[ On Friday, February 5, 1999 at 09:29:56 (+0200), Jukka Marin wrote: ]
> Subject: Re: inn 2.1 dies
>
> > > innd is not setuid.. so I guess this patch wouldn't help. :-/
> > 
> > You never know until you try it!  (There can be many reasons for a core
> > dump to fail, as the patch indicates.)
> 
> Now I have news expire (the expire command) dying as well.  (Yeah, sounds
> bad, maybe a hardware problem (although this system was 100% stable under
> StunOS)).  In the news expire log, I get a message "segmentation violation -
> core dumped" but I can't find the core file anywhere.  Do you know why this
> might be?

That's even more weird.  I can't quite imagine how it could report "core
dumped" and not actually have produced a core file.

Unless you're running -current and have "sysctl kern.shortcorename=0"
then the file will be named "<program>.core".

One thing I lament about in a comment in my patch is that it's not easy
to log the process' current working directory.  I suppose even it's
inode number and device id would be better than nothing....

> This system is about 50 km from here, but I guess I'll have to go there
> and do the usual "pull everything apart and put it back together" kind of
> fix (hopefully, it's just some SIMM become loose..)..

That could be -- however memory errors should be detected unless you've
put faked parity memory into the machine.  I'm not 100% sure the parity
detection works 100% on older sparcstations though -- I wish I had a
test SIMM with a guaranteed bad bit that I could use to excercise it.

-- 
							Greg A. Woods

+1 416 218-0098      VE3TCP      <gwoods@acm.org>      <robohack!woods>
Planix, Inc. <woods@planix.com>; Secrets of the Weird <woods@weird.com>