Received: from punt2.demon.co.uk by oink with SMTP (PP) id <05260-0@oink>;
         Mon, 17 Jun 1996 00:05:05 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 834961463:29464:1; Sun, 16 Jun 96 22:44:23 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa28802; 16 Jun 96 22:43 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA14480 for dbi-users-real; Sun, 16 Jun 1996 14:40:01 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id OAA14473
         for <[email protected]>; Sun, 16 Jun 1996 14:39:47 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id aj16861; 16 Jun 96 22:38 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa09401; 16 Jun 96 22:34 +0100
Received: from toad by oink with SMTP (PP) id <02967-0@oink>;
         Sun, 16 Jun 1996 21:55:30 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA26756;
         Sun, 16 Jun 1996 21:54:48 +0000
Date: Sun, 16 Jun 1996 21:54:48 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: CGI/database
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1861
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424

Forwarded to dbi-users list.

----- Begin Included Message -----

From [email protected]  Sun Jun 16 21:50:58 1996
Date: 15 Jun 96 08:25:30 EDT
From: Peter Gentes <[email protected]>
To: Tim Bonce <[email protected]>
Subject: CGI/database

Tim,

We want to settle up a car-database on our site.  As we are beginners with HTML
and the whole stuff of WWW and databases, please can you give us some usefull
places on the web or elsewhere where we can find more information?

Some data about our ideas for our site:
-------------------
The site exists of several pages (HTML-documents). On one page there is a link
to a page with a form where the user can fill some data: f.e. type of
manufacturer, type of model, max. price and so on. Now (which way I even don't
know how) a query must be started with this data and after the query has ended
the set of resulting records are shown on another page (dynamical build
HTML-document). Thats all. The database consists of only 100-1000 records and
has to be updated one time a week.
------------------
I think up to the point after filling the form we know how to manage the thing.
But - please give us some hints what to do after that. Do we need a "real"
(SQL-?)database? Do we need programming with CGI? In which language, C, PEARL or
another language? Then, is it possible to test all the thing local (incl. the
CGI-queries and output) on my windows computer and then put it easy to the
UNIX-server?

Perhaps so many questions, but after searching some weeks in the web I have a
lot of material, but do not know where to start at all. We have a good knowledge
in programming pascal and delphi and programming would be no problem if I knew
only which progs, tools and apps I have to use to build the car database on the
web.

Thank you in advance
Peter Gentes


----- End Included Message -----

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05259-3@oink>;
         Mon, 17 Jun 1996 00:05:21 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 834961416:20728:1; Sun, 16 Jun 96 22:43:36 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa20056; 16 Jun 96 22:42 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA14471 for dbi-users-real; Sun, 16 Jun 1996 14:38:44 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id OAA14467
         for <[email protected]>; Sun, 16 Jun 1996 14:38:38 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id af16715; 16 Jun 96 22:38 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ae09401; 16 Jun 96 22:34 +0100
Received: from toad by oink with SMTP (PP) id <04896-0@oink>;
         Sun, 16 Jun 1996 22:26:50 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA26868;
         Sun, 16 Jun 1996 22:26:42 +0000
Date: Sun, 16 Jun 1996 22:26:42 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Request for diagnostic advice
X-Sun-Charset: US-ASCII
content-length: 1318
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Parker, Drew)
>
> Thanks to all of you for your advice on diagnostics/stack trace.  I have
> added the stack trace below, followed by my original message.

Thanks (but the three signatures were not really needed :-)

> Please let me know how to proceed on this, or if you can determine
> anything from what is below.

> (gdb) bt
> #0  0x5ffe23dc in dbih_getcom () at DBI.c:238
> #1  0x52c51c in XS_DBD__Oracle__st_execute () at Oracle.c:247

That helps a great deal. It's almost impossible to do anything with
a reported core dump problem without either a stack trace or a small
test case which I can reproduce.

Your problem is that you are calling ora_open or ora_bind with a bad
cursor. You can confirm this by adding the following line *above* your
use Oraperl line

       BEGIN { $Oraperl::safe = 1 }

In other words it's a problem in your script.

I've just added the following lines to Oraperl.pm:

# Help those who get core dumps from non-'safe' Oraperl (bad cursors)
use sigtrap qw(ILL);
$SIG{BUS} = $SIG{SEGV} = sub {
   print STDERR "Add BEGIN { \$Oraperl::safe=1 } above 'use Oraperl'.\n"
       unless $safe;
   goto &sigtrap::trap;
};

I'm very tempted to make safe=1 the default (as mentioned in a recent
message). Can someone do some benchmarks with/without safe=1 for me?

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05611-6@oink>;
         Mon, 17 Jun 1996 01:26:04 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 834967279:22235:1; Mon, 17 Jun 96 00:21:19 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa21547; 17 Jun 96 0:20 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA14596 for dbi-users-real; Sun, 16 Jun 1996 16:16:20 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA14589;
         Sun, 16 Jun 1996 16:16:17 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ab27693; 17 Jun 96 0:13 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa04263; 17 Jun 96 0:09 +0100
Received: from toad by oink with SMTP (PP) id <05319-0@oink>;
         Mon, 17 Jun 1996 00:09:36 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA27113;
         Mon, 17 Jun 1996 00:09:29 +0000
Date: Mon, 17 Jun 1996 00:09:29 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Announce DBI-0.70
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 570
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424

Changes in DBI 0.70,    16 June 1996

   Added extra invalid handle check (dbih_getcom)
   Fixed broken $dbh->quote method.
   Added check for old GCC in Makefile.PL

This is a minor release of the DBI but the next release of
DBD::Oracle will require it so DBD::Oracle users might as well
get it now.

The 'extra invalid handle check' will make non-'safe mode' Oraperl
usage much less likely to core dump (you'll get an "Invalid DBI handle"
croak instead).

I've uploaded it to ftp.demon.co.uk. It'll float round to CPAN
sites via PAUSE over the next few days.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05611-10@oink>;
         Mon, 17 Jun 1996 01:26:17 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 834967748:26320:1; Mon, 17 Jun 96 00:29:08 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa25884; 17 Jun 96 0:28 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA14614 for dbi-users-real; Sun, 16 Jun 1996 16:26:15 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA14610
         for <[email protected]>; Sun, 16 Jun 1996 16:26:00 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id af08245; 16 Jun 96 23:25 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac03045; 17 Jun 96 0:05 +0100
Received: from toad by oink with SMTP (PP) id <05049-0@oink>;
         Sun, 16 Jun 1996 22:50:25 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA26927;
         Sun, 16 Jun 1996 22:50:18 +0000
Date: Sun, 16 Jun 1996 22:50:18 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Oraperl
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1865
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424

Forwarded to dbi-users list.

----- Begin Included Message -----

From [email protected]  Fri Jun 14 11:56:16 1996
Date: Thu, 13 Jun 1996 23:40:16 -0300 (ADT)
X-Sender: [email protected]
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: [email protected]
From: Tim & Kelly MacKinnon <[email protected]>
Subject: Oraperl
Cc: [email protected]

Mr. Bunce:

During the past couple of weeks, I have been trying to do some funky stuff with
Oraperl, and I am not sure if it can be done. I am hoping that you can shed some
light on my problems and help me get unstuck.

In one situation, I am calling a stored pl/sql procedure (or function) that is
stored within an Oracle database. I do this by using the &ora_open command. The
problem lies in that I would like the stored pl/sql procedure to return a value
(i.e. return code) to a variable within the Oraperl script, to be used elsewhere
within the Oraperl script. Is this feasible? Or am I just dreaming.

In another situation, I am trying to run Oracle reports 2.5 (motif) from the
command line within a Oraperl script. This I can accomplish easily with a
straight unix script, however I am trying to use Oraperl. I use the &ora_login
command and then call the Oracle reports runtime by using the system command
(i.e. system(r25runm ...)). This results in a "REP-0001 Unable to access message
sub-system" error. Any idea on how to resolve this? It has been noted, that
the environment variable TWO-TASK gets deleted during the running of an
Oraperl script. I wonder if this could be the cause.

Well that's the problems that I'm trying to overcome, and any help what so ever
would be greatly appreciated. If need be, I would be willing to give you a call
in order to further discuss this.

Thanks,

Jules Leger
Project Developer
(506) 867-4640


----- End Included Message -----

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05611-16@oink>;
         Mon, 17 Jun 1996 01:26:38 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 834969456:10044:1; Mon, 17 Jun 96 00:57:36 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa09575; 17 Jun 96 0:56 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA14647 for dbi-users-real; Sun, 16 Jun 1996 16:52:52 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA14643
         for <[email protected]>; Sun, 16 Jun 1996 16:52:51 -0700
Received: by gw.home.vix.com id QAA12877; Sun, 16 Jun 1996 16:52:46 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ar02762; 17 Jun 96 0:50 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa03045; 17 Jun 96 0:04 +0100
Received: from toad by oink with SMTP (PP) id <05011-0@oink>;
         Sun, 16 Jun 1996 22:42:12 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA26899;
         Sun, 16 Jun 1996 22:42:04 +0000
Date: Sun, 16 Jun 1996 22:42:04 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Urgent survey for Oracle 7.3 users of DBD::Oracle
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424

The number of Oracle 7.3 users with DBD::Oracle build problems
is growing. To help me resolve these I need some information...

*Only* if you have *successfully* built and tested DBD::Oracle
please send a short email message to me (only me) detailing:

       - Exact Oracle version number (7.3.x.y... etc).

       - Exact Pro*C version number (2.1.x.y... etc).

       - Details of any changes to DBD::Oracle you _needed_ to
         make in order to get it to build.

Thanks. Please don't expect a personal reply, I'm too busy. Sorry.

Tim.

p.s. You can get the version numbers from the unix.prd file which should
be in your oracle installation tree (top level or orainst directory).

p.p.s. I may ask some people to send in their Oracle *.mk files later
but please _don't_ send them yet.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <08716-1@oink>;
         Mon, 17 Jun 1996 13:26:35 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835011916:26239:2; Mon, 17 Jun 96 12:45:16 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa26052; 17 Jun 96 12:45 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA15295 for dbi-users-real; Mon, 17 Jun 1996 04:42:17 -0700
Precedence: junk
Sender: [email protected]
Received: from gatekeeper (gatekeeper.imr.no [193.156.92.162])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id EAA15291
         for <[email protected]>; Mon, 17 Jun 1996 04:42:14 -0700
From: [email protected]
Received: by gatekeeper (SMI-8.6/SMI-SVR4) id NAA12263;
         Mon, 17 Jun 1996 13:42:37 +0200
Received: from mailhost(158.37.84.8) by gatekeeper via smap (V1.3) id sma012256;
         Mon Jun 17 13:42:15 1996
Received: from ygdrasil (lofoten) by imr.no (4.1/SMI-4.1) id AA24491;
         Mon, 17 Jun 96 13:39:47 +0200
Received: by ygdrasil (1.37.109.4/Uninett-C-1.4) id AA26382;
         Mon, 17 Jun 96 13:33:55 +0200
Date: Mon, 17 Jun 96 13:33:55 +0200
Message-Id: <9606171333.ZM26380@lofoten>
X-Mailer: Z-Mail (3.2.0 06sep94)
To: [email protected]
Subject: DBD::Ingres development?
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 8bit

Hi there.

I am looking for a tool to make a web-to-Ingres (and maybe some other dbms)
cgi-bin interface. As I read the notes on IngPerl, I found the advice to look
for the coming "standard" for perl5, DBperl, to avoid writing lots of "non
standard" code before its release.

Is DBI the "standard DBperl" I am looking for? (or close enough to use?)

If so,can I expect to find an Ingres module available in the near future?

> DBD::Ingres
> Module Name: DBD::Ingres
> Author: Tim Bunce
> Current Version: Not yet released - Under development

Would I waste my time if I started using IngPerl now (waiting for DBD::Ingres),
or is it "compatible" with DBI?


Thank you in advance.

Arne Vidar Brekke
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <08715-16@oink>;
         Mon, 17 Jun 1996 13:27:34 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835009963:23677:1; Mon, 17 Jun 96 12:12:43 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa23448; 17 Jun 96 12:11 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA15254 for dbi-users-real; Mon, 17 Jun 1996 04:06:08 -0700
Precedence: junk
Sender: [email protected]
Received: from aruba.lerc.nasa.gov (aruba.lerc.nasa.gov [139.88.35.16])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id EAA15250
         for <[email protected]>; Mon, 17 Jun 1996 04:06:03 -0700
Received: from bytor.lerc.nasa.gov by aruba.lerc.nasa.gov
         with ESMTP (NASA LeRC 8.7.4.1/2.01-main) id HAA22627;
         Mon, 17 Jun 1996 07:05:49 -0400 (EDT)
Received: by bytor.lerc.nasa.gov (NASA LeRC 8.7.4.1/2.01-local) id HAA08449;
         Mon, 17 Jun 1996 07:05:48 -0400 (EDT)
From: Gary Gapinski <[email protected]>
Message-Id: <[email protected]>
Date: Mon, 17 Jun 1996 07:05:48 -0400
In-Reply-To: "Willard, Keith" <[email protected]> "multiple SQL statements in a single prepare?" (Jun 14, 4:43pm)
References: <[email protected]>
X-Mailer: Z-Mail (3.2.0 26oct94 MediaMail)
To: "Willard, Keith" <[email protected]>,
   dbilist <[email protected]>
Subject: Re: multiple SQL statements in a single prepare?
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii

Hello, Keith:

No, it is not possible to do this.

You *could* execute multiple PL/SQL statements in an anonymous PL/SQL block,
but this would not return a result set. Each distinct SELECT must be separately
prepared.

Regards,

Gary

--
__________________________________________________________
Gary Gapinski                email: [email protected]
NASA Lewis Research Center   voice: +1 216 433 5251
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <09807-1@oink>;
         Mon, 17 Jun 1996 17:01:02 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835023406:03375:1; Mon, 17 Jun 96 15:56:46 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa03297; 17 Jun 96 15:56 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa23728; 17 Jun 96 15:55 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA15622 for dbi-users-real; Mon, 17 Jun 1996 07:45:59 -0700
Precedence: junk
Sender: [email protected]
Received: from foe.co.uk (amyda.foe.co.uk [193.114.240.82])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id HAA15618
         for <[email protected]>; Mon, 17 Jun 1996 07:45:53 -0700
Received: by foe.co.uk (5.x/SMI-SVR4) id AA28979;
         Mon, 17 Jun 1996 15:50:54 +0100
From: Richard Weatherley <[email protected]>
Message-Id: <[email protected]>
Subject: LONG/RAW Oracle INSERT
To: [email protected]
Date: Mon, 17 Jun 1996 15:50:52 +0100 (BST)
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text

G'day all

Apologies if this is an old one:

I'm trying to INSERT a large LONG into Oracle using &ora_open() and
&ora_bind(), and it's being truncated after $ora_long chars.
I'm happily using $csr->blob_read() to read the LONG.  Is there an
equivalent "blob_insert()" technique, or must I modify $ora_long each
time I want to insert a LONG?

Many thanks

Richard.
--
Some of us wish someone would insert a large long into Microsoft,
but that's another (Java) mailing list...
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <09807-4@oink>;
         Mon, 17 Jun 1996 17:01:17 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835025031:18971:1; Mon, 17 Jun 96 16:23:51 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa18080; 17 Jun 96 16:22 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA15721 for dbi-users-real; Mon, 17 Jun 1996 08:19:32 -0700
Precedence: junk
Sender: [email protected]
Received: from pandemonium.URLabs.com ([email protected] [206.151.14.135])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA15717
         for <[email protected]>; Mon, 17 Jun 1996 08:19:25 -0700
Received: (from priest@localhost) by pandemonium.URLabs.com (8.6.12/8.6.9)
         id LAA00686; Mon, 17 Jun 1996 11:19:39 -0400
Date: Mon, 17 Jun 1996 11:19:39 -0400
Message-Id: <[email protected]>
From: Travis L Priest <[email protected]>
To: [email protected]
Cc: [email protected]
Subject: DBI-0.70 + DBD-Oracle-0.32 Problems


OS:             SunOS server 5.5 Generic_103093-02 sun4m sparc (no patches)
DB:             Oracle 7.3
Perl:           This is perl, version 5.002
DBI:            0.70 (and 0.69)
DBD-Oracle:     0.32
CC:             gcc version 2.7.2
Build:          perl Makefile.PL -m /home/oracle/rdbms/lib/oracle.mk

Connect, disconnect, and select work fine.  However, after disconnect,
I get a bus error whenever the perl script calls exit().  For
instance:

   my($drh) = DBI->install_driver('Oracle');
   my($dbh) = $drh->connect($Db, $User, $Pass);
   die "$ME: Unable to connect to database '$Db' as '$User'" unless $dbh;

   $dbc = select...

   $dbc->finish;
   $dbh->disconnect;
   warn "boy howdy";
   exit(0); # bus error

I have a core file, but alas, no gdb available at the moment to read
it.  Any thoughts?

Travis
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <09807-7@oink>;
         Mon, 17 Jun 1996 17:01:30 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835027147:09338:1; Mon, 17 Jun 96 16:59:07 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa08603; 17 Jun 96 16:57 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA15783 for dbi-users-real; Mon, 17 Jun 1996 08:54:12 -0700
Precedence: junk
Sender: [email protected]
Received: from zeus.socom.com ([email protected] [206.108.56.4])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA15779
         for <[email protected]>; Mon, 17 Jun 1996 08:54:02 -0700
Received: (from lebel@localhost) by zeus.socom.com (8.7.3/8.7.3) id LAA11696;
         Mon, 17 Jun 1996 11:51:46 -0400
To: Travis L Priest <[email protected]>
Cc: [email protected]
Subject: Re: DBI-0.70 + DBD-Oracle-0.32 Problems
References: <[email protected]>
Mime-Version: 1.0 (generated by tm-edit 7.68)
Content-Type: text/plain; charset=US-ASCII
From: David Lebel <[email protected]>
Date: 17 Jun 1996 11:51:46 -0400
In-Reply-To: Travis L Priest's message of Mon, 17 Jun 1996 11:19:39 -0400
Message-ID: <[email protected]>
Lines: 20
X-Mailer: Gnus v5.2.19/XEmacs 19.13

>>>>> "TLP" == Travis L Priest <[email protected]> writes:

TLP: Connect, disconnect, and select work fine.  However, after
TLP: disconnect, I get a bus error whenever the perl script calls
TLP: exit().  For instance:

       I've been having the exact same problem!  I've sent a bug
report to Tim about it, however, since he needed more information I
didn't have the time to provide him a complete backtrace information
on the core dump as he requested.  As soon as I have the time, I'll
send him a copy.  However, if you have the time to do it, then great,
since you have the exact same problem as I have.

       ...David

--
// David Lebel <[email protected]>                 http://www.socom.com/~lebel/
// Analyst / System Administrator    PGP fingerprint: AB 50 15 3A 55 C0 41 FC
// Socom technologies, inc.                           E6 15 5B 9E 1F 14 2E 95
  " I find your lack of faith disturbing. " -- Darth Vader, "Star Wars"
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10472-0@oink>;
         Mon, 17 Jun 1996 17:59:26 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835028685:23842:1; Mon, 17 Jun 96 17:24:45 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa22414; 17 Jun 96 17:23 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA15818 for dbi-users-real; Mon, 17 Jun 1996 09:17:55 -0700
Precedence: junk
Sender: [email protected]
Received: from solutions.solon.com ([email protected] [192.129.84.3])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA15814
         for <[email protected]>; Mon, 17 Jun 1996 09:17:52 -0700
Received: from solutions.solon.com ([email protected] [192.129.84.3])
         by solutions.solon.com (8.6.12/8.6.12) with SMTP id LAA22098;
         Mon, 17 Jun 1996 11:18:34 -0500
Date: Mon, 17 Jun 1996 11:18:32 -0500 (CDT)
From: fenris <[email protected]>
To: Tim Bunce <[email protected]>
cc: [email protected], [email protected]
Subject: Re: CGI/database
In-Reply-To: <[email protected]>
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Peter,


A site I set up for a friend was done like this:

       1: His original data was in an Access database so we export that to
a colon seperated text file.

       2: I wrote a perl script to search through the data, recording
matches.

       3: I then built a cgi based request form so that people could chose
from the given criteria  and then search the base. I then generated a
reply page.

This system worked really well until the database exceeded 800-900
records. It then became a muscle problem and the interface was just too
slow. The best part about this though, is that there are enough
templates/examples and what have you out in the web that it is fairly
easy to figure out what needs to be done. We kept the database updated by
exporting a new database file at the end of the business day and
transfering it to the unix host.


fenris (@solon.com)


We live not alone in the present but also in the past and future. The
radius that circumscribes our lives must necessarily extend backward
indefinitely and forward infinitely.
               G.Warner and C. Foote, preface to History of Hennepin
                       County & the City of Minneapolis


If, then, the past is not simply a stepping-stone to the future, but
a part of our very selves, we can not afford to ignore it, or separate
it from ourselves, as a member might be lopped off from our bodies;
for though the body thus maimed might perform many and perhaps most of
its functions, still it could never again be called complete.

               Ibid.

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10472-3@oink>;
         Mon, 17 Jun 1996 17:59:40 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835029699:03533:1; Mon, 17 Jun 96 17:41:39 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa02500; 17 Jun 96 17:40 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA15882 for dbi-users-real; Mon, 17 Jun 1996 09:36:19 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA15878
         for <[email protected]>; Mon, 17 Jun 1996 09:36:17 -0700
Received: by gw.home.vix.com id JAA09607; Mon, 17 Jun 1996 09:35:44 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id aa24436; 17 Jun 96 16:25 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ag17125; 17 Jun 96 17:01 +0100
Received: from toad by oink with SMTP (PP) id <09194-0@oink>;
         Mon, 17 Jun 1996 13:56:13 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA28982;
         Mon, 17 Jun 1996 13:56:05 +0000
Date: Mon, 17 Jun 1996 13:56:05 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DBD::Ingres development?
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected]
>
> Hi there.
>
> I am looking for a tool to make a web-to-Ingres (and maybe some other dbms)
> cgi-bin interface. As I read the notes on IngPerl, I found the advice to look
> for the coming "standard" for perl5, DBperl, to avoid writing lots of "non
> standard" code before its release.
>
> Is DBI the "standard DBperl" I am looking for? (or close enough to use?)
>
> If so,can I expect to find an Ingres module available in the near future?
>
> > DBD::Ingres
> > Module Name: DBD::Ingres
> > Author: Tim Bunce
> > Current Version: Not yet released - Under development
>
> Would I waste my time if I started using IngPerl now (waiting for DBD::Ingres),
> or is it "compatible" with DBI?

I'm not doing DBD::Ingres any time soon. Some else (I forget who, sorry)
did offer but I've heard nothing since.

Note that IngPerl only works with perl4, and any future DBD::Ingres
would only work with perl5.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10697-0@oink>;
         Mon, 17 Jun 1996 18:39:37 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835031354:19551:0; Mon, 17 Jun 96 18:09:14 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa18896; 17 Jun 96 18:08 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa07891; 17 Jun 96 18:07 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA15949 for dbi-users-real; Mon, 17 Jun 1996 10:01:17 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA15938
         for <[email protected]>; Mon, 17 Jun 1996 10:00:24 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ae02254; 17 Jun 96 18:00 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab05142; 17 Jun 96 17:59 +0100
Received: from toad by oink with SMTP (PP) id <10396-0@oink>;
         Mon, 17 Jun 1996 17:32:33 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA29689;
         Mon, 17 Jun 1996 17:32:26 +0000
Date: Mon, 17 Jun 1996 17:32:26 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DBI-0.70 + DBD-Oracle-0.32 Problems
X-Sun-Charset: US-ASCII
content-length: 1103
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Travis L Priest <[email protected]>
>
> OS:           SunOS server 5.5 Generic_103093-02 sun4m sparc (no patches)
> DB:           Oracle 7.3

7.3.?.?.?

> Perl:         This is perl, version 5.002
> DBI:          0.70 (and 0.69)
> DBD-Oracle:   0.32
> CC:           gcc version 2.7.2
> Build:                perl Makefile.PL -m /home/oracle/rdbms/lib/oracle.mk

You should no longer need to use the -m flag.

> Connect, disconnect, and select work fine.  However, after disconnect,
> I get a bus error whenever the perl script calls exit().  For
> instance:
>
>     my($drh) = DBI->install_driver('Oracle');
>     my($dbh) = $drh->connect($Db, $User, $Pass);
>     die "$ME: Unable to connect to database '$Db' as '$User'" unless $dbh;
>
>     $dbc = select...
>
>     $dbc->finish;
>     $dbh->disconnect;
>     warn "boy howdy";
>     exit(0); # bus error
>
> I have a core file, but alas, no gdb available at the moment to read
> it.  Any thoughts?

Try     echo '$c' | adb /usr/local/bin/perl core

You might need to do a 'make perl' and then './perl testscript.pl'
then echo '$c' | adb ./perl core get adb to give you something useful.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <10698-0@oink>;
         Mon, 17 Jun 1996 18:39:43 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835031535:08084:1; Mon, 17 Jun 96 18:12:15 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa06227; 17 Jun 96 18:10 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA15953 for dbi-users-real; Mon, 17 Jun 1996 10:01:24 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA15945
         for <[email protected]>; Mon, 17 Jun 1996 10:01:11 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ak02254; 17 Jun 96 18:00 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa05142; 17 Jun 96 17:59 +0100
Received: from toad by oink with SMTP (PP) id <10380-0@oink>;
         Mon, 17 Jun 1996 17:26:06 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA29659;
         Mon, 17 Jun 1996 17:25:55 +0000
Date: Mon, 17 Jun 1996 17:25:55 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: LONG/RAW Oracle INSERT
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 494
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Richard Weatherley <[email protected]>
>
> G'day all
>
> Apologies if this is an old one:
>
> I'm trying to INSERT a large LONG into Oracle using &ora_open() and
> &ora_bind(), and it's being truncated after $ora_long chars.
> I'm happily using $csr->blob_read() to read the LONG.  Is there an
> equivalent "blob_insert()" technique, or must I modify $ora_long each
> time I want to insert a LONG?

ora_long is not involved in binding. It must be something else.
Look deeper.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <11381-0@oink>;
         Mon, 17 Jun 1996 21:05:56 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835040955:21647:7; Mon, 17 Jun 96 20:49:15 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa20249; 17 Jun 96 20:48 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA16453 for dbi-users-real; Mon, 17 Jun 1996 12:33:18 -0700
Precedence: junk
Sender: [email protected]
Received: from europa.ikonic.com (mypop.ikonic2.com [204.119.99.106])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id MAA16449
         for <[email protected]>; Mon, 17 Jun 1996 12:33:14 -0700
Received: from 204.119.99.40.ikonic.com ([204.119.99.18])
         by europa.ikonic.com (post.office MTA v1.9.3 ID# 0-11965) with SMTP
         id AAA136; Mon, 17 Jun 1996 12:28:46 -0700
X-Sender: [email protected]
X-Mailer: Windows Eudora Light Version 1.5.2
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: Tim Bunce <[email protected]>
From: [email protected] (Parker, Drew)
Subject: Re: Request for diagnostic advice
Cc: [email protected]
Date: Mon, 17 Jun 1996 12:28:46 -0700
Message-ID: <[email protected]>

Tim,

Thanks for yoru response.  I tried adding:

BEGIN { $Oraperl::safe = 1 }

at the beginning of my script to see where the problem is, and, voila!, with
that line in place, the script no longer core dumps!  When I take it out of
the script, it once again core dumps.  I haven't done thorough testing yet
to make sure all of my other scripts will be so magically fixed.

Does safe=1 provide some benevolent side effects?

The one question I had to your response, is why I would see intermittent
problems with this (it doesn't seem to be consistent where it bombs in the
script).  Is there some "garbage collection" of out-of-scope cursors (either
on the DBI/DBD layer or Oracle layers) that may be kicking in too early?

Thanks,

Drew


At 10:26 PM 6/16/96 +0000, Tim Bunce wrote:

>
>Your problem is that you are calling ora_open or ora_bind with a bad
>cursor. You can confirm this by adding the following line *above* your
>use Oraperl line
>
>       BEGIN { $Oraperl::safe = 1 }
>
>In other words it's a problem in your script.
>
>I've just added the following lines to Oraperl.pm:
>
># Help those who get core dumps from non-'safe' Oraperl (bad cursors)
>use sigtrap qw(ILL);
>$SIG{BUS} = $SIG{SEGV} = sub {
>    print STDERR "Add BEGIN { \$Oraperl::safe=1 } above 'use Oraperl'.\n"
>       unless $safe;
>    goto &sigtrap::trap;
>};
>
>I'm very tempted to make safe=1 the default (as mentioned in a recent
>message). Can someone do some benchmarks with/without safe=1 for me?
>
>Tim.
>
------------------
Drew Parker
[email protected]
Ikonic Interactive
2 Harrison Street, Top Floor
San Francisco, California  94105 USA
(415) 908-8069


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11695-4@oink>;
         Mon, 17 Jun 1996 21:33:56 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835043003:23247:2; Mon, 17 Jun 96 21:23:23 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa22568; 17 Jun 96 21:22 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA15980 for dbi-users-real; Mon, 17 Jun 1996 10:08:29 -0700
Precedence: junk
Sender: [email protected]
Received: from cerberus.launch.com (root@[204.214.62.35])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id KAA15976
         for <[email protected]>; Mon, 17 Jun 1996 10:08:25 -0700
Received: from humpback.launch.com (corp.launch.com [100.1.0.5])
         by cerberus.launch.com (8.6.12/8.6.12) with SMTP id MAA13272
         for <[email protected]>; Mon, 17 Jun 1996 12:10:27 -0500
Received: by humpback.launch.com with Microsoft Exchange (IMC 4.0.837.3)
         id <[email protected]>;
         Mon, 17 Jun 1996 11:57:06 -0500
Message-ID: <c=US%a=_%p=launch%[email protected]>
From: Reilly Maginn <[email protected]>
To: "'[email protected]'" <[email protected]>
Subject: unsubscribe
Date: Mon, 17 Jun 1996 11:56:58 -0500
X-Mailer: Microsoft Exchange Server Internet Mail Connector Version 4.0.837.3
Encoding: 10 TEXT

unsubscribe

I'm sorry I ever subscribed to this users group, you can't get rid of
it!!!!!!!
Reilly Maginn
Catapult Systems Corp.                   Phone:(512) 328-8181
1101 Capital of Texas Hwy.             mailto:[email protected]
Building K Suite 100                        http://www.launch.com
Austin TX  78759

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11695-6@oink>;
         Mon, 17 Jun 1996 21:34:01 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835043133:24668:1; Mon, 17 Jun 96 21:25:33 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa23836; 17 Jun 96 21:24 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA16018 for dbi-users-real; Mon, 17 Jun 1996 10:22:05 -0700
Precedence: junk
Sender: [email protected]
Received: from pandemonium.URLabs.com ([email protected] [206.151.14.138])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id KAA16014
         for <[email protected]>; Mon, 17 Jun 1996 10:21:58 -0700
Received: (from priest@localhost) by pandemonium.URLabs.com (8.6.12/8.6.9)
         id NAA00374; Mon, 17 Jun 1996 13:16:40 -0400
Date: Mon, 17 Jun 1996 13:16:40 -0400
Message-Id: <[email protected]>
From: Travis L Priest <[email protected]>
To: Tim Bunce <[email protected]>
Cc: [email protected]
Subject: Re: DBI-0.70 + DBD-Oracle-0.32 Problems
In-Reply-To: <[email protected]>
References: <[email protected]>

Mon, 17 June 1996, Tim Bunce <[email protected]> wrote:
>
> > From: Travis L Priest <[email protected]>
> >
> > OS:         SunOS server 5.5 Generic_103093-02 sun4m sparc (no patches)
> > DB:         Oracle 7.3
>
> 7.3.?.?.?

Err, Oracle 7.2.2.3 Release of Sun SPARC Solaris2.  Sorry.

> > Build:              perl Makefile.PL -m /home/oracle/rdbms/lib/oracle.mk
>
> You should no longer need to use the -m flag.

server% perl Makefile.PL

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /home/oracle
Oracle sysliblist: -lsocket -lnsl -lm -ldl -laio -lposix4
Unable to locate proc.mk (use -m /path/to/proc.mk to specify)

> > I have a core file, but alas, no gdb available at the moment to read
> > it.  Any thoughts?
>
> Try   echo '$c' | adb /usr/local/bin/perl core
..
> You might need to do a 'make perl' and then './perl testscript.pl'
> then echo '$c' | adb ./perl core get adb to give you something useful.

server% echo '$c' | adb /usr/local/urlabs/bin/perl core
core file = core -- program ``dbload'' on platform SUNW,SPARCstation-5
SIGBUS: Bus Error
init_dummy() + 58
data address not found

I'm not an adb user, so I don't know if the above is useful.  I'd like
to hold off rebuilding perl since this is not the system where our
full perl build was originally done.  I'm pulling down gdb now to see
if it has anything interesting to say.

Travis
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <12257-0@oink>;
         Tue, 18 Jun 1996 00:07:46 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835045630:12192:2; Mon, 17 Jun 96 22:07:10 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa10243; 17 Jun 96 22:05 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA16622 for dbi-users-real; Mon, 17 Jun 1996 14:00:02 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id NAA16612
         for <[email protected]>; Mon, 17 Jun 1996 13:59:55 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ae20469; 17 Jun 96 21:08 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa10684; 17 Jun 96 21:05 +0100
Received: from ig.co.uk by oink id <11012-0@oink>;
         Mon, 17 Jun 1996 19:23:59 +0100
Newsgroups: comp.lang.perl.misc
Subject: Re: problem with dynamic library loading on Solaris 2.5
References: <[email protected]>
Apparently-To: [email protected]
Apparently-To: [email protected]
Date: Mon, 17 Jun 1996 19:23:59 +0100
From: Tim Bunce <[email protected]>
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424
Message-ID: <[email protected]>

In article <[email protected]>,
Jon Adams  <[email protected]> wrote:
> I'm having a major problem with the DBD::Oracle module. I've already
> posted this problem to the DBI mailing list, with no luck (other than a
> response requesting that I inform them if I figure it out)

I don't recall it being left there but then I've been _very_ busy
recently and might have been more curt than my usual sweet self :-).

> and since it
> may not be directly related to the Oracle aspects, I'm posting the
> question here.

It's almost certainly not an oracle problem.

> I am using Perl 5.002, Oracle 7.3.2, and gcc 2.7.0 on an Ultrasparc
> running Solaris 2.5.

That's okay (but I'd recommend upgrading gcc to a non *.0 version).

> >t/base..............install_driver(Oracle) failed: Can't load
> >'./blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle:
> >ld.so.1: /usr/local/bin/perl: fatal: libclntsh.so.1.0: can't open file:
> >errno=2 at /usr/local/lib/perl5/DynaLoader.pm line 140.
>
> libclntsh.so.1.0 is an Oracle library. I've tried such things as playing
> with the path, the library path, specifying the library directories, and
> anything else I could think of. The only thing I've accomplished is
> making it so that the tests do not even run at all.

Try manually setting the LD_LIBRARY_PATH environment variable to the
full path of the directory holding the libclntsh.so.1.0 file.  See the
ld unix man page for more info. I'm assuming your perl isn't set-uid.

I think I forgot to suggest that earlier. Sorry.

I'd be happy to persue this on the dbi-users mailing list (CC'd).

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13565-2@oink>;
         Tue, 18 Jun 1996 07:07:11 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835056126:04156:3; Tue, 18 Jun 96 01:02:06 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa03782; 18 Jun 96 1:01 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA17048 for dbi-users-real; Mon, 17 Jun 1996 16:55:36 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA17044
         for <[email protected]>; Mon, 17 Jun 1996 16:55:35 -0700
Received: by gw.home.vix.com id QAA08709; Mon, 17 Jun 1996 16:55:15 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id aa10895; 18 Jun 96 0:37 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab07097; 18 Jun 96 0:23 +0100
Received: from toad by oink with SMTP (PP) id <12678-0@oink>;
         Tue, 18 Jun 1996 00:21:04 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA00650;
         Tue, 18 Jun 1996 00:20:57 +0000
Date: Tue, 18 Jun 1996 00:20:57 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Request for diagnostic advice
Cc: [email protected]
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Parker, Drew)
>
> At 09:19 PM 6/17/96 +0000, you wrote:
>
> >I presume you're saying that it works and does _not_ get an invalid
> >handle error. That would be rather strange.
>
> I have had a few different examples of crashes with some of my scripts.  For
> the backtrace that I originally sent you, you were right, I did find a
> cursor referenced that had not been properly initialized.

Right.

> Another script, however, gave intermittent "bus errors" when I ran it prior
> to setting safe = 1.  After setting safe =1, I cannot reproduce the problem.
> I did not change any of the code.  This is an intermittent problem, so I am
> wondering if there are any known unix environment issues (e.g., disk space,
> memory, system kernel parameters, etc) that might affect this.

No. (At least not in my code.)

Note that a reproducable test case is very important.

> I did save the core file for the last time it crashed in this context, but
> for that core dump, I cannot even obtain a decent backtrace (below).  I have
> sent you a separate message re: your advice to T.L. Priest about recompiling
> with improved diagnostics, perhaps you can send me the same file.

I will.

> By the way, I think some of my intermittent bus errors may be related to
> what T.L. Priest and David Lebel have reported; I have seen bus errors occur
> on "exit" after a successful &ora_logoff.

Useful.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13565-8@oink>;
         Tue, 18 Jun 1996 07:07:45 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835057796:15095:1; Tue, 18 Jun 96 01:29:56 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa14061; 18 Jun 96 1:27 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA16717 for dbi-users-real; Mon, 17 Jun 1996 14:44:38 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id OAA16713
         for <[email protected]>; Mon, 17 Jun 1996 14:44:25 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id bu23248; 17 Jun 96 21:40 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa18011; 17 Jun 96 21:33 +0100
Received: from toad by oink with SMTP (PP) id <11609-0@oink>;
         Mon, 17 Jun 1996 21:19:27 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA00180;
         Mon, 17 Jun 1996 21:19:20 +0000
Date: Mon, 17 Jun 1996 21:19:20 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Request for diagnostic advice
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1010
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Parker, Drew)
>
> Tim,
>
> Thanks for yoru response.  I tried adding:
>
> BEGIN { $Oraperl::safe = 1 }
>
> at the beginning of my script to see where the problem is, and, voila!, with
> that line in place, the script no longer core dumps!  When I take it out of
> the script, it once again core dumps.

I presume you're saying that it works and does _not_ get an invalid
handle error. That would be rather strange.

> Does safe=1 provide some benevolent side effects?

It trades safety for speed by bypassing the DBI. Until someone does some
useful benchmarking I don't know how much.

> The one question I had to your response, is why I would see intermittent
> problems with this (it doesn't seem to be consistent where it bombs in the
> script).

I don't know. I'd welcome more concrete information if/when you get any.

I think I'll release the next DBD::Oracle with safe (e.g., normal) mode
as the default. At least it'll cut down my work load in the short term :-)

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13565-10@oink>;
         Tue, 18 Jun 1996 07:08:00 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835057925:15854:1; Tue, 18 Jun 96 01:32:05 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15557; 18 Jun 96 1:31 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA16752 for dbi-users-real; Mon, 17 Jun 1996 14:50:40 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id OAA16748;
         Mon, 17 Jun 1996 14:50:32 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ar29419; 17 Jun 96 21:29 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab20723; 17 Jun 96 21:44 +0100
Received: from toad by oink with SMTP (PP) id <11777-0@oink>;
         Mon, 17 Jun 1996 21:38:40 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA00255;
         Mon, 17 Jun 1996 21:38:16 +0000
Date: Mon, 17 Jun 1996 21:38:16 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: unsubscribe
X-Sun-Charset: US-ASCII
content-length: 2037
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Reilly Maginn <[email protected]>
>
> unsubscribe
>
> I'm sorry I ever subscribed to this users group, you can't get rid of
> it!!!!!!!
>
> Reilly Maginn
> Catapult Systems Corp.                   Phone:(512) 328-8181
> 1101 Capital of Texas Hwy.             mailto:[email protected]
> Building K Suite 100                        http://www.launch.com
> Austin TX  78759

As posted here before:

----- Begin Included Message -----

Cc: [email protected]
Subject: HOW TO UNSUBSCRIBE
Date: Tue, 16 Apr 1996 11:20:57 -0700
From: Ted Lemon <[email protected]>


Douglas, when you want to unsubscribe from any mailing list on the
internet, you should expect that if you send mail to the -request
address for that list, you will at least get mail back containing
instructions on how to unsubscribe.

For any mailing list, the -request address is always made up of the
mailing list name, followed by -request, followed by the `@'
character, followed by the fully qualified domain name of the machine
on which that list resides.

In the case of [email protected], the mailing list name is
dbi-users, the fully qualified domain name is fugue.com, and thus the
-request address is [email protected].   If you wanted to
unsubscribe to, e.g., [email protected], you would send mail to
[email protected].

This is very consistent across most mailing lists you will encounter,
although sadly there are some mailing lists where the maintainer has
been too lazy to set up -request addresses.  Still, you should try the
-request address first.

Now, having explained all that, in order to avoid being flamed to a
crisp when you get the response from dbi-users-request and realize
that there's another step, I will tell you that what the
dbi-users-request address actually does is to send you instructions
saying, essentially, that you should go to http://www.fugue.com/dbi
with your favourite web browser (I recommend Netscape or Lynx) to
unsubscribe.

                              _MelloN_


----- End Included Message -----

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13565-11@oink>;
         Tue, 18 Jun 1996 07:08:13 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835057925:15850:2; Tue, 18 Jun 96 01:32:05 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15590; 18 Jun 96 1:31 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA16637 for dbi-users-real; Mon, 17 Jun 1996 14:05:26 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id OAA16633
         for <[email protected]>; Mon, 17 Jun 1996 14:05:20 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id aj23248; 17 Jun 96 21:39 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab18011; 17 Jun 96 21:33 +0100
Received: from toad by oink with SMTP (PP) id <11664-0@oink>;
         Mon, 17 Jun 1996 21:30:21 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA00202;
         Mon, 17 Jun 1996 21:30:14 +0000
Date: Mon, 17 Jun 1996 21:30:14 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: DBI-0.70 + DBD-Oracle-0.32 Problems
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1613
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Travis L Priest <[email protected]>
>
> Mon, 17 June 1996, Tim Bunce <[email protected]> wrote:
> >
> > > From: Travis L Priest <[email protected]>
> > > DB:               Oracle 7.3
> >
> > 7.3.?.?.?
>
> Err, Oracle 7.2.2.3 Release of Sun SPARC Solaris2.  Sorry.

Eh? You said 7.3 above and now 7.2.*?

> > > Build:            perl Makefile.PL -m /home/oracle/rdbms/lib/oracle.mk
> >
> > You should no longer need to use the -m flag.
>
> server% perl Makefile.PL
>
> Unable to locate proc.mk (use -m /path/to/proc.mk to specify)

Where is the (Pro*C) proc.mk file?

> > > I have a core file, but alas, no gdb available at the moment to read
> > > it.  Any thoughts?
> >
> > Try echo '$c' | adb /usr/local/bin/perl core
> ...
> > You might need to do a 'make perl' and then './perl testscript.pl'
> > then echo '$c' | adb ./perl core get adb to give you something useful.
>
> server% echo '$c' | adb /usr/local/urlabs/bin/perl core
> core file = core -- program ``dbload'' on platform SUNW,SPARCstation-5
> SIGBUS: Bus Error
> init_dummy() + 58
> data address not found

Phoey.

> I'm not an adb user, so I don't know if the above is useful.  I'd like
> to hold off rebuilding perl since this is not the system where our
> full perl build was originally done.

You don't need to 'rebuild perl' as such. Just say 'make perl' in the
DBD::Oracle directory.

> I'm pulling down gdb now to see if it has anything interesting to say.

Thanks.

I'll send you an updated Makefile.PL in a seperate message. It'll let
you say  perl Makefile.PL -g  to enable 'proper' debugging and more
informative stack traces.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13565-15@oink>;
         Tue, 18 Jun 1996 07:09:19 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835061523:05605:2; Tue, 18 Jun 96 02:32:03 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa05401; 18 Jun 96 2:31 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id PAA16885 for dbi-users-real; Mon, 17 Jun 1996 15:48:33 -0700
Precedence: junk
Sender: [email protected]
Received: from europa.ikonic.com (mypop.ikonic2.com [204.119.99.106])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id PAA16881
         for <[email protected]>; Mon, 17 Jun 1996 15:48:30 -0700
Received: from 204.119.99.40.ikonic.com ([204.119.99.18])
         by europa.ikonic.com (post.office MTA v1.9.3 ID# 0-11965) with SMTP
         id AAA92; Mon, 17 Jun 1996 15:44:25 -0700
X-Sender: [email protected]
X-Mailer: Windows Eudora Light Version 1.5.2
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: Tim Bunce <[email protected]>
From: [email protected] (Parker, Drew)
Subject: Re: Request for diagnostic advice
Cc: [email protected]
Date: Mon, 17 Jun 1996 15:44:25 -0700
Message-ID: <[email protected]>

At 09:19 PM 6/17/96 +0000, you wrote:

>
>I presume you're saying that it works and does _not_ get an invalid
>handle error. That would be rather strange.
>

I have had a few different examples of crashes with some of my scripts.  For
the backtrace that I originally sent you, you were right, I did find a
cursor referenced that had not been properly initialized.

Another script, however, gave intermittent "bus errors" when I ran it prior
to setting safe = 1.  After setting safe =1, I cannot reproduce the problem.
I did not change any of the code.  This is an intermittent problem, so I am
wondering if there are any known unix environment issues (e.g., disk space,
memory, system kernel parameters, etc) that might affect this.  In the short
run, I will continue to run with safe=1 and see if the error resurfaces.

I did save the core file for the last time it crashed in this context, but
for that core dump, I cannot even obtain a decent backtrace (below).  I have
sent you a separate message re: your advice to T.L. Priest about recompiling
with improved diagnostics, perhaps you can send me the same file.

By the way, I think some of my intermittent bus errors may be related to
what T.L. Priest and David Lebel have reported; I have seen bus errors occur
on "exit" after a successful &ora_logoff.

THANK YOU for all of your efforts on the DBI.  It can't be said often enough.

Drew
------------

GDB is free software and you are welcome to distribute copies of it
under certain conditions; type "show copying" to see the conditions.
There is absolutely no warranty for GDB; type "show warranty" for details.
GDB 4.15.1 (mips-sgi-irix5.3), Copyright 1995 Free Software Foundation, Inc...
Core was generated by `TestBug'.
Program terminated with signal 10, Bus error.
Reading symbols from /usr/lib/libmalloc.so...done.
Reading symbols from /usr/lib/libm.so...done.
Reading symbols from /usr/lib/libc.so.1...done.
Reading symbols from /usr/lib/libcrypt.so...done.
Reading symbols from /usr/lib/libgen.so...done.
Reading symbols from
/usr/local/lib/perl5/site_perl/IP22-irix/auto/DBI/DBI.so...done.
Reading symbols from
/usr/local/lib/perl5/site_perl/IP22-irix/auto/DBD/Oracle/Oracle.so...done.
Reading symbols from /usr/lib/libsocket.so...done.
Reading symbols from /usr/lib/libnsl.so...done.

warning: Hit heuristic-fence-post without finding

warning: enclosing function for address 0xfb7da74
This warning occurs if you are debugging a function without any symbols
(for example, in a stripped executable).  In that case, you may wish to
increase the size of the search with the `set heuristic-fence-post' command.

Otherwise, you told GDB there was a function where there isn't one, or
(more likely) you have encountered a bug in GDB.
#0  0xfb7da74 in _sobuf ()
(gdb) bt
#0  0xfb7da74 in _sobuf ()
warning: Hit heuristic-fence-post without finding
warning: enclosing function for address 0xfb746f8
(gdb)
------------------
Drew Parker
[email protected]
Ikonic Interactive
2 Harrison Street, Top Floor
San Francisco, California  94105 USA
(415) 908-8069


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13676-17@oink>;
         Tue, 18 Jun 1996 07:14:23 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835070312:12709:1; Tue, 18 Jun 96 04:58:32 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa12586; 18 Jun 96 4:57 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa24354; 18 Jun 96 4:56 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id SAA17224 for dbi-users-real; Mon, 17 Jun 1996 18:41:09 -0700
Precedence: junk
Sender: [email protected]
Received: from voro.lbl.gov (voro.lbl.gov [131.243.240.19])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id SAA17220
         for <[email protected]>; Mon, 17 Jun 1996 18:41:08 -0700
Received: from voro.lbl.gov (localhost [127.0.0.1]) by voro.lbl.gov (8.6.12/G)
         with ESMTP id SAA11567 for <[email protected]>;
         Mon, 17 Jun 1996 18:38:36 -0700
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.2 7/18/95
Reply-to: [email protected]
To: [email protected]
Subject: misc 7.3 issues
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 17 Jun 1996 18:38:35 -0700
From: Mark Dedlow <[email protected]>

With Oracle 7.3.2, Solaris 2.4, Perl 5.002, DBD 0.32, DBI 0.70
I now get:

  Bad free() ignored during global destruction.

On exit from scripts.   Otherwise clean build "out of the box".
Passes all tests, and everything else I've tried works.

I've seen discussion of the bad free() problem before,
but can't find any resolution to it in old mail.

Any ideas?

Mark

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13676-18@oink>;
         Tue, 18 Jun 1996 07:14:32 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835070598:13980:0; Tue, 18 Jun 96 05:03:18 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa13918; 18 Jun 96 5:02 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa26839; 18 Jun 96 5:02 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id SAA17200 for dbi-users-real; Mon, 17 Jun 1996 18:30:10 -0700
Precedence: junk
Sender: [email protected]
Received: from inet-user-gw-1.us.oracle.com (inet-user-gw-1.us.oracle.com [192.86.155.82])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id SAA17196
         for <[email protected]>; Mon, 17 Jun 1996 18:30:09 -0700
Received: from ap215sun.us.oracle.com by inet-user-gw-1.us.oracle.com
         with ESMTP (8.6.12/37.7) id SAA17323; Mon, 17 Jun 1996 18:30:06 -0700
Received: by ap215sun.us.oracle.com (8.6.12/37.8) id SAA22338;
         Mon, 17 Jun 1996 18:26:10 -0700
Message-Id: <[email protected]>
Date: Mon, 17 Jun 1996 18:26:10 -0700
From: Dennis Fogg <[email protected]>
To: [email protected], [email protected]
Subject: Garbage Collection bug in Oraperl

Is there a URL or ftp file somewhere that lists known bugs?
The only thing I know about is the DBD link on the DBI home page.

I've run into a bug in Oraperl and it's not listed on that DBD page.
Basically, I get an array back from ora_fetch, assign an
array reference to it, and then try to refer to that
reference after exiting a block.  Turns out that the
array has been garbage collected when I exit the block.
If I copy the array inside the block and make a reference
to the copy, then I can refer to the array outside the block.
I tried the
       BEGIN { $Oraperl::safe = 1 }
modification since it was being discussed, but this did
not fix my bug.
Make a copy is an easy workaround, but it took some time
to figure out what was going on.


