procmail
[Top] [All Lists]

Re: Compile Failure on Linux 2.0.38

2002-07-07 14:13:14

   OK, thanks Phil.

   I am running 2.0.38 with libc-5.4.46.  Although we are working
on upgrading everything to RedHat 7.2, its hard to argue with 2 year
uptimes, and rock stable performance.  Anyhow I have the fix
for the compile and will post it shortly.

   As an aside we are also working on patching the libc-5 to
fix the named vuln recently reported.

   Homer

------------------------------------------------------------------------
Homer Wilson Smith     I would rather die poor    Lightlink Internet
(607) 277-0959 KC2ITF  than suffer the patronage  Ithaca, NY
homer(_at_)lightlink(_dot_)com    of bigots and pinheads.    www.lightlink.com

On Sat, 6 Jul 2002, Philip Guenther wrote:

Homer Wilson Smith <homer(_at_)lightlink(_dot_)com> writes:
   Can someone help me fix this compile failure please.
...
=============================autoconf.h===================================
/* This file was automagically generated by autoconf */

/* 2 moves in 64 steps of size 16384 when reallocing */
#define NOpw_class
#define NOstrlcat
#define NOsetrgid
#define MAX_argc 7588
#define MAILSPOOLDIR "/var/spool/mail/"
#define SENDMAIL "/usr/sbin/sendmail"
#define defPATH "PATH=$HOME/bin:/bin:/usr/bin:/usr/local/bin:/usr/bin/X11"
#define defSPATH "PATH=/bin:/usr/bin:/usr/local/bin:/usr/bin/X11"
#define PM_VERSION "3.22"
/*locktype: 4, countlocks: 0, timeout 0, watchdog 8, /tmp/_locktest*/
/*locktype: 4, countlocks: 0, timeout 0, watchdog 8, ./_locktest*/
/*locktype: 6, countlocks: 0, timeout 0, watchdog 8, /tmp/_locktest*/
/*locktype: 6, countlocks: 0, timeout 0, watchdog 8, ./_locktest*/
/*locktype: 7, countlocks: 80, timeout 0, watchdog 14, /tmp/_locktest*/
/* Hotwire LOCKINGTEST=110 */
/* Procmail will lock via: dotlocking, fcntl(), lockf() */
#define USElockf
/* autoconf completed */
==========================================================================

Somehow, you managed to generate an autoconf.h file that didn't contain
a definition for UDP_protocolno.  That shouldn't be possible unless you
defined NO_COMSAT in the config.h...but then you wouldn't have gotten
a compilation error in comsat.c because all that code would have been
elided.

Hmm, you're also missing some other lines too, including the results
of the strstr() benchmarking and the BIFF_serviceport and IP_localhost
definitions.  Those happen to all be generated by the invocation of a
single C program generated, compiled, and executed by the autoconf script.

I guess the first thing you could do is edit the src/autoconf script
and change line 1415 from:

_autotst >>$ACONF

to

_autotst >>$ACONF ; echo "_autotst returned $?"

Then "make clean" and "make autoconf.h" and see what it says _autotst is
returning.  If it's returning a value larger than 127, then it's actually
coredumping and something is very wrong.  If not, please send the (new)
"make autoconf.h" output to bug(_at_)procmail(_dot_)org

Either way, upgrading to a kernel that's not almost three years old
would be a good thing.  More important, however, is probably the age and
bugginess of the C library and/or utilities.  What versions of libc and
bash does it use?


Philip Guenther
Procmail Maintainer


_______________________________________________
procmail mailing list
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail

<Prev in Thread] Current Thread [Next in Thread>