[Mimedefang] MimeDefang: Perl core dump

[Martin Schweizer]

Hello Kevin

Sorry for the delay but to track down the problem I needed some time.
Now the situation is little bit clearer. Here are the details:

My system is a FreeBSD acsvfbsd06.acutronic.ch 7.2-RELEASE FreeBSD
7.2-RELEASE #0: Thu Sep 17 17:53:59 CEST 2009
martin at acsvfbsd06.acutronic.ch:/usr/obj/usr/src/sys/GENERIC  amd64.

If I use Perl version 5.8.9 then I get in the syslog
(sorry for the wrapping lines)

Dec  3 06:32:43 acsvfbsd06 mimedefang-multiplexor[77805]: Starting
slave 1 (pid 26533) (1 running): Bringing slaves up to minSlaves (2)
Dec  3 06:32:43 acsvfbsd06 mimedefang-multiplexor[77805]: Reap: slave
1 (pid 26533) exited normally with status 1 (SLAVE DIED UNEXPECTEDLY)
Dec  3 06:32:43 acsvfbsd06 mimedefang-multiplexor[77805]: Slave 1
resource usage: req=0, scans=0, user=0.000, sys=0.000, nswap=0,
majflt=0, minflt=12, maxrss=0, bi=0, bo=0


If I use Perl version 5.10.1 then I get in the syslog
(sorry for the wrapping lines)

Dec  3 07:36:09 acsvfbsd06 mimedefang-multiplexor[7703]: Starting
slave 0 (pid 7736) (1 running): Bringing slaves up to minSlaves (2)
Dec  3 07:36:09 acsvfbsd06 kernel: pid 7736 (perl), uid 26: exited on
signal 11 (core dumped)
Dec  3 07:36:09 acsvfbsd06 mimedefang-multiplexor[7703]: Reap: slave 0
(pid 7736) exited due to signal 11 (SLAVE DIED UNEXPECTEDLY)
Dec  3 07:36:09 acsvfbsd06 mimedefang-multiplexor[7703]: Slave 0
resource usage: req=0, scans=0, user=0.168, sys=0.058, nswap=0,
majflt=0, minflt=2832, maxrss=13464, bi=5, bo=18
Dec  3 07:36:10 acsvfbsd06 mimedefang-multiplexor[7703]: Received
SIGTERM: Stopping slaves and terminating

As you can see in this case I get a core dump. When I'm analyzing the
core dump I get the follwing:

(only the relevant part)

[snip]
Loaded symbols for /usr/local/lib/perl5/5.10.1/mach/auto/File/Glob/Glob.so
Reading symbols from /libexec/ld-elf.so.1...(no debugging symbols found)...done.
Loaded symbols for /libexec/ld-elf.so.1
#0  0x00000000023b7033 in Perl_newSTATEOP ()
   from /usr/local/lib/perl5/5.10.1/mach/CORE/libperl.so

With a backtrace I get :

#0  0x00000000023b7033 in Perl_newSTATEOP ()
   from /usr/local/lib/perl5/5.10.1/mach/CORE/libperl.so
#1  0x00000000023e90b5 in Perl_yyparse ()
   from /usr/local/lib/perl5/5.10.1/mach/CORE/libperl.so
#2  0x00000000023c6385 in S_parse_body ()
   from /usr/local/lib/perl5/5.10.1/mach/CORE/libperl.so
#3  0x00000000023c716a in perl_parse ()
   from /usr/local/lib/perl5/5.10.1/mach/CORE/libperl.so
#4  0x0000000000400bbc in main ()

So it seems the problem is around Perl_newSTATEOP (). Since I'm not a
programmer I have no idea where the problem is. Any hints are welcome.



2009/11/30 Kevin A. McGrail <kmcgrail at pccc.com>:
>> In any case I have Unix::Syslog installed and I use -l. So both is
>> possible. Any ideas?
>
> In your features output, it said:
> Unix::Syslog                  : no
> Unix::Syslog                  : Version 1.1
>
>
> My features output says:
>
> Unix::Syslog                  : yes
> Unix::Syslog                  : Version 0.100
>
> Perhaps that's a starting point.
>
> Regards,
> KAM

Regards,

-- 
Martin Schweizer
schweizer.martin at gmail.com
Tel.: +41 32 512 48 54 (VoIP)
Fax: +1 619 3300587