Segfaults

[Nancy Harrison]

From:           	"Nancy Harrison" <vulpia at sonic.net>
To:             	nblug-talk at lists.sonic.net
Date sent:      	Mon, 14 Feb 2000 07:37:16 -800
Subject:        	Segfaults
Send reply to:  	nblug-talk at lists.sonic.net

> Last Sat Devin (aqua) most kindly helped me to try and install
> Debian, but my Pentium Pro seemed to crash out with segmentation
> faults. Devin suggested I have my RAM checked, but later I remembered
> that the RAM had been checked, and came out OK, several months
> ago. Since that time I have not used the box much at all.
>   So - my question is - what else, besides bad RAM chips, causes
> segfaults? I tried checking on Deja, but most of the posts were in
> Czech, Italian or Polish <g> - I can kind of grok the Italian, but
> not the others!  I know programmers can suffer segfaults with
> bad pointers, and that a program can segfault if it tries to
> access memory it has no permission to access, but what does
> this all entail? - Grandma
> -Nancy Harrison, Life Sciences Dept., Santa Rosa JC
> http://www.sonic.net/~vulpia/index.html
> http://www.sonic.net/~vulpia/cnps/mbaker.html (Calif.native plants)
> 

Thanks all, for that long list of what causes segfaults. I have wiped
the drive on the Pentium Pro (except for the first primary DOS partition)
and am going to start experimenting with installs. 
   Before, I was able to install earlier Red Hat dists with no problem; 
also Caldera (with some glitches but it got on there), but both
SuSE (2 different dists) and Debian (last Sat) had the same
problems during the package install phase.
   I'm wondering now if it is how they use memory while installing.
Devin did mention that Debian does it differently from Red Hat -
unwrapping all the packages first, then you configure them afterwards.
There may be some heating problems someplace in my hardware,
but it sure isn't "overclocked" (I wouldn't have a clue how to do that.)

  I have 2 Debian CD's, both Slink. I will try the other one, just in
case something was corrupted on the VA one. I will also try Red Hat
and see what happens. Red Hat 6 had a problem with my uni-processor
on the Tyan Board - I found the solution in a newsgroup (not from RH,
unfortunately) which is to tell LILO "linux -up" instead of "linux" when
you boot. Then you can reconfigure LILO to use the single-processor
mode as the default. This was all done last year, so it's time to
start over, hoping for good luck this time! - G'ma
-Nancy Harrison, Life Sciences Dept., Santa Rosa JC
http://www.sonic.net/~vulpia/index.html
http://www.sonic.net/~vulpia/cnps/mbaker.html (Calif.native plants)