Here's some details:

The following code generates the output given after it:

   while(@row = &ora_fetch($c2)) {
       $row_ref  = \@row;
       print "row_ref (inside) = $$row_ref[1]\n";
   }
       print "row_ref (outside) = $$row_ref[1]\n";


row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
Use of uninitialized value at j.perl line 86.
row_ref (outside) =

(where line 86 is the line that prints "row_ref (outside)"


When I modify the code, I get what I expected (ie: my array ref works):

   while(@row = &ora_fetch($c2)) {

       # Make a copy of row
       @row_copy = @row;
       $row_ref  = \@row_copy;
       print "row_ref (inside) = $$row_ref[1]\n";
   }
       print "row_ref (outside) = $$row_ref[1]\n";


row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
row_ref (outside) = 18-JUN-96
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <16106-0@oink>;
         Tue, 18 Jun 1996 10:07:19 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835083836:28157:2; Tue, 18 Jun 96 08:43:56 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa28031; 18 Jun 96 8:43 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id AAA18348 for dbi-users-real; Tue, 18 Jun 1996 00:33:00 -0700
Precedence: junk
Sender: [email protected]
Received: from inet-user-gw-1.us.oracle.com (inet-user-gw-1.us.oracle.com [192.86.155.82])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id AAA18344
         for <[email protected]>; Tue, 18 Jun 1996 00:32:59 -0700
Received: from mailsun2.us.oracle.com by inet-user-gw-1.us.oracle.com
         with ESMTP (8.6.12/37.7) id XAA16159; Mon, 17 Jun 1996 23:45:56 -0700
Received: by mailsun2.us.oracle.com (8.7.1/37.8) id XAA15772;
         Mon, 17 Jun 1996 23:45:45 -0700 (PDT)
Message-Id: <[email protected]>
Date: 17 Jun 96 23:27:34 -0700
From: "DFOGG.US.ORACLE.COM" <[email protected]>
To: [email protected], [email protected]
Subject: CORRECTION: Garbage Collection bug in Oraperl
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary=Boundary-5434670-0-0


--Boundary-5434670-0-0

Oops, someone pointed out a bug in my bug, so please
ignore my previous posting.  Everything's working just fine.




--Boundary-5434670-0-0
Content-Type: message/rfc822

Date: 17 Jun 96 18:26:10
From:"Dennis Fogg " <[email protected]>
To: [email protected],[email protected]
Subject: Garbage Collection bug in Oraperl
X-Orcl-Application: Precedence:  junk
X-Orcl-Application: Sender:  [email protected]


Is there a URL or ftp file somewhere that lists known bugs?
The only thing I know about is the DBD link on the DBI home page.

I've run into a bug in Oraperl and it's not listed on that DBD page.
Basically, I get an array back from ora_fetch, assign an
array reference to it, and then try to refer to that
reference after exiting a block.  Turns out that the
array has been garbage collected when I exit the block.
If I copy the array inside the block and make a reference
to the copy, then I can refer to the array outside the block.
I tried the
       BEGIN { $Oraperl::safe = 1 }
modification since it was being discussed, but this did
not fix my bug.
Make a copy is an easy workaround, but it took some time
to figure out what was going on.


Here's some details:

The following code generates the output given after it:

   while(@row = &ora_fetch($c2)) {
       $row_ref  = \@row;
       print "row_ref (inside) = $$row_ref[1]\n";
   }
       print "row_ref (outside) = $$row_ref[1]\n";


row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
Use of uninitialized value at j.perl line 86.
row_ref (outside) =

(where line 86 is the line that prints "row_ref (outside)"


When I modify the code, I get what I expected (ie: my array ref works):

   while(@row = &ora_fetch($c2)) {

       # Make a copy of row
       @row_copy = @row;
       $row_ref  = \@row_copy;
       print "row_ref (inside) = $$row_ref[1]\n";
   }
       print "row_ref (outside) = $$row_ref[1]\n";


row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
row_ref (inside) = 18-JUN-96
row_ref (outside) = 18-JUN-96


--Boundary-5434670-0-0--
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <18284-1@oink>;
         Tue, 18 Jun 1996 17:02:37 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835111041:08049:3; Tue, 18 Jun 96 16:17:21 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa07571; 18 Jun 96 16:16 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA18979 for dbi-users-real; Tue, 18 Jun 1996 08:11:21 -0700
Precedence: junk
Sender: [email protected]
Received: from astor.mediabridge.com (astor.mediabridge.com [165.254.109.151])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA18974
         for <[email protected]>; Tue, 18 Jun 1996 08:11:19 -0700
Received: from mediabridge.com (thayer@localhost)
         by astor.mediabridge.com (8.7.3/8.6.9/sed) with ESMTP id LAA27873
         for <[email protected]>; Tue, 18 Jun 1996 11:11:10 -0400 (EDT)
Message-Id: <[email protected]>
To: [email protected]
Subject: Building Sos2.5 Ora7.3
Date: Tue, 18 Jun 1996 11:11:08 -0400
From: Charles Thayer <[email protected]>

 I'm having an unusual error which may or may not be new and unique
to solaris 2.5 on the sparc-ultra.  Any help is greatly appreciated.
It seems that libc is failing when I run the tests:

   libc internal error: _rmutex_unlock: rmutex not held.

/charles
PS. Here's the details.

[metrobeatdb 17:33:12 0] DBD-Oracle-0.32> perl Makefile.PL

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1
Warning: OCI (Pro*C) does not appear to be installed.
Oracle sysliblist: -lsocket -lnsl -lm -ldl
Using /metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/precomp/demo/proc/proc.m
k /metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/precomp/env_precomp.mk
System: perl5.002 sunos metrobeatdb 5.5 generic sun4u sparc sunw,ultra-1
Compiler: gcc -O -I/opt/local/include
Oracle proc.mk would have used these values but we override them:
 CC:       cc
 CFLAGS:   $(GFLAG) $(OPTIMIZE) $(CDEBUG) $(CCFLAGS) $(QACCFLAGS) $(PFLAGS) $(S
HARED_CFLAG) $(ENV_FLAGS)
 LDFLAGS:  -o $@ $(PRODUCT_LIBHOME) -L$(LIBHOME) -L/opt/SUNWcluster/lib -R/opt/
SUNWcluster/lib
Linking with -lclntsh -lsql -lsqlnet -lncr -lsqlnet -lclient -lcommon -lgeneric
-lsqlnet -lncr -lsqlnet -lclient -lcommon -lgeneric $(LIBPLSHACK) -lepc -lnlsrtl
3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lnlsrtl3 $(SPLIBS) $(LOCALLIBS) $(EXOSLIBS)
$(LIBBSD) `cat /metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/rdbms/lib/sysli
blist` $(SECLIBS) $(M6LIBS) -lm -lthread

Checking if your kit is complete...
Looks good

Warning: By default new modules are installed into your 'site_lib'
directories. Since site_lib directories come after the normal library
directories you must delete any old DBD::Oracle files and directories from
your 'privlib' and 'archlib' directories and their auto subdirectories.

Writing Makefile for DBD::Oracle
[metrobeatdb 17:33:24 0] DBD-Oracle-0.32> make test
umask 0 && cp Oraperl.pm blib/lib/Oraperl.pm
umask 0 && cp oraperl.ph blib/lib/oraperl.ph
umask 0 && cp Oracle.pm blib/lib/DBD/Oracle.pm
perl -I/opt/local/lib/perl5/sun4-solaris/5.002 -I/opt/local/lib/perl5 /opt/local
/lib/perl5/ExtUtils/xsubpp  -typemap /opt/local/lib/perl5/ExtUtils/typemap Oracl
e.xs >Oracle.tc && mv Oracle.tc Oracle.c
gcc -c -I. -I/metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/precomp/public -I
/metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/rdbms/demo -I/opt/local/lib/pe
rl5/site_perl/sun4-solaris/DBI -I/opt/local/lib/perl5/sun4-solaris/5.002/DBI -I/
opt/local/include -O -DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" -fpic -I/opt/local
/lib/perl5/sun4-solaris/5.002/CORE -Wall -pedantic -Wno-comment -Wtraditional Or
acle.c
gcc -c -I. -I/metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/precomp/public -I
/metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/rdbms/demo -I/opt/local/lib/pe
rl5/site_perl/sun4-solaris/DBI -I/opt/local/lib/perl5/sun4-solaris/5.002/DBI -I/
opt/local/include -O -DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" -fpic -I/opt/local
/lib/perl5/sun4-solaris/5.002/CORE -Wall -pedantic -Wno-comment -Wtraditional db
dimp.c
dbdimp.c: In function `dbd_db_FETCH':
dbdimp.c:265: warning: value computed is not used
dbdimp.c: In function `dbd_st_FETCH':
dbdimp.c:910: warning: value computed is not used
dbdimp.c: At top level:
dbdimp.c:126: warning: `dump_error_status' defined but not used
Running Mkbootstrap for DBD::Oracle ()
chmod 644 Oracle.bs
LD_RUN_PATH="/lib" ld -o blib/arch/auto/DBD/Oracle/Oracle.so -G -L/opt/local/lib
Oracle.o dbdimp.o -L/metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/lib /metr
obeatdb/d3/oracle/app/oracle/product/7.3.2.1/lib/__fstd.o -lclntsh -lsql -lsqlne
t -lncr -lsqlnet -lclient -lcommon -lgeneric -lsqlnet -lncr -lsqlnet -lclient -l
common -lgeneric  -lepc -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lnlsrtl3
`cat /metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/rdbms/lib/sysliblist`
-lm -lthread  -lsocket -lnsl -lm -ldl
chmod 755 blib/arch/auto/DBD/Oracle/Oracle.so
cp Oracle.bs ./blib/arch/auto/DBD/Oracle/Oracle.bs
chmod 644 blib/arch/auto/DBD/Oracle/Oracle.bs
Manifying ./blib/man3/DBD::Oraperl.3
PERL_DL_NONLAZY=1 /opt/local/bin/perl -I./blib/arch -I./blib/lib -I/opt/local/li
b/perl5/sun4-solaris/5.002 -I/opt/local/lib/perl5 -e 'use Test::Harness qw(&runt
ests $verbose); $verbose=0; runtests @ARGV;' t/*.t
t/base..............ok
All tests successful.
Files=1,  Tests=5,  1 secs ( 0.22 cusr  0.05 csys =  0.27 cpu)
PERL_DL_NONLAZY=1 /opt/local/bin/perl -I./blib/arch -I./blib/lib -I/opt/local/li
b/perl5/sun4-solaris/5.002 -I/opt/local/lib/perl5 test.pl
Oraperl test application $Revision: 1.20 $

Oraperl Emulation Interface version 1.26
Oracle Driver 0.32
DBI-0.69 Switch by Tim Bunce, version 0.69


Connecting
to '' (from command line, else uses ORACLE_SID/TWO_TASK)
as 'crgs/crgs' (via ORACLE_USERID environment var or default)
libc internal error: _rmutex_unlock: rmutex not held.
^@Abort - core dumped
make: *** [test_dynamic] Error 134
[metrobeatdb 17:33:54 2] DBD-Oracle-0.32> make test TEST_VERBOSE=1
PERL_DL_NONLAZY=1 /opt/local/bin/perl -I./blib/arch -I./blib/lib -I/opt/local/li
b/perl5/sun4-solaris/5.002 -I/opt/local/lib/perl5 -e 'use Test::Harness qw(&runt
ests $verbose); $verbose=1; runtests @ARGV;' t/*.t
t/base..............1..5
ok 1
ok 2
ok 3
ok 4
ok 5
ok
All tests successful.
Files=1,  Tests=5,  1 secs ( 0.22 cusr  0.03 csys =  0.25 cpu)
PERL_DL_NONLAZY=1 /opt/local/bin/perl -I./blib/arch -I./blib/lib -I/opt/local/li
b/perl5/sun4-solaris/5.002 -I/opt/local/lib/perl5 test.pl
Oraperl test application $Revision: 1.20 $

Oraperl Emulation Interface version 1.26
Oracle Driver 0.32
DBI-0.69 Switch by Tim Bunce, version 0.69


Connecting
to '' (from command line, else uses ORACLE_SID/TWO_TASK)
as 'crgs/crgs' (via ORACLE_USERID environment var or default)
libc internal error: _rmutex_unlock: rmutex not held.
^@Abort - core dumped
make: *** [test_dynamic] Error 134
[metrobeatdb 17:34:18 2] DBD-Oracle-0.32> perl -V
Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=solaris, osver=2.5, archname=sun4-solaris
   uname='sunos metrobeatdb 5.5 generic sun4u sparc sunw,ultra-1 '
   hint=recommended, useposix=true
 Compiler:
   cc='gcc', optimize='-O', gccversion=2.7.2
   cppflags='-I/opt/local/include'
   ccflags ='-I/opt/local/include'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='gcc', ldflags =' -L/opt/local/lib'
   libpth=/opt/local/lib /lib /usr/lib /usr/ccs/lib
   libs=-lsocket -lnsl -ldl -lm -lc -lcrypt
   libc=/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-fpic', lddlflags='-G -L/opt/local/lib'

@INC: /h/thayer/perl/lib /usr/local/gnu/lib/perl . /opt/local/lib/perl5/sun4-sol
aris/5.002 /opt/local/lib/perl5 /opt/local/lib/perl5/site_perl/sun4-solaris /opt
/local/lib/perl5/site_perl .
[metrobeatdb 17:34:26 0] DBD-Oracle-0.32> echo $LD_LIBRARY_PATH
/metrobeatdb/d3/oracle/app/oracle/product/7.3.2.1/lib

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <20007-14@oink>;
         Tue, 18 Jun 1996 22:58:14 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835125743:05755:3; Tue, 18 Jun 96 20:22:23 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa03901; 18 Jun 96 20:22 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA19024 for dbi-users-real; Tue, 18 Jun 1996 08:21:23 -0700
Precedence: junk
Sender: [email protected]
Received: from dekalb.vf.mmc.com (dekalb.vf.mmc.com [192.35.35.21])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA19020
         for <[email protected]>; Tue, 18 Jun 1996 08:21:21 -0700
Received: from franklin.vf.lmco.com (franklin.vf.lmco.com [166.17.5.51])
         by dekalb.vf.mmc.com (8.7.5/8.7.3) with ESMTP id LAA02171;
         Tue, 18 Jun 1996 11:21:01 -0400 (EDT)
Received: from atl.lmco.com (hannibal.ATL.LMCO.COM [166.20.244.250])
         by franklin.vf.lmco.com (8.7.5/8.7.3) with SMTP id LAA23742;
         Tue, 18 Jun 1996 11:21:00 -0400 (EDT)
Received: from valerie by atl.lmco.com (8.6.13/ATL 1.14) with SMTP id LAA06545;
         Tue, 18 Jun 1996 11:20:57 -0400
Message-ID: <[email protected]>
Date: Tue, 18 Jun 1996 11:20:56 -0400
From: Amy McGovern <[email protected]>
Organization: Lockheed Martin Advanced Technology Laboratory
X-Mailer: Mozilla 2.01 (X11; I; SunOS 5.5 sun4m)
MIME-Version: 1.0
To: Charles Thayer <[email protected]>
CC: [email protected]
Subject: Re: Building Sos2.5 Ora7.3
References: <[email protected]>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Charles Thayer wrote:
>   I'm having an unusual error which may or may not be new and unique
> to solaris 2.5 on the sparc-ultra.  Any help is greatly appreciated.
> It seems that libc is failing when I run the tests:
>
>     libc internal error: _rmutex_unlock: rmutex not held.

This is not an unusual error.  As far as I know, no one has gotten OraPerl to
compile for Solaris 2.5 and Oracle 7.3.2.  We all keep coming back to this
same problem with the mutex!

Tim Bunce is working on a solution.

-Amy
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20009-16@oink>;
         Tue, 18 Jun 1996 22:58:29 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835122320:04035:0; Tue, 18 Jun 96 19:25:20 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa03843; 18 Jun 96 19:25 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa13659; 18 Jun 96 19:23 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA18806 for dbi-users-real; Tue, 18 Jun 1996 06:22:56 -0700
Precedence: junk
Sender: [email protected]
Received: from colorbank.com (colorbank.com [206.20.79.3])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id GAA18802
         for <[email protected]>; Tue, 18 Jun 1996 06:22:54 -0700
Received: from www.colorbank.com by colorbank.com
         via SMTP (940816.SGI.8.6.9/940406.SGI.AUTO)
         for <[email protected]> id CAA02946;
         Tue, 18 Jun 1996 02:25:36 -0700
Message-ID: <[email protected]>
Date: Tue, 18 Jun 1996 02:25:35 -0700
From: Hongyuan Zhang <[email protected]>
X-Mailer: Mozilla 2.0 (X11; I; IRIX 5.3 IP22)
MIME-Version: 1.0
To: [email protected]
Subject: help: how to use SQL*NET to access database?
X-URL: http://www.hermetica.com/technologia/DBI/DBI/index.html
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear Sir:

In PRO*C, you can access database using SQL*NET As the following
statement:

EXEC SQL CONNECT :username IDENTIFIED BY :password
USING :db_string1.

Does any one can help do the same thing in perl?

Thanks.

John
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <20007-20@oink>;
         Tue, 18 Jun 1996 22:58:35 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835129473:11944:1; Tue, 18 Jun 96 21:24:33 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa10317; 18 Jun 96 21:23 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA20153 for dbi-users-real; Tue, 18 Jun 1996 13:04:49 -0700
Precedence: junk
Sender: [email protected]
Received: from astor.mediabridge.com (astor.mediabridge.com [165.254.109.151])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA20149
         for <[email protected]>; Tue, 18 Jun 1996 13:04:45 -0700
Received: from mediabridge.com (thayer@localhost)
         by astor.mediabridge.com (8.7.3/8.6.9/sed) with ESMTP id QAA02342;
         Tue, 18 Jun 1996 16:04:29 -0400 (EDT)
Message-Id: <[email protected]>
To: Amy McGovern <[email protected]>, [email protected]
Cc: Charles Thayer <[email protected]>
CC: [email protected], [email protected]
Subject: Re: Building Oraperl for Sos2.5 Ora7.3
In-reply-to: Your message of "Tue, 18 Jun 1996 11:20:56 EDT." <[email protected]>
Date: Tue, 18 Jun 1996 16:04:27 -0400
From: Charles Thayer <[email protected]>

In message <[email protected]>, Amy McGovern writes:
> Charles Thayer wrote:
> >   I'm having an unusual error which may or may not be new and unique
> > to solaris 2.5 on the sparc-ultra.  Any help is greatly appreciated.
> > It seems that libc is failing when I run the tests:
> >
> >     libc internal error: _rmutex_unlock: rmutex not held.
>
> This is not an unusual error.  As far as I know, no one has gotten OraPerl to
> compile for Solaris 2.5 and Oracle 7.3.2.  We all keep coming back to this
> same problem with the mutex!
>
> Tim Bunce is working on a solution.

I checked some search engines and found that libthread seems to be the
culprit.  Apparently this is a known problem with perl5.002.  One
solution is to add "-lthread" to the file ext.libs in the perl
distribution and recompile.  That made DBD and DBI work for me with my
Oracle installation, but not all tests are passing (for perl that is
-- namely magic is failing now.)

Someone alluded to writing fake stubs for the functions in libthread
as a work-around.  But they pointed to a "SunSolve bug database" which
I don't have access to :-(

Now my oraperl script runs but complains at the end:

       Bad free() ignored during global destruction.

Geez/charles
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20009-34@oink>;
         Tue, 18 Jun 1996 22:59:33 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835131485:04341:1; Tue, 18 Jun 96 21:58:05 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa03581; 18 Jun 96 21:57 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA19501 for dbi-users-real; Tue, 18 Jun 1996 10:35:11 -0700
Precedence: junk
Sender: [email protected]
Received: from colossus.cse.psu.edu ([email protected] [130.203.1.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA19497
         for <[email protected]>; Tue, 18 Jun 1996 10:35:07 -0700
Received: from localhost by colossus.cse.psu.edu with SMTP id <78984>;
         Tue, 18 Jun 1996 13:34:51 -0400
X-Mailer: exmh version 1.6.5 12/11/95
X-uri: <URL:http://www.cse.psu.edu/~groenvel/>
X-work-address: Manufacturing Technology - Systems Development Group PO Box 30
               The Applied Research Lab State College, PA 16804
X-work-phone: +1 814 863 9896 (Voice) +1 814 863 1183 (Fax)
To: Hongyuan Zhang <[email protected]>
cc: [email protected]
Subject: Re: help: how to use SQL*NET to access database?
In-reply-to: Your message of "Tue, 18 Jun 1996 05:25:35 EDT." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 18 Jun 1996 13:33:32 -0400
From: John D Groenveld <[email protected]>
Message-Id: <[email protected]>

$lda = &ora_login("", "scott/tiger\@V1connectstring_or_V2alias","")||die
"ora_login: $ora_errno: $ora_errstr\n";
or you could set TWO_TASK to your V1 connect string or V2 alias.

John

BTW The http://www.hermetica.com/technologia/DBI/ page has links to the
maillist archives. This is a FAQ.


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20009-41@oink>;
         Tue, 18 Jun 1996 23:00:04 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835133392:25067:1; Tue, 18 Jun 96 22:29:52 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa24967; 18 Jun 96 22:29 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa21605; 18 Jun 96 22:27 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA19508 for dbi-users-real; Tue, 18 Jun 1996 10:36:02 -0700
Precedence: junk
Sender: [email protected]
Received: from voro.lbl.gov (voro.lbl.gov [131.243.240.19])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id KAA19503
         for <[email protected]>; Tue, 18 Jun 1996 10:36:00 -0700
Received: from voro.lbl.gov (localhost [127.0.0.1]) by voro.lbl.gov (8.6.12/G)
         with ESMTP id KAA13530; Tue, 18 Jun 1996 10:26:20 -0700
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.2 7/18/95
To: Tim Bunce <[email protected]>
cc: [email protected]
Subject: Re: misc 7.3 issues
In-reply-to: Your message of "Tue, 18 Jun 1996 14:53:37 -0000." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 18 Jun 1996 10:26:20 -0700
From: Mark Dedlow <[email protected]>

>From: Tim Bunce <[email protected]>
>
> > From: Mark Dedlow <[email protected]>
> >
> > With Oracle 7.3.2, Solaris 2.4, Perl 5.002, DBD 0.32, DBI 0.70
> > I now get:
> >
> >    Bad free() ignored during global destruction.
> >
> > On exit from scripts.   Otherwise clean build "out of the box".
> > Passes all tests, and everything else I've tried works.
>
> "Now" being since when? What changed?

Sorry, sometimes I forget you're not a mind reader.  Or maybe
you are, but there's nothing on this end to read.

"Since" meant since install of 0.32/0.70 on 7.3.
I jumped to there from 0.29/0.68 on 7.2.3.

I thought I'd wait out the 7.3 build problems, though it turned
out I didn't have any.


> What's the smallest script which has the problem?

#!/usr/bin/perl

use DBI;

$cs = 'user/pass@db';
$drh = DBI->install_driver('Oracle') || die "$DBI::errstr\n";
$dbh = $drh->connect("", "$cs", '') || die "$DBI::errstr\n";
$stmt = 'select sysdate from dual';
$cursor = $dbh->prepare($stmt) || die "$DBI::errstr\n";
$cursor->execute || die "$DBI::errstr\n";
while (@row = $cursor->fetchrow) { print "@row\n" }
$cursor->finish;
$dbh->disconnect;
exit(0);


> This'll probably suppress it:
>
>  $SIG{__WARN__} = sub { warn $_[0] unless $_[0] =~ /Bad free/ };

Yes. It does.

> but this would be more useful to me:
>
>  $SIG{__WARN__} = sub { $_[0] =~ /Bad free/ ? dump : warn $_[0] };

I don't have DEBUGGING in perl at the moment.  I'll do this when
I get a few minutes later this week.

I did set DBI debugging and got this.  Does not say a heck of a lot
more however.

Loaded DBI.pm at ./DBI_test_bad_free.pl line 3
   DBI dispatch debug level set to 2
DBI->install_driver DBI Oracle at ./DBI_test_bad_free.pl line 8
DBI->install_driver(Oracle) loaded
   New DBI::dr (for DBD::Oracle::dr, parent=, id=)
   dbih_setup_handle(DBI::dr=HASH(0x137ddc)=>DBI::dr=HASH(0x15eb44),
DBD::Oracle::dr, NULL)
   New DBI::dr => DBI::dr=HASH(0x137ddc) (inner=DBI::dr=HASH(0x15eb44)) for
DBD::Oracle::dr
DBI->install_driver(Oracle) = DBI::dr=HASH(0x137ddc)
   >> connect     DISPATCH (DBI::dr=HASH(0x137ddc) @4 g0 a15df48 r1)
   -> connect for DBD::Oracle::dr (DBI::dr=HASH(0x137ddc)~0x15eb44 ''
'pub/cil@dw' '')
   New DBI::db (for DBD::Oracle::db, parent=DBI::dr=HASH(0x15eb44), id=)
   dbih_setup_handle(DBI::db=HASH(0x159ec0)=>DBI::db=HASH(0x1562d8),
DBD::Oracle::db, NULL)
   New DBI::db => DBI::db=HASH(0x159ec0) (inner=DBI::db=HASH(0x1562d8)) for
DBD::Oracle::db
   <- connect= DBI::db=HASH(0x159ec0)
   >> prepare     DISPATCH (DBI::db=HASH(0x159ec0) @2 g0 a15c288 r1)
   -> prepare for DBD::Oracle::db (DBI::db=HASH(0x159ec0)~0x1562d8 'select
sysdate from dual')
   New DBI::st (for DBD::Oracle::st, parent=DBI::db=HASH(0x1562d8), id=)
   dbih_setup_handle(DBI::st=HASH(0x15b940)=>DBI::st=HASH(0x15b958),
DBD::Oracle::st, NULL)
   New DBI::st => DBI::st=HASH(0x15b940) (inner=DBI::st=HASH(0x15b958)) for
DBD::Oracle::st
   <- prepare= DBI::st=HASH(0x15b940)
   >> execute     DISPATCH (DBI::st=HASH(0x15b940) @1 g0 a160088 r1)
   -> execute for DBD::Oracle::st (DBI::st=HASH(0x15b940)~0x15b958)
fbh 1: 'SYSDATE' NULLable, type 12,  dbsize 7, dsize 75, p0 s0
  out: ftype 5, indp 0, bufl 76, rlen 0, rcode 0
   <- execute= '0E0'
   >> fetchrow    DISPATCH (DBI::st=HASH(0x15b940) @1 g1 a0 r1)
   -> fetchrow for DBD::Oracle::st (DBI::st=HASH(0x15b940)~0x15b958)
   <- fetchrow= ( '18-JUN-96' ) [1 items]
18-JUN-96
   >> fetchrow    DISPATCH (DBI::st=HASH(0x15b940) @1 g1 a0 r1)
   -> fetchrow for DBD::Oracle::st (DBI::st=HASH(0x15b940)~0x15b958)
   <- fetchrow= ( ) [0 items]
   >> finish      DISPATCH (DBI::st=HASH(0x15b940) @1 g0 a15e3e8 r1)
   -> finish for DBD::Oracle::st (DBI::st=HASH(0x15b940)~0x15b958)
   <- finish= 1
   >> disconnect  DISPATCH (DBI::db=HASH(0x159ec0) @1 g0 a15c148 r1)
   -> disconnect for DBD::Oracle::db (DBI::db=HASH(0x159ec0)~0x1562d8)
   <- disconnect= 1
DBI::END
DBI::disconnect_all DBI
DBI::disconnect_all for 'Oracle'
   >> disconnect_all DISPATCH (DBI::dr=HASH(0x137ddc) @1 g0 a15dd28 r2)
   -> disconnect_all for DBD::Oracle::dr (DBI::dr=HASH(0x137ddc)~0x15eb44)
   <- disconnect_all= 1
DBI::END complete
   >> DESTROY     DISPATCH (DBI::st=HASH(0x15b940) @1 g0 a15fe68 r0)
      (outer handle DESTROY ignored)
   >> DESTROY     DISPATCH (DBI::st=HASH(0x15b958) @1 g0 a15fe68 r0)
   -> DESTROY for DBD::Oracle::st (DBI::st=HASH(0x15b958))
   <- DESTROY= undef
   dbih_clearcom 0x18f008 done
   >> DESTROY     DISPATCH (DBI::db=HASH(0x159ec0) @1 g0 a15bfa8 r0)
      (outer handle DESTROY ignored)
   >> DESTROY     DISPATCH (DBI::db=HASH(0x1562d8) @1 g0 a15bfa8 r0)
   -> DESTROY for DBD::Oracle::db (DBI::db=HASH(0x1562d8))
   <- DESTROY= undef
   dbih_clearcom 0xbac08 done
   >> DESTROY     DISPATCH (DBI::dr=HASH(0x137ddc) @1 g0 a15dfa8 r0)
      (outer handle DESTROY ignored)
   >> DESTROY     DISPATCH (DBI::dr=HASH(0x15eb44) @1 g0 a15dfa8 r0)
   dbih_clearcom 0x14d188 done
Attempt to free unreferenced scalar during global destruction.


Mark


---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <21310-2@oink>;
         Wed, 19 Jun 1996 00:08:25 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835138449:22221:2; Tue, 18 Jun 96 23:54:09 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa21685; 18 Jun 96 23:53 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA19501 for dbi-users-real; Tue, 18 Jun 1996 10:35:11 -0700
Precedence: junk
Sender: [email protected]
Received: from colossus.cse.psu.edu ([email protected] [130.203.1.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA19497
         for <[email protected]>; Tue, 18 Jun 1996 10:35:07 -0700
Received: from localhost by colossus.cse.psu.edu with SMTP id <78984>;
         Tue, 18 Jun 1996 13:34:51 -0400
X-Mailer: exmh version 1.6.5 12/11/95
X-uri: <URL:http://www.cse.psu.edu/~groenvel/>
X-work-address: Manufacturing Technology - Systems Development Group PO Box 30
               The Applied Research Lab State College, PA 16804
X-work-phone: +1 814 863 9896 (Voice) +1 814 863 1183 (Fax)
To: Hongyuan Zhang <[email protected]>
cc: [email protected]
Subject: Re: help: how to use SQL*NET to access database?
In-reply-to: Your message of "Tue, 18 Jun 1996 05:25:35 EDT." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 18 Jun 1996 13:33:32 -0400
From: John D Groenveld <[email protected]>
Message-Id: <[email protected]>

$lda = &ora_login("", "scott/tiger\@V1connectstring_or_V2alias","")||die
"ora_login: $ora_errno: $ora_errstr\n";
or you could set TWO_TASK to your V1 connect string or V2 alias.

John

BTW The http://www.hermetica.com/technologia/DBI/ page has links to the
maillist archives. This is a FAQ.


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21312-4@oink>;
         Wed, 19 Jun 1996 00:08:42 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835138643:23238:1; Tue, 18 Jun 96 23:57:23 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa22776; 18 Jun 96 23:56 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa24170; 18 Jun 96 23:56 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA19312 for dbi-users-real; Tue, 18 Jun 1996 09:55:21 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net ([158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA19308
         for <[email protected]>; Tue, 18 Jun 1996 09:55:16 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ad04609; 18 Jun 96 17:09 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ae09117; 18 Jun 96 17:02 +0100
Received: from toad by oink with SMTP (PP) id <17666-0@oink>;
         Tue, 18 Jun 1996 14:53:43 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA02901;
         Tue, 18 Jun 1996 14:53:37 +0000
Date: Tue, 18 Jun 1996 14:53:37 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: misc 7.3 issues
X-Sun-Charset: US-ASCII
content-length: 718
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Mark Dedlow <[email protected]>
>
> With Oracle 7.3.2, Solaris 2.4, Perl 5.002, DBD 0.32, DBI 0.70
> I now get:
>
>    Bad free() ignored during global destruction.
>
> On exit from scripts.   Otherwise clean build "out of the box".
> Passes all tests, and everything else I've tried works.

"Now" being since when? What changed?

What's the smallest script which has the problem?

> I've seen discussion of the bad free() problem before,
> but can't find any resolution to it in old mail.

This'll probably suppress it:

$SIG{__WARN__} = sub { warn $_[0] unless $_[0] =~ /Bad free/ };

but this would be more useful to me:

$SIG{__WARN__} = sub { $_[0] =~ /Bad free/ ? dump : warn $_[0] };

:-)

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21843-3@oink>;
         Wed, 19 Jun 1996 01:59:56 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835140421:10887:1; Wed, 19 Jun 96 00:27:01 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa10830; 19 Jun 96 0:26 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa05363; 19 Jun 96 0:25 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA20867 for dbi-users-real; Tue, 18 Jun 1996 16:10:22 -0700
Precedence: junk
Sender: [email protected]
Received: from zeus.carroll.com (zeus.carroll.com [199.224.10.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA20863
         for <[email protected]>; Tue, 18 Jun 1996 16:10:19 -0700
Received: from Phred by zeus.carroll.com with SMTP (8.7.1/0) id TAA04166;
         Tue, 18 Jun 1996 19:12:57 -0400 (EDT)
Message-Id: <[email protected]>
X-Sender: [email protected]
X-Mailer: Windows Eudora Version 1.4.3
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="=====================_835168832==_"
Date: Tue, 18 Jun 1996 19:20:32 -0500
To: [email protected]
From: Jon Adams <[email protected]>
Subject: Re: problem with dynamic library loading on Solaris 2.5
Cc: [email protected]
X-Attachments: C:\WINDOWS\DESKTOP\LOG.OUT;

--=====================_835168832==_
Content-Type: text/plain; charset="us-ascii"

>In article <[email protected]>,
>Jon Adams  <[email protected]> wrote:

>
>Try manually setting the LD_LIBRARY_PATH environment variable to the
>full path of the directory holding the libclntsh.so.1.0 file.  See the
>ld unix man page for more info. I'm assuming your perl isn't set-uid.
>

Okay, I've been working on this long enough that I had remembered trying this
and it not working, but I didn't remember _how_ it was not working. One of
the problems was that I forgot I had to export the variable in ksh. The other
problem was that, after setting LD_LIBRARY_PATH, it just dumps core.
Assuming the library problem is resolved by setting LD_LIBRARY_PATH, and this
is a new problem, I'll give the new details...

Again, any input would be appreciated.

Jon Adams


The results of the make and the test are attached.

I ran base.t (which coredumped) within gdb...here's the trace:

Reading symbols from /usr/local/bin/perl...done.
(gdb) run base.t
Starting program: /usr/local/bin/perl base.t
DynaLoader.pm loaded (/usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/per
l5 /usr/local/lib/perl5/site_perl/sun4-solaris /usr/local/lib/perl5/site_perl /u
sr/local/lib/perl5/sun4-solaris . /export/home/jadams/DBD-Oracle-0.32/blib/lib/
/export/home/jadams/DBD-Oracle-0.32/blib/arch /usr/local/lib/perl5/sun4-solaris/
5.002 /usr/local/lib/perl5, /lib /usr/lib /usr/local/lib /usr/lib /usr/local/lib
/db/oracle/app/oracle/product/7.3.2/lib)
DynaLoader::bootstrap for DBI (auto/DBI/DBI.so)
DynaLoader::bootstrap for DBD::Oracle (auto/DBD/Oracle/Oracle.so)

Program received signal SIGSEGV, Segmentation fault.
0xec683118 in ?? ()
(gdb) where
#0  0xec683118 in ?? ()
#1  0xef53000c in ?? ()
#2  0xb4e60 in XS_DynaLoader_dl_load_file (cv=0x128ef0) at DynaLoader.c:161
#3  0x68e70 in pp_entersub () at pp_hot.c:1756
#4  0x60e3c in run () at run.c:29
#5  0x1d334 in perl_run (sv_interp=0xd6808) at perl.c:557
#6  0x1ba88 in main (argc=2, argv=0xeffffd54, env=0xeffffd60) at perlmain.c:49
(gdb) quit



--=====================_835168832==_
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: attachment; filename="LOG.OUT"



>perl Makefile.PL



Warning: OCI (Pro*C) does not appear to be installed.

Warning: By default new modules are installed into your 'site_lib'
directories. Since site_lib directories come after the normal library
directories you must delete any old DBD::Oracle files and directories from
your 'privlib' and 'archlib' directories and their auto subdirectories.


Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /db/oracle/app/oracle/product/7.3.2
Oracle sysliblist: -lsocket -lnsl -lm -ldl
Using /db/oracle/app/oracle/product/7.3.2/precomp/demo/proc/proc.mk /db/oracle/app/oracle/product/7.3.2/precomp/env_precomp.mk
System: perl5.002 sunos lennon 5.5 generic sun4u sparc sunw,ultra-1
Compiler: gcc -g -I/usr/local/include
Oracle proc.mk would have used these values but we override them:
 CC:       cc
 CFLAGS:   $(GFLAG) $(OPTIMIZE) $(CDEBUG) $(CCFLAGS) $(QACCFLAGS) $(PFLAGS) $(SHARED_CFLAG) $(ENV_FLAGS)
 LDFLAGS:  -o $@ $(PRODUCT_LIBHOME) -L$(LIBHOME) -L/opt/SUNWcluster/lib -R/opt/SUNWcluster/lib
Linking with -lclntsh -lsql -lsqlnet -lncr -lsqlnet -lclient -lcommon -lgeneric -lsqlnet -lncr -lsqlnet -lclient -lcommon -lgeneric $(LIBPLSHACK) -lepc -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lnlsrtl3 $(SPLIBS) $(LOCALLIBS) $(EXOSLIBS) $(LIBBSD) `cat /db/oracle/app/oracle/product/7.3.2/rdbms/lib/sysliblist` $(SECLIBS) $(M6LIBS) -lm -lthread

Checking if your kit is complete...
Looks good
Writing Makefile for DBD::Oracle
umask 0 && cp Oraperl.pm blib/lib/Oraperl.pm
umask 0 && cp oraperl.ph blib/lib/oraperl.ph
umask 0 && cp Oracle.pm blib/lib/DBD/Oracle.pm
perl -I/usr/local/lib/perl5/sun4-solaris/5.002 -I/usr/local/lib/perl5 /usr/local/lib/perl5/ExtUtils/xsubpp  -typemap /usr/local/lib/perl5/ExtUtils/typemap Oracle.xs >Oracle.tc && mv Oracle.tc Oracle.c
gcc -c -I. -I/db/oracle/app/oracle/product/7.3.2/precomp/public -I/db/oracle/app/oracle/product/7.3.2/rdbms/demo -I/usr/local/lib/perl5/site_perl/sun4-solaris/DBI -I/usr/local/lib/perl5/sun4-solaris/5.002/DBI -I/usr/local/include -g -DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" -fpic -I/usr/local/lib/perl5/sun4-solaris/5.002/CORE -Wall -pedantic -Wno-comment -Wtraditional Oracle.c
gcc -c -I. -I/db/oracle/app/oracle/product/7.3.2/precomp/public -I/db/oracle/app/oracle/product/7.3.2/rdbms/demo -I/usr/local/lib/perl5/site_perl/sun4-solaris/DBI -I/usr/local/lib/perl5/sun4-solaris/5.002/DBI -I/usr/local/include -g -DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" -fpic -I/usr/local/lib/perl5/sun4-solaris/5.002/CORE -Wall -pedantic -Wno-comment -Wtraditional dbdimp.c
dbdimp.c: In function `dbd_db_FETCH':
dbdimp.c:265: warning: value computed is not used
dbdimp.c: In function `dbd_st_FETCH':
dbdimp.c:910: warning: value computed is not used
dbdimp.c: At top level:
dbdimp.c:126: warning: `dump_error_status' defined but not used
Running Mkbootstrap for DBD::Oracle ()
chmod 644 Oracle.bs
LD_RUN_PATH="/lib" ld -o blib/arch/auto/DBD/Oracle/Oracle.so -G -L/usr/local/lib Oracle.o dbdimp.o -L/db/oracle/app/oracle/product/7.3.2/lib /db/oracle/app/oracle/product/7.3.2/lib/__fstd.o -lclntsh -lsql -lsqlnet -lncr -lsqlnet -lclient -lcommon -lgeneric -lsqlnet -lncr -lsqlnet -lclient -lcommon -lgeneric  -lepc -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lnlsrtl3     `cat /db/oracle/app/oracle/product/7.3.2/rdbms/lib/sysliblist`   -lm -lthread  -lsocket -lnsl -lm -ldl
chmod 755 blib/arch/auto/DBD/Oracle/Oracle.so
cp Oracle.bs ./blib/arch/auto/DBD/Oracle/Oracle.bs
chmod 644 blib/arch/auto/DBD/Oracle/Oracle.bs
Manifying ./blib/man3/DBD::Oraperl.3
DynaLoader.pm loaded (/usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/sun4-solaris /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/sun4-solaris ., /lib /usr/lib /usr/local/lib /usr/lib /usr/local/lib /db/oracle/app/oracle/product/7.3.2/lib)
PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5/sun4-solaris/5.002 -I/usr/local/lib/perl5 -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t
DynaLoader.pm loaded (./blib/arch ./blib/lib /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/sun4-solaris /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/sun4-solaris ., /lib /usr/lib /usr/local/lib /usr/lib /usr/local/lib /db/oracle/app/oracle/product/7.3.2/lib)
DynaLoader::bootstrap for FileHandle (auto/FileHandle/FileHandle.so)
DynaLoader::bootstrap for Fcntl (auto/Fcntl/Fcntl.so)
t/base..............DynaLoader.pm loaded (./blib/arch ./blib/lib /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/sun4-solaris /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/sun4-solaris . /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/sun4-solaris /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/sun4-solaris . /export/home/jadams/DBD-Oracle-0.32/blib/lib/ /export/home/jadams/DBD-Oracle-0.32/blib/arch /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5, /lib /usr/lib /usr/local/lib /usr/lib /usr/local/lib /db/oracle/app/oracle/product/7.3.2/lib)
DynaLoader::bootstrap for DBI (auto/DBI/DBI.so)
DynaLoader::bootstrap for DBD::Oracle (auto/DBD/Oracle/Oracle.so)
FAILED before any test output arrived
       Test returned status 0 (wstat 139)
FAILED--1 test script could be run, alas -- no output ever seen
make: *** [test_dynamic] Error 29

--=====================_835168832==_
Content-Type: text/plain; charset="us-ascii"

--
Jon Adams                          Promethean Enterprises, Inc.
[email protected]           http://www.pe-netsystems.com
  - Internet and Web Development for the Corporate World -

--=====================_835168832==_--

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21843-6@oink>;
         Wed, 19 Jun 1996 02:00:16 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835140681:13404:0; Wed, 19 Jun 96 00:31:21 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa12809; 19 Jun 96 0:30 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa07029; 19 Jun 96 0:28 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA20919 for dbi-users-real; Tue, 18 Jun 1996 16:21:36 -0700
Precedence: junk
Sender: [email protected]
Received: from ntigate.rich.nt.com ([email protected] [192.135.215.3])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA20915
         for <[email protected]>; Tue, 18 Jun 1996 16:21:31 -0700
X400-Received: by mta NT.COM in /PRMD=NT/ADMD=MCI/C=US/; Relayed;
              Tue, 18 Jun 1996 23:18:01 +0000
X400-Received: by /PRMD=NT/ADMD=MCI/C=US/; Relayed;
              Tue, 18 Jun 1996 20:52:21 +0000
X400-Received: by /PRMD=NT/ADMD=MCI/C=US/; Relayed;
              Tue, 18 Jun 1996 13:45:26 +0000
Date: Tue, 18 Jun 1996 13:45:26 +0000
X400-Originator: [email protected]
X400-Recipients: non-disclosure:;
X400-MTS-Identifier: [/PRMD=NT/ADMD=MCI/C=US/;<[email protected]]
X400-Content-Type: P2-1984 (2)
Content-Identifier: test problems...
From: Greg Giles <[email protected]>
Message-ID: <[email protected]>
To: Tim Bunce <[email protected]>, dbi-users <[email protected]>
Subject: test problems with Oracle D
X-Mailer: Mail*Link SMTP-QM 3.0.2 GM
Mime-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"; Name="Message Body"
Content-Transfer-Encoding: quoted-printable

Hi All,
As you can gather from the title I'm having a few problems with the =
Oracle 0.32 driver. I've enclosed below the output of the build processes =
as requested in the README. I've been trying for the last few days now, =
but I have'nt managed to get any further than the test phase. I'm getting =
pestered by my manager now, so I would be very grateful if you could find =
the time to look at these and point me in the right direction or give me =
some clues as to what I should look at.
Please note, in spite of what it may say in the header my real e-mail =
address is [email protected].

Many thanks
Greg



perl Makefile.PL output
-------------------------------------

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /usr/oracle/dist/7.1.6
Oracle sysliblist: -lsocket -lnsl -lm -ldl -laio
Using /usr/oracle/dist/7.1.6/proc/lib/proc.mk (version 1.9)
System: perl5.002 sunos parzival 5.5 generic sun4m sparc =
sunw,sparcstation-5
Compiler: /opt/FSFgcc/bin/gcc -O
Oracle proc.mk would have used these values but we override them:
 CC:       ld -dy $(COMPOBJS) $(LDSTRING)
 LDFLAGS:  -L$(LIBHOME)
 LDSTRING: -R /opt/SUNWcluster/lib -Y =
P,$(LD_LIBRARY_PATH):/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib -Qy -lc =
$(COMPOBJ)/crtn.o
Linking with -locic $(LIBHOME)/osntab.o -lsqlnet -lora -lsqlnet =
$(LIBPLSHACK) -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 $(LLIBCCOREFUND) =
-lcore3 $(LDLIBS) `cat /usr/oracle/dist/7.1.6/rdbms/lib/sysliblist` =
$(MLSLIBS)

Writing Makefile for DBD::Oracle



make output
--------------------

umask 0 && cp Oraperl.pm blib/lib/Oraperl.pm
umask 0 && cp oraperl.ph blib/lib/oraperl.ph
umask 0 && cp Oracle.pm blib/lib/DBD/Oracle.pm
/opt/gnu/bin/gcc -c -I/usr/oracle/dist/7.1.6/rdbms/demo =
-I/opt/LWperl5/lib/perl5/site_perl/sun4-solaris/DBI =
-I/opt/LWperl5/lib/perl5/sun4-solaris/DBI -O -DVERSION=3D\"0.32\" =
-DXS_VERSION=3D\"0.32\" -fpic -I/opt/LWperl5/lib/perl5/sun4-solaris/CORE  =
Oracle.c
/opt/gnu/bin/gcc -c -I/usr/oracle/dist/7.1.6/rdbms/demo =
-I/opt/LWperl5/lib/perl5/site_perl/sun4-solaris/DBI =
-I/opt/LWperl5/lib/perl5/sun4-solaris/DBI -O -DVERSION=3D\"0.32\" =
-DXS_VERSION=3D\"0.32\" -fpic -I/opt/LWperl5/lib/perl5/sun4-solaris/CORE  =
dbdimp.c
Running Mkbootstrap for DBD::Oracle ()
chmod 644 Oracle.bs
LD_RUN_PATH=3D"/lib" /opt/gnu/bin/gcc -o blib/arch/auto/DBD/Oracle/Oracle.=
so -G Oracle.o  dbdimp.o -L/usr/oracle/dist/7.1.6/lib =
/usr/oracle/dist/7.1.6/lib/__fstd.o -locic =
/usr/oracle/dist/7.1.6/lib/osntab.o -lsqlnet -lora -lsqlnet  -lora =
-lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3  -lcore3  `cat /usr/oracle/dist/7.1.6/r=
dbms/lib/sysliblist`   -lsocket -lnsl -lm -ldl -laio
chmod 755 blib/arch/auto/DBD/Oracle/Oracle.so
cp Oracle.bs ./blib/arch/auto/DBD/Oracle/Oracle.bs
chmod 644 blib/arch/auto/DBD/Oracle/Oracle.bs
Manifying ./blib/man3/DBD::Oraperl.3



make test - stdout
-----------------------------

PERL_DL_NONLAZY=3D1 /usr/opt/LWperl5/bin/perl -I./blib/arch -I./blib/lib =
-I/opt/LWperl5/lib/perl5/sun4-solaris -I/opt/LWperl5/lib/perl5 -e 'use =
Test::Harness qw(&runtests $verbose); $verbose=3D0; runtests @ARGV;' t/*.t=

t/base..............FAILED tests 4-5
       Failed 2/5 tests, 60.00% okay
       Test returned status 2 (wstat 512)
*** Error code 29



make test - stderr
----------------------------

install_driver(Oracle) failed: Can't load =
'./blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle: ld.so.1: =
/usr/opt/LWperl5/bin/perl: fatal: relocation error: symbol not found: =
sslsig: referenced in ./blib/arch/auto/DBD/Oracle/Oracle.so at =
/opt/LWperl5/lib/perl5/DynaLoader.pm line 140.

at blib/lib/DBD/Oracle.pm line 24
       DBI::install_driver called at t/base.t line 16
Failed 1 test script, 0.00% okay. 2/5 subtests failed, 60.00% okay.
make: Fatal error: Command failed for target `test_dynamic'



make test TEST_VERBOSE=3D1, stdout
------------------------------------------------------------

PERL_DL_NONLAZY=3D1 /usr/opt/LWperl5/bin/perl -I./blib/arch -I./blib/lib =
-I/opt/LWperl5/lib/perl5/sun4-solaris -I/opt/LWperl5/lib/perl5 -e 'use =
Test::Harness qw(&runtests $verbose); $verbose=3D1; runtests @ARGV;' =
t/*.t
t/base..............1..5
ok 1
ok 2
ok 3
FAILED tests 4-5
       Failed 2/5 tests, 60.00% okay
       Test returned status 2 (wstat 512)
*** Error code 29



make test TEST_VERBOSE=3D1, stderr
-----------------------------------------------------------

install_driver(Oracle) failed: Can't load =
'./blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle: ld.so.1: =
/usr/opt/LWperl5/bin/perl: fatal: relocation error: symbol not found: =
sslsig: referenced in ./blib/arch/auto/DBD/Oracle/Oracle.so at =
/opt/LWperl5/lib/perl5/DynaLoader.pm line 140.

at blib/lib/DBD/Oracle.pm line 24
       DBI::install_driver called at t/base.t line 16
Failed 1 test script, 0.00% okay. 2/5 subtests failed, 60.00% okay.
make: Fatal error: Command failed for target `test_dynamic'


Oracle and Perl version
--------------------------------------

oracle 7.1.6
perl 5.002



Perl -V
----------

Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=3Dsolaris, osver=3D2.5, archname=3Dsun4-solaris
   uname=3D'sunos parzival 5.5 generic sun4m sparc sunw,sparcstation-5 '
   hint=3Drecommended, useposix=3Dtrue
 Compiler:
   cc=3D'/opt/FSFgcc/bin/gcc', optimize=3D'-O', gccversion=3D2.7.2
   cppflags=3D''
   ccflags =3D''
   stdchar=3D'unsigned char', d_stdstdio=3Ddefine, usevfork=3Dfalse
   voidflags=3D15, castflags=3D0, d_casti32=3Ddefine, d_castneg=3Ddefine
   intsize=3D4, alignbytes=3D8, usemymalloc=3Dy, randbits=3D15
 Linker and Libraries:
   ld=3D'/opt/FSFgcc/bin/gcc', ldflags =3D''
   libpth=3D/lib /usr/lib /usr/ccs/lib
   libs=3D-lsocket -lnsl -ldl -lm -lc -lcrypt
   libc=3D/lib/libc.so, so=3Dso
 Dynamic Linking:
   dlsrc=3Ddl_dlopen.xs, dlext=3Dso, d_dlsymun=3D, ccdlflags=3D' '
   cccdlflags=3D'-fpic', lddlflags=3D'-G'

@INC: /opt/LWperl5/lib/perl5/sun4-solaris /opt/LWperl5/lib/perl5 =
/opt/LWperl5/lib/perl5/site_perl/sun4-solaris =
/opt/LWperl5/lib/perl5/site_perl .



