*BSD News Article 63337


Return to BSD News archive

Newsgroups: comp.unix.bsd.freebsd.misc,comp.mail.smail
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.mel.connect.com.au!munnari.OZ.AU!spool.mu.edu!howland.reston.ans.net!gatech!newsfeed.internetmci.com!in1.uu.net!news.nkn.net!rowdy.lonestar.org!nemesis!uhclem
From: uhclem@nemesis.lonestar.org (Frank Durda IV)
Subject: Re: smailbombs
X-Newsreader: Tin 1.1 PL5
Organization: The Big Blue Box
Message-ID: <DnIvCD.4Ho@nemesis.lonestar.org>
References: <4h01uc$b6q@kryten.awinc.com>
Date: Thu, 29 Feb 1996 05:31:24 GMT
Lines: 42
Xref: euryale.cc.adfa.oz.au comp.unix.bsd.freebsd.misc:15286 comp.mail.smail:2136

Troy Ihmels (troy@awinc.com) wrote:
[0]not sure if this is smail or FreeBSD related... Running FreeBSD port of 
[0]smail 3.1.29.1.  Every few minutes the following happens:-
[0]
[0](from syslog) mail /kernel: pid 8769: smail: uid 19218: exited on signal 11
[0]
[0]obviously this generates a core file in the process.
[0]We've tried swapping the RAM, then the entire machine except hard drive.  No 
[0]joy.

Look in /usr/spool/smail/input (or wherever the input directory is) and
see if there are any zero-length files in there.  If there are, delete them.

The errors should go away, at least until the next system crash that
leaves new zero-length files in the input directory for smail to find.
Strangely, new mail is processed correctly and despite the core dumps.

In my case I run smail -bd -q2m so I get one of those errors every two
minutes until I clean up the zero-length files.  I have 3.1.29 as ported
to FreeBSD and found in the 2.1 tree.

I don't know why smail croaks and I also found that smail under SCO UNIX
didn't get a signal 11 in this same situation.   I don't know if it was
because SCO UNIX SMAIL ignored and deleted the zero-length files or the
these empty files never get created in the first place on SCO.

Perhaps some SMAIL guru will recognize this symptom and have a better
solution.

On a related note, I also have seen cnews sendbatch agents also abort on
signal 10 or 6 depending on the content of the article being processed
on FreeBSD, but the programs ran without complaint (on the same batch
of articles) on SCO.  The program is probably at fault, but cnews never
leaves a core file for me to examine.  This is not a hardware problem.

Hope this helps.

Frank Durda IV <uhclem@nemesis.lonestar.org>|"The Knights who say "LETNi"
or uhclem%nemesis@rwsystr.nkn.net           | demand...  A SEGMENT REGISTER!!!"
					    |"A what?"
or ...letni!rwsys!nemesis!uhclem	    |"LETNi! LETNi! LETNi!"  - 1983