Re: Floating point exception in initdb

Thanks for reply Tom. I am having source code form 8.4.1 version which
is released on 09-09-2009.

I was able to install it on Redhat machine without any issues. But I am
facing this issue while running initdb on Suse Ent. 9 machine.( Linux
2.6.5-7.97-smp #1 SMP Fri Jul 2 14:21:59 UTC 2004 i686 i686 i386
GNU/Linux )

Source code was compiled in following environment:

Gcc : version 4.1.2 20071124 (Red Hat 4.1.2-42)
Kernel Version: 2.6.18-8.el5 #1 SMP

With Configure string:
./configure --prefix=$BIN_DIR --without-zlib --without-readline

I built debug build using CFLAGS="-g" for configuring and now trying to
step through for finding failure location.

Thanks & Regards,
Vikram

-----Original Message-----
From: Tom Lane [mailto:tgl(at)sss(dot)pgh(dot)pa(dot)us]
Sent: Monday, March 29, 2010 4:45 PM
To: Vikram Patil
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: [GENERAL] Floating point exception in initdb

"Vikram Patil" <vpatil(at)actuate(dot)com> writes:
> I am getting an floating exception while running initdb from
custom
> built postgresql . I am using suse linux for installing postgresql.

> initDB.sh: line 14: 20285 Floating point
exception$pgsql_home/bin/initdb
> -A password -U postgres --pwfile="pwFile" -E UTF8 -D postgresql/data
> /log/initDB.log

Hm, dunno what's causing that ...

> Work around gcc bug that causes "floating-point exception" instead of
> "division by zero" on some platforms (Tom)

... but I'm quite sure that patch won't fix it for you, because it was
in code that wouldn't get executed during initdb.

What compiler are you using, for what hardware? Did you use any
nondefault configure or compiler switches? Have you modified the
Postgres sources at all? Which step of initdb gets the failure?
Try running that step under gdb so you can get a stack trace pointing
at the failure location.

regards, tom lane