Core Dump
------------------
No core dump is available



Additional Info
-----------------------
I had to manually edit the makefile to change the path for gcc. Orginally =
it was /opt/FSFgcc/bin/gcc, I changed it to /opt/gnu/bin/gcc.

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21843-9@oink>;
         Wed, 19 Jun 1996 02:00:31 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835141691:21884:0; Wed, 19 Jun 96 00:48:11 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa21459; 19 Jun 96 0:47 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa12561; 19 Jun 96 0:46 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA21008 for dbi-users-real; Tue, 18 Jun 1996 16:37:00 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA21000
         for <[email protected]>; Tue, 18 Jun 1996 16:36:46 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id AAA03531; Wed, 19 Jun 1996 00:27:40 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: Building Oraperl for Sos2.5 Ora7.3
To: Charles Thayer <[email protected]>
Date: Wed, 19 Jun 1996 00:27:39 +0100 (BST)
Cc: [email protected], [email protected], [email protected],
   [email protected], [email protected]
In-Reply-To: <[email protected]> from "Charles Thayer" at Jun 18, 96 04:04:27 pm
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 2127


> > >     libc internal error: _rmutex_unlock: rmutex not held.
> >
> > This is not an unusual error.  As far as I know, no one has gotten OraPerl to
> > compile for Solaris 2.5 and Oracle 7.3.2.  We all keep coming back to this
> > same problem with the mutex!
> >
> > Tim Bunce is working on a solution.
>
> I checked some search engines and found that libthread seems to be the
> culprit.  Apparently this is a known problem with perl5.002.  One
> solution is to add "-lthread" to the file ext.libs in the perl
> distribution and recompile.  That made DBD and DBI work for me with my
> Oracle installation, but not all tests are passing (for perl that is
> -- namely magic is failing now.)
>
> Someone alluded to writing fake stubs for the functions in libthread
> as a work-around.  But they pointed to a "SunSolve bug database" which
> I don't have access to :-(
>
> Now my oraperl script runs but complains at the end:
>
>       Bad free() ignored during global destruction.

Hmmm, this sheds a bit more light on the problem. I, as yet, haven't had
time to upgrade my Solaris box to look at the problem, so I'm wondering
if Sun have unbundled threading out of the linker by default. In 2.3
and 2.4, libthread was sorta tacked on the end, although there were stubs
in libc which would stop linkage failure. A brain-damaged solution, since
all thread code failed miserably, with often no reason why, until you looked
closely at your link line.

They may have done the same again, although, since Solaris threads are now
a wrapper to POSIX-compliant threads, I wonder if adding libthread
explicitly is solving the problem.

I'd recommend it. I wouldn't recommend using ldd to see what's failed to
link. It sort of lies. If someone else could confirm or deny, much
appreciated. Failing that, read the man pages on thr_create ( pthread_create )
and see what libraries they mention you need to link with.

> Geez/charles

--
Alligator Descartes     |       Nexus - Multi-User Virtual Reality
Anathematician          |         http://www.hermetica.com/nexus
[email protected]  | Muse of the Day: "Deathchase" - Mervyn Estcourt
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21843-11@oink>;
         Wed, 19 Jun 1996 02:00:39 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835141754:22419:0; Wed, 19 Jun 96 00:49:14 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa22124; 19 Jun 96 0:48 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa12561; 19 Jun 96 0:46 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA21008 for dbi-users-real; Tue, 18 Jun 1996 16:37:00 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA21000
         for <[email protected]>; Tue, 18 Jun 1996 16:36:46 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id AAA03531; Wed, 19 Jun 1996 00:27:40 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: Building Oraperl for Sos2.5 Ora7.3
To: Charles Thayer <[email protected]>
Date: Wed, 19 Jun 1996 00:27:39 +0100 (BST)
Cc: [email protected], [email protected], [email protected],
   [email protected], [email protected]
In-Reply-To: <[email protected]> from "Charles Thayer" at Jun 18, 96 04:04:27 pm
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 2127


> > >     libc internal error: _rmutex_unlock: rmutex not held.
> >
> > This is not an unusual error.  As far as I know, no one has gotten OraPerl to
> > compile for Solaris 2.5 and Oracle 7.3.2.  We all keep coming back to this
> > same problem with the mutex!
> >
> > Tim Bunce is working on a solution.
>
> I checked some search engines and found that libthread seems to be the
> culprit.  Apparently this is a known problem with perl5.002.  One
> solution is to add "-lthread" to the file ext.libs in the perl
> distribution and recompile.  That made DBD and DBI work for me with my
> Oracle installation, but not all tests are passing (for perl that is
> -- namely magic is failing now.)
>
> Someone alluded to writing fake stubs for the functions in libthread
> as a work-around.  But they pointed to a "SunSolve bug database" which
> I don't have access to :-(
>
> Now my oraperl script runs but complains at the end:
>
>       Bad free() ignored during global destruction.

Hmmm, this sheds a bit more light on the problem. I, as yet, haven't had
time to upgrade my Solaris box to look at the problem, so I'm wondering
if Sun have unbundled threading out of the linker by default. In 2.3
and 2.4, libthread was sorta tacked on the end, although there were stubs
in libc which would stop linkage failure. A brain-damaged solution, since
all thread code failed miserably, with often no reason why, until you looked
closely at your link line.

They may have done the same again, although, since Solaris threads are now
a wrapper to POSIX-compliant threads, I wonder if adding libthread
explicitly is solving the problem.

I'd recommend it. I wouldn't recommend using ldd to see what's failed to
link. It sort of lies. If someone else could confirm or deny, much
appreciated. Failing that, read the man pages on thr_create ( pthread_create )
and see what libraries they mention you need to link with.

> Geez/charles

--
Alligator Descartes     |       Nexus - Multi-User Virtual Reality
Anathematician          |         http://www.hermetica.com/nexus
[email protected]  | Muse of the Day: "Deathchase" - Mervyn Estcourt
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21843-15@oink>;
         Wed, 19 Jun 1996 02:00:54 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835145222:16530:1; Wed, 19 Jun 96 01:47:02 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa16125; 19 Jun 96 1:45 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA21267 for dbi-users-real; Tue, 18 Jun 1996 17:40:55 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id RAA21263
         for <[email protected]>; Tue, 18 Jun 1996 17:40:53 -0700
Received: by gw.home.vix.com id RAA08074; Tue, 18 Jun 1996 17:40:47 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ac16236; 18 Jun 96 23:05 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab26322; 18 Jun 96 22:57 +0100
Received: from toad by oink with SMTP (PP) id <19030-0@oink>;
         Tue, 18 Jun 1996 17:33:43 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA04055;
         Tue, 18 Jun 1996 17:33:31 +0000
Date: Tue, 18 Jun 1996 17:33:31 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Building Sos2.5 Ora7.3
Cc: [email protected], [email protected], [email protected], [email protected]
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Charles Thayer <[email protected]>
>
>   I'm having an unusual error which may or may not be new and unique
> to solaris 2.5 on the sparc-ultra.  Any help is greatly appreciated.
> It seems that libc is failing when I run the tests:
>
>     libc internal error: _rmutex_unlock: rmutex not held.
>
> /charles

Looking back in my mail I see these two messages:

---
Date: Fri, 7 Jun 1996 13:49:19 -0700
From: Parva Oskoui <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Problems installing DBD module

I just took out the reference to clntsh...it is much happier now.
My "lib list" looks like so:
         -lsql -lclient -lsqlnet -lepc \
                 -lncr -lsqlnet -lcommon -lgeneric -lnlsrtl3 \
                 -lcore3 -lnlsrtl3 -lcore3 \
                 -lcv6 -lsocket -lnsl -ldl \
   `cat /app/oracle/product/732/rdbms/lib/sysliblist`    -lm -lthread

If this works, please forward to the group?
---

---
Message-Id: <[email protected]>
Date: Tue, 11 Jun 1996 16:55:15 -0400
From: Amy McGovern <[email protected]>
To: [email protected]
Cc: [email protected], [email protected], [email protected]

[email protected]. wrote:
> Inspite of setting the environment variables ORACLE_SID and ORACLE_USERID,
> the same problem persists.
> Connecting
>  to '' (from command line, else uses ORACLE_SID/TWO_TASK)
>  as 'crgs/crgs' (via ORACLE_USERID environment var or default)
> libc internal error: _rmutex_unlock: rmutex not held.
> Abort - core dumped
> make: *** [test_dynamic] Error 134

Uncomment the following line in test.pl:
$dbname = $ARGV[0];  # $ENV{TWO_TASK} || $ENV{ORACLE_SID} || 'crgs';

make it:
$dbname = $ARGV[0] || $ENV{TWO_TASK} || $ENV{ORACLE_SID} || 'crgs';

It will fix it...
-Amy
---

Please try each seperately and let me know how it goes.

Tim.

p.s. I've only received _two_ replies to my Oracle 7.3 survey.
Please help me to help you. If you're using DBD::Oracle on 7.3
and it works please send me your details.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <22793-8@oink>;
         Wed, 19 Jun 1996 03:23:39 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835150426:13995:2; Wed, 19 Jun 96 03:13:46 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa13566; 19 Jun 96 3:12 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id TAA21888 for dbi-users-real; Tue, 18 Jun 1996 19:02:35 -0700
Precedence: junk
Sender: [email protected]
Received: from voro.lbl.gov (voro.lbl.gov [131.243.240.19])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id TAA21884
         for <[email protected]>; Tue, 18 Jun 1996 19:02:33 -0700
Received: from voro.lbl.gov (localhost [127.0.0.1]) by voro.lbl.gov (8.6.12/G)
         with ESMTP id SAA15671; Tue, 18 Jun 1996 18:59:35 -0700
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.2 7/18/95
To: Tim Bunce <[email protected]>
cc: [email protected]
Subject: Re: misc 7.3 issues
In-reply-to: Your message of "Tue, 18 Jun 1996 23:35:53 -0000." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 18 Jun 1996 18:59:35 -0700
From: Mark Dedlow <[email protected]>


> From: Tim Bunce <[email protected]>
> Is that _really_ the smallest?

Size is in the eye of the beholder :-)

#!/usr/bin/perl
use DBI;
$drh = DBI->install_driver('Oracle') || die "$DBI::errstr\n";

But adding:

$dbh = $drh->connect("", "$cs", '') || die "$DBI::errstr\n";

causes the bad free.

Thanks again,

Mark



---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <23686-0@oink>;
         Wed, 19 Jun 1996 07:10:11 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835153842:27638:2; Wed, 19 Jun 96 04:10:42 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa27296; 19 Jun 96 4:09 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id TAA22058 for dbi-users-real; Tue, 18 Jun 1996 19:58:11 -0700
Precedence: junk
Sender: [email protected]
Received: from netmail.austin.ibm.com (netmail.austin.ibm.com [129.35.208.98])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id TAA22054
         for <[email protected]>; Tue, 18 Jun 1996 19:58:10 -0700
From: [email protected]
Received: from dax.austin.ibm.com (dax.austin.ibm.com [129.35.148.59])
         by netmail.austin.ibm.com (8.6.12/8.6.11) with ESMTP id VAA14426
         for <[email protected]>; Tue, 18 Jun 1996 21:58:10 -0500
Received: (from mhm@localhost)
         by dax.austin.ibm.com (AIX 9610R-UP4.1/UCB 8.7/8.7-client1.0)
         id VAA25846; Tue, 18 Jun 1996 21:58:04 -0500 (CDT)
Message-Id: <[email protected]>
To: [email protected]
Subject: DBD::DB2, its awfully quiet...
Date: Tue, 18 Jun 96 21:58:03 -0500


Is anyone actually using it?  It has been out a month now and I have
not heard anything...good or bad.  And, I know it is not perfect as
I have made two sets of updates.

Mike

------------------------------------------------------------------------------
Michael H. Moran                  | Standard Disclaimer:  The content of
[email protected]                | this posting is independent of
Commercial Performance            | official IBM position.
IBM Corporation, Austin, Texas    |
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <23682-5@oink>;
         Wed, 19 Jun 1996 07:11:29 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835154260:01372:1; Wed, 19 Jun 96 04:17:40 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa00770; 19 Jun 96 4:15 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA21150 for dbi-users-real; Tue, 18 Jun 1996 17:02:25 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id RAA21146
         for <[email protected]>; Tue, 18 Jun 1996 17:02:24 -0700
Received: by gw.home.vix.com id RAA05483; Tue, 18 Jun 1996 17:00:50 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id aa25542; 18 Jun 96 23:38 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ae17148; 19 Jun 96 0:08 +0100
Received: from toad by oink with SMTP (PP) id <21214-0@oink>;
         Tue, 18 Jun 1996 23:39:02 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA08599;
         Tue, 18 Jun 1996 23:38:55 +0000
Date: Tue, 18 Jun 1996 23:38:55 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected], [email protected]
Subject: Re: Building Oraperl for Sos2.5 Ora7.3
Cc: [email protected], [email protected]
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Charles Thayer <[email protected]>
>
> In message <[email protected]>, Amy McGovern writes:
> > Charles Thayer wrote:
> > >   I'm having an unusual error which may or may not be new and unique
> > > to solaris 2.5 on the sparc-ultra.  Any help is greatly appreciated.
> > > It seems that libc is failing when I run the tests:
> > >
> > >     libc internal error: _rmutex_unlock: rmutex not held.
> >
> > This is not an unusual error.  As far as I know, no one has gotten Oraperl
> > to compile for Solaris 2.5 and Oracle 7.3.2.  We all keep coming back to
> > this same problem with the mutex!
>
> I checked some search engines and found that libthread seems to be the
> culprit.  Apparently this is a known problem with perl5.002.

Interesting.

> One solution is to add "-lthread" to the file ext.libs in the perl
> distribution and recompile.  That made DBD and DBI work for me with my
> Oracle installation, but not all tests are passing (for perl that is
> -- namely magic is failing now.)

Not so useful then!

> Someone alluded to writing fake stubs for the functions in libthread
> as a work-around.  But they pointed to a "SunSolve bug database" which
> I don't have access to :-(

Can you reply with the messages your search found?
I've CC'd this to perl5-porters incase someone there can shed any light.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <23682-7@oink>;
         Wed, 19 Jun 1996 07:11:48 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835154390:01910:0; Wed, 19 Jun 96 04:19:50 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa01687; 19 Jun 96 4:18 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa01980; 19 Jun 96 4:17 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id UAA22080 for dbi-users-real; Tue, 18 Jun 1996 20:00:02 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id TAA22070
         for <[email protected]>; Tue, 18 Jun 1996 19:59:55 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id aa16607; 19 Jun 96 3:59 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa15836; 19 Jun 96 3:41 +0100
Received: from toad by oink with SMTP (PP) id <23093-0@oink>;
         Wed, 19 Jun 1996 03:39:55 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA10387;
         Wed, 19 Jun 1996 03:39:34 +0000
Date: Wed, 19 Jun 1996 03:39:34 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: misc 7.3 issues
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 585
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Mark Dedlow <[email protected]>
>
> #!/usr/bin/perl
> use DBI;
> $drh = DBI->install_driver('Oracle') || die "$DBI::errstr\n";
>
> But adding:
>
> $dbh = $drh->connect("", "$cs", '') || die "$DBI::errstr\n";
>
> causes the bad free.

Does it 'bad free' only if the connect works, or fails, or either?

(And then try it using the 0.33 I've just released. See the notes
in the README.) I'm almost certainly going to need a stack trace
to track this down (as per previous mail). At the end of the day
I suspect it'll be an Oracle library bug. Sigh.

Tim [off home to bed].
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <23682-9@oink>;
         Wed, 19 Jun 1996 07:12:06 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835154937:04144:1; Wed, 19 Jun 96 04:28:57 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa03979; 19 Jun 96 4:28 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA21131 for dbi-users-real; Tue, 18 Jun 1996 17:00:54 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id RAA21124
         for <[email protected]>; Tue, 18 Jun 1996 17:00:52 -0700
Received: by gw.home.vix.com id RAA05482; Tue, 18 Jun 1996 17:00:49 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ab24469; 18 Jun 96 23:28 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ad17148; 19 Jun 96 0:08 +0100
Received: from toad by oink with SMTP (PP) id <21206-0@oink>;
         Tue, 18 Jun 1996 23:35:59 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA08590;
         Tue, 18 Jun 1996 23:35:53 +0000
Date: Tue, 18 Jun 1996 23:35:53 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: misc 7.3 issues
Cc: [email protected]
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Mark Dedlow <[email protected]>
>
> >From: Tim Bunce <[email protected]>
> >
> > > From: Mark Dedlow <[email protected]>
> > >
> > > With Oracle 7.3.2, Solaris 2.4, Perl 5.002, DBD 0.32, DBI 0.70
> > > I now get:
> > >
> > >    Bad free() ignored during global destruction.
> > >
> > > On exit from scripts.   Otherwise clean build "out of the box".
> > > Passes all tests, and everything else I've tried works.
> >
> > "Now" being since when? What changed?
>
> Sorry, sometimes I forget you're not a mind reader.  Or maybe
> you are, but there's nothing on this end to read.

:-)

> "Since" meant since install of 0.32/0.70 on 7.3.
> I jumped to there from 0.29/0.68 on 7.2.3.

:-/

> I thought I'd wait out the 7.3 build problems, though it turned
> out I didn't have any.

Other than a Bad free.

> > What's the smallest script which has the problem?
>
> #!/usr/bin/perl
> use DBI;
> $cs = 'user/pass@db';
> $drh = DBI->install_driver('Oracle') || die "$DBI::errstr\n";
> $dbh = $drh->connect("", "$cs", '') || die "$DBI::errstr\n";
> $stmt = 'select sysdate from dual';
> $cursor = $dbh->prepare($stmt) || die "$DBI::errstr\n";
> $cursor->execute || die "$DBI::errstr\n";
> while (@row = $cursor->fetchrow) { print "@row\n" }
> $cursor->finish;
> $dbh->disconnect;
> exit(0);

Is that _really_ the smallest?
Do you _have_ to fetch?
Do you _have_ to execute?
Do you _have_ to prepare?
etc etc.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <23750-0@oink>;
         Wed, 19 Jun 1996 07:14:18 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835164721:21926:1; Wed, 19 Jun 96 07:12:01 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa20927; 19 Jun 96 7:11 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id XAA22534 for dbi-users-real; Tue, 18 Jun 1996 23:06:23 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id XAA22530
         for <[email protected]>; Tue, 18 Jun 1996 23:06:19 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id af06058; 19 Jun 96 6:06 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa15953; 19 Jun 96 3:42 +0100
Received: from toad by oink with SMTP (PP) id <23124-0@oink>;
         Wed, 19 Jun 1996 03:42:29 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA10394;
         Wed, 19 Jun 1996 03:42:14 +0000
Date: Wed, 19 Jun 1996 03:42:14 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: problem with dynamic library loading on Solaris 2.5
X-Sun-Charset: US-ASCII
content-length: 1032
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Jon Adams <[email protected]>
>
> >In article <[email protected]>,
> >Jon Adams  <[email protected]> wrote:
>
> >Try manually setting the LD_LIBRARY_PATH environment variable to the
> >full path of the directory holding the libclntsh.so.1.0 file.  See the
> >ld unix man page for more info. I'm assuming your perl isn't set-uid.
>
> Okay, I've been working on this long enough that I had remembered trying this
> and it not working, but I didn't remember _how_ it was not working. One of
> the problems was that I forgot I had to export the variable in ksh. The other
> problem was that, after setting LD_LIBRARY_PATH, it just dumps core.
> Assuming the library problem is resolved by setting LD_LIBRARY_PATH, and this
> is a new problem, I'll give the new details...
>
> Again, any input would be appreciated.

Try using DBD::Oracle 0.33 (just released). If it core dumps then rebuild
the Makefile using   perl Makefile.PL -g   rerun it and get a stack trace
from the new core file.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <25951-1@oink>;
         Wed, 19 Jun 1996 17:03:14 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835175778:11303:2; Wed, 19 Jun 96 10:16:18 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa10891; 19 Jun 96 10:15 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA22860 for dbi-users-real; Wed, 19 Jun 1996 02:07:01 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id CAA22850;
         Wed, 19 Jun 1996 02:06:58 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ai24005; 19 Jun 96 10:06 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa09475; 19 Jun 96 2:19 +0100
Received: from toad by oink with SMTP (PP) id <22038-0@oink>;
         Wed, 19 Jun 1996 02:19:05 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA10072;
         Wed, 19 Jun 1996 02:18:58 +0000
Date: Wed, 19 Jun 1996 02:18:58 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Announce: DBD::Oracle 0.33
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1070
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


NOTE: This is primarily a release aimed a fixing *some* of the
Oracle 7.3 problems of late. Do not upgrade to this version unless
you have an existing build or core dump problem with DBD::Oracle.

Changes in DBD::Oracle 0.33 (Oraperl 1.27), 19th June 1996

   Added Makefile.PL -g option to enable debugging.
   Added Makefile.PL -s symbol_name option to search for symbols.
   Added README notes about -g option, core files and stack traces.
   Reorganised the way Makefile.PL uses MakeMaker liblist code.
   Oraperl defaults to the 'safe' (normal) mode of using the DBI.
   Oraperl uses sigtrap on SEGV & BUS to give a perl stack trace.
   Small change to $dbname/$user/$passwd logic in connect.

This release should fix the libclntsh.so not found problem.
It _may_ also fix others (but more by luck than judgement).

It's been uploaded (63875 bytes) to

 ftp://ftp.demon.co.uk/pub/perl/db/DBD/Oracle/DBD-Oracle-0.33.tar.Z

and should find it's way to CPAN via PAUSE over the next day or so.

Tim [who really should have been doing "real work" this evening].
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <25952-4@oink>;
         Wed, 19 Jun 1996 17:03:22 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835175771:26141:1; Wed, 19 Jun 96 10:16:11 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa25820; 19 Jun 96 10:15 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA22871 for dbi-users-real; Wed, 19 Jun 1996 02:09:16 -0700
Precedence: junk
Sender: [email protected]
Received: from bsfiles.nerc-bas.ac.uk (bsfiles.nerc-bas.ac.uk [192.171.137.25])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id CAA22865
         for <[email protected]>; Wed, 19 Jun 1996 02:09:04 -0700
Received: from bssmlsa.nerc-bas.ac.uk.nerc-bas.ac.uk
         by bsfiles.nerc-bas.ac.uk (8.6.10/NERC-1.2(Solaris 2.x) id KAA24965;
         Wed, 19 Jun 1996 10:08:50 +0100
Date: Wed, 19 Jun 1996 10:08:47 +0100
Message-Id: <[email protected]>
From: Andrew G Wood <[email protected]>
To: [email protected]
Subject: DBD-Oracle-0.33 ... one small problem.
Reply-to: [email protected]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII


So Tim has pushed out yet another version of the Oracle DBD .... so time
to give it a try.

Got one small problem with the new?? code in Oraperl.pm ....

> $SIG{BUS} = $SIG{SEGV} = sub {
>    print STDERR "Add BEGIN { \$Oraperl::safe=1 } above 'use Oraperl'.\n"
>            unless $safe;
>    goto &sigtrap::trap;
> };

which you can see (if you have good eyes) in the make test.

No such signal: SIGBUS at blib/lib/Oraperl.pm line 50. <++++++++++++++========

This message appears in all scripts (that I have tested so far).

Andy.

+--------------------------------------------+
|   Dr Andy Wood, Database Administrator     |
|   British  Antarctic  Survey               |
|   High Cross, Madingley Road            +----------------------------------+
|   Cambridge,   CB3 0ET,   UK            |    [email protected]              |
|      +44 (0) 1223 361188                |    [email protected]       |
+-----------------------------------------|    [email protected]           |
                                         +----------------------------------+

(BTW ... Oracle 7.1.4)

-------------------------------------% perl -V
Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=irix, osver=5, archname=IP20-irix
   uname='irix bssmlsa 5.3 11091811 ip20 mips '
   hint=previous, useposix=true
 Compiler:
   cc='cc', optimize='-O', gccversion=
   cppflags='-D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -DLANGUAGE_C'
   ccflags ='-D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -Olimit 3000 -DLANGUAGE_C'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=1, d_casti32=define, d_castneg=
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='cc', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /usr/lib /lib
   libs=-lgdbm -lsun -lm -lc -lcrypt -lbsd -lPW
   libc=/usr/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags=' ', lddlflags='-shared -L/usr/local/lib'

@INC: /usr/local/lib/perl5/IP20-irix/5.002 /usr/local/lib/perl5
/usr/local/lib/perl5/site_perl/IP20-irix /usr/local/lib/perl5/site_perl .

-------------------------------------% perl Makefile.PL

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /nerc/packages/oracle/product/7.1.4
Using /nerc/packages/oracle/product/7.1.4/proc/lib/proc.mk (version 1.17)
System: perl5.002 irix bssmlsa 5.3 11091811 ip20 mips
Compiler: cc -O -D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -Olimit 3000 -DLANGUAGE_C
Oracle proc.mk would have used these values but we override them:
 CC:       cc
 CFLAGS:   -I. -O `if [ -d /usr/lib/cmplrs ] ; then echo "-systype svr4 -KPIC" ; fi `
 LDFLAGS:  -L$(LIBHOME) `if [ -x /usr/necbin/uversion ] ; then echo "-Zcommon_pic -znodefs" ;                       elif [ -d /usr/lib/cmplrs ] ; then echo "-systype svr4" ; fi `
Linking with $(LIBHOME)/osntab.o -locic  -lsqlnet -lora -lsqlnet -lora -lnlsrtl -lcv6 -lcore -lnlsrtl -lcv6 -lcore -lc -lsocket -lnsl -lm

Checking if your kit is complete...
Looks good

Warning: By default new modules are installed into your 'site_lib'
directories. Since site_lib directories come after the normal library
directories you must delete any old DBD::Oracle files and directories from
your 'privlib' and 'archlib' directories and their auto subdirectories.

Writing Makefile for DBD::Oracle

-------------------------------------% make

       umask 0 && cp Oraperl.pm ./blib/lib/Oraperl.pm
       umask 0 && cp oraperl.ph ./blib/lib/oraperl.ph
       umask 0 && cp Oracle.pm ./blib/lib/DBD/Oracle.pm
       /bin/perl -I/usr/local/lib/perl5/IP20-irix/5.002 -I/usr/local/lib/perl5 /usr/local/lib/perl5/ExtUtils/xsubpp  -typemap /usr/local/lib/perl5/ExtUtils/typemap Oracle.xs >Oracle.tc && mv Oracle.tc Oracle.c
       cc -c -I/nerc/packages/oracle/product/7.1.4/rdbms/demo -I/usr/local/lib/perl5/site_perl/IP20-irix/DBI -I/usr/local/lib/perl5/IP20-irix/5.002/DBI -D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -Olimit 3000 -DLANGUAGE_C -O -DVERSION=\"0.33\" -DXS_VERSION=\"0.33\"  -I/usr/local/lib/perl5/IP20-irix/5.002/CORE  Oracle.c
       cc -c -I/nerc/packages/oracle/product/7.1.4/rdbms/demo -I/usr/local/lib/perl5/site_perl/IP20-irix/DBI -I/usr/local/lib/perl5/IP20-irix/5.002/DBI -D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -Olimit 3000 -DLANGUAGE_C -O -DVERSION=\"0.33\" -DXS_VERSION=\"0.33\"  -I/usr/local/lib/perl5/IP20-irix/5.002/CORE  dbdimp.c
Running Mkbootstrap for DBD::Oracle ()
       chmod 644 Oracle.bs
       LD_RUN_PATH="/nerc/packages/oracle/product/7.1.4/lib:/usr/lib" cc -o ./blib/arch/auto/DBD/Oracle/Oracle.so -shared -L/usr/local/lib Oracle.o  dbdimp.o /nerc/packages/oracle/product/7.1.4/lib/osntab.o   -L/nerc/packages/oracle/product/7.1.4/lib -locic -lsqlnet -lora -lsqlnet -lora -lnlsrtl -lcv6 -lcore -lnlsrtl -lcv6 -lcore -lc -lsocket -lnsl -lm
ld:
The shared object /usr/lib/libsocket.so did not resolve any symbols.
       You may want to remove it from your link line.
       chmod 755 ./blib/arch/auto/DBD/Oracle/Oracle.so
       cp Oracle.bs ./blib/arch/auto/DBD/Oracle/Oracle.bs
       chmod 644 ./blib/arch/auto/DBD/Oracle/Oracle.bs
Manifying ./blib/man3/DBD::Oraperl.3

-------------------------------------% make test

       PERL_DL_NONLAZY=1 /bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5/IP20-irix/5.002 -I/usr/local/lib/perl5 -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t
t/base..............ok
All tests successful.
Files=1,  Tests=5,  2 secs ( 0.64 cusr  0.50 csys =  1.14 cpu)
       PERL_DL_NONLAZY=1 /bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5/IP20-irix/5.002 -I/usr/local/lib/perl5 test.pl
Oraperl test application $Revision: 1.21 $
No such signal: SIGBUS at blib/lib/Oraperl.pm line 50. <++++++++++++++========

Oraperl Emulation Interface version 1.27
Oracle Driver 0.33
DBI-0.70 Switch by Tim Bunce, version 0.70


Connecting
to '' (from command line, else uses ORACLE_SID/TWO_TASK)
as 'xxx/xxx' (via ORACLE_USERID environment var or default)
Fields:  4
Names:   'NUM_T',       'DATE_T',       'CHAR_T',       'NULL_T'
Lengths: '40',  '75',   '30',   '0'
Types:   '2',   '12',   '1',    '1'
Data rows:
   fetch: 79.2 19-JUN-96 XXX NULL

csr reassigned (forces destruction)...
Fetch list of tables:
Fetched: DUAL
Fetched: SYSTEM_PRIVILEGE_MAP
Fetched: TABLE_PRIVILEGE_MAP
Fetched: STMT_AUDIT_OPTION_MAP
Test ora_do with harmless non-select statement (set transaction read only)
Expect error message:
ora_do: 1453: ORA-01453: SET TRANSACTION must be first statement of transaction (DBD: oexec error)
csr out of scope...
ora_logoff...
lda out of scope...

Testing repetitive connect/open/close/disconnect:
Expect sequence of digits, no other messages:
1 2 3 4 5

Test interaction of explicit close/logoff and implicit DESTROYs
Expect just 'done.', no other messages:
done.

Test complete (3 seconds).

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <25952-15@oink>;
         Wed, 19 Jun 1996 17:04:41 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835185019:22275:2; Wed, 19 Jun 96 12:50:19 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa21542; 19 Jun 96 12:49 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA23049 for dbi-users-real; Wed, 19 Jun 1996 04:42:39 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id EAA23045
         for <[email protected]>; Wed, 19 Jun 1996 04:42:36 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id MAA06823 for [email protected]; Wed, 19 Jun 1996 12:35:01 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: WWW page updates
To: [email protected]
Date: Wed, 19 Jun 1996 12:35:00 +0100 (BST)
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 329


I've finally got round to updating the mailing list archives. Updates to
the WWW pages should happen later today also.

Regards.

--
Alligator Descartes     |       Nexus - Multi-User Virtual Reality
Anathematician          |         http://www.hermetica.com/nexus
[email protected]  | Muse of the Day: "Deathchase" - Mervyn Estcourt
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <28659-3@oink>;
         Thu, 20 Jun 1996 07:07:16 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835216974:21814:2; Wed, 19 Jun 96 21:42:54 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa21274; 19 Jun 96 21:42 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA23587 for dbi-users-real; Wed, 19 Jun 1996 09:38:47 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA23583
         for <[email protected]>; Wed, 19 Jun 1996 09:38:45 -0700
Received: by gw.home.vix.com id JAA06595; Wed, 19 Jun 1996 09:38:41 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ag23281; 19 Jun 96 17:17 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa27429; 19 Jun 96 17:12 +0100
Received: from toad by oink with SMTP (PP) id <23223-0@oink>;
         Wed, 19 Jun 1996 03:52:07 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA10561;
         Wed, 19 Jun 1996 03:51:58 +0000
Date: Wed, 19 Jun 1996 03:51:58 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: test problems with Oracle D
Cc: [email protected]
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Greg Giles <[email protected]>
>
> Hi All,
> As you can gather from the title I'm having a few problems with the Oracle 0.32 driver. I've enclosed below the output of the build processes as requested in the README. I've been trying for the last few days now, but I have'nt managed to get any further than the test phase. I'm getting pestered by my manager now, so I would be very grateful if you could find the time to look at these and point me in the right direction or give me some clues as to what I should look at.
> Please note, in spite of what it may say in the header my real e-mail address is [email protected].

Long lines don't work too well for text.

> LD_RUN_PATH="/lib" /opt/gnu/bin/gcc -o blib/arch/auto/DBD/Oracle/Oracle.so -G Oracle.o  dbdimp.o -L/usr/oracle/dist/7.1.6/lib /usr/oracle/dist/7.1.6/lib/__fstd.o -locic /usr/oracle/dist/7.1.6/lib/osntab.o -lsqlnet -lora -lsqlnet  -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3  -lcore3  `cat /usr/oracle/dist/7.1.6/rdbms/lib/sysliblist`   -lsocket -lnsl -lm -ldl -laio

> install_driver(Oracle) failed: Can't load './blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle: ld.so.1: /usr/opt/LWperl5/bin/perl: fatal: relocation error: symbol not found: sslsig: referenced in ./blib/arch/auto/DBD/Oracle/Oracle.so at /opt/LWperl5/lib/perl5/DynaLoader.pm line 140.

Fetch DBD::Oracle 0.33 and do

       perl Makefile.PL -s sslsig

you should see something like:

Using Oracle in /home/oracle/product/7.1.4
Using /home/oracle/product/7.1.4/proc/lib/proc.mk (version 1.17)
Searching for symbol 'sslsig' in /home/oracle/product/7.1.4/lib ...
 searching libasync.a ...
[31]    |         0|       0|NOTY |GLOB |0    |UNDEF  |sslsig
 searching libc3v6.a ...
 searching libcore.a ...
[14]    |         0|       0|NOTY |GLOB |0    |UNDEF  |sslsig
Symbols from libcore.a[sslsig.o]:
[1]     |         0|       0|FILE |LOCL |0    |ABS    |sslsig.c
[7]     |        16|      24|FUNC |GLOB |0    |3      |sslsig
 searching libcore3.a ...
[14]    |         0|       0|NOTY |GLOB |0    |UNDEF  |sslsig
Symbols from libcore3.a[sslsig.o]:
[1]     |         0|       0|FILE |LOCL |0    |ABS    |sslsig.c
[7]     |        16|      24|FUNC |GLOB |0    |3      |sslsig
 searching libcv6.a ...
 searching libdruid.a ...

Which shows that it's defined in libcore.a and libcore3.a. Since you
link to core3 right at the end of the oracle library list I can't see
what's wrong.

Are you using sun's ld or gnu's?

Does a 'make perl' (in the DBD::Oracle directory) produce a working perl
with DBD::Oracle built in?

Anyway, see if 0.33 works better for you.

Tim.

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <28658-6@oink>;
         Thu, 20 Jun 1996 07:07:32 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835210921:00927:1; Wed, 19 Jun 96 20:02:01 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa28597; 19 Jun 96 20:00 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA23837 for dbi-users-real; Wed, 19 Jun 1996 11:53:26 -0700
Precedence: junk
Sender: [email protected]
Received: from dfw-ix6.ix.netcom.com ([206.214.98.6])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id LAA23833
         for <[email protected]>; Wed, 19 Jun 1996 11:53:24 -0700
Received: from curly.cde.state.co.us ([204.131.28.18])
         by dfw-ix6.ix.netcom.com (8.6.13/8.6.12) with SMTP id KAA29656;
         Wed, 19 Jun 1996 10:09:39 -0700
Message-ID: <[email protected]>
Date: Wed, 19 Jun 1996 11:15:48 -0600
From: Bill Rodgers <[email protected]>
X-Mailer: Mozilla 2.01Gold (Win95; I)
MIME-Version: 1.0
To: [email protected]
CC: [email protected]
Subject: HPUX 10.01--Oracle 7.3.2--ProC 2.2.2.0.0 Problems
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Am attempting to build DBD::Oracle for subject platform.  Found and
corrected following in Makefile.PL:
 1. proc.mk has moved, new dir is: $OH/precomp/demo/proc/proc.mk
Found but not figured out in Makefile.PL:
 1. How to fix problem in fetch_oci_macros() in Makefile.PL: Line 260,
$MK{mkver} never assigned a value.  Leads me to believe that the file
proc.mk has changed.  My knowledge of perl is not extensive enough to
get me much further soon.  Will continue to hack though.  Suspect expert
in perl and proc.mk could fix this quick.  If copyrights not a problem,
would be glad to send you copies of proc.mk and env_precomp.mk (included
in proc.mk).
 2. Can not find file libocic.a in $OH/lib or anywhere else in my
installation.  Currently have version 2.2.2.0.0 of Pro*C installed.
Using newly compiled and tested latest version of Perl (5.002).
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <01179-2@oink>;
         Thu, 20 Jun 1996 11:43:08 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835258628:27966:4; Thu, 20 Jun 96 09:17:08 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa27757; 20 Jun 96 9:16 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id BAA25386 for dbi-users-real; Thu, 20 Jun 1996 01:01:29 -0700
Precedence: junk
Sender: [email protected]
Received: from voro.lbl.gov (voro.lbl.gov [131.243.240.19])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id BAA25382
         for <[email protected]>; Thu, 20 Jun 1996 01:01:28 -0700
Received: from voro.lbl.gov (localhost [127.0.0.1]) by voro.lbl.gov (8.6.12/G)
         with ESMTP id AAA22214; Thu, 20 Jun 1996 00:58:48 -0700
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.2 7/18/95
To: Tim Bunce <[email protected]>
cc: [email protected]
Subject: Re: misc 7.3 issues
In-reply-to: Your message of "Wed, 19 Jun 1996 03:39:34 -0000." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Thu, 20 Jun 1996 00:58:47 -0700
From: Mark Dedlow <[email protected]>


>
> > From: Mark Dedlow <[email protected]>
> >
> > #!/usr/bin/perl
> > use DBI;
> > $drh = DBI->install_driver('Oracle') || die "$DBI::errstr\n";
> >
> > But adding:
> >
> > $dbh = $drh->connect("", "$cs", '') || die "$DBI::errstr\n";
> >
> > causes the bad free.
>
> Does it 'bad free' only if the connect works, or fails, or either?

Either.

I remade perl with debugging, and re-ran my test script, and it
now gives:

Duplicate free() ignored during global destruction.

> (And then try it using the 0.33 I've just released. See the notes
> in the README.)

OK, next...

> I'm almost certainly going to need a stack trace

% dbx /usr/bin/perl core
Reading symbolic information for /usr/bin/perl
core file header read successfully
Reading symbolic information for rtld /usr/lib/ld.so.1
Reading symbolic information for /usr/lib/libsocket.so.1
Reading symbolic information for /usr/lib/libnsl.so.1
Reading symbolic information for /usr/lib/libdl.so.1
Reading symbolic information for /usr/lib/libm.so.1
Reading symbolic information for /usr/lib/libc.so.1
Reading symbolic information for /usr/lib/libintl.so.1
Reading symbolic information for /usr/lib/libw.so.1
Reading symbolic information for /usr/local/lib/perl5/site_perl/sun4-solaris/au
to/DBI/DBI.so
Reading symbolic information for /usr/local/lib/perl5/site_perl/sun4-solaris/au
to/DBD/Oracle/Oracle.so
Reading symbolic information for /opt/oracle/7.3/lib/libclntsh.so.1.0
Reading symbolic information for /lib/libthread.so.1
Reading symbolic information for /usr/lib/nss_files.so.1
Reading symbolic information for /usr/lib/nss_dns.so.1
Reading symbolic information for /usr/lib/libresolv.so.1
detected a multithreaded program
(l@1) terminated by signal ABRT (abort)
(dbx) where

=>[1] epccol_close(0x18941c, 0x189408, 0xb8880, 0x0, 0xdf60b771, 0xdf605a68),
at 0xdf4d2c80
 [2] epc_exit_handler(0x0, 0x0, 0x0, 0xba3f8, 0xc41c4, 0xba3bc), at 0xdf4d7a08
 [3] _exithandle(0x0, 0x1bd58, 0x3, 0xdffff6dc, 0xdf680c30, 0x0), at
0xdf613a48
 [4] exit(0x3, 0xdffff6dc, 0xdffff6ec, 0xb8000, 0x0, 0x0), at 0xdf661d40


Mark

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <02637-2@oink>;
         Thu, 20 Jun 1996 17:01:00 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835270786:00172:2; Thu, 20 Jun 96 12:39:46 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa29345; 20 Jun 96 12:38 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA26764 for dbi-users-real; Thu, 20 Jun 1996 04:30:38 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id EAA26760
         for <[email protected]>; Thu, 20 Jun 1996 04:30:37 -0700
Received: by gw.home.vix.com id EAA14536; Thu, 20 Jun 1996 04:30:33 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from lion.amtrak.co.uk (root@[192.9.201.1])
         by shark.amtrak.co.uk (8.7.3/8.7.3) with SMTP id MAA06607
         for <[email protected]>; Thu, 20 Jun 1996 12:32:05 +0100 (BST)
Received: from redwood by lion.amtrak.co.uk (5.x/SMI-SVR4) id AA03302;
         Thu, 20 Jun 1996 12:30:52 +0100
To: Dbi_Users:;
From: Simon Windsor <[email protected]>
Cc: [email protected]
Subject: Oracle Hints
X-Originating-Host: redwood
Message-Id: <1996Jun20.123723+0000@redwood>
Date: 20 Jun 1996 12:37:23 +0000
Mime-Version: 1.0
X-Mailer: Hummingbird EMail for Windows Version 5.0

Hello

Does the DBD-Oracle/Perl interface handle hints. Within Oracle it is
possible to use the following syntax

select /*+ index(<table> <index>) */
   <cols>
 from <table>
 where <pk col> = <val>

to force the RDBMS to perform an indexed table scan, rather than a
full table scan. Within Sqlplus, this feature allows one of my queries
to return within 3s, but without it the query takes over 8s.

However, when I use this feature within dbperl (Per 5.002, DBI-0.69,
DBD-Oracle-0.32) I see no appreciable benefit.

Simon
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <02637-7@oink>;
         Thu, 20 Jun 1996 17:01:18 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835276426:23729:3; Thu, 20 Jun 96 14:13:46 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa23216; 20 Jun 96 14:13 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA27869 for dbi-users-real; Thu, 20 Jun 1996 06:06:14 -0700
Precedence: junk
Sender: [email protected]
Received: from inet-smtp-gw-1.us.oracle.com (inet-smtp-gw-1.us.oracle.com [192.86.155.81])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id GAA27865
         for <[email protected]>; Thu, 20 Jun 1996 06:06:13 -0700
Received: from behux0.be.oracle.com by inet-smtp-gw-1.us.oracle.com
         with ESMTP (8.6.12/37.7) id GAA30066; Thu, 20 Jun 1996 06:06:11 -0700
Message-Id: <[email protected]>
Received: from b-w001.be.oracle.com by behux0.be.oracle.com
         with SMTP (1.39.111.2/16.2) id AA135295943;
         Thu, 20 Jun 1996 15:05:43 +0200
From: Philippe Vanhaesendonck <[email protected]>
Subject: Re: Oracle Hints
To: [email protected]
Date: Thu, 20 Jun 1996 15:06:09 CES
X-Mailer: Elm [revision: 111.1]

> Does the DBD-Oracle/Perl interface handle hints. Within Oracle it is
> possible to use the following syntax
>
> select /*+ index(<table> <index>) */
>     <cols>
>   from <table>
>   where <pk col> = <val>

Hints are a server feature, not a client spec, so i do not see why it does not
work... (unless DBI/DBD strips comments...)

Note that oracle will ignore the hint if there is a syntax error.

To convince yourseld that the hint is passed to the server, just set sql_trace
to true and look at the trace file.

Phil.

The statements and opinions expressed here are my own and do not necessarily
represent those of Oracle Corporation.

+--------------------------------------+
| Philippe Vanhaesendonck              |
| Senior Technical Support Specialist+-------------------------------------+
| Oracle Belgium - Vuurberg 80       | Internet     [email protected] |
| B-1831 Diegem   (Belgium)          | Compuserve   75611,1504             |
+------------------------------------| Voice        +32-2-719.59.59        |
                                    | Fax          +32-2-719.13.14        |
                                    +-------------------------------------+
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <04847-6@oink>;
         Fri, 21 Jun 1996 07:06:43 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835298726:13834:1; Thu, 20 Jun 96 20:25:26 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa13009; 20 Jun 96 20:24 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA28950 for dbi-users-real; Thu, 20 Jun 1996 12:19:36 -0700
Precedence: junk
Sender: [email protected]
Received: from gate.overcom.com (overcom.com [204.157.28.161])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id MAA28946
         for <[email protected]>; Thu, 20 Jun 1996 12:19:33 -0700
Received: from ray.overcom.com (ray.overcom.com [204.157.28.179])
         by gate.overcom.com (8.6.12/8.6.12) with SMTP id PAA03725
         for <[email protected]>; Thu, 20 Jun 1996 15:19:22 -0400
Message-ID: <[email protected]>
Date: Thu, 20 Jun 1996 15:19:21 -0400
From: Ray Cromwell <[email protected]>
Reply-To: [email protected]
Organization: OverCom
X-Mailer: Mozilla 3.0b4 (Win95; I)
MIME-Version: 1.0
To: [email protected]
Subject: Problem with Oracle DBD 0.32 on Solaris 2.4
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi all,
 Does anyone have a suggestion for this problem? I recently installed
DBI/OracleDBD on a Solaris 2.4 machine. The DBI works fine, but
the DBD fails the test with "rmutex_unlock: mutex not held"
(core dump) I ran it in the exact same environment as I run SQLPLUS,
FYI. I searched the net and saw what looked like a solution --
someone said it was a solaris thread problem and if you just
rebuilt perl5 with libthread linked in, it would fix it, but
it didn't. I believe I'm running Oracle 7.3.2.

Thanks,
-Ray
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <06317-1@oink>;
         Fri, 21 Jun 1996 09:20:33 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835343010:29767:2; Fri, 21 Jun 96 08:43:30 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa29082; 21 Jun 96 8:42 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id AAA00244 for dbi-users-real; Fri, 21 Jun 1996 00:38:29 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id AAA00240
         for <[email protected]>; Fri, 21 Jun 1996 00:38:26 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id aa07447; 21 Jun 96 8:37 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa10160; 21 Jun 96 7:06 +0100
Received: from toad by oink with SMTP (PP) id <01688-0@oink>;
         Thu, 20 Jun 1996 12:11:10 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA15663;
         Thu, 20 Jun 1996 12:11:03 +0000
Date: Thu, 20 Jun 1996 12:11:03 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: misc 7.3 issues
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1374
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Mark Dedlow <[email protected]>
>
> > > From: Mark Dedlow <[email protected]>
> > >
> > > #!/usr/bin/perl
> > > use DBI;
> > > $drh = DBI->install_driver('Oracle') || die "$DBI::errstr\n";
> > >
> > > But adding:
> > >
> > > $dbh = $drh->connect("", "$cs", '') || die "$DBI::errstr\n";
> > >
> > > causes the bad free.
> >
> > Does it 'bad free' only if the connect works, or fails, or either?
>
> Either.

That's interesting. So a failing connect is sufficient.

> I remade perl with debugging, and re-ran my test script, and it
> now gives:
>
> Duplicate free() ignored during global destruction.

Please try a $SIG{__WARN__} = sub { dump if $_[0] =~ /free/ }
and send me (privately) the stack dump from the core file.

> > I'm almost certainly going to need a stack trace
>
> % dbx /usr/bin/perl core

> =>[1] epccol_close(0x18941c, 0x189408, 0xb8880, 0x0, 0xdf60b771, 0xdf605a68),
> at 0xdf4d2c80
>   [2] epc_exit_handler(0x0, 0x0, 0x0, 0xba3f8, 0xc41c4, 0xba3bc), at 0xdf4d7a08
>   [3] _exithandle(0x0, 0x1bd58, 0x3, 0xdffff6dc, 0xdf680c30, 0x0), at
> 0xdf613a48
>   [4] exit(0x3, 0xdffff6dc, 0xdffff6ec, 0xb8000, 0x0, 0x0), at 0xdf661d40

Oh well. That one's not very helpful. Sigh. Still kind'a looks like oracle.

At some point in the next month or so I'll have the time to install
and use the copy of Purify I have sitting on my desk.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <06588-2@oink>;
         Fri, 21 Jun 1996 10:43:46 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835346746:00206:235; Fri, 21 Jun 96 09:45:46 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa04863; 20 Jun 96 13:17 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id FAA27378 for dbi-users-real; Thu, 20 Jun 1996 05:11:50 -0700
Precedence: junk
Sender: [email protected]
Received: from dekalb.vf.mmc.com (dekalb.vf.mmc.com [192.35.35.21])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id FAA27374
         for <[email protected]>; Thu, 20 Jun 1996 05:11:49 -0700
Received: from franklin.vf.lmco.com (franklin.vf.lmco.com [166.17.5.51])
         by dekalb.vf.mmc.com (8.7.5/8.7.3) with ESMTP id IAA28142;
         Thu, 20 Jun 1996 08:09:02 -0400 (EDT)
Received: from atl.lmco.com (hannibal.ATL.LMCO.COM [166.20.244.250])
         by franklin.vf.lmco.com (8.7.5/8.7.3) with SMTP id IAA13960;
         Thu, 20 Jun 1996 08:09:01 -0400 (EDT)
Received: from valerie by atl.lmco.com (8.6.13/ATL 1.14) with SMTP id IAA20177;
         Thu, 20 Jun 1996 08:08:56 -0400
Message-ID: <[email protected]>
Date: Thu, 20 Jun 1996 08:08:53 -0400
From: Amy McGovern <[email protected]>
Organization: Lockheed Martin Advanced Technology Laboratory
X-Mailer: Mozilla 2.01 (X11; I; SunOS 5.5 sun4m)
MIME-Version: 1.0
To: Tim Bunce <[email protected]>
CC: [email protected]
Subject: Re: Announce: DBD::Oracle 0.33
References: <[email protected]>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Tim Bunce wrote:
> NOTE: This is primarily a release aimed a fixing *some* of the
> Oracle 7.3 problems of late. Do not upgrade to this version unless
> you have an existing build or core dump problem with DBD::Oracle.

