crash, that the driver module for your disk(s) and/or
> controller(s) is causing the problem.
> 
> | A sample of two oopses:
> | (Maybe worth noteing is that all oopses I have looked at end with
> | an instruction accessing %e*x, where * is a letter a-d. I do not know
> 
> Those are the designations for general purpose registers on x86
> processors.

Thank you for your reply Derrick. However, I think I just found the
problem. I found a newer version of memtest86 which I got to work.
It revealed some small errors in the far end of one of my memory
modules. I now hope that this was the only problem and am testing the
computer under heavy activities with only the module on which memtest86
didn't find any errors. The thing that worries me is that I recall at
least one freeze with only this module too, when I tested it. But I'm
hopeing that this was because of something else (maybe nvidia's
drivers?). I'm not sure I got the same oopses in that crash now that I
think of it, I blatantly assumed it was the same sort of crash.

I hardly ever read from the NTFS partition, so it seems highly unlikely
that this causes the problem, but thank you for your suggestion. It only
contains windows and windows software since my old life. :)

I'm guessing time will tell if this memory module was the only/real
problem. I will reply here after I've tested the computer and am sure,
so you all get to know how it went.

Regards,
Kristoffer Erlandsson