My original concern was that my machine was suddenly panic dumping after
I had put in Slip and the integer patch from Sun. I posted mu dump
traceback message and asked for help .
kevins@aus.sun.com writes:
> Let me put it this way - *NOBODY* can read these things when they
>are a bunch of numbers - without your kernel, it means nothing. If you
>are going to post a traceback, post a *SYMBOLIC* traceback.
>...
I do apologise but for many of us out here in Sun country, it's the
only info we have in a desperate situation. But once again the net comes
to the rescue. I also got one message saying that I should do savecore
and then run adb as in
echo '$c' | adb -k vmunix.* vmcore.*
on the dumped savecore files. I have tried this, and I understand why
people don't do savecore- it cost 10MB of disk space.
However, Chris.Drake@Corp.Sun.COM came through with a better suggestion,
namely to run adb on your present /vmunix. If teh problem is in the
kernel ( as mine was) that might give you info. If it was in a user
program, it's of course useless.
You run adb /vmunix
Then eneter in the traceback addresses just after "Called from" with ?i
right after
> adb /vmunix
> 0xf81147e4?i
> f8113d7c?i
> etc.
In my case this made it clear that the call was coming from the slip
kernel routines, not the integer patch.
Solution? Get and run the latest version of slip. I had 1.14. The latest
(Dec 89) of Zachariessen's slip is 1.15. available in slipware.tex.Z
from ftp.cs.toronto.edu
I had been testing slip, bringing it up and then killing it, and 1.14
was invoking new ports and leaving old ones not properly killed, so I guess
eventually ping tried to go out one of those improperly killed ones
(why, I don't know, but am also not going to try to figure it out).
Thanks to Chris.Drake@Corp.Sun.COM and to
Fabrice Le Metayer DOMAIN : fabrice@sj.ate.slb.com
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:18 CDT