I installed DBI 0.70 and DBD 0.33 yesterday and they work fine for Oracle
7.3.2 Solaris 2.5 with some extra help.

To make them work (just in case anyone else needs to do this):

1) We recompiled Perl with -lthread as suggested earlier
2) We installed DBI just fine
3) We installed DBD changing the test script to ignore bad free warnings (as
also suggested earlier) because it was still complaining.

It works just fine in the tests I have done so far except for one thing.  If I
have the environment variable ORA_NLS32 set (right now it is set to:
/home/oracle/app/oracle/product/7.3.2/ocomon/nls/admin/data), then DBD::Oracle
crashes with the following error:

Can't call method "prepare" without a package or object reference at
/opt/gnu/lib/perl5/site_perl/Oraperl.pm line 102.

unsetting the environment variable causes the script to execute normally.
Wierd!

-Amy
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <06907-0@oink>;
         Fri, 21 Jun 1996 11:18:46 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835351661:25575:16; Fri, 21 Jun 96 11:07:41 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa25355; 21 Jun 96 11:07 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa08188; 21 Jun 96 11:03 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA00432 for dbi-users-real; Fri, 21 Jun 1996 02:51:10 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id CAA00428
         for <[email protected]>; Fri, 21 Jun 1996 02:51:05 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ab21713; 21 Jun 96 10:45 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa08468; 21 Jun 96 10:43 +0100
Received: from toad by oink with SMTP (PP) id <06565-0@oink>;
         Fri, 21 Jun 1996 10:42:25 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA18594;
         Fri, 21 Jun 1996 10:42:18 +0000
Date: Fri, 21 Jun 1996 10:42:18 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: How to help me find/fix bad free()'s
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 3117
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Bill Rodgers <[email protected]>
>
> I have just gotten DBI&DBD working with Oracle.  Am running HPUX
> 10.01,perl 5.002, DBI 0.69 and DBD::Oracle 0.32.  All works well except
> that I see a "Bad free() ignored during global destruction" error when
> the script terminates.  I have seen oblique references to this problem
> in the 2 days I've been on the mailing list but no real answers. This
> situation will cause problems with my cgi-bin scripts unless I fix it or
> I find a way to trash the error message and not send it to the www
> server.

The README in 0.33 documents a way to suppress the message.

The README in 0.34 will have extra instructions like this:

IF YOU HAVE PROBLEMS:
...
5. If you still get a core dump rebuild DBD::Oracle with debugging
   enabled by executing: perl Makefile.PL -g  (note the -g option)
   then rerun the code to get a new core dump file, finally use a
   debugger (gdb, sdb, dbx, adb etc) to get a stack trace from it.
   NOTE: I may not be able to help you much without a stack trace!
   It is worth fetching and building the GNU GDB debugger (4.15) if
   you don't have a good debugger on your system. If desparate try:
     make perl; ./perl script; echo '$c' | adb ./perl core
...
-------------------------------------------------------------------------
Oracle 7.3 on Solaris 2.5 (maybe others):

   libc internal error: _rmutex_unlock: rmutex not held.

- Try passing explicit db sid name to ora_login or connect.
- Try removing clntsh from the list of libraries to link to (edit Makefile).
- Try adding -lthread to the list of libraries to link to (edit Makefile).

If any of these work please let me know (include full version info).

-------------------------------------------------------------------------
Bad free() warnings:

These are generally cause by problems in Oracle's own library code.
You can use this code to hide them:

   $SIG{__WARN__} = sub { warn $_[0] unless $_[0] =~ /^Bad free/ }

But please let me know if you have this problem (include full version info).
Rather than simply hiding it it would be much better to help me fix it!
Follow the instructions in step 5 above (for generating a version of
DBD::Oracle with debugging enabled) and then use the following code
to generate a core dump at the point the bad free happens.

   $SIG{__WARN__} = sub { $_[0] =~ /^Bad free/ ? dump : warn $_[0] }

Follow the instructions in step 5 above for getting a stack trace from
the core file and then send it to me along with the other information
requested in the 'If you have problems' section.


> Also, is there an archive of this mailing list somewhere that I
> could get and comb thru old messages?  Thanks,  Bill Rodgers, Sirius
> Inc., Denver CO.

There are gzip'd tar files in the Perl DB archive which cover just about
everything: ftp://ftp.demon.co.uk/pub/perl/db/DBI/perldb-interest/
It starts on 16 Oct 1992. It totals over 2Mb _gzipped_!

I think there's a WWW archive somewhere but I can't remember where.
It'll probably be listed at http://www.hermetica.com/technologia/DBI.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <06907-2@oink>;
         Fri, 21 Jun 1996 11:18:57 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835351852:27487:3; Fri, 21 Jun 96 11:10:52 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa22734; 21 Jun 96 11:03 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA00423 for dbi-users-real; Fri, 21 Jun 1996 02:49:50 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id CAA00419
         for <[email protected]>; Fri, 21 Jun 1996 02:49:39 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ah21850; 21 Jun 96 10:47 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac10160; 21 Jun 96 7:06 +0100
Received: from toad by oink with SMTP (PP) id <03088-0@oink>;
         Thu, 20 Jun 1996 17:27:38 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA16435;
         Thu, 20 Jun 1996 17:27:31 +0000
Date: Thu, 20 Jun 1996 17:27:31 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Announce: DBD::Oracle 0.33
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1812
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Amy McGovern <[email protected]>
>
> Tim Bunce wrote:
> > NOTE: This is primarily a release aimed a fixing *some* of the
> > Oracle 7.3 problems of late. Do not upgrade to this version unless
> > you have an existing build or core dump problem with DBD::Oracle.
>
> I installed DBI 0.70 and DBD 0.33 yesterday and they work fine for Oracle
> 7.3.2 Solaris 2.5 with some extra help.

Good news. Thanks.

> To make them work (just in case anyone else needs to do this):
>
> 1) We recompiled Perl with -lthread as suggested earlier

Please try *not* rebuilding perl with -lthread (use a standard perl)
but add -lthread to the DBD::Oracle Makefile. I don't know if it
will help but it would certainly make life *much* easier for people
if it did. If you try this please double check that you're not using
the perl binary with -lthread.

> 3) We installed DBD changing the test script to ignore bad free warnings
> (as also suggested earlier) because it was still complaining.

Anyone want to bet that Oracle don't use Purify, or a similar tool? Sigh.

> It works just fine in the tests I have done so far except for one thing.
> If I have the environment variable ORA_NLS32 set (right now it is set to:
> /home/oracle/app/oracle/product/7.3.2/ocomon/nls/admin/data), then DBD::Oracle
> crashes with the following error:
>
> Can't call method "prepare" without a package or object reference at
> /opt/gnu/lib/perl5/site_perl/Oraperl.pm line 102.
>
> unsetting the environment variable causes the script to execute normally.
> Wierd!

That's:
       sub ora_open {
           my($lda, $stmt, $cache) = @_;
==>         my $csr = $lda->prepare($stmt) or return undef;

so please check that the ora_login succeeded and returned a handle.
I suspect it failed but your didn't catch it and print the error message.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <07992-0@oink>;
         Fri, 21 Jun 1996 17:00:54 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835355082:00416:4; Fri, 21 Jun 96 12:04:42 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa29461; 21 Jun 96 12:03 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id DAA00508 for dbi-users-real; Fri, 21 Jun 1996 03:56:52 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id DAA00504
         for <[email protected]>; Fri, 21 Jun 1996 03:56:50 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ac26079; 21 Jun 96 10:56 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa19889; 21 Jun 96 11:18 +0100
Received: from toad by oink with SMTP (PP) id <06819-0@oink>;
         Fri, 21 Jun 1996 11:04:52 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA18663;
         Fri, 21 Jun 1996 11:04:45 +0000
Date: Fri, 21 Jun 1996 11:04:45 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: misc 7.3 issues
X-Sun-Charset: US-ASCII
content-length: 1865
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Mark Dedlow <[email protected]>
>
> > Please try a $SIG{__WARN__} = sub { dump if $_[0] =~ /free/ }
> > and send me (privately) the stack dump from the core file.
>
> Still on 0.32 here.
>
>   [1] 0xdf7d4bf0(0xdf313cb2, 0xdf7f2cf4, 0xdfffecf4, 0xdf6077a8, 0x41d,
> 0xc9cdf77), at 0xdf7d4bef
>   [2] 0xdf7d4548(0xdf313cb2, 0xdf790758, 0xdf79069c, 0xdf7f2760, 0xdfffecf4,
> 0x0), at 0xdf7d4547
>   [3] 0xdf7d8860(0xdf32f67c, 0xdf3128b8, 0xdf79069c, 0xdf319080, 0xdf313cb2,
> 0x51c), at 0xdf7d885f
>   [4] 0xdf7d3b8c(0xdf319080, 0x1f, 0x47406b, 0x65, 0xdf6e00e4, 0x1f), at
> 0xdf7d3b8b
>   [5] 0xdf32f134(0x0, 0x0, 0x0, 0x0, 0x6f736500, 0x1), at 0xdf32f133
>   [6] _sigon(0x186033, 0x186032, 0x18602e, 0x18602f, 0x1, 0x18602d), at
> 0xdf319080
>   [7] _lrw_unlock(0x0, 0x37d34b, 0x80, 0xdf337088, 0xdf337070, 0xdf337098), at
> 0xdf317a8c
>   [8] 0xdf7d89b8(0xdf56bb08, 0xdf4d2c80, 0xdf7904e4, 0xdf4d7a08, 0xdf35982f,
> 0xdf350fe4), at 0xdf7d89b7
>   [9] 0xdf7d3b8c(0xdf4d7a08, 0x18a814, 0x18a818, 0xc0008, 0x1a44, 0x0), at
> 0xdf7d3b8b
>   [10] 0xdf5684d0(0x18a81c, 0x18a808, 0xb8880, 0x0, 0xdf60b771, 0xdf605a68),
> at 0xdf5684cf
>   [11] epc_exit_handler(0x0, 0x0, 0x0, 0xba3f8, 0xc4654, 0xba3bc), at
> 0xdf4d7a08
>   [12] _exithandle(0x0, 0x1bd58, 0x3, 0xdffff0ec, 0xdf680c30, 0x0), at
> 0xdf613a48
>   [13] exit(0x3, 0xdffff0ec, 0xdffff0fc, 0xb8000, 0x0, 0x0), at 0xdf661d40
>
> Any better?

Interesting... a search for _lrw_unlock found it in... /usr/lib/libthread.so!
Same for _sigon.

Without a better stack trace (e.g., from GDB with perl and DBD::Oracle built
with -g) I can't tell what the 5 deeper functions are and thus who is
really responsible. The lack of symbols implies a dynamically loaded
function. The libthread implication is interesting.

Did you rebuild perl with -lthread, DBD::Oracle with -lthread or neither?

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <07992-9@oink>;
         Fri, 21 Jun 1996 17:01:35 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835362293:14100:3; Fri, 21 Jun 96 14:04:53 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa13835; 21 Jun 96 14:04 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa14413; 21 Jun 96 14:03 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA00417 for dbi-users-real; Fri, 21 Jun 1996 02:49:05 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id CAA00413
         for <[email protected]>; Fri, 21 Jun 1996 02:46:42 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ab21713; 21 Jun 96 10:45 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa08468; 21 Jun 96 10:43 +0100
Received: from toad by oink with SMTP (PP) id <06565-0@oink>;
         Fri, 21 Jun 1996 10:42:25 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA18594;
         Fri, 21 Jun 1996 10:42:18 +0000
Date: Fri, 21 Jun 1996 10:42:18 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: How to help me find/fix bad free()'s
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 3117
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Bill Rodgers <[email protected]>
>
> I have just gotten DBI&DBD working with Oracle.  Am running HPUX
> 10.01,perl 5.002, DBI 0.69 and DBD::Oracle 0.32.  All works well except
> that I see a "Bad free() ignored during global destruction" error when
> the script terminates.  I have seen oblique references to this problem
> in the 2 days I've been on the mailing list but no real answers. This
> situation will cause problems with my cgi-bin scripts unless I fix it or
> I find a way to trash the error message and not send it to the www
> server.

The README in 0.33 documents a way to suppress the message.

The README in 0.34 will have extra instructions like this:

IF YOU HAVE PROBLEMS:
...
5. If you still get a core dump rebuild DBD::Oracle with debugging
   enabled by executing: perl Makefile.PL -g  (note the -g option)
   then rerun the code to get a new core dump file, finally use a
   debugger (gdb, sdb, dbx, adb etc) to get a stack trace from it.
   NOTE: I may not be able to help you much without a stack trace!
   It is worth fetching and building the GNU GDB debugger (4.15) if
   you don't have a good debugger on your system. If desparate try:
     make perl; ./perl script; echo '$c' | adb ./perl core
...
-------------------------------------------------------------------------
Oracle 7.3 on Solaris 2.5 (maybe others):

   libc internal error: _rmutex_unlock: rmutex not held.

- Try passing explicit db sid name to ora_login or connect.
- Try removing clntsh from the list of libraries to link to (edit Makefile).
- Try adding -lthread to the list of libraries to link to (edit Makefile).

If any of these work please let me know (include full version info).

-------------------------------------------------------------------------
Bad free() warnings:

These are generally cause by problems in Oracle's own library code.
You can use this code to hide them:

   $SIG{__WARN__} = sub { warn $_[0] unless $_[0] =~ /^Bad free/ }

But please let me know if you have this problem (include full version info).
Rather than simply hiding it it would be much better to help me fix it!
Follow the instructions in step 5 above (for generating a version of
DBD::Oracle with debugging enabled) and then use the following code
to generate a core dump at the point the bad free happens.

   $SIG{__WARN__} = sub { $_[0] =~ /^Bad free/ ? dump : warn $_[0] }

Follow the instructions in step 5 above for getting a stack trace from
the core file and then send it to me along with the other information
requested in the 'If you have problems' section.


> Also, is there an archive of this mailing list somewhere that I
> could get and comb thru old messages?  Thanks,  Bill Rodgers, Sirius
> Inc., Denver CO.

There are gzip'd tar files in the Perl DB archive which cover just about
everything: ftp://ftp.demon.co.uk/pub/perl/db/DBI/perldb-interest/
It starts on 16 Oct 1992. It totals over 2Mb _gzipped_!

I think there's a WWW archive somewhere but I can't remember where.
It'll probably be listed at http://www.hermetica.com/technologia/DBI.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <07992-11@oink>;
         Fri, 21 Jun 1996 17:01:55 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835364643:07592:1; Fri, 21 Jun 96 14:44:03 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa07094; 21 Jun 96 14:43 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa21783; 21 Jun 96 14:42 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA00692 for dbi-users-real; Fri, 21 Jun 1996 06:37:21 -0700
Precedence: junk
Sender: [email protected]
Received: from dekalb.vf.mmc.com (dekalb.vf.mmc.com [192.35.35.21])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id GAA00688
         for <[email protected]>; Fri, 21 Jun 1996 06:37:20 -0700
Received: from franklin.vf.lmco.com (franklin.vf.lmco.com [166.17.5.51])
         by dekalb.vf.mmc.com (8.7.5/8.7.3) with ESMTP id JAA26243;
         Fri, 21 Jun 1996 09:34:49 -0400 (EDT)
Received: from atl.lmco.com (hannibal.ATL.LMCO.COM [166.20.244.250])
         by franklin.vf.lmco.com (8.7.5/8.7.3) with SMTP id JAA02180;
         Fri, 21 Jun 1996 09:34:47 -0400 (EDT)
Received: from husnock by atl.lmco.com (8.6.13/ATL 1.14) with SMTP id JAA14706;
         Fri, 21 Jun 1996 09:34:45 -0400
Message-ID: <[email protected]>
Date: Fri, 21 Jun 1996 09:34:43 -0400
From: Amy McGovern <[email protected]>
Organization: Lockheed Martin Advanced Technology Laboratory
X-Mailer: Mozilla 2.01 (X11; I; SunOS 5.5 sun4m)
MIME-Version: 1.0
To: Tim Bunce <[email protected]>
CC: [email protected], [email protected]
Subject: Re: Announce: DBD::Oracle 0.33
References: <[email protected]>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I answered this mail once already, so if you see this twice, I'm sorry!  My
mail reader crashed when I sent it, so I'm not sure if it ever went out.

Tim Bunce wrote:
> Please try *not* rebuilding perl with -lthread (use a standard perl)
> but add -lthread to the DBD::Oracle Makefile. I don't know if it
> will help but it would certainly make life *much* easier for people
> if it did. If you try this please double check that you're not using
> the perl binary with -lthread.

I'll leave this to other people trying to install DBD::Oracle because I don't
want to reinstall stuff now that it works!  And I'd like not to drive the sys
admin too nuts (maybe he'll give me root and then I could do it?)

> > Can't call method "prepare" without a package or object reference at
> > /opt/gnu/lib/perl5/site_perl/Oraperl.pm line 102.
> so please check that the ora_login succeeded and returned a handle.
> I suspect it failed but your didn't catch it and print the error message.

You were correct, I was not checking to see if ora_login failed.  However,
once I added error checking, the following line:

$lda = ora_login($mysid, $myname, $mypass) || die $ora_errstr;

dies with the following error:
ORA-01019: unable to allocate memory in the user side (DBD: login failed) at
/oratest.pl line 10.

unsetting the ORA_NLS32 environment variable lets it work again.  Very odd!

-Amy
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <07992-12@oink>;
         Fri, 21 Jun 1996 17:02:09 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835364647:07619:0; Fri, 21 Jun 96 14:44:07 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa07252; 21 Jun 96 14:43 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa21759; 21 Jun 96 14:42 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA00698 for dbi-users-real; Fri, 21 Jun 1996 06:38:58 -0700
Precedence: junk
Sender: [email protected]
Received: from x4u2.desy.de (x4u2.desy.de [131.169.30.42])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id GAA00686
         for <[email protected]>; Fri, 21 Jun 1996 06:36:20 -0700
Received: (from hfalken@localhost) by x4u2.desy.de (8.6.10/8.6.9) id NAA19843
         for [email protected]; Fri, 21 Jun 1996 13:34:36 GMT
From: Harald Falkenberg <[email protected]>
Message-Id: <[email protected]>
Subject: DBD-Oracle-0.32: can't link on HP/UX 9.05
To: [email protected]
Date: Fri, 21 Jun 1996 15:34:35 +0200 (MST)
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 460

Hi,

I tried to link DBD-Oracle-0.32 on a HP/UX 9.05 using Oracle 7.1.6,
DBI-0.69 and Perl 5.002. Unfortunately I got an error while linking and a list
of unresolved symbols. But, so far I can see, this searched functions are
included in the Oracle libs. This is a problem, which I earlier also saw
with Oraperl.

Is there someone who has experience building DBD on a HP and can tell me
how to manage this problem?

Thank you in advance
       Harald Falkenberg
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <07992-18@oink>;
         Fri, 21 Jun 1996 17:03:04 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835368127:14578:2; Fri, 21 Jun 96 15:42:07 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa13571; 21 Jun 96 15:41 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA00857 for dbi-users-real; Fri, 21 Jun 1996 07:37:26 -0700
Precedence: junk
Sender: [email protected]
Received: from gateway1.srs.gov (gateway1.srs.gov [192.33.240.10])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id HAA00850
         for <[email protected]>; Fri, 21 Jun 1996 07:36:23 -0700
Received: by gateway1.srs.gov id AA19598 (InterLock SMTP Gateway 3.0
         for [email protected]); Fri, 21 Jun 1996 10:33:28 -0400
Message-Id: <[email protected]>
Received: by gateway1.srs.gov (Internal Mail Agent-1);
         Fri, 21 Jun 1996 10:33:28 -0400
Date: Fri, 21 Jun 1996 10:33:07 -0400
From: James Taylor <[email protected]>
Subject: Partial Solution to Solaris 2.5 and Oracle 7.3.2.1
To: [email protected]
Cc: [email protected]
Mime-Version: 1.0
X-Mailer: Mozilla 3.0b4 (X11; I; SunOS 5.4 sun4m)
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I had the hardest time getting past the mutex error.

This is what finally worked for me.  Using DBD-Oracle-0.33 and DBI-0.70.
After running "perl Makefile.PL" the -lthread was already in the
Makefile.  I replaced all "-lthread" with "-mt".  I did this as a test
because of Solaris bug #1224467 (found in the Installation Notes for
Solaris 2.5).  This is what it says:

"If you use -lthread, your program can cause a core dump.

If you have a C++ program that uses threads, do not use -lthread to link
the program.

Workaround:  Use the -mt option to compile and link your application.
Using this option ensures that libthread is linked in the correct
order."

I still have the "Bad free() ignored during global destruction." error.

Later,

James
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <07992-25@oink>;
         Fri, 21 Jun 1996 17:04:32 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835370974:14853:1; Fri, 21 Jun 96 16:29:34 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa14307; 21 Jun 96 16:28 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa09150; 21 Jun 96 16:27 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA00982 for dbi-users-real; Fri, 21 Jun 1996 08:21:37 -0700
Precedence: junk
Sender: [email protected]
Received: from voro.lbl.gov (voro.lbl.gov [131.243.240.19])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA00978
         for <[email protected]>; Fri, 21 Jun 1996 08:21:36 -0700
Received: from voro.lbl.gov (localhost [127.0.0.1]) by voro.lbl.gov (8.6.12/G)
         with ESMTP id IAA29161; Fri, 21 Jun 1996 08:16:46 -0700
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.2 7/18/95
To: Tim Bunce <[email protected]>
cc: [email protected]
Subject: Re: misc 7.3 issues
In-reply-to: Your message of "Fri, 21 Jun 1996 11:04:45 -0000." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 21 Jun 1996 08:16:46 -0700
From: Mark Dedlow <[email protected]>

> From: Tim Bunce <[email protected]>
>
> Interesting... a search for _lrw_unlock found it in... /usr/lib/libthread.so!
> Same for _sigon.
>
> Without a better stack trace (e.g., from GDB with perl and DBD::Oracle built
> with -g) I can't tell what the 5 deeper functions are and thus who is
> really responsible. The lack of symbols implies a dynamically loaded
> function. The libthread implication is interesting.
>
> Did you rebuild perl with -lthread, DBD::Oracle with -lthread or neither?

Neither.  The perl installation is all defaults, except for debugging,
and the DBD::DBI builds are: 'perl Makefile.pl; make; make install'

Mark


---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <09471-1@oink>;
         Fri, 21 Jun 1996 20:13:25 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835384382:06743:2; Fri, 21 Jun 96 20:13:02 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa02536; 21 Jun 96 20:11 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA00802 for dbi-users-real; Fri, 21 Jun 1996 07:10:43 -0700
Precedence: junk
Sender: [email protected]
Received: from fleet.britain.eu.net (fleet.Britain.EU.net [192.91.199.7])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id HAA00797
         for <[email protected]>; Fri, 21 Jun 1996 07:10:39 -0700
From: [email protected]
Message-Id: <[email protected]>
Received: from videotron.co.uk by fleet.britain.eu.net with UUCP
         id <[email protected]>; Fri, 21 Jun 1996 15:10:08 +0100
Date: Fri, 21 Jun 1996 15:10:06 +0100
To: [email protected], [email protected]
Subject: RE: bad free
MIME-version: 1.0 (Created by TFS)
Content-Type: text/plain ; charset=ISO-8859-1
Content-transfer-encoding: quoted-printable
X-Mailer: TFS Gateway V210U0782M



[ apologies if some of the braces get garbled ... something to do with  =20
Microsh?te Exchange ]

Ok .. think I did what I was told :P~

compiled perl with -g
compiled DBI 0.70 -g
compiled DBD oracle 0.33 -g
compiled GDB

system is Digital Unix 3.2
Oracle 7.2.3

SQL> select * from v$version;

BANNER
----------------------------------------------------------------
Oracle7 Server Release 7.2.3.0.0 - Production Release
PL/SQL Release 2.2.3.0.0 - Production
CORE Version 3.4.3.0.0 - Production
TNS for DEC OSF/1 AXP: Version 2.2.3.0.0 - Production
NLSRTL Version 3.1.4.6.0 - Production

Simple script gives bad free:

alpha2:/usr/users/dsmith/DBD-Oracle-0.33# cat t.pl
use Oraperl;

$SIG{__WARN__} =3D sub { $_[0] =3D~ /^Bad free/ ? dump : warn $_[0] };
ora_login($lda,'system/manager@P:WEB','','') or die "Cant logon\n";
ora_logoff($lda);

btrace:
alpha2:/usr/users/dsmith/DBD-Oracle-0.33# gdb /usr/bin/perl core
GDB is free software and you are welcome to distribute copies of it
under certain conditions; type "show copying" to see the conditions.
There is absolutely no warranty for GDB; type "show warranty" for  =20
details.
GDB 4.16 (alpha-dec-osf3.2), Copyright 1996 Free Software Foundation,  =20
Inc...
Core was generated by `perl'.
Program terminated with signal 6, IOT/Abort trap.
Reading symbols from /usr/shlib/libm.so...done.
Reading symbols from /usr/shlib/libc.so...done.
Reading symbols from  =20
/usr/local/lib/perl5/site_perl/alpha-dec_osf/auto/DBI/DBI.so...done.
Reading symbols from  =20
/usr/local/lib/perl5/site_perl/alpha-dec_osf/auto/DBD/Oracle/Oracle.so...d=20=
=20
one.
Reading symbols from /usr/shlib/librt.so...done.
#0  0x3ff8010cf28 in kill ()
(gdb) bt
#0  0x3ff8010cf28 in kill ()
#1  0x12001f848 in my_unexec () at perl.c:1251
#2  0x120059648 in pp_goto () at pp_ctl.c:1836
#3  0x120058678 in pp_dump () at pp_ctl.c:1585
#4  0x120093558 in run () at run.c:54
#5  0x12001d9b4 in perl_call_sv (sv=3D0x140024768, flags=3D2) at perl.c:777
#6  0x120019f4c in warn (pat=3D0x140009230 "%s free() ignored",  =20
..=3D0x1400172a0)
   at util.c:1031
#7  0x120080f24 in free (mp=3D0x1401a4808) at malloc.c:276
#8  0x300000779dc in osndcx ()
#9  0x30000058180 in osnpds ()
#10 0x300000f6d8c in upidhs ()
#11 0x300000f6688 in upilof ()
#12 0x3000004eb40 in ocilof ()
#13 0x30000049eb0 in dbd_db_disconnect ()
#14 0x30000047278 in XS_DBD__Oracle__db_DESTROY ()
#15 0x3ffbfff605c in XS_DBI_dispatch ()
#16 0x120039ab0 in pp_entersub () at pp_hot.c:1756
#17 0x12001d990 in perl_call_sv (sv=3D0x140129968, flags=3D22) at perl.c:77=
5
#18 0x12007b720 in sv_clear (sv=3D0x140127788) at sv.c:2296
#19 0x12007bca0 in sv_free (sv=3D0x140127788) at sv.c:2442
#20 0x12007b944 in sv_clear (sv=3D0x140124e58) at sv.c:2344
#21 0x12007bca0 in sv_free (sv=3D0x140124e58) at sv.c:2442
---Type <return> to continue, or q <return> to quit---
#22 0x12006f5d0 in mg_free (sv=3D0x14015cfa8) at mg.c:240
#23 0x12007b7e4 in sv_clear (sv=3D0x14015cfa8) at sv.c:2310
#24 0x12007bca0 in sv_free (sv=3D0x14015cfa8) at sv.c:2442
#25 0x12007b944 in sv_clear (sv=3D0x14001e068) at sv.c:2344
#26 0x12007bca0 in sv_free (sv=3D0x14001e068) at sv.c:2442
#27 0x120081bb0 in free_tmps () at scope.c:105
#28 0x120031b24 in pp_nextstate () at pp_hot.c:35
#29 0x120093558 in run () at run.c:54
#30 0x12001cba0 in perl_run (sv_interp=3D0x14001d818) at perl.c:557
#31 0x120017ce0 in main (argc=3D2, argv=3D0x11ffffb08, env=3D0x11ffffb20)
   at perlmain.c:49

Hope this helps.

We also got the same bad frees with DBI 0.69 and DBD 0.29

Dan
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <09499-0@oink>;
         Fri, 21 Jun 1996 20:17:39 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835384454:07505:1; Fri, 21 Jun 96 20:14:14 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa06221; 21 Jun 96 20:12 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA00842 for dbi-users-real; Fri, 21 Jun 1996 07:33:35 -0700
Precedence: junk
Sender: [email protected]
Received: from netmail1.austin.ibm.com (netmail1.austin.ibm.com [129.35.208.96])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id HAA00835;
         Fri, 21 Jun 1996 07:33:33 -0700
From: [email protected]
Received: from dax.austin.ibm.com (dax.austin.ibm.com [129.35.148.59])
         by netmail1.austin.ibm.com (8.6.12/8.6.11) with ESMTP id JAA64005;
         Fri, 21 Jun 1996 09:33:32 -0500
Received: (from mhm@localhost)
         by dax.austin.ibm.com (AIX 9610R-UP4.1/UCB 8.7/8.7-client1.0)
         id JAA21988; Fri, 21 Jun 1996 09:33:31 -0500 (CDT)
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.7 5/3/96
To: [email protected]
cc: [email protected]
Subject: ANNOUNCE: DBD::DB2-0.59
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 21 Jun 96 09:33:31 -0500


DBD::DB2-0.59 is making its way around CPAN.  This version fixes some
bugs I introduced when attempting to change the namespace utilization
of DBD::DB2::Constants


------------------------------------------------------------------------------
Michael H. Moran                  | Standard Disclaimer:  The content of
[email protected]                | this posting is independent of
Commercial Performance            | official IBM position.
IBM Corporation, Austin, Texas    |


---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <09499-2@oink>;
         Fri, 21 Jun 1996 20:17:47 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835384593:08850:2; Fri, 21 Jun 96 20:16:33 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa07377; 21 Jun 96 20:15 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA01475 for dbi-users-real; Fri, 21 Jun 1996 12:10:32 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id MAA01471
         for <[email protected]>; Fri, 21 Jun 1996 12:10:31 -0700
Received: by gw.home.vix.com id MAA07882; Fri, 21 Jun 1996 12:10:24 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ag13364; 21 Jun 96 18:54 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa26236; 21 Jun 96 19:37 +0100
Received: from toad by oink with SMTP (PP) id <09232-0@oink>;
         Fri, 21 Jun 1996 19:33:30 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA21757;
         Fri, 21 Jun 1996 19:33:23 +0000
Date: Fri, 21 Jun 1996 19:33:23 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Partial Solution to Solaris 2.5 and Oracle 7.3.2.1
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: James Taylor <[email protected]>
>
> James Taylor wrote:
> >
> >I replaced all "-lthread" with "-mt".  I did this as a test
>
> You know, since my ld DOESN'T HAVE a -mt option, all this really did was
> to remove all occurences of -lthread from the Makefile!
>
> However, this seems to work.  So, my fix is to remove -lthread from the
> Makefile and, for whatever reason, the rmutex error and core dump is
> gone.  Bad free() is still there.
>
> Sorry for the error,

No problem. But didn't your ld complain about the -mt flag?
Umm, perhaps it read it as -m & -t (both valid flags).

Could you retry (make clean etc) with _just_ removing -lthread.
That way we'd know for sure.

Once I hear confirmation from you I'll make another DBD::Oracle
release with that workaround in it.

Meanwhile others with the _rmutex_unlock problem might like
to try the same thing (edit Makefile and delete all instances
of -lthread).

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <10159-2@oink>;
         Fri, 21 Jun 1996 21:54:55 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835386542:29901:1; Fri, 21 Jun 96 20:49:02 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa27664; 21 Jun 96 20:46 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA01534 for dbi-users-real; Fri, 21 Jun 1996 12:41:10 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id MAA01530
         for <[email protected]>; Fri, 21 Jun 1996 12:40:50 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id af10761; 21 Jun 96 18:27 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ad22396; 21 Jun 96 19:25 +0100
Received: from toad by oink with SMTP (PP) id <09029-0@oink>;
         Fri, 21 Jun 1996 19:22:55 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA21717;
         Fri, 21 Jun 1996 19:22:48 +0000
Date: Fri, 21 Jun 1996 19:22:48 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Bad free() using Oracle
X-Sun-Charset: US-ASCII
content-length: 2191
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected]
>
> [ apologies if some of the braces get garbled ... something to do with
> Microsh?te Exchange ]

:-)

> Ok .. think I did what I was told :P~
>
> compiled perl with -g
> compiled DBI 0.70 -g
> compiled DBD oracle 0.33 -g
> compiled GDB
>
> system is Digital Unix 3.2
> Oracle 7.2.3
>
> SQL> select * from v$version;
>
> BANNER
>  ----------------------------------------------------------------
> Oracle7 Server Release 7.2.3.0.0 - Production Release
> PL/SQL Release 2.2.3.0.0 - Production
> CORE Version 3.4.3.0.0 - Production
> TNS for DEC OSF/1 AXP: Version 2.2.3.0.0 - Production
> NLSRTL Version 3.1.4.6.0 - Production
>
> Simple script gives bad free:
>
> alpha2:/usr/users/dsmith/DBD-Oracle-0.33# cat t.pl
> use Oraperl;
>
> $SIG{__WARN__} = sub { $_[0] =~ /^Bad free/ ? dump : warn $_[0] };
> ora_login($lda,'system/manager@P:WEB','','') or die "Cant logon\n";
> ora_logoff($lda);
>
> btrace:
> alpha2:/usr/users/dsmith/DBD-Oracle-0.33# gdb /usr/bin/perl core
..
> #6  0x120019f4c in warn (pat=0x140009230 "%s free() ignored",
> ...=0x1400172a0)
>     at util.c:1031
> #7  0x120080f24 in free (mp=0x1401a4808) at malloc.c:276
> #8  0x300000779dc in osndcx ()
> #9  0x30000058180 in osnpds ()
> #10 0x300000f6d8c in upidhs ()
> #11 0x300000f6688 in upilof ()
> #12 0x3000004eb40 in ocilof ()
> #13 0x30000049eb0 in dbd_db_disconnect ()
..
> Hope this helps.

Helps? It nails it on the head! Thank you.

I hope one of the Oracle people on this list can get this logged
as a bug (at a high priority). Or maybe tell us what version it
was fixed in.

There is really no excuse for a company like Oracle shipping
code which does bad free's. Tools like Purify and Pure Coverage
should make this kind of bug a thing of the past.

I'll see if I can put in a workaround. What I'm thinking of will
only work if you actually call the Oraperl ora_logoff() function.

I note that you're bad free just might not be the same as some
others have reported so I'd still like to have some other people
(especially 7.3 users getting Bad free's) do the same as Dan so
I can see what's happening on your setup and compare them.

Thanks again Dan.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <10159-5@oink>;
         Fri, 21 Jun 1996 21:55:09 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835387974:15947:1; Fri, 21 Jun 96 21:12:54 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa12448; 21 Jun 96 21:09 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA01076 for dbi-users-real; Fri, 21 Jun 1996 09:20:30 -0700
Precedence: junk
Sender: [email protected]
Received: from gateway1.srs.gov (gateway1.srs.gov [192.33.240.10])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA01066
         for <[email protected]>; Fri, 21 Jun 1996 09:19:23 -0700
Received: by gateway1.srs.gov id AA03363 (InterLock SMTP Gateway 3.0
         for [email protected]); Fri, 21 Jun 1996 12:16:38 -0400
Message-Id: <[email protected]>
Received: by gateway1.srs.gov (Internal Mail Agent-1);
         Fri, 21 Jun 1996 12:16:38 -0400
Date: Fri, 21 Jun 1996 12:16:14 -0400
From: James Taylor <[email protected]>
Subject: Re: Partial Solution to Solaris 2.5 and Oracle 7.3.2.1
To: [email protected], [email protected]
Mime-Version: 1.0
X-Mailer: Mozilla 3.0b4 (X11; I; SunOS 5.4 sun4m)
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
References: <[email protected]>

James Taylor wrote:
>
>I replaced all "-lthread" with "-mt".  I did this as a test

You know, since my ld DOESN'T HAVE a -mt option, all this really did was
to remove all occurences of -lthread from the Makefile!

However, this seems to work.  So, my fix is to remove -lthread from the
Makefile and, for whatever reason, the rmutex error and core dump is
gone.  Bad free() is still there.

Sorry for the error,

James
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <10159-9@oink>;
         Fri, 21 Jun 1996 21:55:21 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835388260:19317:0; Fri, 21 Jun 96 21:17:40 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa18653; 21 Jun 96 21:16 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa25260; 21 Jun 96 21:16 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA01607 for dbi-users-real; Fri, 21 Jun 1996 13:12:44 -0700
Precedence: junk
Sender: [email protected]
Received: from mercury.Sun.COM (mercury.Sun.COM [192.9.25.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA01603
         for <[email protected]>; Fri, 21 Jun 1996 13:12:43 -0700
Received: by mercury.Sun.COM (Sun.COM) id NAA11463;
         Fri, 21 Jun 1996 13:10:07 -0700
Received: from suntri.East.Sun.COM by East.Sun.COM (5.x/SMI-5.3) id AA02764;
         Fri, 21 Jun 1996 16:09:59 -0400
Received: from utopia.East.Sun.COM by suntri.East.Sun.COM (SMI-8.6/SMI-SVR4)
         id QAA12727; Fri, 21 Jun 1996 16:09:56 -0400
Received: by utopia.East.Sun.COM (5.x/SMI-SVR4) id AA00939;
         Fri, 21 Jun 1996 16:06:23 -0400
Date: Fri, 21 Jun 1996 16:06:23 -0400
From: Dave Anderson - SMCC Reseller SE - Cary NC <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: ?:Help
Cc: [email protected]
Mime-Version: 1.0
Content-Type: multipart/mixed;boundary=5e9d_52cd-4cdd_2b2-64bd_6b1b


--5e9d_52cd-4cdd_2b2-64bd_6b1b
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-MD5: Hv+JS98PVt1tepiQ5gN+uw==
X-Sun-Data-Type: text

All,

Has anyone every seen this problem.  See attached files

TIA,
Dave
--5e9d_52cd-4cdd_2b2-64bd_6b1b
Content-Type: application/octet-stream
Content-Transfer-Encoding: 7bit
Content-MD5: a3HOL/jG4dhI2Nq78M0/sA==
Content-Description: x1

PERL_DL_NONLAZY=1 /opt/local/bin/perl -I./blib/arch -I./blib/lib -I/opt/local/lib/perl5/sun4-solaris/5.002 -I/opt/local/lib/perl5 -e 'use Test::Harness qw(&runtests $verbose); $verbose=1; runtests @ARGV;' t/*.t
t/base..............install_driver(Oracle) failed: Can't load './blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle: ld.so.1: /opt/local/bin/perl: fatal: libclntsh.so.1.0: can't open file: errno=2 at /opt/local/lib/perl5/DynaLoader.pm line 140.

at blib/lib/DBD/Oracle.pm line 24
       DBI::install_driver called at t/base.t line 16
1..5
ok 1
ok 2
ok 3
FAILED tests 4-5
       Failed 2/5 tests, 60.00% okay
       Test returned status 2 (wstat 512)
Failed 1 test script, 0.00% okay. 2/5 subtests failed, 60.00% okay.
*** Error code 29
make: Fatal error: Command failed for target `test_dynamic'
--5e9d_52cd-4cdd_2b2-64bd_6b1b
Content-Type: application/octet-stream
Content-Transfer-Encoding: 7bit
Content-MD5: FG/xypd2hlCHIzO+kTqQgA==
Content-Description: x2

Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=solaris, osver=2.4, archname=sun4-solaris
   uname='sunos utopia 5.4 generic_101945-32 sun4m sparc '
   hint=recommended, useposix=true
 Compiler:
   cc='cc', optimize='-O', gccversion=
   cppflags=''
   ccflags =''
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=n, randbits=15
 Linker and Libraries:
   ld='cc', ldflags =' -L/usr/local/lib -L/opt/local/lib'
   libpth=/usr/local/lib /opt/local/lib /lib /usr/lib /usr/ccs/lib
   libs=-lsocket -lnsl -ldl -lm -lc -lcrypt
   libc=/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-Kpic', lddlflags='-G -L/usr/local/lib -L/opt/local/lib'

@INC: /opt/local/lib/perl5/sun4-solaris/5.002 /opt/local/lib/perl5 /opt/local/lib/perl5/site_perl/sun4-solaris /opt/local/lib/perl5/site_perl /opt/local/lib/perl5/sun4-solaris .
--5e9d_52cd-4cdd_2b2-64bd_6b1b--
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10188-12@oink>;
         Fri, 21 Jun 1996 21:56:24 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835379352:00783:2134; Fri, 21 Jun 96 18:49:12 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa10110; 20 Jun 96 21:34 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA28661 for dbi-users-real; Thu, 20 Jun 1996 10:08:47 -0700
Precedence: junk
Sender: [email protected]
Received: from mail13.digital.com (mail13.digital.com [192.208.46.30])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id KAA28657
         for <[email protected]>; Thu, 20 Jun 1996 10:08:46 -0700
Received: from us1rmc.bb.dec.com by mail13.digital.com (8.7.5/UNX 1.2/1.0/WV)
         id NAA13677; Thu, 20 Jun 1996 13:00:27 -0400 (EDT)
Received: from decna.lkg.dec.com by us1rmc.bb.dec.com (5.65/rmc-22feb94)
         id AA26305; Thu, 20 Jun 96 13:00:01 -0400
Received: by decna.lkg.dec.com (UCX V3.3-7A); Thu, 20 Jun 1996 12:59:25 -0400
Message-Id: <[email protected]>
Date: Thu, 20 Jun 1996 13:02:41 -0400
From: Chris Perreault <[email protected]>
Reply-To: [email protected]
Organization: Digital Equipment Corporation
X-Mailer: Mozilla 3.0b4Gold (WinNT; I)
Mime-Version: 1.0
To: [email protected]
Subject: DSBD:Ingres
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi Tim


I was wondering about the availability of the driver for Ingres. I have a
project rapidly approaching and I'm going to have to use University
Ingres :(

Even if it were in really rough form, it would still be a huge help. I
promise not to bother you with bug reports (unless you want them)

Thanks a lot.


Chris


--
_______________________________________________
Christopher Perreault
Web Guy, ISBU Technical Support
Digital Equipment Corporation - LKG2-2 Pole Q9
[email protected] - 508.486.6849
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10188-17@oink>;
         Fri, 21 Jun 1996 21:56:43 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835379471:00782:2393; Fri, 21 Jun 96 18:51:11 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa10851; 20 Jun 96 16:30 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id FAA27356 for dbi-users-real; Thu, 20 Jun 1996 05:07:28 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net ([158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id FAA27352
         for <[email protected]>; Thu, 20 Jun 1996 05:07:19 -0700
Received: by relay-2.mail.demon.net id aa01077; 20 Jun 96 13:03 +0100
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ae29406; 20 Jun 96 12:55 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa03767; 20 Jun 96 11:58 +0100
Received: from toad by oink with SMTP (PP) id <01284-0@oink>;
         Thu, 20 Jun 1996 11:57:04 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA15622;
         Thu, 20 Jun 1996 11:56:57 +0000
Date: Thu, 20 Jun 1996 11:56:57 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Quotes -- again
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1265
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: "Kathleen F. Duncan" <[email protected]>
>
> I've installed version DBI-0.70 and am using DBI-mSQL-0.60pl10.  The
> quote function is still causing problems.  When I invoke the function, I
> get the message :
>
> Can't locate object method "quote" via package "DBI::db"

Ooops.

> I've looked at the code but don't know perl well enough to figure out how
> to correct it.  I have modified the Database Session Class interface to
> include a line for the quote as below.  At least it finds it, but what is
> quoted and returned is 'msql'.

Your change was almost perfect. Here's my 'oficial' patch (which I'll
include in the next release).


*** 1.56        1996/06/16 23:01:32
--- DBI.pm      1996/06/20 10:28:42
***************
*** 89,94 ****
--- 89,95 ----
       errmsg     =>   { U =>[1,1], O=>0x04 },
       disconnect =>   { U =>[1,1] },
       tables     =>   { U =>[1,1] },
+       quote      =>   { U =>[2,2, '$str'] },
       @Common_IF,
       @TieHash_IF,
     },

> Also, in the quote function you replaced "'" with "''".  Is that standard
> for most SQL's?  For mSQL it needs to be "escaped"  (i.e. "\'").

The mSQL driver isn't supplying it own quote method so you're getting
the inherited default (SQL92) method that the DBI supplies.

The mSQL driver needs fixing.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10188-35@oink>;
         Fri, 21 Jun 1996 21:58:07 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835381646:00783:5256; Fri, 21 Jun 96 19:27:26 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa23435; 21 Jun 96 7:01 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id WAA29998 for dbi-users-real; Thu, 20 Jun 1996 22:57:59 -0700
Precedence: junk
Sender: [email protected]
Received: from zeus.carroll.com (zeus.carroll.com [199.224.10.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id WAA29994
         for <[email protected]>; Thu, 20 Jun 1996 22:57:57 -0700
Received: from Phred by zeus.carroll.com with SMTP (8.7.1/0) id CAA30313;
         Fri, 21 Jun 1996 02:00:32 -0400 (EDT)
Message-Id: <[email protected]>
X-Sender: [email protected]
X-Mailer: Windows Eudora Version 1.4.3
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Fri, 21 Jun 1996 02:09:41 -0500
To: [email protected]
From: [email protected] (Jon Adams)
Subject: Re: problem with dynamic library loading on Solaris 2.5
Cc: Tim Bunce <[email protected]>

>
>> From: Jon Adams <[email protected]>
>
>Try using DBD::Oracle 0.33 (just released). If it core dumps then rebuild
>the Makefile using   perl Makefile.PL -g   rerun it and get a stack trace
>from the new core file.
>
>Tim.
>

Okay, sorry it took awhile to respond. Had to find a debugger that handled ELF
correctly. I installed both the new version of the DBI (0.70) and DBD::Oracle
0.33. This also dumps core. Here's the trace from ADB...


core file = core -- program ``perl'' on platform SUNW,Ultra-1
SIGSEGV: Segmentation Fault

               object: /usr/local/bin/perl
               dump: /export/home/jadams/DBD-Oracle-0.33/core

               (10$d) number base
radix=16 base ten

               ($m) address map
? map
b1 =    10000 e1 =    bd055 f1 =      0 `/usr/local/bin/perl'
b2 =    cd058 e2 =    d53f0 f2 =  ad058 `/usr/local/bin/perl'
b3 = ef782efc e3 = ef78c1a0 f3 =   2efc `/lib/libsocket.so.1'
b4 = ef691890 e4 = ef6d923c f4 =  11890 `/lib/libnsl.so.1'
b5 = ef7b0440 e5 = ef7b0470 f5 =    440 `/lib/libdl.so.1'
b6 = ef7435d8 e6 = ef752414 f6 =   35d8 `/lib/libm.so.1'
b7 = ef596938 e7 = ef5f3ed4 f7 =  16938 `/lib/libc.so.1'
b8 = ef720d84 e8 = ef722438 f8 =    d84 `/lib/libintl.so.1'
b9 = ef660b60 e9 = ef662994 f9 =    b60 `/lib/libmp.so.1'
b10 = ef6425a8 e10 = ef6465c0 f10 =   25a8 `/lib/libw.so.1'
b11 = ef620448 e11 = ef62330c f11 =    448
`/usr/platform/SUNW,Ultra-1/lib/libc_psr.so.1'
b12 = ef55202c e12 = ef55adac f12 =   202c
`/usr/local/lib/perl5/site_perl/sun4-solaris/auto/DBI/DBI.so'
b13 = ef532ca0 e13 = ef539d04 f13 =   2ca0
`./blib/arch/auto/DBD/Oracle/Oracle.so'
b14 = ef2b3018 e14 = ef48de84 f14 =  33018
`/db/oracle/app/oracle/product/7.3.2/lib/libclntsh.so.1.0'
b15 = ef4f6de0 e15 = ef504560 f15 =   6de0 `/lib/libthread.so.1'
/ map
b1 =    cc000 e1 =    d6000 f1 =   1174 `core'
b2 =    d6000 e2 =   18a000 f2 =   b174 `core'
b3 = ee802000 e3 = ee804000 f3 =  bf174 `core'
b4 = ee904000 e4 = ee906000 f4 =  c1174 `core'
b5 = eea06000 e5 = eea08000 f5 =  c3174 `core'
b6 = eeb08000 e6 = eeb0a000 f6 =  c5174 `core'
b7 = eec0a000 e7 = eec0c000 f7 =  c7174 `core'
b8 = eed0c000 e8 = eed0e000 f8 =  c9174 `core'
b9 = eee0e000 e9 = eee10000 f9 =  cb174 `core'
b10 = eef0c000 e10 = eef0e000 f10 =  cd174 `core'
b11 = eef10000 e11 = eef12000 f11 =  cf174 `core'
b12 = ef00e000 e12 = ef010000 f12 =  d1174 `core'
b13 = ef4bc000 e13 = ef4cc000 f13 =  d3174 `core'
b14 = ef4cc000 e14 = ef4d0000 f14 =  e3174 `core'
b15 = ef514000 e15 = ef518000 f15 =  e7174 `core'
b16 = ef518000 e16 = ef51e000 f16 =  eb174 `core'
b17 = ef54a000 e17 = ef54c000 f17 =  f1174 `core'
b18 = ef56c000 e18 = ef56e000 f18 =  f3174 `core'
b19 = ef60c000 e19 = ef616000 f19 =  f5174 `core'
b20 = ef616000 e20 = ef618000 f20 =  ff174 `core'
b21 = ef632000 e21 = ef634000 f21 = 101174 `core'
b22 = ef656000 e22 = ef658000 f22 = 103174 `core'
b23 = ef672000 e23 = ef674000 f23 = 105174 `core'
b24 = ef6ee000 e24 = ef6f8000 f24 = 107174 `core'
b25 = ef6f8000 e25 = ef700000 f25 = 111174 `core'
b26 = ef732000 e26 = ef734000 f26 = 119174 `core'
b27 = ef764000 e27 = ef766000 f27 = 11b174 `core'
b28 = ef79c000 e28 = ef79e000 f28 = 11d174 `core'
b29 = ef7c0000 e29 = ef7c2000 f29 = 11f174 `core'
b30 = ef7f4000 e30 = ef7f6000 f30 = 121174 `core'
b31 = efffc000 e31 = f0000000 f31 = 123174 `core'

               ($v) b: data start, d: data length, e: entry point,
                       m: magic number, t: text length
variables
0 = 10
b = cd058
d = 964c
e = 7f454c46
m = 7f454c46
t = ad055

               ($?) pid, signal, registers, & pgm ctr
no process
SIGSEGV: Segmentation Fault
g0    0x0                               l0      0x0
g1    0xa3                              l1      0x0
g2    0xd5400       nextf+0x8           l2      0x0
g3    0x6c000000                        l3      0x0
g4    0x0                               l4      0x0
g5    0x0                               l5      0x0
g6    0x0                               l6      0x0
g7    0x188000                          l7      0xef5147f8  _GLOBAL_OFFSET_TABLE_
o0    0x0                               i0      0xb
o1    0xb                               i1      0xeffff820
o2    0x0                               i2      0xeffff568
o3    0x0                               i3      0xeffff4a8
o4    0xff00                            i4      0xff00
o5    0xff                              i5      0xef51829c  _siguflags
sp    0xeffff3c8                        fp      0xeffff448
o7    0xef503030    __libthread_segvhdlr+0x110  i7    0xef502418
sigacthandler+0x64c
y     0x0
tstate: 4482001a07  (ccr=0x44, asi=0x82, pstate=0x1a, cwp=0x7)
pstate: ag:0 ie:1 priv:0 am:1 pef:1 mm:0 tle:0 cle:0 mg:0 ig:0
pc    0xef5b890c __lwp_kill+8:  bgeu,a  __lwp_kill + 0x1c
npc   0xef5b8910 __lwp_kill+0xc:                clr     %o0

               ($c) c stack backtrace
__lwp_kill(0xb,0xeffff820,0xeffff568,0xeffff4a8,0xff00,0xef51829c) + 8

               ($q) quit adb
--
Jon Adams                          Promethean Enterprises, Inc.
[email protected]           http://www.pe-netsystems.com
  - Internet and Web Development for the Corporate World -

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10188-37@oink>;
         Fri, 21 Jun 1996 21:58:18 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835381903:00782:5707; Fri, 21 Jun 96 19:31:43 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa25139; 21 Jun 96 0:37 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA28910 for dbi-users-real; Thu, 20 Jun 1996 11:45:12 -0700
Precedence: junk
Sender: [email protected]
Received: from YANG.EARLHAM.EDU (YANG.EARLHAM.EDU [159.28.1.1])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id LAA28906
         for <[email protected]>; Thu, 20 Jun 1996 11:45:09 -0700
Received: from polycarp.math.earlham.edu by YANG.EARLHAM.EDU (MX V4.1 VAX)
         with SMTP; Thu, 20 Jun 1996 13:44:36 EST
Received: from litoria by polycarp.math.earlham.edu (NX5.67e/NX3.0M) id AA13312;
         Thu, 20 Jun 96 13:44:35 -0500
Received: by litoria.math.earlham.edu (NX5.67e/NX3.0X) id AA25303;
         Thu, 20 Jun 96 13:44:33 -0500
Date: Thu, 20 Jun 1996 13:44:32 -0500 (EST)
From: Karl Kuntz <[email protected]>
To: [email protected]
Subject: postgres95
Message-ID: <Pine.NXT.3.91.960620133939.25299B-100000@litoria>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


Hello,  has anyone know used  pgperl with postgres95 and if so does it
support joins?

thanks

Karl


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10188-50@oink>;
         Fri, 21 Jun 1996 21:59:09 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835383176:00782:7444; Fri, 21 Jun 96 19:52:56 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id ab04143; 21 Jun 96 2:45 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA29596 for dbi-users-real; Thu, 20 Jun 1996 17:01:53 -0700
Precedence: junk
Sender: [email protected]
Received: from gateway1.srs.gov (gateway1.srs.gov [192.33.240.10])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id RAA29592
         for <[email protected]>; Thu, 20 Jun 1996 17:00:47 -0700
Received: by gateway1.srs.gov id AA07614 (InterLock SMTP Gateway 3.0
         for [email protected]); Thu, 20 Jun 1996 19:58:08 -0400
Message-Id: <[email protected]>
Received: by gateway1.srs.gov (Internal Mail Agent-1);
         Thu, 20 Jun 1996 19:58:08 -0400
Date: Thu, 20 Jun 1996 19:58:02 -0400
From: James Taylor <[email protected]>
Subject: Re: DBD-Oracle bad free() error and web scripts
To: Bill Rodgers <[email protected]>
Cc: [email protected], [email protected]
Mime-Version: 1.0
X-Mailer: Mozilla 3.0b4 (X11; I; SunOS 5.4 sun4m)
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
References: <[email protected]>

Bill Rodgers wrote:
>

Isn't a fix for that in the README along with a request for full info
from Tim?

