Re: Sun Solaris 2.5.1 Seg Faults PostgreSQL7.1.3 build com

From: "Riendeau, Mike" <mike(dot)riendeau(at)analog(dot)com>
To: "Riendeau, Mike" <Mike(dot)Riendeau(at)exchange(dot)analog(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: Re: Sun Solaris 2.5.1 Seg Faults PostgreSQL7.1.3 build com
Date: 2002-03-04 19:12:45
Message-ID: D84A32D1A483D211A38F0060B06993D00875FBF2@wilmexm1.analog.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs


Follow up:

PGv7.2 still suffers from the Seg. Fault on exit problem reported below
on that system.

However, Ran the new v7.2 on a new machine with the following config:

Architecture (example: Intel Pentium) : Sun sparc, Ultra-2

Operating System (example: Linux 2.0.26 ELF) : Solaris 2.8

PostgreSQL version (example: PostgreSQL-7.1.3): PostgreSQL-7.2

Compiler used (example: gcc 2.95.2) : gcc 2.95.2

On, the new machine, the problem goes away. After some investigation,
it appears to be some problem with our run-time linker cleaning up
on the older systems.

Regards,

M Riendeau

-----Original Message-----
From: Riendeau, Mike [mailto:Mike(dot)Riendeau(at)exchange(dot)analog(dot)com]
Sent: Friday, February 01, 2002 4:04 PM
To: Tom Lane; Riendeau, Mike
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: RE: [BUGS] Sun Solaris 2.5.1 Seg Faults PostgreSQL7.1.3 build
com mands

Your name : Mike Riendeau
Your email address : mike(dot)riendeau(at)analog(dot)com

System Configuration
---------------------
Architecture (example: Intel Pentium) : Sun Sparc 20

Operating System (example: Linux 2.0.26 ELF) : Solaris 2.5.1

PostgreSQL version (example: PostgreSQL-7.1.3): PostgreSQL-7.1.3

Compiler used (example: gcc 2.95.2) : gcc 2.95.2

Please enter a FULL description of your problem:
------------------------------------------------

I am having a problem with v7.1.3 PostgreSQL commands generating
Seg. Faults on exit. (Continued)

Steps taken on this pass:

1) built readline4.2a successfully.

2) Ran PGv7.1.3 configure as follows with a clean cache.

./configure --prefix=/home/mriendea --with-tcl
--with-tclconfig=/home/mriendea/lib --with-tkconfig=/home/mriendea/lib
--with-includes=/home/mriendea/include --with-pgport=8000 --with-odbc
--enable-debug --with-libs=/home/mriendea/lib --enable-debug

Without 'gmake install' of the new readline lib, the script sees
the same readline used to build PGv7.0.2 and completes successfully.
Of course the Seg Fault behavior is apparent.

3) 'gmake install' of the new readline lib.

4) Ran PGv7.1.3 configure with a clean cache again. The script now sees
the new readline4.2a lib and crashes with the identical
problem reported by Holger Mittewald bug #490. (test following optreset
bails the config script for some reason.)

rpath is the same (at least in all Makes in the source root)
Both config runs have the same parameters. So, something other
than the setting of rpath is going on here. Or I do not understand
how it all works (Definite possibility).

Synopsys of configure with above switches:

PROCESS

Tested readline Config gmake installed run
version libs
-----------------------------------------------------
Pgv7.0.2 system yes yes v7.0.2 yes

Pgv7.1.3 system yes yes v7.0.2 v7.1.3 pgsql,
createdb seem to
work with new db.
(brief testing)

Pgv7.1.3 system yes yes v7.1.3 Seg. Faults on exit
for pgsql, createdb ...

Pgv7.1.3 v4.2a bails n/a n/a n/a
after
optreset

Giving up soon,

Mike Riendeau

Browse pgsql-bugs by date

  From Date Subject
Next Message pgsql-bugs 2002-03-04 19:21:10 Bug #607: to_date() function bug
Previous Message Reinhard Max 2002-03-04 11:10:39 Re: Indexes not always used after inserts/updates/vacuum