1 |
Kyle Liddell wrote: |
2 |
|
3 |
>Here's just an idea: You took out half your RAM, and then more recently |
4 |
>you tested the other half of the RAM that you left in the computer? How |
5 |
>about you swap the RAM around and run memtest86+ or some such. Perhaps |
6 |
>some of the RAM you took out of your system is bad, and you've got a |
7 |
>corrupted file or something lying around? I had this happen on my old |
8 |
>server...had very occasional weird errors, but worked enough to compile |
9 |
>a stage1 install, then when the RAM completely died and things went nuts |
10 |
>I checked it with memtest86 and found that a stick was bad, dumped it, |
11 |
>but still had strange crashes every once in a while. Completely |
12 |
>reinstalling (switching to debian actually) stopped the problems. |
13 |
> |
14 |
|
15 |
Interesting idea, that. The other day in my clumsiness I managed to |
16 |
knock the spare sticks onto the tiled floor, so now I can't trust them |
17 |
not to have PCB cracks. Secondly, I'm still getting segmentation faults |
18 |
and that one bus error I mentioned, and in the meantime I've zapped the |
19 |
root partition of my Xfce system and rebuilt it from scratch. So on |
20 |
balance I think I won't pursue your suggestion, thanks all the same. I |
21 |
did enjoy reading about your experience though. |
22 |
|
23 |
-- |
24 |
Rgds |
25 |
Peter Humphrey |
26 |
Linux Counter 5290, Aug 93. |
27 |
|
28 |
-- |
29 |
gentoo-amd64@g.o mailing list |