> that I see a "Bad free() ignored during global destruction" error when
> the script terminates.  I have seen oblique references to this problem
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10188-61@oink>;
         Fri, 21 Jun 1996 21:59:52 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835385156:00783:10129; Fri, 21 Jun 96 20:25:56 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id ab21480; 21 Jun 96 0:29 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa24404; 21 Jun 96 0:29 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA29528 for dbi-users-real; Thu, 20 Jun 1996 16:20:16 -0700
Precedence: junk
Sender: [email protected]
Received: from dfw-ix7.ix.netcom.com (dfw-ix7.ix.netcom.com [206.214.98.7])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA29524
         for <[email protected]>; Thu, 20 Jun 1996 16:20:15 -0700
Received: from curly.cde.state.co.us ([204.131.28.18])
         by dfw-ix7.ix.netcom.com (8.6.13/8.6.12) with SMTP id QAA00382;
         Thu, 20 Jun 1996 16:17:13 -0700
Message-ID: <[email protected]>
Date: Thu, 20 Jun 1996 17:16:09 -0600
From: Bill Rodgers <[email protected]>
X-Mailer: Mozilla 2.01Gold (Win95; I)
MIME-Version: 1.0
To: [email protected]
CC: [email protected]
Subject: DBD-Oracle bad free() error and web scripts
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I have just gotten DBI&DBD working with Oracle.  Am running HPUX
10.01,perl 5.002, DBI 0.69 and DBD::Oracle 0.32.  All works well except
that I see a "Bad free() ignored during global destruction" error when
the script terminates.  I have seen oblique references to this problem
in the 2 days I've been on the mailing list but no real answers.  This
situation will cause problems with my cgi-bin scripts unless I fix it or
I find a way to trash the error message and not send it to the www
server.  Also, is there an archive of this mailing list somewhere that I
could get and comb thru old messages?  Thanks,  Bill Rodgers, Sirius
Inc., Denver CO.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10250-6@oink>;
         Fri, 21 Jun 1996 22:00:10 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835385827:00783:11044; Fri, 21 Jun 96 20:37:07 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id ab27863; 20 Jun 96 19:34 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA28337 for dbi-users-real; Thu, 20 Jun 1996 08:00:47 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA28333
         for <[email protected]>; Thu, 20 Jun 1996 08:00:43 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id PAA16700; Thu, 20 Jun 1996 15:48:40 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: Quotes -- again
To: [email protected] (Tim Bunce)
Date: Thu, 20 Jun 1996 15:48:40 +0100 (BST)
Cc: [email protected]
In-Reply-To: <[email protected]> from "Tim Bunce" at Jun 20, 96 11:56:57 am
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 467


> The mSQL driver isn't supplying it own quote method so you're getting
> the inherited default (SQL92) method that the DBI supplies.
>
> The mSQL driver needs fixing.

Very true. I'll put it on the ToDo list, unless anyone fancies supplying
a patch? (hint)

> Tim.

--
Alligator Descartes     |       Nexus - Multi-User Virtual Reality
Anathematician          |         http://www.hermetica.com/nexus
[email protected]  | Muse of the Day: "Deathchase" - Mervyn Estcourt
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10188-67@oink>;
         Fri, 21 Jun 1996 22:00:21 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835386606:00783:12151; Fri, 21 Jun 96 20:50:06 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id ac11495; 21 Jun 96 7:59 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id XAA00164 for dbi-users-real; Thu, 20 Jun 1996 23:57:59 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id XAA00160
         for <[email protected]>; Thu, 20 Jun 1996 23:57:53 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ac02197; 21 Jun 96 6:56 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab10160; 21 Jun 96 7:06 +0100
Received: from toad by oink with SMTP (PP) id <03072-0@oink>;
         Thu, 20 Jun 1996 17:20:21 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA16416;
         Thu, 20 Jun 1996 17:20:14 +0000
Date: Thu, 20 Jun 1996 17:20:14 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Oracle Hints
X-Sun-Charset: US-ASCII
content-length: 665
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Philippe Vanhaesendonck <[email protected]>
>
> > Does the DBD-Oracle/Perl interface handle hints. Within Oracle it is
> > possible to use the following syntax
> >
> > select /*+ index(<table> <index>) */
> >     <cols>
> >   from <table>
> >   where <pk col> = <val>
>
> Hints are a server feature, not a client spec, so i do not see why it
> does not work...

Should do.

> (unless DBI/DBD strips comments...)

It doesn't.

> Note that oracle will ignore the hint if there is a syntax error.
>
> To convince yourseld that the hint is passed to the server, just set sql_trace
> to true and look at the trace file.
>
> Phil.

Thanks Phil.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10250-11@oink>;
         Fri, 21 Jun 1996 22:00:34 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835387253:00973:10432; Fri, 21 Jun 96 21:00:53 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id ac24277; 21 Jun 96 0:35 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA29163 for dbi-users-real; Thu, 20 Jun 1996 13:58:46 -0700
Precedence: junk
Sender: [email protected]
Received: from relay1.UU.NET (relay1.UU.NET [192.48.96.5])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA29159
         for <[email protected]>; Thu, 20 Jun 1996 13:58:45 -0700
Received: from gateway.deere.com by relay1.UU.NET
         with SMTP (peer crosschecked as: gateway.deere.com [192.43.69.34])
         id QQauyt05060; Thu, 20 Jun 1996 16:58:33 -0400 (EDT)
Received: by gateway.deere.com; id PAA20365; Thu, 20 Jun 1996 15:58:43 -0500
Received: from deere.com(192.43.1.3) by gateway.deere.com via smap (g3.0.1)
         id xma020271; Thu, 20 Jun 96 15:58:21 -0500
Received: from ou79138 ([204.53.32.66]) by deere.dx.deere.com (4.1/SMI-4.0)
         id AA06531; Thu, 20 Jun 96 15:58:19 CDT
Message-Id: <[email protected]>
Date: Thu, 20 Jun 1996 15:58:15 -0700
From: "Dr. Blue" <[email protected]>
Organization: John Deere Credit
X-Mailer: Mozilla 3.0b4 (Win16; I)
Mime-Version: 1.0
To: [email protected]
Subject: How to specify TNS string for Oracle
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I am using DBI/DBD-Oracle (downloaded a week ago).
My problem is how to access a database using the
TNS string as in
    user/passwd@TNS_string

I have no problem connecting to local databases using:
    &ora_login ("TEST", "my_id/my_passwd", ' ');
where TEST is a local database.

I would like to say
    &ora_login ('', 'my_id/[email protected]', '');
or
    &ora_login ('PROD.aix2.com', 'my_id/mypasswd', ' ');
or something similar.

Needless to say, I have tried as many variations as I could
but nothing seems to work. I tried setting ORACLE_SID and
TWO_TASK and it either doesn't work or I did it wrong.

Can anyone help? Thanks.

Richard
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <12694-0@oink>;
         Fri, 21 Jun 1996 23:25:37 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835395693:10616:1; Fri, 21 Jun 96 23:21:33 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa09619; 21 Jun 96 23:20 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id PAA01883 for dbi-users-real; Fri, 21 Jun 1996 15:17:15 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id PAA01879
         for <[email protected]>; Fri, 21 Jun 1996 15:17:09 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id bg00505; 21 Jun 96 22:42 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa11807; 21 Jun 96 22:11 +0100
Received: from toad by oink with SMTP (PP) id <10408-0@oink>;
         Fri, 21 Jun 1996 22:11:21 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22363;
         Fri, 21 Jun 1996 22:11:13 +0000
Date: Fri, 21 Jun 1996 22:11:13 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: problem with dynamic library loading on Solaris 2.5
X-Sun-Charset: US-ASCII
content-length: 759
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Jon Adams)
>
> >> From: Jon Adams <[email protected]>
> >
> >Try using DBD::Oracle 0.33 (just released). If it core dumps then rebuild
> >the Makefile using   perl Makefile.PL -g   rerun it and get a stack trace
> >from the new core file.
>
> Okay, sorry it took awhile to respond. Had to find a debugger that handled ELF
> correctly. I installed both the new version of the DBI (0.70) and DBD::Oracle
> 0.33. This also dumps core. Here's the trace from ADB...

GDB is the way to go for free debuggers.

> b15 = ef4f6de0 e15 = ef504560 f15 =   6de0 `/lib/libthread.so.1'
> __lwp_kill(0xb,0xeffff820,0xeffff568,0xeffff4a8,0xff00,0xef51829c) + 8

Upgrade to 0.34 (being uploaded now). Announcement to follow.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <12730-0@oink>;
         Fri, 21 Jun 1996 23:35:22 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835396426:18322:1; Fri, 21 Jun 96 23:33:46 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa17096; 21 Jun 96 23:32 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id PAA01898 for dbi-users-real; Fri, 21 Jun 1996 15:27:31 -0700
Precedence: junk
Sender: [email protected]
Received: from gonzo.tamu.edu (gonzo.tamu.edu [165.91.72.200])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id PAA01894
         for <[email protected]>; Fri, 21 Jun 1996 15:27:26 -0700
Received: from gonzo.tamu.edu (localhost [127.0.0.1])
         by gonzo.tamu.edu (8.7.5/8.7.5) with ESMTP id RAA19370;
         Fri, 21 Jun 1996 17:22:18 -0500 (CDT)
From: [email protected] (Philip Kizer)
To: Tim Bunce <[email protected]>
cc: [email protected], [email protected]
Subject: Re: Partial Solution to Solaris 2.5 and Oracle 7.3.2.1
Reply-To: [email protected]
X-Office: 326E Teague
X-Office-Phone: 409.862.4120
X-Organization: TAMU CIS Operating Systems Group, Unix
X-Pgp-Key: Get key 0x11989061 by sending message with subject: "get pgp key"
X-Url: http://gonzo.tamu.edu/
X-Reposting-Policy: redistribute only with permission
X-Face: _B<mzj1U/]zP}h)]5#BSta@y22b`"<mJtYKCPN{4BF^@S<[rOZL*pyXkh3jeqjeJHJrb%ld
       f'dU)hBy9
In-reply-to: Your message of "Fri, 21 Jun 1996 19:33:23 -0000." <[email protected]>
Date: Fri, 21 Jun 1996 17:22:18 -0500
Message-ID: <[email protected]>


Tim Bunce <[email protected]> wrote:
>> However, this seems to work.  So, my fix is to remove -lthread from the
>> Makefile and, for whatever reason, the rmutex error and core dump is
>> gone.  Bad free() is still there.

>No problem. But didn't your ld complain about the -mt flag?
>Umm, perhaps it read it as -m & -t (both valid flags).

Well, we don't have 7.3.x yet, so I cannot try the threading, but that
should be 'cc' instead of 'ld'.  From 'cc -flags':

[ ... ]
-mt             Specify options needed when compiling multi-threaded code
[ ... ]

Sun wants 'cc' to be called as the linker, and it will pass the appropriate
-l<lib> commands to 'ld' to get threading done 'right' whatever their
definition of 'right' may be today.


-philip

____________________________________________________________ Philip Kizer ___
Texas A&M CIS Operating Systems Group, Unix ( 409.862.4120 ) [email protected]
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <12747-0@oink>;
         Fri, 21 Jun 1996 23:40:26 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835396767:21597:0; Fri, 21 Jun 96 23:39:27 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa21028; 21 Jun 96 23:38 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa21233; 21 Jun 96 23:35 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id PAA01914 for dbi-users-real; Fri, 21 Jun 1996 15:30:04 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id PAA01907
         for <[email protected]>; Fri, 21 Jun 1996 15:29:43 -0700
Received: by relay-4.mail.demon.net id aa27924; 21 Jun 96 21:19 GMT
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ad26833; 21 Jun 96 21:09 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa10265; 21 Jun 96 22:05 +0100
Received: from toad by oink with SMTP (PP) id <10352-0@oink>;
         Fri, 21 Jun 1996 22:05:40 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22349;
         Fri, 21 Jun 1996 22:05:34 +0000
Date: Fri, 21 Jun 1996 22:05:34 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: ?:Help
X-Sun-Charset: US-ASCII
content-length: 196
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Dave Anderson - SMCC Reseller SE - Cary NC <[email protected]>
>
> Has anyone every seen this problem.  See attached files

Yes. Upgrade to 0.33 or 0.34 in a day or so.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <12785-0@oink>;
         Fri, 21 Jun 1996 23:50:33 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835397290:27051:1; Fri, 21 Jun 96 23:48:10 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-2.mail.demon.net id aa26542; 21 Jun 96 23:47 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa23042; 21 Jun 96 23:45 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA01403 for dbi-users-real; Fri, 21 Jun 1996 11:42:22 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id LAA01399
         for <[email protected]>; Fri, 21 Jun 1996 11:42:18 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ab11223; 21 Jun 96 18:31 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac22396; 21 Jun 96 19:25 +0100
Received: from toad by oink with SMTP (PP) id <08951-0@oink>;
         Fri, 21 Jun 1996 18:44:05 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA21375;
         Fri, 21 Jun 1996 18:43:58 +0000
Date: Fri, 21 Jun 1996 18:43:58 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: DBD::Ingres
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 687
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Paul Lindner <[email protected]>
>
> I'd be able to donate some time towards implementing/debugging the
> DBD::Ingres driver.

Thanks. Someone else did awhile ago (a few months maybe) but I've
heard nothing since (hence CC'ing this to dbi-users).

> Or are you just waiting to move everything over to an ODBC
> infrastructure?

More a case of being too busy. I'd certainly be very grateful if
someone did a DBD::Ingres, even it it was only just sufficient
to implement an Ingperl emulation layer (a _vital_ thing to have).

It should not be very difficult to do a basic DBD::Ingres.
Find an existing DBD that's based on embedded SQL (if any?)
and copy it.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <13378-0@oink>;
         Sat, 22 Jun 1996 02:41:22 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835400325:28645:3; Sat, 22 Jun 96 00:38:45 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa28006; 22 Jun 96 0:38 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA01595 for dbi-users-real; Fri, 21 Jun 1996 13:05:23 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id NAA01591
         for <[email protected]>; Fri, 21 Jun 1996 13:03:38 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ai05791; 21 Jun 96 19:32 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ae22396; 21 Jun 96 19:25 +0100
Received: from toad by oink with SMTP (PP) id <09035-0@oink>;
         Fri, 21 Jun 1996 19:23:12 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA21721;
         Fri, 21 Jun 1996 19:22:59 +0000
Date: Fri, 21 Jun 1996 19:22:59 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Partial Solution to Solaris 2.5 and Oracle 7.3.2.1
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1232
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: James Taylor <[email protected]>
>
> I had the hardest time getting past the mutex error.
>
> This is what finally worked for me.  Using DBD-Oracle-0.33 and DBI-0.70.
> After running "perl Makefile.PL" the -lthread was already in the
> Makefile.  I replaced all "-lthread" with "-mt".  I did this as a test
> because of Solaris bug #1224467 (found in the Installation Notes for
> Solaris 2.5).  This is what it says:
>
> "If you use -lthread, your program can cause a core dump.
>
> If you have a C++ program that uses threads, do not use -lthread to link
> the program.
>
> Workaround:  Use the -mt option to compile and link your application.
> Using this option ensures that libthread is linked in the correct
> order."

Thanks. This sounds like a good solution but it also sounds like
it's specific to sun's own C compiler (the ld linker doesn't have
a -mt option). Can you get the compiler to show you what args it's
passing to the linker with and without -mt? That way I might be able
to replicate the change via Makefile.PL.

Can someone see if Sun have generated a patch for this bug yet?

> I still have the "Bad free() ignored during global destruction." error.

See my reply to Dan Smith's mail.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <13378-4@oink>;
         Sat, 22 Jun 1996 02:41:47 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835401361:08980:2; Sat, 22 Jun 96 00:56:01 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa07840; 22 Jun 96 0:55 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA02126 for dbi-users-real; Fri, 21 Jun 1996 16:50:04 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA02119
         for <[email protected]>; Fri, 21 Jun 1996 16:49:57 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ag12485; 22 Jun 96 0:19 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac07423; 21 Jun 96 21:54 +0100
Received: from ig.co.uk by oink id <10141-0@oink>;
         Fri, 21 Jun 1996 21:53:44 +0100
Newsgroups: comp.lang.perl.misc
Subject: Re: Web to Oracle interface
References: <[email protected]> <[email protected]> <[email protected]>
Apparently-To: [email protected]
Apparently-To: [email protected]
Date: Fri, 21 Jun 1996 21:53:44 +0100
From: Tim Bunce <[email protected]>
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424
Message-ID: <[email protected]>

In article <[email protected]>,
Bill Rodgers  <[email protected]> wrote:
> Hi Folks, I am in the process of getting DBI and DBD-Oracle working with
> HPUX 10.01 and Oracle 7.3.2.  I found that the latest perl compiled
> perfectly under HPUX 10.01 on the second try.  (what a surprise!)  Now
> it appears that the DBD stuff doesn't want to compile correctly with
> this version of Oracle.  It may be as simple as location of some files.
> Hopefully, I can get it working within next few hours and will let you
> know.
> If anyone has experience with DBI-DBD and HPUX-10, I'd love to hear
> about it.  I've been building a rather simple web application that
> requires a half dozen SQL calls to Oracle and the biggest problem to
> arise has been getting those SQL calls working.  Good luck to all!


In dbi-users you wrote:
> Am attempting to build DBD::Oracle for subject platform.  Found and
> corrected following in Makefile.PL:
>   1. proc.mk has moved, new dir is: $OH/precomp/demo/proc/proc.mk
> Found but not figured out in Makefile.PL:
>   1. How to fix problem in fetch_oci_macros() in Makefile.PL: Line 260,
> $MK{mkver} never assigned a value.  Leads me to believe that the file
> proc.mk has changed.  My knowledge of perl is not extensive enough to
> get me much further soon.  Will continue to hack though.  Suspect expert
> in perl and proc.mk could fix this quick.  If copyrights not a problem,
> would be glad to send you copies of proc.mk and env_precomp.mk (included
> in proc.mk).
>   2. Can not find file libocic.a in $OH/lib or anywhere else in my
> installation.  Currently have version 2.2.2.0.0 of Pro*C installed.
> Using newly compiled and tested latest version of Perl (5.002).

I've not had time to get back to you yet because I've been too busy
(mainly fighting Solaris -lthread and Oracle Bad free() problems for
dbi-users people).

The Makefile.PL prints the message:

       Remember to actually *READ* the README file!

The README asks for some specific information which you've not given.
Especially the version of DBD::Oracle you used. You should be using
at least 0.32. 0.33 is out and worth a try if you have library problems.
0.34 will be out very soon (mainly for Solaris -lthread bugs).

The $MK{mkver} is not set when using Oracle 7.3 and that does not
cause any problems for newer versions of DBD::Oracle so I guess
you've an old one.

The fact it can't find libocic.a is strange (since DBD::Oracle reads
Oracle own makefiles). The fact you've not included the build log
which the README asks for means I can't help you any more right now.

Please get the latest version from ftp://ftp.demon.co.uk/pub/perl/DBD/Oracle
(go for 0.33, or 0.34 if it's there by the time you take a look) and
then send any problems (with full information) to the dbi-users mailing
list and I, or someone else, will help you ASAP.

Tim.
[posted and mailed]
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13376-8@oink>;
         Sat, 22 Jun 1996 02:42:06 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835400759:20286:0; Sat, 22 Jun 96 00:45:59 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa20152; 22 Jun 96 0:45 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa04523; 22 Jun 96 0:44 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA01632 for dbi-users-real; Fri, 21 Jun 1996 13:19:35 -0700
Precedence: junk
Sender: [email protected]
Received: from zeus.carroll.com (zeus.carroll.com [199.224.10.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA01628
         for <[email protected]>; Fri, 21 Jun 1996 13:19:33 -0700
Received: from Phred by zeus.carroll.com with SMTP (8.7.1/0) id QAA04714;
         Fri, 21 Jun 1996 16:22:25 -0400 (edt)
Message-Id: <[email protected]>
X-Sender: [email protected]
X-Mailer: Windows Eudora Version 1.4.3
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Fri, 21 Jun 1996 16:25:01 -0500
To: [email protected]
From: Jon Adams <[email protected]>
Subject: Re: problem with dynamic library loading on Solaris 2.5
Cc: Tim Bunce <[email protected]>


The trace I did looks like it's the thread libraries that are having problems.
The sysadmin did some checking and came up with this...

>Page 24 of "Solaris 2.5 SPARC: Installation
>Notes" (Part # 802-2006-10, Revision A, November 1995).

> +
> + If you use -lthread, your program can cause a core dump.
> +
> + If you have a C++ program that uses threads, do not use -lthread to
> + link the program.
> +
> + Workaround: Use the -mt option to compile and link your application.
> + Using this option ensures that libthread is linked in the correct
> + order.
> +

I don't know if this may be part of the problem, but gcc has no -mt option, and
I'm not sure what the 'correct order' would be for linking with -lthread. If
anyone has the Sunsoft C++ compiler docs running around and can ''reveal the
truth' about -mt, I'd appreciate it.

Jon




[the core file again]

core file = core -- program ``perl'' on platform SUNW,Ultra-1
SIGSEGV: Segmentation Fault

               object: /usr/local/bin/perl
               dump: /export/home/jadams/DBD-Oracle-0.33/core

               (10$d) number base
radix=16 base ten

               ($m) address map
? map
b1 =    10000 e1 =    bd055 f1 =      0 `/usr/local/bin/perl'
b2 =    cd058 e2 =    d53f0 f2 =  ad058 `/usr/local/bin/perl'
b3 = ef782efc e3 = ef78c1a0 f3 =   2efc `/lib/libsocket.so.1'
b4 = ef691890 e4 = ef6d923c f4 =  11890 `/lib/libnsl.so.1'
b5 = ef7b0440 e5 = ef7b0470 f5 =    440 `/lib/libdl.so.1'
b6 = ef7435d8 e6 = ef752414 f6 =   35d8 `/lib/libm.so.1'
b7 = ef596938 e7 = ef5f3ed4 f7 =  16938 `/lib/libc.so.1'
b8 = ef720d84 e8 = ef722438 f8 =    d84 `/lib/libintl.so.1'
b9 = ef660b60 e9 = ef662994 f9 =    b60 `/lib/libmp.so.1'
b10 = ef6425a8 e10 = ef6465c0 f10 =   25a8 `/lib/libw.so.1'
b11 = ef620448 e11 = ef62330c f11 =    448
`/usr/platform/SUNW,Ultra-1/lib/libc_psr.so.1'
b12 = ef55202c e12 = ef55adac f12 =   202c
`/usr/local/lib/perl5/site_perl/sun4-solaris/auto/DBI/DBI.so'
b13 = ef532ca0 e13 = ef539d04 f13 =   2ca0
`./blib/arch/auto/DBD/Oracle/Oracle.so'
b14 = ef2b3018 e14 = ef48de84 f14 =  33018
`/db/oracle/app/oracle/product/7.3.2/lib/libclntsh.so.1.0'
b15 = ef4f6de0 e15 = ef504560 f15 =   6de0 `/lib/libthread.so.1'
/ map
b1 =    cc000 e1 =    d6000 f1 =   1174 `core'
b2 =    d6000 e2 =   18a000 f2 =   b174 `core'
b3 = ee802000 e3 = ee804000 f3 =  bf174 `core'
b4 = ee904000 e4 = ee906000 f4 =  c1174 `core'
b5 = eea06000 e5 = eea08000 f5 =  c3174 `core'
b6 = eeb08000 e6 = eeb0a000 f6 =  c5174 `core'
b7 = eec0a000 e7 = eec0c000 f7 =  c7174 `core'
b8 = eed0c000 e8 = eed0e000 f8 =  c9174 `core'
b9 = eee0e000 e9 = eee10000 f9 =  cb174 `core'
b10 = eef0c000 e10 = eef0e000 f10 =  cd174 `core'
b11 = eef10000 e11 = eef12000 f11 =  cf174 `core'
b12 = ef00e000 e12 = ef010000 f12 =  d1174 `core'
b13 = ef4bc000 e13 = ef4cc000 f13 =  d3174 `core'
b14 = ef4cc000 e14 = ef4d0000 f14 =  e3174 `core'
b15 = ef514000 e15 = ef518000 f15 =  e7174 `core'
b16 = ef518000 e16 = ef51e000 f16 =  eb174 `core'
b17 = ef54a000 e17 = ef54c000 f17 =  f1174 `core'
b18 = ef56c000 e18 = ef56e000 f18 =  f3174 `core'
b19 = ef60c000 e19 = ef616000 f19 =  f5174 `core'
b20 = ef616000 e20 = ef618000 f20 =  ff174 `core'
b21 = ef632000 e21 = ef634000 f21 = 101174 `core'
b22 = ef656000 e22 = ef658000 f22 = 103174 `core'
b23 = ef672000 e23 = ef674000 f23 = 105174 `core'
b24 = ef6ee000 e24 = ef6f8000 f24 = 107174 `core'
b25 = ef6f8000 e25 = ef700000 f25 = 111174 `core'
b26 = ef732000 e26 = ef734000 f26 = 119174 `core'
b27 = ef764000 e27 = ef766000 f27 = 11b174 `core'
b28 = ef79c000 e28 = ef79e000 f28 = 11d174 `core'
b29 = ef7c0000 e29 = ef7c2000 f29 = 11f174 `core'
b30 = ef7f4000 e30 = ef7f6000 f30 = 121174 `core'
b31 = efffc000 e31 = f0000000 f31 = 123174 `core'

               ($v) b: data start, d: data length, e: entry point,
                       m: magic number, t: text length
variables
0 = 10
b = cd058
d = 964c
e = 7f454c46
m = 7f454c46
t = ad055

               ($?) pid, signal, registers, & pgm ctr
no process
SIGSEGV: Segmentation Fault
g0    0x0                               l0      0x0
g1    0xa3                              l1      0x0
g2    0xd5400       nextf+0x8           l2      0x0
g3    0x6c000000                        l3      0x0
g4    0x0                               l4      0x0
g5    0x0                               l5      0x0
g6    0x0                               l6      0x0
g7    0x188000                          l7      0xef5147f8  _GLOBAL_OFFSET_TABLE_
o0    0x0                               i0      0xb
o1    0xb                               i1      0xeffff820
o2    0x0                               i2      0xeffff568
o3    0x0                               i3      0xeffff4a8
o4    0xff00                            i4      0xff00
o5    0xff                              i5      0xef51829c  _siguflags
sp    0xeffff3c8                        fp      0xeffff448
o7    0xef503030    __libthread_segvhdlr+0x110  i7    0xef502418
sigacthandler+0x64c
y     0x0
tstate: 4482001a07  (ccr=0x44, asi=0x82, pstate=0x1a, cwp=0x7)
pstate: ag:0 ie:1 priv:0 am:1 pef:1 mm:0 tle:0 cle:0 mg:0 ig:0
pc    0xef5b890c __lwp_kill+8:  bgeu,a  __lwp_kill + 0x1c
npc   0xef5b8910 __lwp_kill+0xc:                clr     %o0

               ($c) c stack backtrace
__lwp_kill(0xb,0xeffff820,0xeffff568,0xeffff4a8,0xff00,0xef51829c) + 8

               ($q) quit adb
--
Jon Adams                          Promethean Enterprises, Inc.
[email protected]           http://www.pe-netsystems.com
  - Internet and Web Development for the Corporate World -

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13376-25@oink>;
         Sat, 22 Jun 1996 02:43:30 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835404491:22302:1; Sat, 22 Jun 96 01:48:11 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa21884; 22 Jun 96 1:47 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id PAA01875 for dbi-users-real; Fri, 21 Jun 1996 15:16:45 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id PAA01864
         for <[email protected]>; Fri, 21 Jun 1996 15:16:40 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id aw00505; 21 Jun 96 22:41 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa17117; 21 Jun 96 22:33 +0100
Received: from toad by oink with SMTP (PP) id <10605-0@oink>;
         Fri, 21 Jun 1996 22:33:11 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22594;
         Fri, 21 Jun 1996 22:33:04 +0000
Date: Fri, 21 Jun 1996 22:33:04 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DSBD:Ingres
X-Sun-Charset: US-ASCII
content-length: 445
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Chris Perreault <[email protected]>
>
> Hi Tim
>
> I was wondering about the availability of the driver for Ingres. I have a
> project rapidly approaching and I'm going to have to use University
> Ingres :(
>
> Even if it were in really rough form, it would still be a huge help. I
> promise not to bother you with bug reports (unless you want them)

I'm not doing it but somebody else here is talking about it.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13376-28@oink>;
         Sat, 22 Jun 1996 02:43:40 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835404762:24341:1; Sat, 22 Jun 96 01:52:42 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa23663; 22 Jun 96 1:51 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA01822 for dbi-users-real; Fri, 21 Jun 1996 14:58:16 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id OAA01818
         for <[email protected]>; Fri, 21 Jun 1996 14:58:15 -0700
Received: by gw.home.vix.com id OAA17611; Fri, 21 Jun 1996 14:58:12 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id aa28729; 21 Jun 96 21:26 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa15307; 21 Jun 96 22:25 +0100
Received: from toad by oink with SMTP (PP) id <10559-0@oink>;
         Fri, 21 Jun 1996 22:25:41 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22564;
         Fri, 21 Jun 1996 22:25:35 +0000
Date: Fri, 21 Jun 1996 22:25:35 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Announce: DBD::Oracle 0.34
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


Changes in DBD::Oracle 0.34 (Oraperl 1.28), 21st June 1996

   Workaround Solaris 2 bug #1224467 (_rmutex_unlock).
   With many thanks to James Taylor.

   Further additions to the README about Bad free()'s.

   Added 'Bad free()' warning suppression to ora_logon and
   ora_logoff(). Setting the DBD_DUMP environment variable
   will trigger a (handy for me) core dump if a Bad free
   warning is detected.

This version is recommended for those using Solaris 2.5 and/or
Oracle 7.3. Others are welcome to try it. It should be stable.

Uploaded to ftp.demon.co.uk as usual (64825 bytes).
On CPAN in a day or two.

Tim [hoping this release brings some peace!]
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13376-46@oink>;
         Sat, 22 Jun 1996 02:44:46 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835407201:10786:1; Sat, 22 Jun 96 02:33:21 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa10448; 22 Jun 96 2:32 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA02048 for dbi-users-real; Fri, 21 Jun 1996 16:15:58 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA02037
         for <[email protected]>; Fri, 21 Jun 1996 16:15:52 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ac05398; 21 Jun 96 22:30 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa19144; 21 Jun 96 22:41 +0100
Received: from toad by oink with SMTP (PP) id <10658-0@oink>;
         Fri, 21 Jun 1996 22:41:30 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22616;
         Fri, 21 Jun 1996 22:41:23 +0000
Date: Fri, 21 Jun 1996 22:41:23 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Announce: DBD::Oracle 0.33
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 533
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Amy McGovern <[email protected]>
>
>
> You were correct, I was not checking to see if ora_login failed.  However,
> once I added error checking, the following line:
>
> $lda = ora_login($mysid, $myname, $mypass) || die $ora_errstr;
>
> dies with the following error:
> ORA-01019: unable to allocate memory in the user side (DBD: login failed) at
> ./oratest.pl line 10.
>
> unsetting the ORA_NLS32 environment variable lets it work again.  Very odd!

Very odd true, but not a DBD::Oracle problem. Thankfully!

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13376-49@oink>;
         Sat, 22 Jun 1996 02:44:57 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835407556:13005:2; Sat, 22 Jun 96 02:39:16 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa11595; 22 Jun 96 2:38 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA02059 for dbi-users-real; Fri, 21 Jun 1996 16:20:42 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA02055
         for <[email protected]>; Fri, 21 Jun 1996 16:20:35 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ai08394; 21 Jun 96 23:00 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa16954; 21 Jun 96 22:32 +0100
Received: from toad by oink with SMTP (PP) id <10595-0@oink>;
         Fri, 21 Jun 1996 22:32:24 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22590;
         Fri, 21 Jun 1996 22:32:18 +0000
Date: Fri, 21 Jun 1996 22:32:18 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: problem with dynamic library loading on Solaris 2.5
X-Sun-Charset: US-ASCII
content-length: 545
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Jon Adams)
>
> The trace I did looks like it's the thread libraries that are having problems.
>
> I don't know if this may be part of the problem, but gcc has no -mt option, and
> I'm not sure what the 'correct order' would be for linking with -lthread. If
> anyone has the Sunsoft C++ compiler docs running around and can ''reveal the
> truth' about -mt, I'd appreciate it.

DBD:Oracle 0.34 (just released) simply deletes -lthread from the library
list if building on a Solaris 2.5 system. Try that.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <15633-7@oink>;
         Sat, 22 Jun 1996 06:44:08 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835414440:18713:1; Sat, 22 Jun 96 04:34:00 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa18482; 22 Jun 96 4:33 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id UAA02481 for dbi-users-real; Fri, 21 Jun 1996 20:30:29 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id UAA02474;
         Fri, 21 Jun 1996 20:30:24 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ac18006; 22 Jun 96 4:30 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa03662; 22 Jun 96 3:16 +0100
Received: from toad by oink with SMTP (PP) id <15015-0@oink>;
         Sat, 22 Jun 1996 03:16:02 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA23614;
         Sat, 22 Jun 1996 03:15:48 +0000
Date: Sat, 22 Jun 1996 03:15:48 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Announce: DBD::Oracle 0.35 (Phoey! I hate it when that happens!)
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 317
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424

Changes in DBD::Oracle 0.35 (Oraperl 1.28), 21st June 1996

   Fixed broken Solaris 2.5 check in Makefile.PL.
   Added ld path to the log.

A one char typo (osver instead of osvers) means that the
Solaris 2.5 test in verson 0.34 won't work. Sigh.
Fixed in this release.

Uploaded to ftp.demon.co.uk as usual.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <15633-9@oink>;
         Sat, 22 Jun 1996 06:44:18 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835415186:21872:1; Sat, 22 Jun 96 04:46:26 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa21795; 22 Jun 96 4:46 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id UAA02510 for dbi-users-real; Fri, 21 Jun 1996 20:43:37 -0700
Precedence: junk
Sender: [email protected]
Received: from yoda.coe.missouri.edu (rhino.coe.missouri.edu [128.206.76.25])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id UAA02506
         for <[email protected]>; Fri, 21 Jun 1996 20:43:35 -0700
Received: from localhost (pixel@localhost)
         by yoda.coe.missouri.edu (8.7.3/8.7.3) with SMTP id WAA04634
         for <[email protected]>; Fri, 21 Jun 1996 22:43:33 -0500 (CDT)
X-Authentication-Warning: yoda.coe.missouri.edu: pixel owned process doing -bs
Date: Fri, 21 Jun 1996 22:43:33 -0500 (CDT)
From: James Cooper <[email protected]>
To: [email protected]
Subject: Accessing Oracle from another machine
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


Hello,

I know this answer has been answered in the past, but I've never quite
understood the process.

I have Oracle on a machine named yoda.
I have a web server with a CGI script that needs access to yoda's data
(server name is busboy).

Since busboy does not run Oracle, it does not have the SQL*NET installed.
It's actually running a different OS than yoda.

Can I use DBI in a CGI script on busboy to talk to Oracle on yoda?  I
realize that compiling DBD::Oracle requires the Oracle libraries, so I'm
pretty sure I'm out of luck (since I don't have libraries for busboy).

Thanks

-- James

........................................................................
                       James Paul Cooper * Pixel
                        MU College of Education
      [email protected] - http://tiger.coe.missouri.edu/~pixel/
........................................................................

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <22100-18@oink>;
         Mon, 24 Jun 1996 07:08:08 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835563974:03799:1; Sun, 23 Jun 96 22:06:14 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa02653; 23 Jun 96 22:05 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA04783 for dbi-users-real; Sun, 23 Jun 1996 13:46:04 -0700
Precedence: junk
Sender: [email protected]
Received: from colossus.cse.psu.edu ([email protected] [130.203.1.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id NAA04779
         for <[email protected]>; Sun, 23 Jun 1996 13:46:03 -0700
Received: from localhost by colossus.cse.psu.edu with SMTP id <78849>;
         Sun, 23 Jun 1996 16:45:15 -0400
X-Mailer: exmh version 1.6.5 12/11/95
X-uri: <URL:http://www.cse.psu.edu/~groenvel/>
X-work-address: Manufacturing Technology - Systems Development Group PO Box 30
               The Applied Research Lab State College, PA 16804
X-work-phone: +1 814 863 9896 (Voice) +1 814 863 1183 (Fax)
To: [email protected] (Jon Adams)
cc: [email protected]
Subject: Re: problem with dynamic library loading on Solaris 2.5
In-reply-to: Your message of "Fri, 21 Jun 1996 17:25:01 EDT." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Sun, 23 Jun 1996 16:43:36 -0400
From: John D Groenveld <[email protected]>
Message-Id: <[email protected]>

From 'man -s 1 CC' (SPARCompiler C++ 4.1/Solaris 2.5):
  -mt   (Solaris 2.x) Compiles and links a multithreaded  pro-
         gram,  and  passes   - D_REENTRANT to the preprocessor.
         Passed to ld is the command:
         -lsunmath_mt -lm -lC -lw -lthread -lcx -lc
         instead of:
         -lsunmath -lm -lC -lC_mtstubs -lw -lcx -lc
         Note: Use this option, rather than - lthread,  to  link
         with libthread to ensure proper library linking order.

Hope this helps,
John

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <22100-27@oink>;
         Mon, 24 Jun 1996 07:10:47 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835590139:01899:1; Mon, 24 Jun 96 05:22:19 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa01643; 24 Jun 96 5:21 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id VAA05175 for dbi-users-real; Sun, 23 Jun 1996 21:14:19 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id VAA05171
         for <[email protected]>; Sun, 23 Jun 1996 21:14:16 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ae06009; 24 Jun 96 5:14 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa18747; 21 Jun 96 22:40 +0100
Received: from toad by oink with SMTP (PP) id <10637-0@oink>;
         Fri, 21 Jun 1996 22:39:56 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22611;
         Fri, 21 Jun 1996 22:39:49 +0000
Date: Fri, 21 Jun 1996 22:39:49 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DBD-Oracle-0.32: can't link on HP/UX 9.05
X-Sun-Charset: US-ASCII
content-length: 628
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Harald Falkenberg <[email protected]>
>
> Hi,
>
> I tried to link DBD-Oracle-0.32 on a HP/UX 9.05 using Oracle 7.1.6,
> DBI-0.69 and Perl 5.002. Unfortunately I got an error while linking and a list
> of unresolved symbols. But, so far I can see, this searched functions are
> included in the Oracle libs. This is a problem, which I earlier also saw
> with Oraperl.
>
> Is there someone who has experience building DBD on a HP and can tell me
> how to manage this problem?

Please read the README (as Makefile.PL asked you to) and send all
the information it says you should include in problem reports.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <25622-7@oink>;
         Mon, 24 Jun 1996 17:03:07 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835616847:17228:3; Mon, 24 Jun 96 12:47:27 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa16869; 24 Jun 96 12:46 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA05521 for dbi-users-real; Mon, 24 Jun 1996 04:40:06 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id EAA05514
         for <[email protected]>; Mon, 24 Jun 1996 04:40:00 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ah22347; 24 Jun 96 12:35 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ad05603; 24 Jun 96 11:50 +0100
Received: from toad by oink with SMTP (PP) id <24586-0@oink>;
         Mon, 24 Jun 1996 11:43:13 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA29712;
         Mon, 24 Jun 1996 11:43:07 +0000
Date: Mon, 24 Jun 1996 11:43:07 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: problem with dynamic library loading on Solaris 2.5
Cc: [email protected], [email protected]
X-Sun-Charset: US-ASCII
content-length: 1319
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: John D Groenveld <[email protected]>
>
> >From 'man -s 1 CC' (SPARCompiler C++ 4.1/Solaris 2.5):
>    -mt   (Solaris 2.x) Compiles and links a multithreaded  pro-
>           gram,  and  passes   - D_REENTRANT to the preprocessor.
>           Passed to ld is the command:
>           -lsunmath_mt -lm -lC -lw -lthread -lcx -lc
>           instead of:
>           -lsunmath -lm -lC -lC_mtstubs -lw -lcx -lc
>           Note: Use this option, rather than - lthread,  to  link
>           with libthread to ensure proper library linking order.
>
> Hope this helps,

Thanks.  My reading of all this -lthread and mutex error business
(on Solaris 2.5) is that if you dynamically load a shared object into
an executable then either both or neither should be linked with -lthread.

For the time being deleting -lthread from DBD::Oracle's library
list seems to be a viable way forward. That's what v0.34 does.

At some point MakeMaker's liblist function could be modified to
automatically delete -lthread from a liblist if perl is not linked
with -lthread, and automatically add -lthread to a liblist if
perl was built with -lthread but the liblist doesn't include it.

I'm sure they'll be many other threading issues to address in time
but this has certainly bitten us.

CC'd to the makemaker mailing list.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <25622-21@oink>;
         Mon, 24 Jun 1996 17:04:17 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835628148:00477:1; Mon, 24 Jun 96 15:55:48 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa29746; 24 Jun 96 15:54 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA05817 for dbi-users-real; Mon, 24 Jun 1996 07:47:15 -0700
Precedence: junk
Sender: [email protected]
Received: from appliedinfo.com (appliedinfo.com [199.254.240.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id HAA05813
         for <[email protected]>; Mon, 24 Jun 1996 07:47:14 -0700
Received: by appliedinfo.com (SMI-8.6/SMI-SVR4) id KAA24817;
         Mon, 24 Jun 1996 10:39:01 -0400
Date: Mon, 24 Jun 1996 10:39:01 -0400
From: [email protected] (Howard Yin)
Message-Id: <[email protected]>
To: [email protected], [email protected], [email protected]
Subject: DBD-Oracle 0.30, DBI 0.69, Oracle 7.0.16, perl 5.002 Bad Free()

I'm getting this Bad Free() Ignore... error at this platform:
       Sun Sparc Server 2000
       Cygnus  gcc 2.3.3
       perl 5.002
       dbd-oracle 0.30
       dbi 0.69
       oracle 7.0.16

Anyone know how to fix this error?

Howard
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <27045-0@oink>;
         Mon, 24 Jun 1996 22:15:48 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835636386:24423:2; Mon, 24 Jun 96 18:13:06 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa22977; 24 Jun 96 18:12 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA06108 for dbi-users-real; Mon, 24 Jun 1996 10:02:01 -0700
Precedence: junk
Sender: [email protected]
Received: from mccoy.cs.twsu.edu (mccoy.cs.twsu.edu [156.26.10.14])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA06104
         for <[email protected]>; Mon, 24 Jun 1996 10:02:00 -0700
Received: by mccoy.cs.twsu.edu (4.1/SMI-4.1) id AA14405;
         Mon, 24 Jun 96 11:55:42 CDT
Date: Mon, 24 Jun 1996 11:55:42 -0500 (CDT)
From: WebDB Project <[email protected]>
To: "'[email protected]'" <[email protected]>
Subject: Connection to ORacle takes time..
Message-Id: <Pine.SUN.3.91.960624115132.14353A-100000@mccoy>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Hi all,
       I am using dbPerl on SunOs 4.1.3
       DBI::Oracle 0.32
       DBD 0.69
       PERL 5.002
       Oracle 7


       Connecting to oracle takes lot of time. I am working on a web
project and there is a need to connect everytime to oracle and I feel
this to be really slow.. Is there any posibility to speed up the
connecting process.

thankz
-Narayanan

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <27049-14@oink>;
         Mon, 24 Jun 1996 22:17:17 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835641457:16813:0; Mon, 24 Jun 96 19:37:37 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa16257; 24 Jun 96 19:36 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa28309; 24 Jun 96 19:36 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA05848 for dbi-users-real; Mon, 24 Jun 1996 08:04:42 -0700
Precedence: junk
Sender: [email protected]
Received: from mail.industry.net (mail.industry.net [206.150.196.12])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA05844
         for <[email protected]>; Mon, 24 Jun 1996 08:04:41 -0700
Received: from dns2.industry.net (dns2.industry.net [206.150.196.30])
         by mail.industry.net (8.6.9/8.6.8) with SMTP id LAA16517
         for <[email protected]>; Mon, 24 Jun 1996 11:10:27 -0400
Message-ID: <[email protected]>
Date: Sun, 23 Jun 1996 04:33:42 -0400
From: "Rujith S. de Silva" <[email protected]>
Organization: Industry.Net
X-Mailer: Mozilla 2.0 (X11; I; SunOS 4.1.3_U1 sun4m)
MIME-Version: 1.0
To: [email protected]
Subject: Closed connections
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I have an Oracle connection used for retrieving data that may be kept
open for a long period of time, as in several days.  If the database
happens to go down and be brought back up during that time, it looks
like any attempted operations on the connection return either

       ORA-03113: end-of-file on communication channel
or      ORA-03114: not connected to ORACLE

So I can then close the connection and re-open it, and things work fine.
In fact, I can even do a `dummy' operation (like SELECT SYSDATE FROM
DUAL) to see if the connection is still open, before doing any `real'
retrievals.

My question is: is there a better or more `approved' way to find out if
an Oracle connection is still really connected to the database?

Later,
Rujith.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <28321-7@oink>;
         Tue, 25 Jun 1996 00:59:59 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835657609:25800:2; Tue, 25 Jun 96 00:06:49 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa24116; 25 Jun 96 0:05 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA06365 for dbi-users-real; Mon, 24 Jun 1996 12:50:32 -0700
Precedence: junk
Sender: [email protected]
Received: from colossus.cse.psu.edu ([email protected] [130.203.1.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id MAA06361
         for <[email protected]>; Mon, 24 Jun 1996 12:50:30 -0700
Received: from localhost by colossus.cse.psu.edu with SMTP id <78990>;
         Mon, 24 Jun 1996 15:50:12 -0400
X-Mailer: exmh version 1.6.5 12/11/95
X-uri: <URL:http://www.cse.psu.edu/~groenvel/>
X-work-address: Manufacturing Technology - Systems Development Group PO Box 30
               The Applied Research Lab State College, PA 16804
X-work-phone: +1 814 863 9896 (Voice) +1 814 863 1183 (Fax)
To: WebDB Project <[email protected]>
cc: [email protected]
Subject: Re: Connection to ORacle takes time..
In-reply-to: Your message of "Mon, 24 Jun 1996 12:55:42 EDT." <Pine.SUN.3.91.960624115132.14353A-100000@mccoy>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 24 Jun 1996 15:48:32 -0400
From: John D Groenveld <[email protected]>
Message-Id: <[email protected]>

Database connections are always going to be a source of time delay and
overhead. Write a daemon to accept SQL statements from your CGI. You can also
look at Doug MacEachern's Apache modules (http://www.osf.org/~dougm/apache/)
which embeds the Perl interpretter into Apache. Finally, take a look at the
mail archive for this list (somewhere under http://www.hermetica.com/technologi
a/DBI/) -- this is a recurring thread.
John

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <29402-4@oink>;
         Tue, 25 Jun 1996 07:06:14 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835672313:02157:1; Tue, 25 Jun 96 04:11:53 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa01683; 25 Jun 96 4:10 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id TAA07310 for dbi-users-real; Mon, 24 Jun 1996 19:51:55 -0700
Precedence: junk
Sender: [email protected]
Received: from vega.math.ualberta.ca ([email protected] [129.128.88.12])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id TAA07306
         for <[email protected]>; Mon, 24 Jun 1996 19:51:53 -0700
Received: from ellpspace.math.ualberta.ca by vega.math.ualberta.ca
         with SMTP (1.37.109.16/16.2) id AA284850778;
         Mon, 24 Jun 1996 20:46:18 -0600
Received: by ellpspace.math.ualberta.ca (Linux Smail3.1.29.1 #3)
         id m0uYOEc-000MjeC; Mon, 24 Jun 96 20:52 MDT
Message-Id: <[email protected]>
From: [email protected] (Michal Jaegermann)
Subject: Memory leak
To: [email protected] ([email protected])
Date: Mon, 24 Jun 1996 20:52:10 -0600 (MDT)
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text
Content-Length: 3026


I wonder if somebody knows something about the following problem.

Using Oraperl and Perl5 I wrote, on DEC OSF1 Alpha machine, a daemon
which logs once to Oracle, handles database querries which come from
another machine and sends results back.  Everything is fine, roughly,
with an exception of one nasty gotcha.  The daemon is loosing around
150K of memory per incoming request.

I checked my code for possible "lost references" and rewrote it a bit
to make sure that memory is not kept that way.  It did not help.  So I
rewrote the daemon using Perl4 and oraperl 2.4 - which I stil
have. (This turned out to be more laborious than expected; one is
getting used to good things really fast :-).  That new daemon not only
does not have a memory leak, but it also has quite a bit smaller
footprint, which is not without signifince in a program of such
kind. Surprise, as in an old oraperl everything is linked statically.

The daemon does not fork any subprocess to handle incoming connection.
This is done on purpose as to "serialize" responses.  Currently it
seems likely that a queue on a socket should be more than enough to
handle congestions in an incoming traffic.

A kind of a "companion daemon", also written in Perl5 but which runs
on Linux machine and which, obviously, does not use Oraperl, does not
show any memory leaks and in general behaves like one would want it
to.  These two things (oraperl 2.4 and Linux) together seem to
indicate that the problem is indeed inside Oraperl/DBI, but make your
own interpretations.

I have installed currently DBD-Oracle-0.34 and DBI-0.70 but I have
seen exactly the same thing with DBD-Oracle-0.32 and DBI-0.69.
Details of my Perl configuration are included below.

So, the question is - can anybody shed a light on what is going on?
Similar experiences or experiences to the contrary?

  Thanks,
  Michal


Summary of my perl5 (5.0 patchlevel 2 subversion 1) configuration:
 Platform:
   osname=dec_osf, osver=3.2, archname=alpha-dec_osf
   uname='osf1 skogen.env.gov.ab.ca v3.2 148 alpha '
   hint=previous, useposix=true, d_sigaction=define
 Compiler:
   cc='gcc', optimize='-O2', gccversion=2.7.0
   cppflags='-DSTANDARD_C -I/usr/local/include -D__LANGUAGE_C__ -D_NO_PROTO'
   ccflags ='-DSTANDARD_C -I/usr/local/include -D__LANGUAGE_C__ -D_NO_PROTO'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=n, randbits=15
 Linker and Libraries:
   ld='ld', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /usr/shlib /lib /usr/lib /usr/ccs/lib
   libs=-ldbm -lm -lc -lbsd -lPW
   libc=/usr/shlib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-fpic', lddlflags='-shared -expect_unresolved "*" -s -hidden -L/usr/local/lib'

@INC: /usr/local/lib/perl5/alpha-dec_osf/5.00201 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/alpha-dec_osf /usr/local/lib/perl5/site_perl .

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <29402-6@oink>;
         Tue, 25 Jun 1996 07:06:26 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835674762:13133:1; Tue, 25 Jun 96 04:52:42 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa12814; 25 Jun 96 4:51 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id UAA07358 for dbi-users-real; Mon, 24 Jun 1996 20:45:08 -0700
Precedence: junk
Sender: [email protected]
Received: from undergrad.math.uwaterloo.ca ([email protected] [129.97.204.13])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id UAA07354
         for <[email protected]>; Mon, 24 Jun 1996 20:45:01 -0700
Received: from mobius09.math.uwaterloo.ca ([email protected] [129.97.204.29])
         by undergrad.math.uwaterloo.ca (8.7.5/8.7.3) with ESMTP id XAA16922
         for <[email protected]>; Mon, 24 Jun 1996 23:44:56 -0400 (EDT)
From: Amy Fong <[email protected]>
Received: (from afong@localhost) by mobius09.math.uwaterloo.ca (8.7.5/8.7.3)
         id XAA28119 for [email protected]; Mon, 24 Jun 1996 23:44:53 -0400
Message-Id: <[email protected]>
Subject: Please UNSUBSCRIBE
To: [email protected]
Date: Mon, 24 Jun 1996 23:44:51 -0400 (EDT)
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

Please unsubscribe

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <29403-12@oink>;
         Tue, 25 Jun 1996 07:07:06 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835672473:12932:1; Tue, 25 Jun 96 04:14:33 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa12448; 25 Jun 96 4:13 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id UAA07323 for dbi-users-real; Mon, 24 Jun 1996 20:02:48 -0700
Precedence: junk
Sender: [email protected]
Received: from yoda.coe.missouri.edu (rhino.coe.missouri.edu [128.206.76.25])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id UAA07319
         for <[email protected]>; Mon, 24 Jun 1996 20:02:47 -0700
Received: from localhost (pixel@localhost)
         by yoda.coe.missouri.edu (8.7.3/8.7.3) with SMTP id WAA08954
         for <[email protected]>; Mon, 24 Jun 1996 22:02:44 -0500 (CDT)
X-Authentication-Warning: yoda.coe.missouri.edu: pixel owned process doing -bs
Date: Mon, 24 Jun 1996 22:02:44 -0500 (CDT)
From: James Cooper <[email protected]>
To: [email protected]
Subject: Indexed hypermail archive
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


Hello,

I realize that Alligator just updated some of the hypermail archives for
this list, but I just wanted to let everyone know that I've got another
one setup at this URL:

http://www.coe.missouri.edu/~faq/lists/dbiusers/

It has a searchable index that re-indexes 3 times a week.  I've put the
last month or so in, and it's updating itself each time a new message is
posted to the list (thanks to procmail)

I don't mean to step on any toes here, but I did this mainly so I could
quickly search what others have posted before repeating the same question.

I figured others may find this useful as well (would also help Tim ward
off the 10+ posts a week from people who can't figure out how to compile
DBI :-)

Thanks

-- James

BTW, I'm doing this for a few other lists as well, such as mod_perl,
postgres95, and fastcgi.  There're at the same location (just knock off
the dbiusers in the url)

........................................................................
                       James Paul Cooper * Pixel
                        MU College of Education
      [email protected] - http://tiger.coe.missouri.edu/~pixel/
........................................................................

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <01555-0@oink>;
         Tue, 25 Jun 1996 10:55:11 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835691192:06745:2; Tue, 25 Jun 96 09:26:32 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa00264; 25 Jun 96 9:26 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id BAA07671 for dbi-users-real; Tue, 25 Jun 1996 01:10:23 -0700
Precedence: junk
Sender: [email protected]
Received: from spider.ilink.fr (spider.ilink.fr [194.98.79.225])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id BAA07666
         for <[email protected]>; Tue, 25 Jun 1996 01:10:16 -0700
Received: from pentium by spider.ilink.fr
         via SMTP (940816.SGI.8.6.9/940406.SGI.AUTO)
         for <[email protected]> id KAA19188;
         Tue, 25 Jun 1996 10:11:56 +0200
Message-Id: <[email protected]>
X-Sender: [email protected]
X-Mailer: Windows Eudora Light Version 1.5.4 (32)
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Date: Tue, 25 Jun 1996 10:14:27 -0500
To: [email protected]
From: Alexandre COSTE <[email protected]>
Subject: Please UNSUBSCRIBE

Please unsubscribe

Alexandre COSTE - Ing=E9nieur syst=E8me Ilink      - [email protected]
Ilink           - Fournisseur d'acc=E9s Internet - [email protected]
-http://www.ilink.fr

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <01828-2@oink>;
         Tue, 25 Jun 1996 12:24:45 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835701796:14200:1; Tue, 25 Jun 96 12:23:16 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa13580; 25 Jun 96 12:22 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA07793 for dbi-users-real; Tue, 25 Jun 1996 04:11:40 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id EAA07789
         for <[email protected]>; Tue, 25 Jun 1996 04:11:38 -0700
Received: by gw.home.vix.com id EAA12403; Tue, 25 Jun 1996 04:11:21 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ap18626; 25 Jun 96 12:03 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa03416; 25 Jun 96 10:58 +0100
Received: from toad by oink with SMTP (PP) id <01566-0@oink>;
         Tue, 25 Jun 1996 10:58:21 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA04292;
         Tue, 25 Jun 1996 10:58:13 +0000
Date: Tue, 25 Jun 1996 10:58:13 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: HOW TO UNSUBSCRIBE
X-Sun-Charset: ISO-8859-1
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Alexandre COSTE <[email protected]>
> Subject: Please UNSUBSCRIBE
>
> Please unsubscribe
>
> Alexandre COSTE - Ingnieur systme Ilink      - [email protected]
> Ilink           - Fournisseur d'accs Internet - [email protected]
> -http://www.ilink.fr

----- Begin Included Message -----

From [email protected]  Wed Apr 17 07:24:59 1996
Sender: [email protected]
To: Douglas Sur <[email protected]>
Cc: [email protected]
Subject: HOW TO UNSUBSCRIBE
In-Reply-To: Your message of "Tue, 16 Apr 1996 08:51:31 -0100." <[email protected]>
Date: Tue, 16 Apr 1996 11:20:57 -0700
From: Ted Lemon <[email protected]>


Douglas, when you want to unsubscribe from any mailing list on the
internet, you should expect that if you send mail to the -request
address for that list, you will at least get mail back containing
instructions on how to unsubscribe.

For any mailing list, the -request address is always made up of the
mailing list name, followed by -request, followed by the `@'
character, followed by the fully qualified domain name of the machine
on which that list resides.

In the case of [email protected], the mailing list name is
dbi-users, the fully qualified domain name is fugue.com, and thus the
-request address is [email protected].   If you wanted to
unsubscribe to, e.g., [email protected], you would send mail to
[email protected].

This is very consistent across most mailing lists you will encounter,
although sadly there are some mailing lists where the maintainer has
been too lazy to set up -request addresses.  Still, you should try the
-request address first.

Now, having explained all that, in order to avoid being flamed to a
crisp when you get the response from dbi-users-request and realize
that there's another step, I will tell you that what the
dbi-users-request address actually does is to send you instructions
saying, essentially, that you should go to http://www.fugue.com/dbi
with your favourite web browser (I recommend Netscape or Lynx) to
unsubscribe.

                              _MelloN_


----- End Included Message -----

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <02974-0@oink>;
         Tue, 25 Jun 1996 16:32:00 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835711680:04029:2; Tue, 25 Jun 96 15:08:00 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa03397; 25 Jun 96 15:07 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA07961 for dbi-users-real; Tue, 25 Jun 1996 06:55:21 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id GAA07956
         for <[email protected]>; Tue, 25 Jun 1996 06:51:48 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ak12275; 25 Jun 96 14:42 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa05829; 25 Jun 96 12:50 +0100
Received: from toad by oink with SMTP (PP) id <01978-0@oink>;
         Tue, 25 Jun 1996 12:50:30 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA04963;
         Tue, 25 Jun 1996 12:50:21 +0000
Date: Tue, 25 Jun 1996 12:50:21 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Memory leak
X-Sun-Charset: US-ASCII
content-length: 2515
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Michal Jaegermann)
>
> I wonder if somebody knows something about the following problem.
>
> Using Oraperl and Perl5 I wrote, on DEC OSF1 Alpha machine, a daemon
> which logs once to Oracle, handles database querries which come from
> another machine and sends results back.  Everything is fine, roughly,
> with an exception of one nasty gotcha.  The daemon is loosing around
> 150K of memory per incoming request.

That's a ridiculous amount. What does one "incoming request" 'do' that
would involve 150K of memory? Is that approximately how much data is
returned by the query(s)? How many queries are there per incoming request?

> I checked my code for possible "lost references" and rewrote it a bit
> to make sure that memory is not kept that way.  It did not help.  So I
> rewrote the daemon using Perl4 and oraperl 2.4 - which I stil
> have. (This turned out to be more laborious than expected; one is
> getting used to good things really fast :-).  That new daemon not only
> does not have a memory leak

Take that code and add the line

      eval 'use Oraperl; 1' || die $@ if $] >= 5;

and then see if it leaks under Perl5.

> but it also has quite a bit smaller
> footprint, which is not without signifince in a program of such
> kind. Surprise, as in an old oraperl everything is linked statically.

Not really a surprise. Life gets better for Oracle 7.3 but otherwise
you should probably link statically where memory usage and/or startup
time are critical.

> I have installed currently DBD-Oracle-0.34 and DBI-0.70 but I have
> seen exactly the same thing with DBD-Oracle-0.32 and DBI-0.69.
> Details of my Perl configuration are included below.
>
> So, the question is - can anybody shed a light on what is going on?
> Similar experiences or experiences to the contrary?

A handy trick, which has worked well for me in the past, is to let the
program leak till it's very large, then core dump it (kill -QUIT pid).
You can then use strings and/or a binary file viewer (less is good) to
simply look at the bloat in the core file (about 75% of the way through).

You'll have to do this yourself as it's simply not possible for me to
find a leak without a small test case.

*If* you can give me a small test case that demonstrates that my code
leaks I'll fix it at once. Till then you're the only person reporting
a memory leak with the current release and, given the amount, I find
it hard to believe that it's the fault of the DBI or DBD::Oracle.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <02972-4@oink>;
         Tue, 25 Jun 1996 16:32:16 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835711692:15382:2; Tue, 25 Jun 96 15:08:12 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa14896; 25 Jun 96 15:07 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA07974 for dbi-users-real; Tue, 25 Jun 1996 06:57:38 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id GAA07963
         for <[email protected]>; Tue, 25 Jun 1996 06:57:11 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ak12275; 25 Jun 96 14:42 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa05829; 25 Jun 96 12:50 +0100
Received: from toad by oink with SMTP (PP) id <01978-0@oink>;
         Tue, 25 Jun 1996 12:50:30 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA04963;
         Tue, 25 Jun 1996 12:50:21 +0000
Date: Tue, 25 Jun 1996 12:50:21 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Memory leak
X-Sun-Charset: US-ASCII
content-length: 2515
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Michal Jaegermann)
>
> I wonder if somebody knows something about the following problem.
>
> Using Oraperl and Perl5 I wrote, on DEC OSF1 Alpha machine, a daemon
> which logs once to Oracle, handles database querries which come from
> another machine and sends results back.  Everything is fine, roughly,
> with an exception of one nasty gotcha.  The daemon is loosing around
> 150K of memory per incoming request.

That's a ridiculous amount. What does one "incoming request" 'do' that
would involve 150K of memory? Is that approximately how much data is
returned by the query(s)? How many queries are there per incoming request?

> I checked my code for possible "lost references" and rewrote it a bit
> to make sure that memory is not kept that way.  It did not help.  So I
> rewrote the daemon using Perl4 and oraperl 2.4 - which I stil
> have. (This turned out to be more laborious than expected; one is
> getting used to good things really fast :-).  That new daemon not only
> does not have a memory leak

Take that code and add the line

      eval 'use Oraperl; 1' || die $@ if $] >= 5;

and then see if it leaks under Perl5.

> but it also has quite a bit smaller
> footprint, which is not without signifince in a program of such
> kind. Surprise, as in an old oraperl everything is linked statically.

Not really a surprise. Life gets better for Oracle 7.3 but otherwise
you should probably link statically where memory usage and/or startup
time are critical.

> I have installed currently DBD-Oracle-0.34 and DBI-0.70 but I have
> seen exactly the same thing with DBD-Oracle-0.32 and DBI-0.69.
> Details of my Perl configuration are included below.
>
> So, the question is - can anybody shed a light on what is going on?
> Similar experiences or experiences to the contrary?

A handy trick, which has worked well for me in the past, is to let the
program leak till it's very large, then core dump it (kill -QUIT pid).
You can then use strings and/or a binary file viewer (less is good) to
simply look at the bloat in the core file (about 75% of the way through).

You'll have to do this yourself as it's simply not possible for me to
find a leak without a small test case.

*If* you can give me a small test case that demonstrates that my code
leaks I'll fix it at once. Till then you're the only person reporting
a memory leak with the current release and, given the amount, I find
it hard to believe that it's the fault of the DBI or DBD::Oracle.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <02972-11@oink>;
         Tue, 25 Jun 1996 16:32:53 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835714750:14465:2; Tue, 25 Jun 96 15:59:10 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa14027; 25 Jun 96 15:58 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA08068 for dbi-users-real; Tue, 25 Jun 1996 07:45:22 -0700
Precedence: junk
Sender: [email protected]
Received: from vega.math.ualberta.ca ([email protected] [129.128.88.12])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id HAA08064
         for <[email protected]>; Tue, 25 Jun 1996 07:45:20 -0700
Received: from ellpspace.math.ualberta.ca by vega.math.ualberta.ca
         with SMTP (1.37.109.16/16.2) id AA049223568;
         Tue, 25 Jun 1996 08:39:28 -0600
Received: by ellpspace.math.ualberta.ca (Linux Smail3.1.29.1 #3)
         id m0uYZMV-000MjeC; Tue, 25 Jun 96 08:45 MDT
Message-Id: <[email protected]>
From: [email protected] (Michal Jaegermann)
Subject: Re: Memory leak
To: [email protected] (Tim Bunce)
Date: Tue, 25 Jun 1996 08:45:03 -0600 (MDT)
Cc: [email protected]
In-Reply-To: <[email protected]> from "Tim Bunce" at Jun 25, 96 12:50:21 pm
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text
Content-Length: 1274

>
> > The daemon is loosing around
> > 150K of memory per incoming request.
>
> That's a ridiculous amount.

I agree.

> What does one "incoming request" 'do' that
> would involve 150K of memory?

Really no idea.  A request is one line of numbers representing
time intervals.  A response sends backs all data retrieved from a single
table which fall into these intervals.

> Is that approximately how much data is
> returned by the query(s)?

No, an amount of data returned varies.  In limit cases it may range
from 0 (plus small identifying header not coming from a database)
to around 1 Meg in practice (in theory this is limited only by a size
of the table in question).

> How many queries are there per incoming request?

One, but repeated as many times in a loop a number of time intervals
present (typically five).

>
> Take that code and add the line
>
>        eval 'use Oraperl; 1' || die $@ if $] >= 5;
>
> and then see if it leaks under Perl5.

Ok, I will try that, although the biggest visible differences are
in a code which handles sockets.

> A handy trick, which has worked well for me in the past, is to let the
> program leak till it's very large, then core dump it (kill -QUIT pid).

Thanks for the tip.  I will try that and we will see.

  Michal
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <04763-3@oink>;
         Tue, 25 Jun 1996 22:38:03 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835732156:12658:1; Tue, 25 Jun 96 20:49:16 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa12354; 25 Jun 96 20:48 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa11957; 25 Jun 96 20:47 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA08222 for dbi-users-real; Tue, 25 Jun 1996 09:14:17 -0700
Precedence: junk
Sender: [email protected]
Received: from dhirst.demon.co.uk ([email protected] [158.152.72.211])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA08218
         for <[email protected]>; Tue, 25 Jun 1996 09:13:54 -0700
Received: from DAVIDB.demon.co.uk ([192.168.2.10])
         by dhirst.demon.co.uk (8.6.12/8.6.9) with SMTP id JAA15787
         for <[email protected]>; Tue, 25 Jun 1996 09:11:35 GMT
Received: by DAVIDB.demon.co.uk with Microsoft Mail
         id <[email protected]>;
         Tue, 25 Jun 1996 09:11:54 +-100
Message-ID: <[email protected]>
From: David Blackburn <[email protected]>
To: "'[email protected]'" <[email protected]>
Subject: unsubscribe
Date: Tue, 25 Jun 1996 09:11:48 +-100
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

unsubscribe

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <04763-10@oink>;
         Tue, 25 Jun 1996 22:38:29 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835736464:28378:1; Tue, 25 Jun 96 22:01:04 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa28041; 25 Jun 96 22:00 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA08449 for dbi-users-real; Tue, 25 Jun 1996 11:04:12 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id LAA08445
         for <[email protected]>; Tue, 25 Jun 1996 11:04:08 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ah05581; 25 Jun 96 17:07 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa02517; 25 Jun 96 10:55 +0100
Received: from toad by oink with SMTP (PP) id <01521-0@oink>;
         Tue, 25 Jun 1996 10:52:16 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA04267;
         Tue, 25 Jun 1996 10:52:08 +0000
Date: Tue, 25 Jun 1996 10:52:08 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Indexed hypermail archive
X-Sun-Charset: US-ASCII
content-length: 882
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: James Cooper <[email protected]>
>
> Hello,
>
> I realize that Alligator just updated some of the hypermail archives for
> this list, but I just wanted to let everyone know that I've got another
> one setup at this URL:
>
> http://www.coe.missouri.edu/~faq/lists/dbiusers/
>
> It has a searchable index that re-indexes 3 times a week.  I've put the
> last month or so in, and it's updating itself each time a new message is
> posted to the list (thanks to procmail)
>
> I don't mean to step on any toes here, but I did this mainly so I could
> quickly search what others have posted before repeating the same question.
>
> I figured others may find this useful as well (would also help Tim ward
> off the 10+ posts a week from people who can't figure out how to compile
> DBI :-)

Thanks!

Aligator, can you add this (prominently) to your DBI pages?

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <06139-0@oink>;
         Wed, 26 Jun 1996 01:51:07 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835750116:14267:2; Wed, 26 Jun 96 01:48:36 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa13943; 26 Jun 96 1:48 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA08858 for dbi-users-real; Tue, 25 Jun 1996 13:57:00 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA08853
         for <[email protected]>; Tue, 25 Jun 1996 13:56:58 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id VAA15390; Tue, 25 Jun 1996 21:44:55 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: Indexed hypermail archive
To: [email protected] (Tim Bunce)
Date: Tue, 25 Jun 1996 21:44:55 +0100 (BST)
Cc: [email protected], [email protected]
In-Reply-To: <[email protected]> from "Tim Bunce" at Jun 25, 96 10:52:08 am
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 896


> > I don't mean to step on any toes here, but I did this mainly so I could
> > quickly search what others have posted before repeating the same question.
> >
> > I figured others may find this useful as well (would also help Tim ward
> > off the 10+ posts a week from people who can't figure out how to compile
> > DBI :-)
>
> Thanks!
>
> Aligator, can you add this (prominently) to your DBI pages?

Yup, the hypermail archive is permanently moving from Hermetica as soon
as I get the nod from James that he's got all the old stuff I've sent him
fully online. Takes a lot of weight off Hermetica's server, and frees me up
a bit to get on with squashing bugs and writing the damn FAQ.

> Tim.

--
Alligator Descartes     |       Nexus - Multi-User Virtual Reality
Anathematician          |         http://www.hermetica.com/nexus
[email protected]  | Muse of the Day: "Deathchase" - Mervyn Estcourt
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <06958-14@oink>;
         Wed, 26 Jun 1996 07:08:54 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835754642:03718:2; Wed, 26 Jun 96 03:04:02 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa03434; 26 Jun 96 3:03 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA09276 for dbi-users-real; Tue, 25 Jun 1996 16:50:34 -0700
Precedence: junk
Sender: [email protected]
Received: from idiom.com ([email protected] [140.174.82.4])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA09272
         for <[email protected]>; Tue, 25 Jun 1996 16:50:33 -0700
Received: (from muir@localhost) by idiom.com (8.7.5/8.6.12) id QAA17856
         for [email protected]; Tue, 25 Jun 1996 16:50:25 -0700 (PDT)
Date: Tue, 25 Jun 1996 16:50:25 -0700 (PDT)
From: David Muir Sharnoff <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: How do you tell the difference between 0 rows modifed and an error?



I've read http://www.hermetica.com/technologia/DBI/ a dozen time and
either I'm blind or it doesn't say it.

In execute(), how do you tell the difference between an error and
zero rows modifed.

What I would expect is that with an error, the return code should
be undef and with zero rows selected/modified/deleted, etc, it
should be 0.

As far as I can tell, the documentation doesn't say.

Is there some other way to tell?  Look at $DBI::errstr perhaps?

Thanks,

-Dave

P.S.  I wrote a DBD for a mSQL-like database that a friend of mine
wrote.  He hasn't released the database yet so there's no point in
me releasing the DBD either.  Besides it's mostly a copy of DBD-mSQL.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <06958-21@oink>;
         Wed, 26 Jun 1996 07:10:09 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835761365:04772:1; Wed, 26 Jun 96 04:56:05 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa04699; 26 Jun 96 4:55 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id UAA09583 for dbi-users-real; Tue, 25 Jun 1996 20:42:44 -0700
Precedence: junk
Sender: [email protected]
Received: from algw2.att.com (algw2.att.com [205.147.213.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id UAA09579
         for <[email protected]>; Tue, 25 Jun 1996 20:42:42 -0700
From: [email protected]
Received: from ihgp12.ih.lucent.com by alig2.att.att.com (SMI-8.6/EMS-1.2 sol2)
         id LAA21487; Tue, 25 Jun 1996 11:01:06 -0400
Received: by ihgp12.ih.lucent.com (SMI-8.6/EMS-L sol2) id JAA17077;
         Tue, 25 Jun 1996 09:55:13 -0500
Received: from fvigilpc.ih.att.com by ihgp12.ih.lucent.com (SMI-8.6/EMS-L sol2)
         id JAA17071; Tue, 25 Jun 1996 09:55:10 -0500
Received: by fvigilpc.ih.att.com with Microsoft Mail
         id <[email protected]>;
         Tue, 25 Jun 1996 09:54:35 -0500
Message-ID: <[email protected]>
Original-From: Fernando Vigil <[email protected]>
To: "'[email protected]'" <[email protected]>
Date: Tue, 25 Jun 1996 09:54:33 -0500
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable


Any suggestions?

I am trying to build a web interface to an Ingres database.  I know that =
the Ingres extension for DBperl is not out yet.  Is there anything else =
for Perl5 or should I just use Perl4 and ingperl?  Any feedback is =
appreciated.


                                       Fernando Vigil
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <10444-0@oink>;
         Wed, 26 Jun 1996 14:38:07 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835792004:14917:2; Wed, 26 Jun 96 13:26:44 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa12765; 26 Jun 96 13:25 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id FAA09919 for dbi-users-real; Wed, 26 Jun 1996 05:05:50 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id FAA09915
         for <[email protected]>; Wed, 26 Jun 1996 05:05:17 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ac29574; 26 Jun 96 13:05 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac08339; 26 Jun 96 12:49 +0100
Received: from toad by oink with SMTP (PP) id <10063-0@oink>;
         Wed, 26 Jun 1996 11:54:57 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA15454;
         Wed, 26 Jun 1996 11:54:50 +0000
Date: Wed, 26 Jun 1996 11:54:50 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: How do you tell the difference between 0 rows modifed and an error?
X-Sun-Charset: US-ASCII
content-length: 623
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: David Muir Sharnoff <[email protected]>
>
> I've read http://www.hermetica.com/technologia/DBI/ a dozen time and
> either I'm blind or it doesn't say it.
>
> In execute(), how do you tell the difference between an error and
> zero rows modifed.
>
> What I would expect is that with an error, the return code should
> be undef and with zero rows selected/modified/deleted, etc, it
> should be 0.

Very close:

       error   = undef         (false)
       0       = 0E0           (zero but true)
       1       = 1             (true)
       ...

> As far as I can tell, the documentation doesn't say.
>
> Is there some other way to tell?  Look at $DBI::errstr perhaps?

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <11123-0@oink>;
         Wed, 26 Jun 1996 16:25:00 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835802617:21714:1; Wed, 26 Jun 96 16:23:37 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa19848; 26 Jun 96 16:22 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA10076 for dbi-users-real; Wed, 26 Jun 1996 08:11:18 -0700
Precedence: junk
Sender: [email protected]
Received: from metric.inch.com ([email protected] [204.178.32.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA10072
         for <[email protected]>; Wed, 26 Jun 1996 08:11:17 -0700
Received: (from jw@localhost) by metric.inch.com (8.6.12/8.6.12) id LAA15983
         for [email protected]; Wed, 26 Jun 1996 11:21:41 -0400
From: Jeff Milton <[email protected]>
Message-Id: <[email protected]>
Subject: DBI::Postgres95
To: [email protected]
Date: Wed, 26 Jun 1996 11:21:40 -0400 (EDT)
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 169

I have looked for a DBI for Postgres95.  Befire I waste any more time
searching for one, can someone confirm that one exists, of meybe even where
to get it?

Thanks,
jw
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <13619-4@oink>;
         Thu, 27 Jun 1996 07:06:44 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835814849:24081:1; Wed, 26 Jun 96 19:47:29 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa23104; 26 Jun 96 19:46 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA10171 for dbi-users-real; Wed, 26 Jun 1996 08:53:32 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id IAA10167
         for <[email protected]>; Wed, 26 Jun 1996 08:53:17 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id do26736; 26 Jun 96 14:33 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab08339; 26 Jun 96 12:49 +0100
Received: from toad by oink with SMTP (PP) id <10055-0@oink>;
         Wed, 26 Jun 1996 11:52:09 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA15357;
         Wed, 26 Jun 1996 11:52:02 +0000
Date: Wed, 26 Jun 1996 11:52:02 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
X-Sun-Charset: US-ASCII
content-length: 348
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected]
>
> Any suggestions?
>
> I am trying to build a web interface to an Ingres database.  I know that the Ingres extension for DBperl is not out yet.  Is there anything else for Perl5


Some talk, no action yet.

> or should I just use Perl4 and ingperl?  Any feedback is appreciated.

Perl4 ingperl for now.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <13619-7@oink>;
         Thu, 27 Jun 1996 07:07:20 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835816035:04791:1; Wed, 26 Jun 96 20:07:15 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa04500; 26 Jun 96 20:06 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA10613 for dbi-users-real; Wed, 26 Jun 1996 11:56:15 -0700
Precedence: junk
Sender: [email protected]
Received: from algw2.att.com (algw2.att.com [205.147.213.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id LAA10609
         for <[email protected]>; Wed, 26 Jun 1996 11:56:14 -0700
From: [email protected]
Received: from ihgp12.ih.lucent.com by alig2.att.att.com (SMI-8.6/EMS-1.2 sol2)
         id LAA16750; Wed, 26 Jun 1996 11:23:37 -0400
Received: by ihgp12.ih.lucent.com (SMI-8.6/EMS-L sol2) id KAA01122;
         Wed, 26 Jun 1996 10:09:10 -0500
Received: from fvigilpc.ih.att.com by ihgp12.ih.lucent.com (SMI-8.6/EMS-L sol2)
         id KAA01109; Wed, 26 Jun 1996 10:09:07 -0500
Received: by fvigilpc.ih.att.com with Microsoft Mail
         id <[email protected]>;
         Wed, 26 Jun 1996 10:08:34 -0500
Message-ID: <[email protected]>
Original-From: Fernando Vigil <[email protected]>
To: "'[email protected]'" <[email protected]>
Subject: Ingperl and Perl5
Date: Wed, 26 Jun 1996 10:08:32 -0500
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Does anyone know of the problems associated with using ingperl under =
Perl5?  I'm running Perl5 but need inperl and do not want to move back =
to Perl4.  Can it be done without problems?  Any responses appreciated?

                                                              Fernando
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <13619-14@oink>;
         Thu, 27 Jun 1996 07:10:07 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835816884:12670:2; Wed, 26 Jun 96 20:21:24 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa11791; 26 Jun 96 20:20 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA10200 for dbi-users-real; Wed, 26 Jun 1996 09:01:21 -0700
Precedence: junk
Sender: [email protected]
Received: from idiom.com ([email protected] [140.174.82.4])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA10196
         for <[email protected]>; Wed, 26 Jun 1996 09:01:18 -0700
Received: (from muir@localhost) by idiom.com (8.7.5/8.6.12) id IAA25271;
         Wed, 26 Jun 1996 08:58:22 -0700 (PDT)
Date: Wed, 26 Jun 1996 08:58:22 -0700 (PDT)
From: David Muir Sharnoff <[email protected]>
Message-Id: <[email protected]>
To: Tim Bunce <[email protected]>
Cc: [email protected]
Subject: Re: Re: How do you tell the difference between 0 rows modifed and an
        error?

* Very close:
*
*       error   = undef         (false)
*       0       = 0E0           (zero but true)
*       1       = 1             (true)
*       ...

That's very clever.  Thanks for the info.

-Dave
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13620-11@oink>;
         Thu, 27 Jun 1996 07:12:34 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835830978:21951:0; Thu, 27 Jun 96 00:16:18 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa21456; 27 Jun 96 0:15 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa04241; 27 Jun 96 0:15 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA10696 for dbi-users-real; Wed, 26 Jun 1996 13:25:30 -0700
Precedence: junk
Sender: [email protected]
Received: from yoda.coe.missouri.edu (rhino.coe.missouri.edu [128.206.76.25])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA10692
         for <[email protected]>; Wed, 26 Jun 1996 13:25:28 -0700
Received: from localhost (pixel@localhost)
         by yoda.coe.missouri.edu (8.7.3/8.7.3) with SMTP id PAA14707;
         Wed, 26 Jun 1996 15:25:20 -0500 (CDT)
X-Authentication-Warning: yoda.coe.missouri.edu: pixel owned process doing -bs
Date: Wed, 26 Jun 1996 15:25:19 -0500 (CDT)
From: James Cooper <[email protected]>
To: Jeff Milton <[email protected]>
cc: [email protected]
Subject: Re: DBI::Postgres95
In-Reply-To: <[email protected]>
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

> I have looked for a DBI for Postgres95.  Befire I waste any more time
> searching for one, can someone confirm that one exists, of meybe even where
> to get it?

Currently no driver exists for Postgres95, although one is planned for
development this summer.  If you want, email Paul Walmsley
([email protected]), and tell him you're interested in hearing
about updates.

Otherwise stick with the pg95perl interface for Postgres.  It works just
fine.  But I use Oracle and Postgres, and would like to have a common
interface for both.

Paul is a *bright* fellow, and knows his way around Postgres95 pretty well
(he did the IRIX port).  So I expect a driver sometime this summer.

Cheers,

-- James

........................................................................
                       James Paul Cooper * Pixel
                        MU College of Education
      [email protected] - http://tiger.coe.missouri.edu/~pixel/
........................................................................

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13689-14@oink>;
         Thu, 27 Jun 1996 07:13:26 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835834311:17714:3; Thu, 27 Jun 96 01:11:51 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa17366; 27 Jun 96 1:11 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA11037 for dbi-users-real; Wed, 26 Jun 1996 17:05:50 -0700
Precedence: junk
Sender: [email protected]
Received: from algw2.att.com (algw2.att.com [205.147.213.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id RAA11033
         for <[email protected]>; Wed, 26 Jun 1996 17:05:49 -0700
From: [email protected]
Received: from ihgp12.ih.lucent.com by alig2.att.att.com (SMI-8.6/EMS-1.2 sol2)
         id SAA21920; Wed, 26 Jun 1996 18:04:28 -0400
Received: by ihgp12.ih.lucent.com (SMI-8.6/EMS-L sol2) id QAA05832;
         Wed, 26 Jun 1996 16:54:18 -0500
Received: from fvigilpc.ih.att.com by ihgp12.ih.lucent.com (SMI-8.6/EMS-L sol2)
         id QAA05827; Wed, 26 Jun 1996 16:54:15 -0500
Received: by fvigilpc.ih.att.com with Microsoft Mail
         id <[email protected]>;
         Wed, 26 Jun 1996 16:53:42 -0500
Message-ID: <[email protected]>
Original-From: Fernando Vigil <[email protected]>
To: "'[email protected]'" <[email protected]>
Date: Wed, 26 Jun 1996 16:53:40 -0500
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

A couple of questions:

Is DBperl by itself able to communicate with Ingres and other databases =
or are database specific modules absolutely necessary?

Basically, is my only option if I want to write Perl scripts for an =
Ingres database to use Perl4 and ingperl? or can I use some other =
combination the uses Perl5?

Thanks,

Fernando
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <13689-23@oink>;
         Thu, 27 Jun 1996 07:15:00 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835842893:00344:1; Thu, 27 Jun 96 03:34:53 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa28838; 27 Jun 96 3:33 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id TAA11202 for dbi-users-real; Wed, 26 Jun 1996 19:23:18 -0700
Precedence: junk
Sender: [email protected]
Received: from ns.cdv.com (cdv.com [206.105.122.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id TAA11198
         for <[email protected]>; Wed, 26 Jun 1996 19:23:16 -0700
Received: from local by ns.cdv.com via SMTP (940816.SGI.8.6.9/940406.SGI.AUTO)
         id WAA19033; Wed, 26 Jun 1996 22:43:24 -0700
Message-ID: <[email protected]>
Date: Wed, 26 Jun 1996 22:24:50 -0400
From: Syme Kutz <[email protected]>
Organization: Cyber Concepts
X-Mailer: Mozilla 3.0b4Gold (WinNT; I)
MIME-Version: 1.0
To: [email protected]
CC: [email protected]
Subject: install trouble with DBI and irix 5.3
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Here Is what I'm getting and as I'm new to this little of it makes sence to me so any help would be
appreciated. I was having trouble using the standard cc that came with Irix so I got and made gcc and
rebuilt perl with that the rebuilt DBI both sucessfully but I get stuck here. I'm not sure if it could be
a Pro C problem.... Lemme know what ya think...

Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=irix, osver=5, archname=IP22-irix
   uname='irix ns 5.3 11091812 ip22 mips '
   hint=recommended, useposix=true
 Compiler:
   cc='gcc', optimize='-O', gccversion=2.7.2
   cppflags='-D_BSD_TYPES -I/usr/local/include -DLANGUAGE_C'
   ccflags ='-D_BSD_TYPES -I/usr/local/include -DLANGUAGE_C'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='ld', ldflags =' -L/usr/local/lib'
   libpth=/usr/lib /usr/oracle/lib /usr/local/lib
   libs=-lsun -lm -lc -lcrypt -lbsd -lPW
   libc=/usr/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-fpic', lddlflags='-shared -L/usr/local/lib'

@INC: /usr/local/lib/perl5/IP22-irix/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/IP22-irix
/usr/local/lib/perl5/site_perl .
###############################################################################################
OUTPUT FROM make test TEST_VERBOSE=1

PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5/IP22-irix/5.002
-I/usr/local/lib/perl5 -e 'use Test::Harness qw(&runtests $verbose); $verbose=1; runtests @ARGV;' t/*.t
t/base..............install_driver(Oracle) failed: Can't load './blib/arch/auto/DBD/Oracle/Oracle.so' for
module DBD::Oracle: 18938:/usr/local/bin/perl: rld: Fatal Error: unresolvable symbol in
/blib/arch/auto/DBD/Oracle/Oracle.so: licbas at /usr/local/lib/perl5/DynaLoader.pm line 140.

at blib/lib/DBD/Oracle.pm line 24
       DBI::install_driver called at t/base.t line 16
1..5
ok 1
ok 2
ok 3
FAILED tests 4-5
       Failed 2/5 tests, 60.00% okay
       Test returned status 2 (wstat 512)
Failed 1 test script, 0.00% okay. 2/5 subtests failed, 60.00% okay.
*** Error code 2 (bu21)
############################################################################

/usr/oracle/proc/lib/proc.mk

#
# $Header: proc.mk.pp 1.4 95/04/26 10:28:11 jboyce Osd<unix> $ proc.mk.pp
#
#
# proc.mk - Command file for "make" to compile and load Pro*C 2.0 programs.
#
# Usage to build any program:
#       make -f proc.mk EXE=xyz OBJS="a.o b.o c.o"
#
# Special syntax to link demo program:
#       make -f proc.mk sample1
#
# Special syntax to link and install proc executable:
#       make -f proc.mk install
#
# NOTE:   ORACLE_HOME must be either:
#               . set in the user's environment
#               . passed in on the command line
#               . defined in a modified version of this makefile
#
# Pro*C programs are assumed to have the extension ".pc"
#

SHELL=/bin/sh

LIBHOME=$(ORACLE_HOME)/lib

I_SYM= -I

CC=cc

CFLAGS=-I. -O

CFLAGS=-I. -O `if [ -d /usr/lib/cmplrs ] ; then echo "-systype svr4 -KPIC" ; fi `

LDFLAGS=-L$(LIBHOME) `if [ -x /usr/necbin/uversion ] ; then echo "-Zcommon_pic -znodefs" ; \
                     elif [ -d /usr/lib/cmplrs ] ; then echo "-systype svr4" ; fi `

ARLOCAL=

AR=ar $(ARLOCAL)
ARCREATE=ar cr$(ARLOCAL)
ARDELETE=ar d$(ARLOCAL)
ARREPLACE=ar r$(ARLOCAL)

ECHO=$(ORACLE_HOME)/bin/echodo

OTHERLIBS=`cat $(ORACLE_HOME)/rdbms/lib/sysliblist` $(MLSLIBS)

LLIBPSO=

LLIBPSO=`cat $(ORACLE_HOME)/rdbms/lib/psoliblist`

XLIBS=-lXaw -lXmu -lXext -lXt -lX11 -lm

XLIBHOME = /usr/lib
MOTIFLIBHOME = /usr/lib

MOTIFLIBS = -L$(MOTIFLIBHOME) -lXm -L$(XLIBHOME) -lXt -lX11 -lm

MOTIFLIBS = -L$(MOTIFLIBHOME) -lXm -L$(XLIBHOME) -lXt -lX11 -lgen -lm

CLIBS=$(OTHERLIBS)

SO=a

NLSRTLHOME= $(SRCHOME)/nlsrtl3

NLSRTLLIB= -lnlsrtl3.a

LIBNLSRTL= $(LIBHOME)/libnlsrtl3.$(SO)
LLIBNLSRTL=-lnlsrtl3

NLSRTLINC = $(I_SYM)$(NLSRTLHOME)/public $(I_SYM)$(NLSRTLHOME)/sosd/public \
$(I_SYM)$(NLSRTLHOME)/include $(I_SYM)$(NLSRTLHOME)/src/nlsdata

NLSRTLPUBLIC = $(I_SYM)$(NLSRTLHOME)/public \
$(I_SYM)$(NLSRTLHOME)/sosd/public $(STDINCLUDE)

COREHOME= $(SRCHOME)/oracore3

LIBCORE= $(LIBHOME)/libcore3.$(SO)
LIBCV6= $(LIBHOME)/libc3v6.$(SO)
LLIBCORE= -lcore3
LLIBCV6= -lc3v6

CORELIBD= $(LIBNLSRTL) $(LIBCV6) $(LIBCORE)
CORELIBS= $(LLIBNLSRTL) $(LLIBCV6) $(LLIBCORE) $(LLIBNLSRTL) \
       $(LLIBCCOREFUND) $(LLIBCORE)

LIBNETWORK= $(LIBHOME)/libsqlnet.a
NTCONTAB= $(LIBHOME)/ntcontab.o

TNSLIBS= -lsqlnet
TNSLIBD= $(LIBSQLNET)

LIBSQLNET=$(LIBHOME)/libsqlnet.a
LLIBSQLNET=-lsqlnet
NETLIBD= $(OSNTAB) $(LIBSQLNET)
NETLIBS=  $(OSNTAB) $(LLIBSQLNET)

OSNTAB= $(LIBHOME)/osntab.o
OSNTABST= $(LIBHOME)/osntabst.o
LIBSQLNET= $(LIBHOME)/libsqlnet.a

SQLNETLIBS= -lsqlnet

LIBPLS= $(LIBHOME)/libpls.a
LLIBPLS= -lpls

ROSHOME= $(TK2HOME)/ros

ROSPUBLIC= $(I_SYM)$(ROSHOME)/include
ROSDEFS= $(I_SYM)$(ROSHOME)/defs

LIBROS= $(LIBHOME)/libros.a
LLIBROS= -lros

LIBTK2C= $(LIBHOME)/libtk2c.a
LLIBTK2C= -ltk2c
LIBTK2M= $(LIBHOME)/libtk2m.a
LLIBTK2M= -ltk2m

LIBQAP= $(LIBHOME)/libqapstub.a
LLIBQAP= -lqapstub

LIBTK2P= $(LIBHOME)/libtk2p.a
LLIBTK2P= -ltk2p
LIBTK2UC= $(LIBHOME)/libuc.a
LLIBTK2UC= -luc
LIBTK2REM= $(LIBHOME)/librem.a
LLIBTK2REM= -lrem
LIBTK2OT= $(LIBHOME)/libot.a
LLIBTK2OT= -lot
LIBTK2OTX= $(LIBHOME)/libotx.a

LLIBTK2OTX= -lotx

LIBTK2UT= $(LIBHOME)/libut.a

LLIBTK2UT= -lut

LIBTK2SL= $(LIBHOME)/libsl.a
LLIBTK2SL= -lsl
LIBTK2RE= $(LIBHOME)/libre.a

LLIBTK2RE= -lre

TK2UICLIBD= $(LIBTK2RE) $(LIBTK2UC) $(LIBTK2C) $(LIBTK2OT) $(LIBTK2REM) \
       $(LIBROS) $(LIBTK2P) $(LIBTK2UT) $(LIBTK2SL)

TK2UICLIBS= $(DYNAMIC_ON_TK2) \
       $(LLIBTK2SL) $(LLIBTK2RE) $(LLIBTK2UC) $(LLIBTK2C) $(LLIBTK2OT) \
       $(LLIBTK2RE) $(LLIBTK2REM) $(LLIBROS) $(LLIBTK2C) $(LLIBTK2P) \
       $(LLIBTK2C) $(LLIBTK2P) -lm $(LLIBTK2REM) $(LLIBTK2C) $(LLIBTK2OT) \
       $(LLIBTK2UT) $(LLIBTK2UC) $(LLIBTK2SL) $(LLIBTK2C) \
       $(DYNAMIC_OFF_TK2)

TK2UIMLIBD= $(LIBTK2OT) $(LIBTK2RE) $(LIBTK2UC) $(LIBTK2REM) $(LIBROS) \
       $(LIBTK2M) $(LIBQAP) $(LIBTK2P) $(LIBTK2UT) $(LIBTK2SL)

TK2UIMLIBS= $(DYNAMIC_ON_TK2) \
       $(LLIBTK2OT) $(LLIBTK2RE) $(LLIBTK2UC) $(LLIBTK2REM) $(LLIBROS) \
       $(LLIBTK2M) $(LLIBQAP) $(LLIBTK2P) $(LLIBTK2M) $(LLIBTK2P) \
       $(LLIBTK2M) $(LLIBTK2OT) $(LLIBTK2UT) $(LLIBTK2SL) $(DYNAMIC_OFF_TK2)

PLSPECFILES=`cat $(ORACLE_HOME)/rdbms/lib/knlsyslibs`

LIBOCIC= $(LIBHOME)/libocic.a
LLIBOCIC= -locic

TTLIBD= $(NETLIBD) $(LIBORA) $(CORELIBD)

TTLIBS= $(NETLIBS) $(LLIBORA) $(LLIBSQLNET) $(LLIBORA) $(LIBPLSHACK) $(CORELIBS) $(LDLIBS) $(CLIBS)

STLIBDNOPLS= $(OSNTABST) $(CONFIG) $(CORELIBD) $(NETLIBD) $(RDBMSLIBDNOPLS)
STLIBD= $(OSNTABST) $(CONFIG) $(CORELIBD) $(NETLIBD) $(RDBMSLIBD)
STLIBSNOPLS= $(OSNTABST) $(CONFIG) $(RDBMSLIBSNOPLS) $(RDBMSLIBSNOPLS) \
$(PLSPECFILES) $(LLIBSQLNET) $(LLIBORA) $(LLIBSQLNET) $(CORELIBS) $(LLIBPSO) \
$(LDLIBS) $(CLIBS)
STLIBS= $(OSNTABST) $(CONFIG) $(RDBMSLIBS) $(RDBMSLIBS) $(PLSPECFILES) \
$(LLIBSQLNET) $(LLIBORA) $(LLIBSQLNET) $(CORELIBS) $(LLIBPSO) $(LDLIBS) $(CLIBS)

LLIBNETV2= $(LIBNETV2)
LLIBNETWORK= $(LIBNETWORK)

CONFIG= $(LIBHOME)/config.o
OPIMAI= $(LIBHOME)/opimai.o
LIBORA= $(LIBHOME)/libora.a
LIBKNL= $(LIBHOME)/libknl.a
LIBKNLOPT= $(LIBHOME)/libknlopt.a

LLIBORA= -lora

LLIBKNL= -lknl
LLIBKNLOPT= -lknlopt

RDBMSLIBDNOPLS= $(LIBORA) $(LIBKNLOPT) $(LIBKNL)
RDBMSLIBD= $(LIBORA) $(LIBKNLOPT) $(LIBPLS) $(LIBKNL)
RDBMSLIBSNOPLS= $(LLIBORA) $(LLIBKNLOPT) $(LLIBKNL)
RDBMSLIBS= $(LLIBORA) $(LLIBKNLOPT) $(LLIBPLS) $(LLIBKNL)

LIBSOSD=
LLIBSOSD=

LIBPCORE=$(LIBHOME)/libpcore.a
LLIBPCORE= -lpcore

LIBCGEN= $(LIBHOME)/pdc.o
LIBFORGEN= $(LIBHOME)/pdf.o
LIBCOBGEN= $(LIBHOME)/pdb.o
LIBPASGEN= $(LIBHOME)/pdp.o
LIBADAGEN= $(LIBHOME)/pda.o
LIBPLIGEN= $(LIBHOME)/pd1.o
LIBOSDGEN= $(LIBHOME)/pds.o

CODEGENHOME= $(SRCHOME)/codegen

LIBSLAX=$(LIBHOME)/pxslax.o

SLAX=/usr/local/slax/newsx

y.c:
       $(SLAX) -p -d -v $*.y

SQLHOME= $(ORACLE_HOME)/sqllib

SQLPUBLIC=$(I_SYM)$(SQLHOME)/public
LIBSQL= $(LIBHOME)/libsql.a
SQLLIBS= -lsql

LIBPROC20=$(LIBHOME)/libproc.a
LLIBPROC20= -lproc
LIBPRO= $(LIBPROC20)
LLIBPRO= $(LLIBPROC20)
PROLDLIBS= $(SQLLIBS) $(TTLIBS)

PROC=$(ORACLE_HOME)/bin/proc

PROCOBJS=$(ORACLE_HOME)/proc/lib/main.o \
       $(ORACLE_HOME)/proc/lib/pcdsfv.o \
       $(ORACLE_HOME)/proc/lib/pcdlut.o
PROCPLSFLAGS= sqlcheck=full userid=$(USERID) dbms=v6_char
USERID= scott/tiger
EXE= sample1
OBJS= sample1.o

SAMPLES=sample1 sample2 sample3 sample4 sample5 sample6 sample7 sample8 sample9 sample10 oraca sqlvcp
cv_demo

all: build

# Rule to compile any program (specify EXE= and OBJS= on command line)
build: $(OBJS)
       @if [ "$(ORA_CLIENT_LIB)" = "shared" ]; then \
       $(ECHO) $(CC) $(LDFLAGS) -o $(EXE) $(OBJS) -L$(LIBHOME) -lclntsh $(TTLIBS); \
       else \
       $(ECHO) $(CC) $(LDFLAGS) -o $(EXE) $(OBJS) $(PROLDLIBS) $(OTHERLIBS); \
       fi

# "Shorthand" Rule to compile all the sample program
samples: $(SAMPLES)

$(SAMPLES):
       $(MAKE) -f $(ORACLE_HOME)/proc/lib/proc.mk [email protected] EXE=$@ build

sample6.o: sample6.pc
       $(PROC) dbms=v6_char iname=$*.pc
       $(CC) $(CFLAGS) $(SQLPUBLIC) -c $*.c

sample9.o: sample9.pc
       $(PROC) $(PROCPLSFLAGS) iname=$*.pc
       $(CC) $(CFLAGS) $(SQLPUBLIC) -c $*.c

cv_demo.o: cv_demo.pc
       $(PROC) $(PROCPLSFLAGS) iname=$*.pc
       $(CC) $(CFLAGS) $(SQLPUBLIC) -c $*.c

# Suffix rules
SUFFIXES: .exe .o .c .pc

pc.c:
       $(PROC) $(PROFLAGS) iname=$*.pc

pc.o:
       $(PROC) $(PROFLAGS) iname=$*.pc
       $(CC) $(CFLAGS) $(SQLPUBLIC) -c $*.c

c.o:
       $(CC) $(CFLAGS) $(SQLPUBLIC) -c $*.c

LIBDIR= $(ORACLE_HOME)/proc/lib
DEMODIR= $(ORACLE_HOME)/proc/demo

install_files:
       -rm -f $(DEMODIR)/proc.mk
       -ln $(LIBDIR)/proc.mk $(DEMODIR)/proc.mk

# Rules to relink the proc executable
install: clean proc
       -chmod 755 $(ORACLE_HOME)/bin/proc
       -mv proc $(ORACLE_HOME)/bin/proc
       -chmod 755 $(ORACLE_HOME)/bin/proc

clean:
       -rm -f proc

LIBPLSHACK= $(LLIBPLS)
proc:
       @$(ECHO) $(CC) $(LDFLAGS) -o proc $(PROCOBJS) $(LLIBPROC20) \
       $(LIBCGEN) $(LIBSLAX) $(LIBOSDGEN) $(LIBPCORE) $(LLIBPLS) $(TTLIBS)
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <17493-18@oink>;
         Thu, 27 Jun 1996 14:48:24 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835881058:28539:1; Thu, 27 Jun 96 14:10:58 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa27152; 27 Jun 96 14:09 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id FAA11685 for dbi-users-real; Thu, 27 Jun 1996 05:55:58 -0700
Precedence: junk
Sender: [email protected]
Received: from dealer.cards.com ([128.126.162.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id FAA11681
         for <[email protected]>; Thu, 27 Jun 1996 05:55:56 -0700
Received: from solitaire (solitaire.cards.com)
         by dealer.cards.com (4.1/mls/3.2) id AA19622;
         Thu, 27 Jun 96 08:55:47 EDT
Message-Id: <[email protected]>
Date: Thu, 27 Jun 1996 08:55:08 -0400
From: Brian Curfman <[email protected]>
X-Mailer: Mozilla 2.01 (X11; I; SunOS 4.1.3_U1 sun4m)
Mime-Version: 1.0
To: [email protected]
Cc: [email protected], [email protected]
Subject: Perl5, Oraperl Emulation problem
X-Url: http://www.hermetica.com/technologia/DBI/DBD/index.html#Oracle
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hello,
   I apologize if this problem is not relevant to this list.

   I recently upgraded from Oracle 7.0.16 to 7.2.3 using
Perl 5 and Oraperl Emulation.  It worked fine under 7.0.16
but now I am having problems - inconsistently.

I have some scripts that work always and some that hang
for a minute - always - and then fall thru sucessfully ...
no errors - no warnings - just SLOW success.

I am setting ORACLE_HOME and ORACLE_SID and using
those in my ora_login function.

I've found a few messages with a similar problem - but no
solution yet.

Any ideas or a direction would be greatly appreciated.

Thanks
--
Brian Curfman
Senior Software Engineer
Strictly Business Computer Systems, Inc.
Software Engineering Division

PROJECTS:
National Software Data and Information Repository (NSDIR)

CONTACT INFORMATION:
email : [email protected]
       [email protected]
SBI   phone : (304) 366-8246
SBI   fax   : (304) 366-8350
WVHTC phone : (304) 363-1731

URL:  http://www.cards.com/~curfman
MAIL: mailto:[email protected]
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <17635-0@oink>;
         Thu, 27 Jun 1996 15:43:28 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 835884424:28433:1; Thu, 27 Jun 96 15:07:04 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa23421; 27 Jun 96 14:59 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA11715 for dbi-users-real; Thu, 27 Jun 1996 06:47:34 -0700
Precedence: junk
Sender: [email protected]
Received: from x4u2.desy.de (x4u2.desy.de [131.169.30.42])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id GAA11710
         for <[email protected]>; Thu, 27 Jun 1996 06:41:43 -0700
Received: (from hfalken@localhost) by x4u2.desy.de (8.6.10/8.6.9) id NAA22380
         for [email protected]; Thu, 27 Jun 1996 13:40:46 GMT
From: Harald Falkenberg <[email protected]>
Message-Id: <[email protected]>
Subject: can't build DBD on HP
To: [email protected]
Date: Thu, 27 Jun 1996 15:40:45 +0200 (MST)
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 9898

Hi,

I have problems to link DBD on a HP/UX with the OS 9.05. I tried it several
times with dynamic and static link options (LINKTYPE=...), but it failed.
For some reasons, some functions are not found at link time.
I append my full log to this mail. If someone has an idea or has already
solved this problem - please give me an answer.

regards
       Harald Falkenberg

-------------
perl Makefile.PL

   Remember to actually *read* the README file!

   Build, TEST and INSTALL Perl 5 first. Use version 5.002beta2 or later.

   DO NOT build the DBI under the Perl source tree.

   Use 'make test' to execute self tests.

   Install DBI (using 'make install' before building DBD::* drivers.

MakeMaker (v5.21)
Checking if your kit is complete...
Looks good
       NAME => q[DBI]
       VERSION_FROM => q[DBI.pm]
       clean => { FILES=>q[$(DISTVNAME)/] }
       dist => { DIST_DEFAULT=>q[clean distcheck disttest ci tardist], PREOP=>q[$(MAKE) -f Makefile.old distdir] }
DBI VERSION is 0.69 (from DBI.pm)
Using PERL=/usr/local/bin/perl
Writing Makefile for DBI
hfalken - ~/perl/db/DBI-0.69 :> make
       umask 0 && cp DBIXS.h ./blib/arch/DBI/DBIXS.h
       umask 0 && cp lib/DBD/ExampleP.pm ./blib/lib/DBD/ExampleP.pm
       umask 0 && cp lib/DBD/NullP.pm ./blib/lib/DBD/NullP.pm
       umask 0 && cp lib/DBD/Sponge.pm ./blib/lib/DBD/Sponge.pm
       umask 0 && cp DBI.pm ./blib/lib/DBI.pm
       /usr/local/bin/perl -I/usr/local/lib/perl5.002 -I/usr/local/lib/perl5.002 /usr/local/lib/perl5.002/ExtUtils/xsubpp  -typemap /usr/local/lib/perl5.002/ExtUtils/typemap DBI.xs >DBI.tc && mv DBI.tc DBI.c
       cc -c  -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.69\" -DXS_VERSION=\"0.69\" +z -I/usr/local/lib/perl5.002/CORE  DBI.c
Running Mkbootstrap for DBI ()
       chmod 644 DBI.bs
       LD_RUN_PATH="" ld -o ./blib/arch/auto/DBI/DBI.sl -b -L/usr/local/lib DBI.o
       chmod 755 ./blib/arch/auto/DBI/DBI.sl
       cp DBI.bs ./blib/arch/auto/DBI/DBI.bs
       chmod 644 ./blib/arch/auto/DBI/DBI.bs
hfalken - ~/perl/db/DBI-0.69 :> make test
       PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5.002 -I/usr/local/lib/perl5.002 -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t
t/min...............DBD::ExampleP in /usr/local/lib/perl5.002 is hidden by DBD::ExampleP in ./blib/lib
at t/min.t line 16
DBD::NullP in /usr/local/lib/perl5.002 is hidden by DBD::NullP in ./blib/lib
at t/min.t line 16
DBD::Sponge in /usr/local/lib/perl5.002 is hidden by DBD::Sponge in ./blib/lib
at t/min.t line 16
ok
All tests successful.
Files=1,  Tests=12,  1 secs ( 0.24 cusr  0.06 csys =  0.30 cpu)
       PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5.002 -I/usr/local/lib/perl5.002 test.pl
test.pl
DBI test application $Revision: 1.14 $
Use of uninitialized value at /usr/local/lib/perl5.002/Getopt/Long.pm line 602.
Switch: DBI-0.69 Switch by Tim Bunce, 0.69
DebugDispatch: 0
DBD::ExampleP in /usr/local/lib/perl5.002 is hidden by DBD::ExampleP in ./blib/lib
at test.pl line 49
DBD::NullP in /usr/local/lib/perl5.002 is hidden by DBD::NullP in ./blib/lib
at test.pl line 49
DBD::Sponge in /usr/local/lib/perl5.002 is hidden by DBD::Sponge in ./blib/lib
at test.pl line 49
Available Drivers: ExampleP, NullP, Sponge
Read DBI special variables (expect 0, 99, 99):
err:
err:    99
errstr: 99
Attempt to modify DBI special variables.
Expect a 'Can't modify' error message:
Can't modify $DBI::&rows special variable at test.pl line 59

Connected as DBI::db=HASH(0x400fb5a0)

Test error handling: prepare()'ing an invalid query.
Expect prepare to return UNDEF and an error message:
prepare returned UNDEF. $DBI::err=Unknown field names: unknown_field_name

Cursor prepare'd as DBI::st=HASH(0x400fb6c0)
Prepared as DBI::st=HASH(0x400fb774)
Number of fields: 3
Data type of first field: 1
Driver name: ExampleP

Fetching data from both cursors.
Expect several rows of data:
16749 78080 ., 1 1024 .
16877 2 .., 1 1024 ..
16749 79360 lib, 1 24 rexd
16749 144646 bin, 1 1024 cron
16895 108841 tmp, 1 1024 sockets
16895 512688150 man, 1 1024 rwho
16749 38417 spool, 1 1024 mqueue
16749 104967 etc, 1 1024 lp
16749 116516 netdemo, 1 24 uucppublic
16895 512688136 include, 1 1024 uucp
16749 7 nettest, 1 24 ppl

Automatic method parameter usage check.
Expect a 'DBI ... invalid' error and a 'Usage: ...' message:
DBI commit: invalid number of parameters: handle + 1
Usage: $h->commit() at test.pl line 124.

Preparing new $cursor_a to replace current $cursor_a.
We enable debugging on current to watch it's destruction.
Expect several lines of DBI trace information:
   DBI::st=HASH(0x400fb798) debug level set to 2
   -> DESTROY for DBD::ExampleP::st (DBI::st=HASH(0x400fb798))
   <- DESTROY= undef

Executing via func redirect: $h->func(..., 'execute')

Binding columns of $cursor_a to variables.

Fetching one row from new $cursor_a with a bound column.
Expect a large number follwed by a dot:
835871032 .
bind_col worked

Cursor tests done (scoped objects will be destroyed now)
test.pl done (global destruction will follow)

hfalken - ~/perl/db/DBI-0.69 :> cd ../DBD-Oracle-0.32
DBD-Oracle-0.32.tar.Z     DBD-Oracle-0.32/
hfalken - ~/perl/db/DBI-0.69 :> cd ../DBD-Oracle-0.32
hfalken - ~/perl/db/DBD-Oracle-0.32 :> perl Makefile.PL

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6
Oracle sysliblist: -lcl -lm
Using /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/proc/lib/proc.mk (version 1.17)
Warning: forced to build static not dynamic on hpux
        See README and Makefile.PL for more information.
System: perl5.002 hp-ux dirac a.09.05 a 9000735 2005669792 two-user license
Compiler: cc -O -D_HPUX_SOURCE -Aa -I/usr/local/include
Oracle proc.mk would have used these values but we override them:
 CC:       cc
 CFLAGS:   -I. -O -Aa -D_HPUX_SOURCE +ESsfc +ESlit
 LDFLAGS:  -Wl,-aarchive -L$(LIBHOME)
Linking with -locic $(LIBHOME)/osntab.o -lsqlnet -lora $(LIBPLSHACK) -lsqlnet -lora -lnlsrtl -lcv6 -lcore -lnlsrtl -lcv6 -lcore $(LDLIBS) `cat /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/rdbms/lib/sysliblist` $(MLSLIBS)

Writing Makefile for DBD::Oracle
hfalken - ~/perl/db/DBD-Oracle-0.32 :> make
       umask 0 && cp Oraperl.pm ./blib/lib/Oraperl.pm
       umask 0 && cp oraperl.ph ./blib/lib/oraperl.ph
       umask 0 && cp Oracle.pm ./blib/lib/DBD/Oracle.pm
       cc -c -I/afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/rdbms/demo -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" +z -I/usr/local/lib/perl5.002/CORE  Oracle.c
       cc -c -I/afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/rdbms/demo -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" +z -I/usr/local/lib/perl5.002/CORE  dbdimp.c
       ar cr ./blib/arch/auto/DBD/Oracle/Oracle.a Oracle.o  dbdimp.o && : ./blib/arch/auto/DBD/Oracle/Oracle.a
       chmod 755 ./blib/arch/auto/DBD/Oracle/Oracle.a
Manifying ./blib/man3/DBD::Oraperl.3
hfalken - ~/perl/db/DBD-Oracle-0.32 :> make test
Writing "Makefile.aperl" for this perl

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6
Oracle sysliblist: -lcl -lm
Using /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/proc/lib/proc.mk (version 1.17)
Warning: forced to build static not dynamic on hpux
        See README and Makefile.PL for more information.
System: perl5.002 hp-ux dirac a.09.05 a 9000735 2005669792 two-user license
Compiler: cc -O -D_HPUX_SOURCE -Aa -I/usr/local/include
Oracle proc.mk would have used these values but we override them:
 CC:       cc
 CFLAGS:   -I. -O -Aa -D_HPUX_SOURCE +ESsfc +ESlit
 LDFLAGS:  -Wl,-aarchive -L$(LIBHOME)
Linking with -locic $(LIBHOME)/osntab.o -lsqlnet -lora $(LIBPLSHACK) -lsqlnet -lora -lnlsrtl -lcv6 -lcore -lnlsrtl -lcv6 -lcore $(LDLIBS) `cat /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/rdbms/lib/sysliblist` $(MLSLIBS)

Writing Makefile.aperl for DBD::Oracle
       make -f Makefile.aperl static perl
       umask 0 && cp Oraperl.pm ./blib/lib/Oraperl.pm
       umask 0 && cp oraperl.ph ./blib/lib/oraperl.ph
       umask 0 && cp Oracle.pm ./blib/lib/DBD/Oracle.pm
Writing ./perlmain.c
       cd . && cc -c  -I/usr/local/lib/perl5.002/CORE  -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.32\" -DXS_VERSION=\"0.32\"  -I/usr/local/lib/perl5.002/CORE perlmain.c
       cat ./blib/arch/auto/DBD/Oracle/extralibs.ld >> ./blib/arch/auto/DBD/Oracle/extralibs.all
       cat /afs/desy.de/products/perl/perl5.002/spec/hp700_ux90/lib/auto/DBI/extralibs.ld >> ./blib/arch/auto/DBD/Oracle/extralibs.all
       cat /afs/desy.de/products/perl/perl5.002/spec/hp700_ux90/lib/auto/DynaLoader/extralibs.ld >> ./blib/arch/auto/DBD/Oracle/extralibs.all
       cc -L/usr/local/lib -Wl,-E  -o perl ./perlmain.o /usr/local/lib/perl5.002/CORE/libperl.a /afs/desy.de/products/perl/perl5.002/spec/hp700_ux90/lib/auto/DynaLoader/DynaLoader.a  /afs/desy.de/products/perl/perl5.002/spec/hp700_ux90/lib/auto/DBI/DBI.a  ./blib/arch/auto/DBD/Oracle/Oracle.a `cat ./blib/arch/auto/DBD/Oracle/extralibs.all` -lnet -lndbm -ldld -lm -lc -lndir -lcrypt
/bin/ld: Unsatisfied symbols:
  sosngpgn (code)
  osnqtg (code)
  osnqig1 (code)
  osnqrn (code)
  nldtr1 (code)
  sosngun (code)
  nscontrol (code)
  snlfngenv (code)
  snlfnexed (code)
*** Error code 1

Stop.
*** Error code 1

Stop.
hfalken - ~/perl/db/DBD-Oracle-0.32 :>
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18358-5@oink>;
         Thu, 27 Jun 1996 17:02:37 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835890014:19852:1; Thu, 27 Jun 96 16:40:14 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa18657; 27 Jun 96 16:39 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA11856 for dbi-users-real; Thu, 27 Jun 1996 08:32:16 -0700
Precedence: junk
Sender: [email protected]
Received: from mail.industry.net (mail.industry.net [206.150.196.12])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA11852
         for <[email protected]>; Thu, 27 Jun 1996 08:32:13 -0700
Received: from dns2.industry.net (dns2.industry.net [206.150.196.30])
         by mail.industry.net (8.6.9/8.6.8) with SMTP id LAA09408;
         Thu, 27 Jun 1996 11:36:41 -0400
Message-ID: <[email protected]>
Date: Wed, 26 Jun 1996 04:59:56 -0400
From: "Rujith S. de Silva" <[email protected]>
Organization: Industry.Net
X-Mailer: Mozilla 2.0 (X11; I; SunOS 4.1.3_U1 sun4m)
MIME-Version: 1.0
To: Brian Curfman <[email protected]>
CC: [email protected], [email protected], [email protected]
Subject: Re: Perl5, Oraperl Emulation problem
References: <[email protected]>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Brian Curfman wrote:
> I have some scripts that work always and some that hang
> for a minute - always - and then fall thru sucessfully ...
> no errors - no warnings - just SLOW success.
>
> I am setting ORACLE_HOME and ORACLE_SID and using
> those in my ora_login function.

Are you setting %ENV in Perl?  It looks like Perl/Oraperl is unhappy,
for some strange reason, if you set %ENV yourself before including
Oraperl.  It's actually much faster to write a five-line shell program
that sets the environment, and then calls the Perl/Oraperl program so
that %ENV inherits the right values from the parent shell process.

Later,
Rujith.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20342-3@oink>;
         Fri, 28 Jun 1996 07:06:11 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835895311:07682:1; Thu, 27 Jun 96 18:08:31 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa05995; 27 Jun 96 18:07 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA12192 for dbi-users-real; Thu, 27 Jun 1996 10:02:52 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA12188
         for <[email protected]>; Thu, 27 Jun 1996 10:02:02 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id bd29029; 27 Jun 96 17:17 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ai04099; 27 Jun 96 17:09 +0100
Received: from toad by oink with SMTP (PP) id <18292-0@oink>;
         Thu, 27 Jun 1996 16:51:01 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA01869;
         Thu, 27 Jun 1996 16:50:53 +0000
Date: Thu, 27 Jun 1996 16:50:53 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Perl5, Oraperl Emulation problem
Cc: [email protected], [email protected]
X-Sun-Charset: US-ASCII
content-length: 750
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Brian Curfman <[email protected]>
>
> Hello,
>     I apologize if this problem is not relevant to this list.
>
>     I recently upgraded from Oracle 7.0.16 to 7.2.3 using
> Perl 5 and Oraperl Emulation.  It worked fine under 7.0.16
> but now I am having problems - inconsistently.
>
> I have some scripts that work always and some that hang
> for a minute - always - and then fall thru sucessfully ...
> no errors - no warnings - just SLOW success.
>
> I am setting ORACLE_HOME and ORACLE_SID and using
> those in my ora_login function.
>
> I've found a few messages with a similar problem - but no
> solution yet.
>
> Any ideas or a direction would be greatly appreciated.

Read the README for some ideas about slow connects.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20342-19@oink>;
         Fri, 28 Jun 1996 07:08:34 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835904330:12681:1; Thu, 27 Jun 96 20:38:50 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa08557; 27 Jun 96 20:36 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA13000 for dbi-users-real; Thu, 27 Jun 1996 12:26:15 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id MAA12996
         for <[email protected]>; Thu, 27 Jun 1996 12:26:12 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ak20417; 27 Jun 96 17:46 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac04099; 27 Jun 96 17:08 +0100
Received: from toad by oink with SMTP (PP) id <16980-0@oink>;
         Thu, 27 Jun 1996 11:00:41 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA27458;
         Thu, 27 Jun 1996 11:00:33 +0000
Date: Thu, 27 Jun 1996 11:00:33 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: (please use the subject line)
X-Sun-Charset: US-ASCII
content-length: 418
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected]
>
> A couple of questions:
>
> Is DBperl by itself able to communicate with Ingres and other databases or are database specific modules absolutely necessary?

Absolutely necessary.

> Basically, is my only option if I want to write Perl scripts for an Ingres database to use Perl4 and ingperl? or can I use some other combination the uses Perl5?

Perl 4 and ingperl for now.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20342-24@oink>;
         Fri, 28 Jun 1996 07:09:34 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835908401:25138:1; Thu, 27 Jun 96 21:46:41 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa25113; 27 Jun 96 21:46 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa20126; 27 Jun 96 21:45 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA12295 for dbi-users-real; Thu, 27 Jun 1996 10:42:08 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA12291
         for <[email protected]>; Thu, 27 Jun 1996 10:42:05 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ab11377; 27 Jun 96 18:38 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aj04099; 27 Jun 96 17:10 +0100
Received: from toad by oink with SMTP (PP) id <18305-0@oink>;
         Thu, 27 Jun 1996 16:53:06 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA01877;
         Thu, 27 Jun 1996 16:52:59 +0000
Date: Thu, 27 Jun 1996 16:52:59 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: can't build DBD on HP
X-Sun-Charset: US-ASCII
content-length: 463
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Harald Falkenberg <[email protected]>
>
> Hi,
>
> I have problems to link DBD on a HP/UX with the OS 9.05. I tried it several
> times with dynamic and static link options (LINKTYPE=...), but it failed.
> For some reasons, some functions are not found at link time.
> I append my full log to this mail. If someone has an idea or has already
> solved this problem - please give me an answer.

Upgrade from to DBD::Oracle 0.34 and try again.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20342-28@oink>;
         Fri, 28 Jun 1996 07:10:17 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835912435:07435:1; Thu, 27 Jun 96 22:53:55 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa02785; 27 Jun 96 22:52 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA11992 for dbi-users-real; Thu, 27 Jun 1996 09:36:12 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA11988
         for <[email protected]>; Thu, 27 Jun 1996 09:36:04 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ad16040; 27 Jun 96 16:22 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id af04099; 27 Jun 96 17:09 +0100
Received: from toad by oink with SMTP (PP) id <17075-0@oink>;
         Thu, 27 Jun 1996 11:40:43 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA27954;
         Thu, 27 Jun 1996 11:40:35 +0000
Date: Thu, 27 Jun 1996 11:40:35 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Memory leak
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 2206
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Michal Jaegermann <[email protected]>
>
> > Take that code and add the line
> >
> >        eval 'use Oraperl; 1' || die $@ if $] >= 5;
> >
> > and then see if it leaks under Perl5.
>
> Well, you are right.  Doing that and watching results accross a number
> of requests gives something as follows:
>
>    PID %CPU %MEM   VSZ  RSS TTY      S    STARTED         TIME
>  18626  0.0  1.5 6.85M 1.4M ttyp5    S    15:40:52     0:00.07
.. 5 deleted...
>  18626  0.0  2.0 6.92M 1.9M ttyp5    S    15:40:52     0:09.73
>
> An initial jump (from the first to the second line) is expected,
> as login to Oracle does not happen until the first request comes.
> TIME field also reflects that (the first request was small).
>
> On the other hand, running the same code under oraperl 2.4 looks
> like this:
>
>  32169  0.0  0.7 5.28M 632K ttyp5    S    15:53:31     0:00.05
.. 5 deleted...
>  32169  0.0  0.5 5.37M 488K ttyp5    I    15:53:31     0:05.09

Please rerun the benckmark with $Oraperl::safe=0; before the eval
(assuming your using version 0.33 or later). That should make a
difference and I'd be interested in how much.

It would also be worth compiling a perl with the DBI, DBD::Oracle
and the oracle libs statically linked into it (like oraperl 2.4 has).
That would be *especially* important if you have more than one
perl + DBD::Oracle loaded at one time.

When you rerun the benckmark please just send the 2nd and 12th (say)
ps lines for each config.

The other significant difference is that I've not got around to
implementing a row cache for DBD::Oracle (oraperl v2.4 has one).
I plan to implement a row cache in a month or two (no promises).

> Now, an intereresting question is what was leaking in the original. :-)

Good question. Your problem :-)

> I do not want to bother you with this, but if you would feel so
> inclined I may send you this sample.  It is not very long, but it
> will be not very useful without a context, I am afraid.

No samples please. I just don't have the time.

Carefully rerunning the benchmarks as described above would be
*very* intereting to me and probably others on dbi-users (CC'd).

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <20454-23@oink>;
         Fri, 28 Jun 1996 07:15:02 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835940045:15333:1; Fri, 28 Jun 96 06:34:05 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa14941; 28 Jun 96 6:32 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id WAA15327 for dbi-users-real; Thu, 27 Jun 1996 22:22:16 -0700
Precedence: junk
Sender: [email protected]
Received: from dns2.noc.best.net (dns2.noc.best.net [206.86.0.21])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id WAA15323
         for <[email protected]>; Thu, 27 Jun 1996 22:22:15 -0700
Received: from newpath.com.best.com (harvey.vip.best.com [206.86.7.68])
         by dns2.noc.best.net (8.6.12/8.6.5) with SMTP id WAA05196;
         Thu, 27 Jun 1996 22:22:10 -0700
Message-Id: <[email protected]>
Date: Thu, 27 Jun 96 22:09:14 -0700
From: Dennis Harvey <[email protected]>
MIME-Version: 1.0
To: [email protected]
CC: [email protected]
Subject: Problem running DBD Oraperl emulation with Perl 5.003
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

This is probably a stupid question but ...

I am converting an Oracle mail agent written in Oraperl v2.4/Perl 4.036 to
DBD Oraperl emulation with Perl 5.003.  When I run my program, all I get is:

install_driver(Oracle) failed: Can't find loadable object for module DBD::Oracle
in @INC (/usr/local/lib/perl5/PA-RISC1.1/5.003 /usr/local/lib/perl5 /usr/local/
lib/perl5/site_perl/PA-RISC1.1 /usr/local/lib/perl5/site_perl .) at /usr/local/l
ib/perl5/site_perl/DBD/Oracle.pm line 24
       DBI::install_driver called at /usr/local/lib/perl5/site_perl/Oraperl.pm
line 62
       require Oraperl.pm called at (eval 1) line 1
       main::BEGIN called at /usr/local/lib/perl5/site_perl/Oraperl.pm line 0
       eval {...} called at /usr/local/lib/perl5/site_perl/Oraperl.pm line 0
       eval 'use Oraperl; 1
;' called at ../omg.p line 300
       main::main called at ../omg.p line 17
BEGIN failed--compilation aborted at (eval 1) line 1.

My program includes the line:

eval 'use Oraperl; 1' || die $@ if $] >=5;

My versions are:

Perl 5.003
DBI-0.70
DBD-Oracle-0.35
Oracle 7.3.2.1
HP-UX 10.10

Output from erl -v:
       This is perl, version 5.003 with EMBED
       built under hpux at Jun 27 1996 11:55:34
       + suidperl security patch

Copyright 1987-1996, Larry Wall

My installation passed all of the tests.  Here is my output from make test for DBD-Oracle-0.35:

       make -f Makefile.aperl perl
`perl' is up to date.
       PERL_DL_NONLAZY=1 ./perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5/PA-RISC1.1/5.003 -I/usr/local/lib/perl5 -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t
t/base..............ok
All tests successful.
Files=1,  Tests=5,  0 secs ( 0.42 cusr  0.09 csys =  0.51 cpu)
       PERL_DL_NONLAZY=1 ./perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5/PA-RISC1.1/5.003 -I/usr/local/lib/perl5 test.pl
Oraperl test application $Revision: 1.21 $

Oraperl Emulation Interface version 1.28
Oracle Driver 0.35
DBI-0.70 Switch by Tim Bunce, version 0.70


Connecting
to '' (from command line, else uses ORACLE_SID/TWO_TASK)
as 'harveydr/harveydr' (via ORACLE_USERID environment var or default)
Fields:  4
Names:   'NUM_T',       'DATE_T',       'CHAR_T',       'NULL_T'
Lengths: '40',  '75',   '30',   '0'
Types:   '2',   '12',   '1',    '1'
Data rows:
   fetch: 79.2 27-JUN-96 HARVEYDR NULL

csr reassigned (forces destruction)...
Fetch list of tables:
Fetched: CDEF$
Fetched: TAB$
Fetched: IND$
Fetched: CLU$
Test ora_do with harmless non-select statement (set transaction read only)
Expect error message:
ora_do: 1453: ORA-01453: SET TRANSACTION must be first statement of transaction (DBD: oexec error)
csr out of scope...
ora_logoff...
lda out of scope...

Testing repetitive connect/open/close/disconnect:
Expect sequence of digits, no other messages:
1 2 3 4 5

Test interaction of explicit close/logoff and implicit DESTROYs
Expect just 'done.', no other messages:
done.

Test complete (2 seconds).
Bad free() ignored during global destruction.

What am I doing wrong?? I've worked with Perl4 a good bit, but I'm new to Perl5.

Thanks.  ;-(

Dennis Harvey
[email protected]
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <25725-0@oink>;
         Fri, 28 Jun 1996 17:00:57 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835965855:23734:4; Fri, 28 Jun 96 13:44:15 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa23316; 28 Jun 96 13:43 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id FAA19078 for dbi-users-real; Fri, 28 Jun 1996 05:40:21 -0700
Precedence: junk
Sender: [email protected]
Received: from colossus.cse.psu.edu ([email protected] [130.203.1.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id FAA19074
         for <[email protected]>; Fri, 28 Jun 1996 05:40:19 -0700
Received: from localhost by colossus.cse.psu.edu with SMTP id <78359>;
         Fri, 28 Jun 1996 08:40:13 -0400
X-Mailer: exmh version 1.6.5 12/11/95
X-uri: <URL:http://www.cse.psu.edu/~groenvel/>
X-work-address: Manufacturing Technology - Systems Development Group PO Box 30
               The Applied Research Lab State College, PA 16804
X-work-phone: +1 814 863 9896 (Voice) +1 814 863 1183 (Fax)
To: Dennis Harvey <[email protected]>
cc: [email protected]
Subject: Re: Problem running DBD Oraperl emulation with Perl 5.003
In-reply-to: Your message of "Fri, 28 Jun 1996 01:09:14 EDT." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 28 Jun 1996 08:38:04 -0400
From: John D Groenveld <[email protected]>
Message-Id: <[email protected]>

Read the bit in the README's about suppressing Bad free() warnings.
John

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <25725-3@oink>;
         Fri, 28 Jun 1996 17:01:12 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 835966128:26286:1; Fri, 28 Jun 96 13:48:48 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa23775; 28 Jun 96 13:47 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id FAA19068 for dbi-users-real; Fri, 28 Jun 1996 05:38:42 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id FAA19064
         for <[email protected]>; Fri, 28 Jun 1996 05:38:41 -0700
Received: by gw.home.vix.com id FAA23202; Fri, 28 Jun 1996 05:38:39 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ah01439; 28 Jun 96 12:33 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa14298; 28 Jun 96 13:25 +0100
Received: from toad by oink with SMTP (PP) id <24833-0@oink>;
         Fri, 28 Jun 1996 12:26:23 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA05443;
         Fri, 28 Jun 1996 12:26:16 +0000
Date: Fri, 28 Jun 1996 12:26:16 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Problem running DBD Oraperl emulation with Perl 5.003
X-Sun-Charset: US-ASCII
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Dennis Harvey <[email protected]>
>
> This is probably a stupid question but ...
>
> I am converting an Oracle mail agent written in Oraperl v2.4/Perl 4.036 to
> DBD Oraperl emulation with Perl 5.003.

By 'converting' I hope you mean *just* adding the one 'use Oraperl;' line
(as described at the top of Oraperl.pm).

> When I run my program, all I get is:
>
> install_driver(Oracle) failed: Can't find loadable object for module DBD::Oracle
>  in @INC (/usr/local/lib/perl5/PA-RISC1.1/5.003 /usr/local/lib/perl5 /usr/local/
> lib/perl5/site_perl/PA-RISC1.1 /usr/local/lib/perl5/site_perl .) at /usr/local/l
> ib/perl5/site_perl/DBD/Oracle.pm line 24

> My program includes the line:
>
> eval 'use Oraperl; 1' || die $@ if $] >=5;
>
> My versions are:
>
> Perl 5.003
> DBI-0.70
> DBD-Oracle-0.35
> Oracle 7.3.2.1
> HP-UX 10.10

Mort probably you didn't run 'make install'. Do so and what the paths printed
and compare them to the paths shown in the error message above.

> My installation passed all of the tests.  Here is my output from make test for DBD-Oracle-0.35:
>
>       make -f Makefile.aperl perl

Why did it make a perl binary? Your system should support dynamic loading.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <04204-13@oink>;
         Mon, 1 Jul 1996 07:15:09 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836187414:08951:1; Mon, 01 Jul 96 03:16:54 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa08866; 1 Jul 96 3:16 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id TAA22973 for dbi-users-real; Sun, 30 Jun 1996 19:03:17 -0700
Precedence: junk
Sender: [email protected]
Received: from pearl.aims.gov.AU (pearl.aims.gov.au [138.7.32.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id TAA22969
         for <[email protected]>; Sun, 30 Jun 1996 19:03:06 -0700
Received: from harry.aims.gov.au by pearl.aims.gov.AU (SMI-8.6/SMI-SVR4)
         id MAA09370; Mon, 1 Jul 1996 12:02:32 +1000
Received: by harry.aims.gov.au (5.x/SMI-SVR4) id AA02158;
         Mon, 1 Jul 1996 12:04:20 +1000
From: [email protected] (Melanie Ralph)
Message-Id: <[email protected]>
Subject: Problems with make test DBD-Oracle-0.35
To: [email protected]
Date: Mon, 1 Jul 1996 12:04:19 +1000 (EST)
Cc: [email protected]
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text

Hi
       I have recently been trying to add the DBD-Oracle 'extension' onto
Perl.  I am using Perl version 5.0 patch level 2.  I have successfully
installed DBI-0.70 with no problems.
       The problem arises when trying to install DBD-Oracle-0.35.  Everything goes
well until the make test, when tests 4 and 5 fail trying to install a driver.
       I have read some of the old mailing lists I found on the net and others
seem to have had similar problems, but none of the fix-its have worked!!
       All the details required are included below (I think!!!)

I hope someone out there can help!!

Thanks in advance

Melanie Ralph

#########################################################
Out put of ~/perl/bin/perl Makefile.PL -m $ORACLE_HOME/rdbms/lib/oracle.mk
#######################################
Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /oracle/oracle
Using /oracle/oracle/rdbms/lib/oracle.mk (version 1.19)
System: perl5.002 sunos pearl 5.5 generic_103093-02 sun4m sparc sunw,sparcstation-20
Compiler: gcc -O -I/usr/local/include
Linker:   /usr/ccs/bin/ld
Oracle proc.mk would have used these values but we override them:
 CC:       ld -dy $(COMPOBJS)
 LDFLAGS:  -L$(LIBHOME)
 LDSTRING: -R /opt/SUNWcluster/lib:/usr/ucblib -Y P,$(LD_LIBRARY_PATH):/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib -Qy -lc $(COMPOBJ)/crtn.o
Linking with $(LIBHOME)/osntab.o /oracle/oracle/lib/crtn.o $(LIBHOME)/libocic.a  -lsqlnet -lora -lsqlnet -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lsocket -lnsl -lm -ldl -laio -R /opt/SUNWcluster/lib:/usr/ucblib -YP,:/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib -Qy -lc  -lcore3

Checking if your kit is complete...
Looks good
Unrecognized argument in LIBS ignored: '$(LIBHOME)/libocic.a'
Unrecognized argument in LIBS ignored: '/opt/SUNWcluster/lib:/usr/ucblib'
Unrecognized argument in LIBS ignored: '-YP,:/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib'
Unrecognized argument in LIBS ignored: '-Qy'
Writing Makefile for DBD::Oracle

#############################
OUTPUT from make
############################
umask 0 && cp Oraperl.pm blib/lib/Oraperl.pm
umask 0 && cp oraperl.ph blib/lib/oraperl.ph
umask 0 && cp Oracle.pm blib/lib/DBD/Oracle.pm
gcc -c -I/oracle/oracle/rdbms/demo  -I/oracle/oracle/rdbms/demo -I/export/home/sbainbri/perl/lib/site_perl/sun4-solaris/DBI -I/export/home/sbainbri/perl/lib/sun4-solaris/5.002/DBI -I/usr/local/include -O -DVERSION=\"0.35\" -DXS_VERSION=\"0.35\" -fpic -I.. -Wall -pedantic -Wno-comment -Wtraditional Oracle.c
gcc -c -I/oracle/oracle/rdbms/demo  -I/oracle/oracle/rdbms/demo -I/export/home/sbainbri/perl/lib/site_perl/sun4-solaris/DBI -I/export/home/sbainbri/perl/lib/sun4-solaris/5.002/DBI -I/usr/local/include -O -DVERSION=\"0.35\" -DXS_VERSION=\"0.35\" -fpic -I.. -Wall -pedantic -Wno-comment -Wtraditional dbdimp.c
dbdimp.c:126: warning: `dump_error_status' defined but not used
Running Mkbootstrap for DBD::Oracle ()
chmod 644 Oracle.bs
LD_RUN_PATH="/oracle/oracle/lib:/lib" gcc -o blib/arch/auto/DBD/Oracle/Oracle.so -G -L/usr/local/lib Oracle.o  dbdimp.o /oracle/oracle/lib/osntab.o /oracle/oracle/lib/crtn.o /oracle/oracle/lib/__fstd.o  -L/oracle/oracle/lib -lsqlnet -lora -lsqlnet -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lsocket -lnsl -lm -ldl -laio -lc -lcore3
chmod 755 blib/arch/auto/DBD/Oracle/Oracle.so
cp Oracle.bs ./blib/arch/auto/DBD/Oracle/Oracle.bs
chmod 644 blib/arch/auto/DBD/Oracle/Oracle.bs
Manifying ./blib/man3/DBD::Oraperl.3


############################
OUTPUT from 'make test' (this is where the problems are!!)
###########################
PERL_DL_NONLAZY=1 ~/perl/bin/perl -I./blib/arch -I./blib/lib -I../lib -I../lib -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t
t/base..............install_driver(Oracle) failed: Can't load './blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle: ld.so.1: /export/home/sbainbri/perl/bin/perl: fatal: relocation error: symbol not found: oerhms: referenced in ./blib/arch/auto/DBD/Oracle/Oracle.so at ../lib/DynaLoader.pm line 140.

at blib/lib/DBD/Oracle.pm line 24
       DBI::install_driver called at t/base.t line 16
FAILED tests 4-5
       Failed 2/5 tests, 60.00% okay
       Test returned status 2 (wstat 512)
Failed 1 test script, 0.00% okay. 2/5 subtests failed, 60.00% okay.
*** Error code 29
make: Fatal error: Command failed for target `test_dynamic'


#########################
OUTPUT from 'make test TEST_VERBOSE=1'
#########################
PERL_DL_NONLAZY=1 ~/perl/bin/perl -I./blib/arch -I./blib/lib -I../lib -I../lib -e 'use Test::Harness qw(&runtests $verbose); $verbose=1; runtests @ARGV;' t/*.t
t/base..............install_driver(Oracle) failed: Can't load './blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle: ld.so.1: /export/home/sbainbri/perl/bin/perl: fatal: relocation error: symbol not found: oerhms: referenced in ./blib/arch/auto/DBD/Oracle/Oracle.so at ../lib/DynaLoader.pm line 140.

at blib/lib/DBD/Oracle.pm line 24
       DBI::install_driver called at t/base.t line 16
1..5
ok 1
ok 2
ok 3
FAILED tests 4-5
       Failed 2/5 tests, 60.00% okay
       Test returned status 2 (wstat 512)
Failed 1 test script, 0.00% okay. 2/5 subtests failed, 60.00% okay.
*** Error code 29
make: Fatal error: Command failed for target `test_dynamic'

I am using Perl versoin 5.0 patch level 2 and Oracle version 7.2.3.0.0

The output from a perl -V is below
##############################
Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=solaris, osver=2.5, archname=sun4-solaris
   uname='sunos pearl 5.5 generic_103093-02 sun4m sparc sunw,sparcstation-20 '
   hint=recommended, useposix=true
 Compiler:
   cc='gcc', optimize='-O', gccversion=2.6.3
   cppflags='-I/usr/local/include'
   ccflags ='-I/usr/local/include'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='gcc', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /lib /usr/lib /usr/ccs/lib
   libs=-lsocket -lnsl -ldl -lm -lc -lcrypt
   libc=/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-fpic', lddlflags='-G -L/usr/local/lib'

@INC: /export/home/sbainbri/perl/lib/sun4-solaris/5.002 /export/home/sbainbri/perl/lib /export/home/sbainbri/perl/lib/site_perl/sun4-solaris /export/home/sbainbri/perl/lib/site_perl .


--

----------------------------oOo----------------------------
Melanie Ralph
Australian Institute of Marine Science
PMB No. 3
Townsville MC
Queensland 4810

email [email protected]

Tel. (077) 534 298 (direct)     (077) 534 211 (switchboard)
Fax. (077) 725 852
----------------------------oOo----------------------------
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <08775-0@oink>;
         Mon, 1 Jul 1996 12:08:45 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836217272:09199:2; Mon, 01 Jul 96 11:34:32 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa08870; 1 Jul 96 11:33 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id DAA23405 for dbi-users-real; Mon, 1 Jul 1996 03:28:40 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id DAA23401
         for <[email protected]>; Mon, 1 Jul 1996 03:28:36 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id au03475; 1 Jul 96 10:22 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac06798; 1 Jul 96 10:44 +0100
Received: from toad by oink with SMTP (PP) id <08314-0@oink>;
         Mon, 1 Jul 1996 10:43:26 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA16469;
         Mon, 1 Jul 1996 10:43:20 +0000
Date: Mon, 1 Jul 1996 10:43:20 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Problems with make test DBD-Oracle-0.35
X-Sun-Charset: US-ASCII
content-length: 2036
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (Melanie Ralph)
>
> Hi
>       I have recently been trying to add the DBD-Oracle 'extension' onto
> Perl.  I am using Perl version 5.0 patch level 2.  I have successfully
> installed DBI-0.70 with no problems.
>       The problem arises when trying to install DBD-Oracle-0.35.  Everything goes
>  well until the make test, when tests 4 and 5 fail trying to install a driver.
>       I have read some of the old mailing lists I found on the net and others
> seem to have had similar problems, but none of the fix-its have worked!!

> Output of ~/perl/bin/perl Makefile.PL -m $ORACLE_HOME/rdbms/lib/oracle.mk

You should always try to use Pro*C's proc.mk file and not oracle.mk.
Where is Pro*C installed? If it's not installed then install it.

> Linking with $(LIBHOME)/osntab.o /oracle/oracle/lib/crtn.o $(LIBHOME)/libocic.a  -lsqlnet -lora
-lsqlnet -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lsocket -lnsl -lm -ldl -laio -R /opt/SUNWcluster/lib:/usr/ucblib -YP,:/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib -Qy -lc  -lcore3

> Unrecognized argument in LIBS ignored: '$(LIBHOME)/libocic.a'

I think that's the problem and that's my fault.

Here's a patch for Makefile.PL which should help:

diff -c -r1.45 Makefile.PL
*** 1.45        1996/06/22 02:09:29
--- Makefile.PL 1996/07/01 09:40:00
***************
*** 333,339 ****
     my $backtick  = shift;
     my $default = $strip ? '' : "\$($var)";
     # warn "Expand var: $var\n";
!     return '$(LIBHOME)' if $var eq 'LIBHOME'; # gets too noisy
     return $ENV{ORACLE_HOME} if $var eq 'ORACLE_HOME';
     $MK{$var} ? expand_mkvars($MK{$var}, $strip, $backtick) : $default;
 }
--- 334,340 ----
     my $backtick  = shift;
     my $default = $strip ? '' : "\$($var)";
     # warn "Expand var: $var\n";
!     return '$(LIBHOME)' if $var eq 'LIBHOME' and !$strip; # gets noisy
     return $ENV{ORACLE_HOME} if $var eq 'ORACLE_HOME';
     $MK{$var} ? expand_mkvars($MK{$var}, $strip, $backtick) : $default;
 }

Please let me know how it works out ASAP.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <09685-2@oink>;
         Mon, 1 Jul 1996 14:07:08 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836220792:17820:1; Mon, 01 Jul 96 12:33:12 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15566; 1 Jul 96 12:30 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA23446 for dbi-users-real; Mon, 1 Jul 1996 04:26:21 -0700
Precedence: junk
Sender: [email protected]
Received: from epiphore.francenet.fr ([email protected] [194.2.144.10])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id EAA23442
         for <[email protected]>; Mon, 1 Jul 1996 04:26:16 -0700
Received: from epiphore.francenet.fr ([email protected] [127.0.0.1])
         by epiphore.francenet.fr (8.7.5/8.7.3) with ESMTP id NAA02101;
         Mon, 1 Jul 1996 13:26:05 +0200 (MET DST)
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.5 12/8/95
To: [email protected], [email protected], [email protected],
   [email protected], [email protected], [email protected],
   [email protected], [email protected], [email protected]
Subject: DBD-Oracle-0.32 test failed on HP/UX 10.10
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 01 Jul 1996 13:25:52 +0200
From: Gildas Perrot <[email protected]>

Hi,

I am trying to install DBD-Oracle-0.32 on HP/UX 10.10.

For the moment, I have installed current version of Perl 5.002 (all tests was
successful) and here is the perl-V output :

Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=hpux, osver=10, archname=PA-RISC1.1
   uname='hp-ux emphase b.10.10 e 9000831 2013407021 8-user license '
   hint=recommended, useposix=true
 Compiler:
   cc='cc', optimize='-O', gccversion=
   cppflags='-D_HPUX_SOURCE -Aa -I/usr/local/include'
   ccflags ='-D_HPUX_SOURCE -Aa -I/usr/local/include'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='ld', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /lib/pa1.1 /lib /usr/lib /usr/ccs/lib
   libs=-lnet -lnsl_s -lndbm -lgdbm -ldld -lm -lc -lndir -lcrypt
   libc=/lib/libc.sl, so=sl
 Dynamic Linking:
   dlsrc=dl_hpux.xs, dlext=sl, d_dlsymun=, ccdlflags='-Wl,-E '
   cccdlflags='+z', lddlflags='-b -L/usr/local/lib'

@INC: /usr/local/lib/perl5/PA-RISC1.1/5.002 /usr/local/lib/perl5
/usr/local/lib/perl5/site_perl/PA-RISC1.1 /usr/local/lib/perl5/site_perl .

I have installed DBI-0.69 first dynamically and then statiscally since
DBD-Oracle was complaining about that. Here was my procedure to install DBI :

perl5 Makefile.PL LINKTYPE=static
make
make perl
make test FULLPERL=./perl

Here is the output of the test step :

       PERL_DL_NONLAZY=1 ./perl -I./blib/arch -I./blib/lib -I../../lib
-I../../lib -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests
@ARGV;' t/*.t
t/min...............DBD::ExampleP in /usr/local/lib/perl5/site_perl is hidden
by DBD::ExampleP in ./blib/lib
at t/min.t line 16
DBD::NullP in /usr/local/lib/perl5/site_perl is hidden by DBD::NullP in
/blib/lib
at t/min.t line 16
DBD::Sponge in /usr/local/lib/perl5/site_perl is hidden by DBD::Sponge in
/blib/lib
at t/min.t line 16
ok
All tests successful.
Files=1,  Tests=12,  1 secs ( 0.25 cusr  0.10 csys =  0.35 cpu)
       PERL_DL_NONLAZY=1 ./perl -I./blib/arch -I./blib/lib -I../../lib
-I../../lib test.pl
test.pl
DBI test application $Revision: 1.14 $
Use of uninitialized value at /opt/pck/perl5.002/ext/DBI-0.69/../../lib/Getopt/
Long.pm line 602.
Switch: DBI-0.69 Switch by Tim Bunce, 0.69
DebugDispatch: 0
DBD::ExampleP in /usr/local/lib/perl5/site_perl is hidden by DBD::ExampleP in
/blib/lib
at test.pl line 49
DBD::NullP in /usr/local/lib/perl5/site_perl is hidden by DBD::NullP in
/blib/lib
at test.pl line 49
DBD::Sponge in /usr/local/lib/perl5/site_perl is hidden by DBD::Sponge in
/blib/lib
at test.pl line 49
Available Drivers: ExampleP, NullP, Sponge
Read DBI special variables (expect 0, 99, 99):
err:
err:    99
errstr: 99
Attempt to modify DBI special variables.
Expect a 'Can't modify' error message:
Can't modify $DBI::&rows special variable at test.pl line 59

Connected as DBI::db=HASH(0x40105c80)

Test error handling: prepare()'ing an invalid query.
Expect prepare to return UNDEF and an error message:
prepare returned UNDEF. $DBI::err=Unknown field names: unknown_field_name

Cursor prepare'd as DBI::st=HASH(0x40105da0)
Prepared as DBI::st=HASH(0x40105e54)
Number of fields: 3
Data type of first field: 1
Driver name: ExampleP

Fetching data from both cursors.
Expect several rows of data:
16749 2 ., 1 1024 .
16877 2 .., 1 1024 ..
16877 3 lost+found, 0 96 lost+found
16749 4 bin, 1 1024 uucp
16749 6 sbin, 0 96 sockets
16749 54 lib, 1 1024 lp
16749 109 newconfig, 0 96 cron
16749 129 share, 0 96 pcnfs
16749 138 lbin, 0 96 rexd
16749 145 contrib, 1 1024 mqueue
16749 148 conf, 0 96 rwho
16749 152 ccs, 0 96 locks
16749 155 obam, 0 96 uucppublic
16749 697 include, 0 96 sw

Automatic method parameter usage check.
Expect a 'DBI ... invalid' error and a 'Usage: ...' message:
DBI commit: invalid number of parameters: handle + 1
Usage: $h->commit() at test.pl line 124.

Preparing new $cursor_a to replace current $cursor_a.
We enable debugging on current to watch it's destruction.
Expect several lines of DBI trace information:
   DBI::st=HASH(0x40105e78) debug level set to 2
   -> DESTROY for DBD::ExampleP::st (DBI::st=HASH(0x40105e78))
   <- DESTROY= undef

Executing via func redirect: $h->func(..., 'execute')

Binding columns of $cursor_a to variables.

Fetching one row from new $cursor_a with a bound column.
Expect a large number follwed by a dot:
836219748 .
bind_col worked

Cursor tests done (scoped objects will be destroyed now)
test.pl done (global destruction will follow)

Now, after installing DBI with make install, I am trying to install DBD-Oracle
with :

./DBI-0.69/perl Makefile.PL
make
make perl
make test FULLPERL=./perl (or make test)

I have the following message :

Writing "Makefile.aperl" for this perl
Can't load module DBI, dynamic loading not available in this perl.
 (You may need to build a new perl executable which either supports
 dynamic loading or has the DBI module statically linked into it.)
at /usr/local/lib/perl5/site_perl/DBI.pm line 32
BEGIN failed--compilation aborted at Makefile.PL line 14.
*** Error exit code 2

Any idea about this problem. It's strange to get this problem since Perl 5 as
been compiled with dynamic loading support and DBI module is statically linked
into it.

Thanks in advance for your help.        Gildas.


--
Gildas PERROT, [email protected]         __o
FranceNet, 28 rue Desaix, 75015 Paris ---_ \<,_
                                  ---- (_)/ (_)


---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <10239-0@oink>;
         Mon, 1 Jul 1996 15:41:49 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836227594:19735:1; Mon, 01 Jul 96 14:26:34 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa19435; 1 Jul 96 14:26 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA23573 for dbi-users-real; Mon, 1 Jul 1996 06:21:20 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id GAA23569
         for <[email protected]>; Mon, 1 Jul 1996 06:21:18 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id aa15060; 1 Jul 96 13:20 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac09059; 1 Jul 96 14:07 +0100
Received: from toad by oink with SMTP (PP) id <09353-0@oink>;
         Mon, 1 Jul 1996 12:56:14 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA16994;
         Mon, 1 Jul 1996 12:56:01 +0000
Date: Mon, 1 Jul 1996 12:56:01 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Problem running DBD Oraperl emulation with Perl 5.003
X-Sun-Charset: US-ASCII
content-length: 186
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Tim Bunce <[email protected]>
>
> Mort probably you didn't run 'make install'. Do so and what the paths printed

Sigh. I should try reading what I write sometimes.

:-)

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <10239-2@oink>;
         Mon, 1 Jul 1996 15:41:57 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836228976:03538:1; Mon, 01 Jul 96 14:49:36 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa02789; 1 Jul 96 14:48 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA23612 for dbi-users-real; Mon, 1 Jul 1996 06:45:47 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id GAA23608
         for <[email protected]>; Mon, 1 Jul 1996 06:45:44 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ad01492; 1 Jul 96 14:45 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab09059; 1 Jul 96 14:07 +0100
Received: from toad by oink with SMTP (PP) id <09331-0@oink>;
         Mon, 1 Jul 1996 12:55:19 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA16989;
         Mon, 1 Jul 1996 12:55:09 +0000
Date: Mon, 1 Jul 1996 12:55:09 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: FW: Problem running DBD Oraperl emulation with Perl 5.003
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 530
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: HarveyDR <[email protected]>
>
> Tim:
>
> Thanks for your rapid response.  The whole Perl package is a great productivity
> tool for the programming world, and it's terrific having people like you to
> support it.
>
> I was able to solve the problem by building DBI and DBD-Oracle layers
> statically.  In fact, the DB-Oracle-0.35 README indicates that Oracle does not
> supply the required OCI shared libraries for HP-UX.

Umm, I thought HP-UX 10 should still be able to build a dynamic extension.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <10239-7@oink>;
         Mon, 1 Jul 1996 15:42:24 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836231477:29372:3; Mon, 01 Jul 96 15:31:17 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa29197; 1 Jul 96 15:30 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA23652 for dbi-users-real; Mon, 1 Jul 1996 07:24:46 -0700
Precedence: junk
Sender: [email protected]
Received: from x4u2.desy.de (x4u2.desy.de [131.169.30.42])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id HAA23647
         for <[email protected]>; Mon, 1 Jul 1996 07:23:39 -0700
Received: (from hfalken@localhost) by x4u2.desy.de (8.6.10/8.6.9) id OAA10439;
         Mon, 1 Jul 1996 14:17:19 GMT
From: Harald Falkenberg <[email protected]>
Message-Id: <[email protected]>
Subject: Re: can't build DBD on HP
To: [email protected] (Tim Bunce)
Date: Mon, 1 Jul 1996 16:17:17 +0200 (MST)
Cc: [email protected]
In-Reply-To: <[email protected]> from "Tim Bunce" at Jun 27, 96 04:52:59 pm
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 9783

Hi,

I followed your advice, but there are still problems during the 'make test'
step for DBD 0.35 on our HP/UX (OS 9.05). Therefore I appended again the
log of my screen to this mail. I hope, that help is possibile.

regards
       Harald Falkenberg

:>
:>
:>> From: Harald Falkenberg <[email protected]>
:>>
:>> Hi,
:>>
:>> I have problems to link DBD on a HP/UX with the OS 9.05. I tried it several
:>> times with dynamic and static link options (LINKTYPE=...), but it failed.
:>> For some reasons, some functions are not found at link time.
:>> I append my full log to this mail. If someone has an idea or has already
:>> solved this problem - please give me an answer.
:>
:>Upgrade from to DBD::Oracle 0.34 and try again.
:>
:>Tim.
:>


perl Makefile.PL  
   Remember to actually *read* the README file!

   Build, TEST and INSTALL Perl 5 first. Use version 5.002beta2 or later.

   DO NOT build the DBI under the Perl source tree.

   Use 'make test' to execute self tests.

   Install DBI (using 'make install' before building DBD::* drivers.

MakeMaker (v5.21)
Checking if your kit is complete...
Looks good
       NAME => q[DBI]
       VERSION_FROM => q[DBI.pm]
       clean => { FILES=>q[$(DISTVNAME)/] }
       dist => { DIST_DEFAULT=>q[clean distcheck disttest ci tardist], PREOP=>q[$(MAKE) -f Makefile.old distdir] }
DBI VERSION is 0.70 (from DBI.pm)
Using PERL=/usr/local/bin/perl
Writing Makefile for DBI
hfalken - /afs/.desy.de/products/source/oraperl/db/DBI-0.70 :> make
       umask 0 && cp DBIXS.h ./blib/arch/DBI/DBIXS.h
       umask 0 && cp lib/DBD/ExampleP.pm ./blib/lib/DBD/ExampleP.pm
       umask 0 && cp lib/DBD/NullP.pm ./blib/lib/DBD/NullP.pm
       umask 0 && cp lib/DBD/Sponge.pm ./blib/lib/DBD/Sponge.pm
       umask 0 && cp DBI.pm ./blib/lib/DBI.pm
       /usr/local/bin/perl -I/usr/local/lib/perl5.002 -I/usr/local/lib/perl5.002 /usr/local/lib/perl5.002/ExtUtils/xsubpp  -typemap /usr/local/lib/perl5.002/ExtUtils/typemap DBI.xs >DBI.tc && mv DBI.tc DBI.c
       cc -c  -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.70\" -DXS_VERSION=\"0.70\" +z -I/usr/local/lib/perl5.002/CORE  DBI.c
Running Mkbootstrap for DBI ()
       chmod 644 DBI.bs
       LD_RUN_PATH="" ld -o ./blib/arch/auto/DBI/DBI.sl -b -L/usr/local/lib DBI.o
       chmod 755 ./blib/arch/auto/DBI/DBI.sl
       cp DBI.bs ./blib/arch/auto/DBI/DBI.bs
       chmod 644 ./blib/arch/auto/DBI/DBI.bs
hfalken - /afs/.desy.de/products/source/oraperl/db/DBI-0.70 :> make test
       PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5.002 -I/usr/local/lib/perl5.002 -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t
t/min...............DBD::ExampleP in /usr/local/lib/perl5.002 is hidden by DBD::ExampleP in ./blib/lib
at t/min.t line 16
DBD::NullP in /usr/local/lib/perl5.002 is hidden by DBD::NullP in ./blib/lib
at t/min.t line 16
DBD::Sponge in /usr/local/lib/perl5.002 is hidden by DBD::Sponge in ./blib/lib
at t/min.t line 16
ok
All tests successful.
Files=1,  Tests=12,  1 secs ( 0.23 cusr  0.08 csys =  0.31 cpu)
       PERL_DL_NONLAZY=1 /usr/local/bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5.002 -I/usr/local/lib/perl5.002 test.pl
test.pl
DBI test application $Revision: 1.14 $
Use of uninitialized value at /usr/local/lib/perl5.002/Getopt/Long.pm line 602.
Switch: DBI-0.70 Switch by Tim Bunce, 0.70
DebugDispatch: 0
DBD::ExampleP in /usr/local/lib/perl5.002 is hidden by DBD::ExampleP in ./blib/lib
at test.pl line 49
DBD::NullP in /usr/local/lib/perl5.002 is hidden by DBD::NullP in ./blib/lib
at test.pl line 49
DBD::Sponge in /usr/local/lib/perl5.002 is hidden by DBD::Sponge in ./blib/lib
at test.pl line 49
Available Drivers: ExampleP, NullP, Sponge
Read DBI special variables (expect 0, 99, 99):
err:
err:    99
errstr: 99
Attempt to modify DBI special variables.
Expect a 'Can't modify' error message:
Can't modify $DBI::&rows special variable at test.pl line 59

Connected as DBI::db=HASH(0x400fbdb8)

Test error handling: prepare()'ing an invalid query.
Expect prepare to return UNDEF and an error message:
prepare returned UNDEF. $DBI::err=Unknown field names: unknown_field_name

Cursor prepare'd as DBI::st=HASH(0x400fbed8)
Prepared as DBI::st=HASH(0x400fbf8c)
Number of fields: 3
Data type of first field: 1
Driver name: ExampleP

Fetching data from both cursors.
Expect several rows of data:
16749 78080 ., 1 1024 .
16877 2 .., 1 1024 ..
16749 79360 lib, 1 24 rexd
16749 144646 bin, 1 1024 cron
16895 108841 tmp, 1 1024 sockets
16895 512688150 man, 1 1024 rwho
16749 38417 spool, 1 1024 mqueue
16749 104967 etc, 1 1024 lp
16749 116516 netdemo, 1 24 uucppublic
16895 512688136 include, 1 1024 uucp
16749 7 nettest, 1 24 ppl

Automatic method parameter usage check.
Expect a 'DBI ... invalid' error and a 'Usage: ...' message:
DBI commit: invalid number of parameters: handle + 1
Usage: $h->commit() at test.pl line 124.

Preparing new $cursor_a to replace current $cursor_a.
We enable debugging on current to watch it's destruction.
Expect several lines of DBI trace information:
   DBI::st=HASH(0x400fbfb0) debug level set to 2
   -> DESTROY for DBD::ExampleP::st (DBI::st=HASH(0x400fbfb0))
   <- DESTROY= undef

Executing via func redirect: $h->func(..., 'execute')

Binding columns of $cursor_a to variables.

Fetching one row from new $cursor_a with a bound column.
Expect a large number follwed by a dot:
836227190 .
bind_col worked

Cursor tests done (scoped objects will be destroyed now)
test.pl done (global destruction will follow)

hfalken - /afs/.desy.de/products/source/oraperl/db/DBI-0.70 :> cd ../DBD-Oracle-0.3
DBD-Oracle-0.32/          DBD-Oracle-0.35/
DBD-Oracle-0.32.tar.Z     DBD-Oracle-0.35.tar.Z
hfalken - /afs/.desy.de/products/source/oraperl/db/DBI-0.70 :> cd ../DBD-Oracle-0.35/
hfalken - /afs/.desy.de/products/source/oraperl/db/DBD-Oracle-0.35 :> perl Makefile.PL

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6
Using /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/proc/lib/proc.mk (version 1.17)
Warning: forced to build static not dynamic on hpux
        See README and Makefile.PL for more information.
System: perl5.002 hp-ux dirac a.09.05 a 9000735 2005669792 two-user license
Compiler: cc -O -D_HPUX_SOURCE -Aa -I/usr/local/include
Linker:   /bin/ld
Oracle proc.mk would have used these values but we override them:
 CC:       cc
 CFLAGS:   -I. -O -Aa -D_HPUX_SOURCE +ESsfc +ESlit
 LDFLAGS:  -Wl,-aarchive -L$(LIBHOME)
Linking with $(LIBHOME)/osntab.o -locic  -lsqlnet -lora -lsqlnet -lora -lnlsrtl -lcv6 -lcore -lnlsrtl -lcv6 -lcore -lcl -lm

Checking if your kit is complete...
Looks good
Writing Makefile for DBD::Oracle
hfalken - /afs/.desy.de/products/source/oraperl/db/DBD-Oracle-0.35 :> perl Makefile.PLmake        umask 0 && cp Oraperl.pm ./blib/lib/Oraperl.pm
       umask 0 && cp oraperl.ph ./blib/lib/oraperl.ph
       umask 0 && cp Oracle.pm ./blib/lib/DBD/Oracle.pm
       /usr/local/bin/perl -I/usr/local/lib/perl5.002 -I/usr/local/lib/perl5.002 /usr/local/lib/perl5.002/ExtUtils/xsubpp  -typemap /usr/local/lib/perl5.002/ExtUtils/typemap Oracle.xs >Oracle.tc && mv Oracle.tc Oracle.c
       cc -c -I/afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/rdbms/demo -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.35\" -DXS_VERSION=\"0.35\" +z -I/usr/local/lib/perl5.002/CORE  Oracle.c
       cc -c -I/afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/rdbms/demo -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -I/afs/.desy.de/products/perl/perl5.002/lib/DBI -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.35\" -DXS_VERSION=\"0.35\" +z -I/usr/local/lib/perl5.002/CORE  dbdimp.c
       ar cr ./blib/arch/auto/DBD/Oracle/Oracle.a Oracle.o  dbdimp.o && : ./blib/arch/auto/DBD/Oracle/Oracle.a
       chmod 755 ./blib/arch/auto/DBD/Oracle/Oracle.a
Manifying ./blib/man3/DBD::Oraperl.3
hfalken - /afs/.desy.de/products/source/oraperl/db/DBD-Oracle-0.35 :> make
hfalken - /afs/.desy.de/products/source/oraperl/db/DBD-Oracle-0.35 :> make test
Writing "Makefile.aperl" for this perl

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6
Using /afs/desy.de/products/oracle/hp700_ux90/product/7.1.6/proc/lib/proc.mk (version 1.17)
Warning: forced to build static not dynamic on hpux
        See README and Makefile.PL for more information.
System: perl5.002 hp-ux dirac a.09.05 a 9000735 2005669792 two-user license
Compiler: cc -O -D_HPUX_SOURCE -Aa -I/usr/local/include
Linker:   /bin/ld
Oracle proc.mk would have used these values but we override them:
 CC:       cc
 CFLAGS:   -I. -O -Aa -D_HPUX_SOURCE +ESsfc +ESlit
 LDFLAGS:  -Wl,-aarchive -L$(LIBHOME)
Linking with $(LIBHOME)/osntab.o -locic  -lsqlnet -lora -lsqlnet -lora -lnlsrtl -lcv6 -lcore -lnlsrtl -lcv6 -lcore -lcl -lm

Writing Makefile.aperl for DBD::Oracle
       make -f Makefile.aperl static perl
       umask 0 && cp Oraperl.pm ./blib/lib/Oraperl.pm
       umask 0 && cp oraperl.ph ./blib/lib/oraperl.ph
       umask 0 && cp Oracle.pm ./blib/lib/DBD/Oracle.pm
Writing ./perlmain.c
       cd . && cc -c  -I/usr/local/lib/perl5.002/CORE  -D_HPUX_SOURCE -Aa -I/usr/local/include -O -DVERSION=\"0.35\" -DXS_VERSION=\"0.35\"  -I/usr/local/lib/perl5.002/CORE perlmain.c
       cat ./blib/arch/auto/DBD/Oracle/extralibs.ld >> ./blib/arch/auto/DBD/Oracle/extralibs.all
       cat /afs/desy.de/product
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11570-0@oink>;
         Mon, 1 Jul 1996 17:31:28 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836238417:16450:1; Mon, 01 Jul 96 17:26:57 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa15950; 1 Jul 96 17:26 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa22855; 1 Jul 96 17:25 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA23970 for dbi-users-real; Mon, 1 Jul 1996 09:18:29 -0700
Precedence: junk
Sender: [email protected]
Received: from netserve.mar.lmco.com (netserve.lasc.lockheed.com [134.5.45.28])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA23966
         for <[email protected]>; Mon, 1 Jul 1996 09:18:27 -0700
Message-Id: <[email protected]>
Received: from mip.mar.lmco.com by netserve.mar.lmco.com
         with SMTP (1.39.111.2/16.2) id AA214377855;
         Mon, 1 Jul 1996 12:17:35 -0400
Received: by mip.mar.lmco.com (1.38.193.4/16.2) id AA17771;
         Mon, 1 Jul 1996 12:17:38 -0400
Date: Mon, 1 Jul 1996 12:17:38 -0400
From: Terry Greenlaw <[email protected]>
To: [email protected]
Subject: Re: FW: Problem running DBD Oraperl emulation with Perl 5.003

> > From: HarveyDR <[email protected]>
> >
> > Tim:
> >
> > Thanks for your rapid response.  The whole Perl package is a great productivity
> > tool for the programming world, and it's terrific having people like you to
> > support it.
> >

Ditto tenfold...

> > I was able to solve the problem by building DBI and DBD-Oracle layers
> > statically.  In fact, the DB-Oracle-0.35 README indicates that Oracle does not
> > supply the required OCI shared libraries for HP-UX.

> Umm, I thought HP-UX 10 should still be able to build a dynamic extension.

> Tim.

HP-UX 10 and Oracle 7.2.x do work together when creating dynamic libraries.
The problem was older Oracle libraries were built without the +z flag to cc,
and were therefore position-dependent libraries that can't be linked
dynamically. Newer Oracle releases don't have this problem and it may be
possible to even use the newer Oracle libraries under HP-UX 9. Oracle 7.3
will ONLY work under HP-UX 10, however.

tog

-----------------------------v------------v------------------------------
Onsite @ Lockheed Martin     |   Terry    |        Encompass Technologies
[email protected]      |            |       [email protected]
http://www.lmco.com          |  Greenlaw  |  http://www.encompasstech.com
-----------------------------^------------^------------------------------

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <11714-0@oink>;
         Mon, 1 Jul 1996 17:50:47 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836239664:27007:4; Mon, 01 Jul 96 17:47:44 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa26850; 1 Jul 96 17:47 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA24036 for dbi-users-real; Mon, 1 Jul 1996 09:40:40 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA24032
         for <[email protected]>; Mon, 1 Jul 1996 09:40:35 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id af29901; 1 Jul 96 16:40 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa04697; 1 Jul 96 17:13 +0100
Received: from toad by oink with SMTP (PP) id <11472-0@oink>;
         Mon, 1 Jul 1996 17:12:59 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA19833;
         Mon, 1 Jul 1996 17:12:46 +0000
Date: Mon, 1 Jul 1996 17:12:46 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: can't build DBD on HP
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 382
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Harald Falkenberg <[email protected]>
>
> Hi,
>
> I followed your advice, but there are still problems during the 'make test'
> step for DBD 0.35 on our HP/UX (OS 9.05). Therefore I appended again the
> log of my screen to this mail. I hope, that help is possibile.

The message was truncated. Please resend (don't bother including the large
DBI test output).

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <11714-2@oink>;
         Mon, 1 Jul 1996 17:51:05 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836239736:27900:1; Mon, 01 Jul 96 17:48:56 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa26890; 1 Jul 96 17:47 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA24042 for dbi-users-real; Mon, 1 Jul 1996 09:41:00 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA24038
         for <[email protected]>; Mon, 1 Jul 1996 09:40:57 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ag22720; 1 Jul 96 17:40 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa11673; 1 Jul 96 17:36 +0100
Received: from toad by oink with SMTP (PP) id <11591-0@oink>;
         Mon, 1 Jul 1996 17:35:23 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA20189;
         Mon, 1 Jul 1996 17:35:13 +0000
Date: Mon, 1 Jul 1996 17:35:13 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: HP-UX DBD::Oracle
X-Sun-Charset: US-ASCII
content-length: 511
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Terry Greenlaw <[email protected]>
>
> HP-UX 10 and Oracle 7.2.x do work together when creating dynamic libraries.
> The problem was older Oracle libraries were built without the +z flag to cc,
> and were therefore position-dependent libraries that can't be linked
> dynamically. Newer Oracle releases don't have this problem and it may be
> possible to even use the newer Oracle libraries under HP-UX 9. Oracle 7.3
> will ONLY work under HP-UX 10, however.

Thanks for the clarification.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <19565-0@oink>;
         Tue, 2 Jul 1996 17:20:58 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836324300:05889:1; Tue, 02 Jul 96 17:18:20 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa05161; 2 Jul 96 17:16 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA25875 for dbi-users-real; Tue, 2 Jul 1996 08:57:12 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id IAA25871
         for <[email protected]>; Tue, 2 Jul 1996 08:57:03 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id am20140; 2 Jul 96 16:56 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ae16416; 2 Jul 96 16:21 +0100
Received: from toad by oink with SMTP (PP) id <17536-0@oink>;
         Tue, 2 Jul 1996 12:49:37 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA23566;
         Tue, 2 Jul 1996 12:49:31 +0000
Date: Tue, 2 Jul 1996 12:49:31 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DBD-Oracle-0.32 test failed on HP/UX 10.10
X-Sun-Charset: US-ASCII
content-length: 1280
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Gildas Perrot <[email protected]>
>
> Hi,
>
> I am trying to install DBD-Oracle-0.32 on HP/UX 10.10.

Try DBD-Oracle-0.35.

> I have installed DBI-0.69 first dynamically and then statiscally since
> DBD-Oracle was complaining about that.

There should be no need. Dynamic should work fine on HP-UX 10.
I suggest you get that working first.

> Now, after installing DBI with make install, I am trying to install DBD-Oracle
> with :
>
> ../DBI-0.69/perl Makefile.PL
> make
> make perl
> make test FULLPERL=./perl (or make test)
>
> I have the following message :
>
> Writing "Makefile.aperl" for this perl
> Can't load module DBI, dynamic loading not available in this perl.
>   (You may need to build a new perl executable which either supports
>   dynamic loading or has the DBI module statically linked into it.)
>  at /usr/loc

> Any idea about this problem. It's strange to get this problem since Perl 5 as
> been compiled with dynamic loading support and DBI module is statically linked
> into it.

al/lib/perl5/site_perl/DBI.pm line 32

The perl you use to build DBD::Oracle when not using dynamic loading must be
a perl binary with DBI statically linked in. But since you're on HP-UX 10
I suggest you get dynamic loading of the DBI working first.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21835-0@oink>;
         Wed, 3 Jul 1996 07:10:20 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836350038:15891:0; Wed, 03 Jul 96 00:27:18 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa15750; 3 Jul 96 0:26 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa03911; 3 Jul 96 0:22 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA26633 for dbi-users-real; Tue, 2 Jul 1996 16:10:25 -0700
Precedence: junk
Sender: [email protected]
Received: from lightning.synoptics.com (lightning.synoptics.com [134.177.3.18])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA26629
         for <[email protected]>; Tue, 2 Jul 1996 16:10:24 -0700
Received: from pobox ([134.177.1.95]) by lightning.synoptics.com (4.1/SMI-4.1)
         id AA03652; Tue, 2 Jul 96 16:09:50 PDT
Received: from sc-mail1.corpwest.BayNetworks.com by pobox (4.1/SMI-4.1)
         id AA07454; Tue, 2 Jul 96 16:09:49 PDT
Received: from mailless (mailless.corpwest.baynetworks.com [134.177.59.194])
         by sc-mail1.corpwest.BayNetworks.com (post.office MTA v2.0 0529 ID# 0-13459)
         with SMTP id AAA13786; Tue, 2 Jul 1996 16:08:55 -0700
Message-Id: <[email protected]>
Date: Tue, 02 Jul 1996 16:08:04 -0700
From: David LeVine <[email protected]>
Organization: Bay Networks
X-Mailer: Mozilla 3.0b4Gold (X11; I; SunOS 5.5 sun4m)
Mime-Version: 1.0
To: [email protected]
Cc: [email protected]
Subject: DBD-Oracle-0.35 problem
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi,

I just downloaded perl5.002, DBI 0.69 and DBD-Oracle-0.35.  The
environment is Solaris 2.4 on a sparc.  Perl and DBI built, tested
and installed fine, however, Oracle 0.35 complains that it wants DBI
0.70.  The README says it needs DBI 0.69 or better, but the Makefile.PL
is pretty explicit with a 'use DBI 0.70' line.  Is the README correct
and the Makefile.PL in error or is there a DBI 0.70 available to
download?

Many thanks.

[email protected]
-----------------------
% perl -V
Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=solaris, osver=2.4, archname=sun4-solaris
   uname='sunos simba 5.4 generic_101945-34 sun4m sparc '
   hint=recommended, useposix=true
 Compiler:
   cc='gcc', optimize='-O', gccversion=2.7.0
   cppflags=''
   ccflags =''
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='gcc', ldflags =''
   libpth=/lib /usr/lib /usr/ccs/lib
   libs=-lsocket -lnsl -ldl -lm -lc -lcrypt
   libc=/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-fpic', lddlflags='-G'

@INC: /import/appdist/pkgs/perl5.002/solaris/lib/sun4-solaris/5.002
/import/appdist/pkgs/perl5.002/solaris/lib
/import/appdist/pkgs/perl5.002/solaris/lib/site_perl/sun4-solaris
/import/appdist/pkgs/perl5.002/solaris/lib/site_perl .
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21835-7@oink>;
         Wed, 3 Jul 1996 07:11:33 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836356054:01947:1; Wed, 03 Jul 96 02:07:34 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa01506; 3 Jul 96 2:06 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id SAA26894 for dbi-users-real; Tue, 2 Jul 1996 18:02:16 -0700
Precedence: junk
Sender: [email protected]
Received: from cs.pdx.edu ([email protected] [204.203.64.22])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id SAA26890
         for <[email protected]>; Tue, 2 Jul 1996 18:02:15 -0700
Received: (xuhao@localhost) by cs.pdx.edu (8.7.3/CATastrophe-2/10/96-P)
         id SAA13607; Tue, 2 Jul 1996 18:02:05 -0700 (PDT) for
From: [email protected] (Hao Xu)
Message-Id: <[email protected]>
Subject: Question (can't compile )
Date: Tue, 2 Jul 1996 18:02:04 -0700 (PDT)
Cc: [email protected]
X-Mailer: ELM [version 2.4 PL13]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

Hello:

I downloaded DBI-0.69, and got Makefile.PL

I changed the first line of the script to: "#/usr/bin/perl5 -w"
But I failed even for the first thing, and here is the log:

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
%perl5 Makefile.PL
Argument "(undef)" isn't numeric for numeric lt at /usr/contrib/lib/perl5/Exporter.pm line 177
       Exporter::require_version called at /usr/contrib/lib/perl5/Exporter.pm line 125
       Exporter::export called at /usr/contrib/lib/perl5/Exporter.pm line 163
       Exporter::import called at Makefile.pl line 6
       main::BEGIN called at /usr/contrib/lib/perl5/ExtUtils/MakeMaker.pm line 0
       (eval) called at /usr/contrib/lib/perl5/ExtUtils/MakeMaker.pm line 0
ExtUtils::MakeMaker 5.16 required--this is only version (undef) at Makefile.pl line 6
BEGIN failed--compilation aborted at Makefile.pl line 6.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

I got nearly the same message when execute the Makefile.PL in
DBD-oracle-0.32 directory.


I would appreciate much if you can give me a hint!

Thank you very much!

-Hao
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21835-24@oink>;
         Wed, 3 Jul 1996 07:14:25 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836358961:20468:1; Wed, 03 Jul 96 02:56:01 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa20173; 3 Jul 96 2:55 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA26734 for dbi-users-real; Tue, 2 Jul 1996 16:44:00 -0700
Precedence: junk
Sender: [email protected]
Received: from lightning.synoptics.com (lightning.synoptics.com [134.177.3.18])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id QAA26730
         for <[email protected]>; Tue, 2 Jul 1996 16:43:59 -0700
Received: from pobox ([134.177.1.95]) by lightning.synoptics.com (4.1/SMI-4.1)
         id AA04428; Tue, 2 Jul 96 16:42:05 PDT
Received: from sc-mail1.corpwest.BayNetworks.com by pobox (4.1/SMI-4.1)
         id AA08674; Tue, 2 Jul 96 16:42:05 PDT
Received: from mailless (mailless.corpwest.baynetworks.com [134.177.59.194])
         by sc-mail1.corpwest.BayNetworks.com (post.office MTA v2.0 0529 ID# 0-13459)
         with SMTP id AAA15730; Tue, 2 Jul 1996 16:41:10 -0700
Message-Id: <[email protected]>
Date: Tue, 02 Jul 1996 16:40:20 -0700
From: David LeVine <[email protected]>
Organization: Bay Networks
X-Mailer: Mozilla 3.0b4Gold (X11; I; SunOS 5.5 sun4m)
Mime-Version: 1.0
To: [email protected], [email protected]
Subject: Re: DBD-Oracle-0.35 problem
References: <[email protected]>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Please ignore my question.  I've already received appropriate
pointers to a newer version of DBI and perl for that matter.

David LeVine wrote:
>
> Hi,
>
> I just downloaded perl5.002, DBI 0.69 and DBD-Oracle-0.35.  The
> environment is Solaris 2.4 on a sparc.  Perl and DBI built, tested
> and installed fine, however, Oracle 0.35 complains that it wants DBI
> 0.70.  The README says it needs DBI 0.69 or better, but the Makefile.PL
> is pretty explicit with a 'use DBI 0.70' line.  Is the README correct
> and the Makefile.PL in error or is there a DBI 0.70 available to
> download?
>
> Many thanks.
>
> [email protected]
> -----------------------
> % perl -V
> Summary of my perl5 (5.0 patchlevel 2) configuration:
>   Platform:
>     osname=solaris, osver=2.4, archname=sun4-solaris
>     uname='sunos simba 5.4 generic_101945-34 sun4m sparc '
>     hint=recommended, useposix=true
>   Compiler:
>     cc='gcc', optimize='-O', gccversion=2.7.0
>     cppflags=''
>     ccflags =''
>     stdchar='unsigned char', d_stdstdio=define, usevfork=false
>     voidflags=15, castflags=0, d_casti32=define, d_castneg=define
>     intsize=4, alignbytes=8, usemymalloc=y, randbits=15
>   Linker and Libraries:
>     ld='gcc', ldflags =''
>     libpth=/lib /usr/lib /usr/ccs/lib
>     libs=-lsocket -lnsl -ldl -lm -lc -lcrypt
>     libc=/lib/libc.so, so=so
>   Dynamic Linking:
>     dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
>     cccdlflags='-fpic', lddlflags='-G'
>
> @INC: /import/appdist/pkgs/perl5.002/solaris/lib/sun4-solaris/5.002
> /import/appdist/pkgs/perl5.002/solaris/lib
> /import/appdist/pkgs/perl5.002/solaris/lib/site_perl/sun4-solaris
> /import/appdist/pkgs/perl5.002/solaris/lib/site_perl .
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <24815-0@oink>;
         Wed, 3 Jul 1996 10:01:53 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836380704:23857:2; Wed, 03 Jul 96 08:58:24 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa23496; 3 Jul 96 8:57 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id AAA27329 for dbi-users-real; Wed, 3 Jul 1996 00:49:39 -0700
Precedence: junk
Sender: [email protected]
Received: from anna.in-berlin.de ([193.175.21.66])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id AAA27325
         for <[email protected]>; Wed, 3 Jul 1996 00:49:37 -0700
Received: by anna.in-berlin.de (940816.SGI.8.6.9/940406.SGI) id JAA03341;
         Wed, 3 Jul 1996 09:43:45 +0200
Date: Wed, 3 Jul 1996 09:43:45 +0200
Message-Id: <[email protected]>
From: Andreas Koenig <[email protected]>
To: [email protected]
CC: [email protected], [email protected]
In-reply-to: <[email protected]> (message from David LeVine on Tue, 02 Jul 1996 16:08:04 -0700)
Subject: Re: DBD-Oracle-0.35 problem
Reply-to: [email protected]

CPAN/modules/01modules.index.html contains the line (broken for readablility):

DBI
<A HREF="../authors/Tim_Bunce">TIMB</A>
<A HREF="../authors/id/TIMB/DBI/DBI-0.70.tar.gz">DBI-0.70.tar.gz</A>
34k  16 Jun 1996


Which means, that DBI 0.70 is on CPAN since Jun, 16. If your CPAN site
doesn't have it, please ask the local administrator, why. You find the
local administrator's address in CPAN/MIRRORED.BY.


andreas
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <26109-14@oink>;
         Wed, 3 Jul 1996 14:41:51 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836398813:16938:17; Wed, 03 Jul 96 14:00:13 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15813; 3 Jul 96 13:59 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id FAA27612 for dbi-users-real; Wed, 3 Jul 1996 05:50:52 -0700
Precedence: junk
Sender: [email protected]
Received: from urges.urlabs.com (urges.urlabs.com [206.161.152.65])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id FAA27608
         for <[email protected]>; Wed, 3 Jul 1996 05:50:50 -0700
Received: from pandemonium.urlabs.com (ppp-nhr-1-6.nhr.com [206.151.14.134])
         by urges.urlabs.com (8.7.5/8.7.3/urlabs) with SMTP id IAA03644;
         Wed, 3 Jul 1996 08:46:24 -0400 (EDT)
Message-ID: <[email protected]>
Date: Wed, 03 Jul 1996 08:46:53 -0400
From: Travis L Priest <[email protected]>
Organization: Unified Research Laboratories, Inc.
X-Mailer: Mozilla 3.0b4 (Win95; I)
MIME-Version: 1.0
To: Tim Bunce <[email protected]>
CC: [email protected]
Subject: Re: DBD::Oracle
References: <[email protected]>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Tim Bunce wrote:
> > It would be very convenient to have the embedded sql command describe
> > usable from oraperl, as in:
> >
> > $csr = &ora_open($dbh, 'describe user.table');
> > while ( ($name, $type, $size) = &ora_fetch($csr) ) {
> >   print "Column name $name has type $type of size $size\n";
> > }
> >
> > I could use this functionality right now to create a generic cgi script
> > to allow you to modify fields on any table (a WWW table editor).
> > It would also be convient for making a reasonably generic db loading
> > script.
>
> Please send mail via the dbi-users list.
>
> Describe is an SQL*Plus command not embedded sql.

Ah, sorry, there's a documented embedded SQL describe statement, but
upon closer inspection it is not exactly what I want (the sql*plus behavior
is what I'm after).

What's done in dbdimp.c:dbd_describe()?  It looks to me like you are getting
some information from the database to determine how much storage to allocate
for returned rows, as well as type information.  Is there any way to mold
that information into a call like the one described above?

Travis
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <26236-0@oink>;
         Wed, 3 Jul 1996 14:58:35 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836402220:18416:1; Wed, 03 Jul 96 14:57:00 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa17716; 3 Jul 96 14:55 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA27678 for dbi-users-real; Wed, 3 Jul 1996 06:49:40 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id GAA27674
         for <[email protected]>; Wed, 3 Jul 1996 06:49:38 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ad03818; 3 Jul 96 14:49 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa04001; 3 Jul 96 14:45 +0100
Received: from toad by oink with SMTP (PP) id <26189-0@oink>;
         Wed, 3 Jul 1996 14:44:24 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA01758;
         Wed, 3 Jul 1996 14:44:10 +0000
Date: Wed, 3 Jul 1996 14:44:10 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: DBD::Oracle
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1283
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Travis L Priest <[email protected]>
>
> Tim Bunce wrote:
> > > It would be very convenient to have the embedded sql command describe
> > > usable from oraperl, as in:
> > >
> > > $csr = &ora_open($dbh, 'describe user.table');
> > > while ( ($name, $type, $size) = &ora_fetch($csr) ) {
> > >   print "Column name $name has type $type of size $size\n";
> > > }
> > >
> > > I could use this functionality right now to create a generic cgi script
> > > to allow you to modify fields on any table (a WWW table editor).
> > > It would also be convient for making a reasonably generic db loading
> > > script.
> >
> > Please send mail via the dbi-users list.
> >
> > Describe is an SQL*Plus command not embedded sql.
>
> Ah, sorry, there's a documented embedded SQL describe statement, but
> upon closer inspection it is not exactly what I want (the sql*plus behavior
> is what I'm after).
>
> What's done in dbdimp.c:dbd_describe()?  It looks to me like you are getting
> some information from the database to determine how much storage to allocate
> for returned rows, as well as type information.  Is there any way to mold
> that information into a call like the one described above?

select * from user.table then read the Oraperl manual for ora_types etc.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <26711-0@oink>;
         Wed, 3 Jul 1996 15:24:47 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836403698:03721:1; Wed, 03 Jul 96 15:21:38 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa02105; 3 Jul 96 15:20 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA27706 for dbi-users-real; Wed, 3 Jul 1996 07:12:48 -0700
Precedence: junk
Sender: [email protected]
Received: from bsfiles.nerc-bas.ac.uk (bsfiles.nerc-bas.ac.uk [192.171.137.25])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id HAA27702
         for <[email protected]>; Wed, 3 Jul 1996 07:12:45 -0700
Received: from bssmlsa.nerc-bas.ac.uk.nerc-bas.ac.uk
         by bsfiles.nerc-bas.ac.uk (8.6.10/NERC-1.2(Solaris 2.x) id PAA29106;
         Wed, 3 Jul 1996 15:11:47 +0100
Date: Wed, 3 Jul 1996 15:11:46 +0100
Message-Id: <[email protected]>
From: Andrew G Wood <[email protected]>
To: [email protected], [email protected]
In-reply-to: <[email protected]> (message from Tim Bunce on Wed, 3 Jul 1996 14:44:10 +0000)
Subject: Re: DBD::Oracle
Reply-to: [email protected]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII


   >> From: Travis L Priest <[email protected]>
   >>
   [...]
   >>
   >> Ah, sorry, there's a documented embedded SQL describe
   >> statement, but upon closer inspection it is not exactly what I
   >> want (the sql*plus behavior is what I'm after).
   >>
   >> What's done in dbdimp.c:dbd_describe()?  It looks to me like
   >> you are getting some information from the database to determine
   >> how much storage to allocate for returned rows, as well as type
   >> information.  Is there any way to mold that information into a
   >> call like the one described above?

   > select * from user.table then read the Oraperl manual for
   > ora_types etc.

You might also be interested in the following "select" statement
which was intended to simulate an sql*plus describe .....

     $selstr =
     "select column_name \"Name\",".
     " decode(nullable,\'N\',\'NOT NULL\',\'Y\',null) \"Null\",".
     " decode(data_type,\'DATE\',data_type,".
     "\'NUMBER\',data_type||\'(\'||data_precision||\',\'||data_scale||\')\',".
     "data_type||\'(\'||data_length||\')\' ) \"Type\"".
     " from all_tab_columns".
     " where (owner,table_name) = (select table_owner,table_name".
     " from user_synonyms where synonym_name = $NAME)".
     " or (owner = $OWNER and table_name = $NAME)";

Andy.

+--------------------------------------------+
|   Dr Andy Wood, Database Administrator     |
|   British  Antarctic  Survey               |
|   High Cross, Madingley Road            +----------------------------------+
|   Cambridge,   CB3 0ET,   UK            |    [email protected]              |
|      +44 (0) 1223 361188                |    [email protected]       |
+-----------------------------------------|    [email protected]           |
                                         +----------------------------------+
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <27034-12@oink>;
         Wed, 3 Jul 1996 17:02:28 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836408101:18343:1; Wed, 03 Jul 96 16:35:01 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa18188; 3 Jul 96 16:34 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa12689; 3 Jul 96 16:33 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA27817 for dbi-users-real; Wed, 3 Jul 1996 08:27:46 -0700
Precedence: junk
Sender: [email protected]
Received: from pt.Cyanamid.COM (igate.cyanamid.com [141.173.63.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id IAA27813
         for <[email protected]>; Wed, 3 Jul 1996 08:27:45 -0700
Received: from jmsun by apbr1.pt.Cyanamid.COM (4.1/SMI-4.1) id AA25322;
         Wed, 3 Jul 96 11:27:11 EDT
Date: Wed, 3 Jul 1996 11:27:10 -0400 (EDT)
From: Jon Meek <[email protected]>
Reply-To: Jon Meek <[email protected]>
Subject: Re: DBD::Oracle - Describe
To: [email protected]
In-Reply-To: "Your message with ID" <[email protected]>
Message-Id: <Roam.3.0.1.836407630.23718.meekj@apbr1>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII

This is what I use for a SQL*Plus describe. I make two queries, the second is
just for Index information.


$sqlQuery1 =<<"EOsqlQuery1";
select COLUMN_NAME, COLUMN_ID, DATA_TYPE, DATA_LENGTH, NULLABLE
from all_tab_columns where TABLE_NAME = '$table'
order by COLUMN_ID
EOsqlQuery1

$sqlQuery2 =<<"EOsqlQuery2";
select COLUMN_NAME from all_ind_columns where TABLE_NAME = '$table'
EOsqlQuery2

Jon

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <27276-0@oink>;
         Wed, 3 Jul 1996 17:16:14 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836410475:12629:1; Wed, 03 Jul 96 17:14:35 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa12043; 3 Jul 96 17:13 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA27918 for dbi-users-real; Wed, 3 Jul 1996 09:02:14 -0700
Precedence: junk
Sender: [email protected]
Received: from voro.lbl.gov (voro.lbl.gov [131.243.240.19])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA27914
         for <[email protected]>; Wed, 3 Jul 1996 09:02:13 -0700
Received: from voro.lbl.gov (localhost [127.0.0.1]) by voro.lbl.gov (8.6.12/G)
         with ESMTP id IAA18291; Wed, 3 Jul 1996 08:58:43 -0700
Message-Id: <[email protected]>
X-Mailer: exmh version 1.6.2 7/18/95
To: Travis L Priest <[email protected]>
cc: [email protected]
Subject: Re: DBD::Oracle
In-reply-to: Your message of "Wed, 03 Jul 1996 08:46:53 EDT." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Wed, 03 Jul 1996 08:58:42 -0700
From: Mark Dedlow <[email protected]>

> From: Travis L Priest <[email protected]>
> Tim Bunce wrote:
> > > It would be very convenient to have the embedded sql command describe
> > > usable from oraperl, as in:
> > >
> > > $csr = &ora_open($dbh, 'describe user.table');
> > > while ( ($name, $type, $size) = &ora_fetch($csr) ) {
> > >   print "Column name $name has type $type of size $size\n";
> > > }
> > >
> > > I could use this functionality right now to create a generic cgi script
> > > to allow you to modify fields on any table (a WWW table editor).
> > > It would also be convient for making a reasonably generic db loading
> > > script.
> >
> > Please send mail via the dbi-users list.
> >
> > Describe is an SQL*Plus command not embedded sql.
>
> Ah, sorry, there's a documented embedded SQL describe statement, but
> upon closer inspection it is not exactly what I want (the sql*plus behavior
> is what I'm after).
>
> What's done in dbdimp.c:dbd_describe()?  It looks to me like you are getting
> some information from the database to determine how much storage to allocate
> for returned rows, as well as type information.  Is there any way to mold
> that information into a call like the one described above?
>
> Travis

Don't you just want:

select COLUMN_NAME, DATA_TYPE, DATA_LENGTH, DATA_PRECISION
from ALL_TAB_COLUMNS where TABLE_NAME = ':1'

A generic table browser/editor can be made using this.

Mark


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <27793-0@oink>;
         Wed, 3 Jul 1996 17:49:00 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836410687:14805:3; Wed, 03 Jul 96 17:18:07 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa14658; 3 Jul 96 17:17 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA27936 for dbi-users-real; Wed, 3 Jul 1996 09:10:23 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA27932
         for <[email protected]>; Wed, 3 Jul 1996 09:10:20 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ad27067; 3 Jul 96 16:09 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa12154; 3 Jul 96 17:06 +0100
Received: from toad by oink with SMTP (PP) id <27153-0@oink>;
         Wed, 3 Jul 1996 17:06:24 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA02699;
         Wed, 3 Jul 1996 17:05:59 +0000
Date: Wed, 3 Jul 1996 17:05:59 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Oracle Installation question
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 606
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424

Forwarded to dbi-users

----- Begin Included Message -----

From [email protected]  Wed Jul  3 17:03:48 1996
Date: Wed, 3 Jul 1996 09:35:52 -0600 (MDT)
From: Jonathan Olson <[email protected]>
To: Tim Bunce <[email protected]>
Subject: Oracle Installation question
In-Reply-To: <[email protected]>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset="US-ASCII"


Hello,

What is the minimum installation one must do in order to be able to use
DBD/DBI?

Do you need to set up an Oracle Server?  (Or is SQL*Net enough?)

Thanks,

Jonathan

----- End Included Message -----

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <29782-7@oink>;
         Thu, 4 Jul 1996 07:08:12 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836424421:15349:2; Wed, 03 Jul 96 21:07:01 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15219; 3 Jul 96 21:06 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA27784 for dbi-users-real; Wed, 3 Jul 1996 08:10:20 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id IAA27780
         for <[email protected]>; Wed, 3 Jul 1996 08:10:16 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ad18152; 3 Jul 96 15:09 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa24962; 3 Jul 96 16:08 +0100
Received: from toad by oink with SMTP (PP) id <26821-0@oink>;
         Wed, 3 Jul 1996 15:41:38 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA02278;
         Wed, 3 Jul 1996 15:41:32 +0000
Date: Wed, 3 Jul 1996 15:41:32 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected], [email protected]
Subject: Re: DBD::Oracle
X-Sun-Charset: US-ASCII
content-length: 1453
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Andrew G Wood <[email protected]>

>     > select * from user.table then read the Oraperl manual for
>     > ora_types etc.
>
> You might also be interested in the following "select" statement
> which was intended to simulate an sql*plus describe .....
>
>       $selstr =
>       "select column_name \"Name\",".
>       " decode(nullable,\'N\',\'NOT NULL\',\'Y\',null) \"Null\",".
>       " decode(data_type,\'DATE\',data_type,".
>       "\'NUMBER\',data_type||\'(\'||data_precision||\',\'||data_scale||\')\',".
>       "data_type||\'(\'||data_length||\')\' ) \"Type\"".
>       " from all_tab_columns".
>       " where (owner,table_name) = (select table_owner,table_name".
>       " from user_synonyms where synonym_name = $NAME)".
>       " or (owner = $OWNER and table_name = $NAME)";

Thanks.

Backwacking all the quotes is noisy. Personally I'd have phrased it more like:

      $selstr = q{
        select column_name "Name",
          decode(nullable, 'N','NOT NULL', 'Y',null) "Null",
          decode(data_type, 'DATE',data_type,
           'NUMBER',data_type||'('||data_precision||','||data_scale||')',
            data_type||'('||data_length||')' ) "Type"
          from all_tab_columns
          where (owner = $OWNER and table_name = $NAME)
             or (owner, table_name) = (
                  select table_owner,table_name
                  from user_synonyms where synonym_name = $NAME)
      };

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <29783-24@oink>;
         Thu, 4 Jul 1996 07:12:06 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836439100:23425:1; Thu, 04 Jul 96 01:11:40 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa22954; 4 Jul 96 1:10 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA28661 for dbi-users-real; Wed, 3 Jul 1996 17:04:56 -0700
Precedence: junk
Sender: [email protected]
Received: from cais.cais.com (cais.com [199.0.216.4])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id RAA28657
         for <[email protected]>; Wed, 3 Jul 1996 17:04:54 -0700
Received: from cais.cais.com (cais.com [199.0.216.4])
         by cais.cais.com (8.6.10/8.6.5) with SMTP id UAA22980
         for <[email protected]>; Wed, 3 Jul 1996 20:04:47 -0400
Date: Wed, 3 Jul 1996 20:04:47 -0400 (EDT)
From: Ric Foster <[email protected]>
Reply-To: [email protected]
To: [email protected]
Subject: DB2 disconnect problem
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


WARNING: Naive user!

I am in transition from an OS/2 & REXX environment to AIX & PERL.
DB2/6000 access is critical.  I installed PERL 5.002, DBI 0.70, and
DBD-DB2-0.59.  When I run the query.pl test program I get the following
on $dbh->disconnect;

  disconnect(DBI::db=HASH(0x20018b80)) invalidates 1 associated cursor(s)
  at query.pl line 23.

followed by:

  IOT/Abort trap(coredump)

Any clue on what I did wrong? (the fetch worked fine).

Also, there is a paragraph in the CAVEATS and HISTORY file that I
don't understand at all:

> Immediately prior to calling $sth->execute() you must call
> $sth->bind_param(index, name, type_hash_ref).   Note, that a
> minimal set of type_hash_refs are provided by DB2.pm and they
> may be used to create new type_hash_refs
>    $new_type = { %$existing_type, 'Stype' => SQL_"NewTYPE" };


Ric Foster <[email protected]>
National Gallery of Art, Washington

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <03117-0@oink>;
         Thu, 4 Jul 1996 11:56:31 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836477599:12676:1; Thu, 04 Jul 96 11:53:19 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa12242; 4 Jul 96 11:52 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id DAA29133 for dbi-users-real; Thu, 4 Jul 1996 03:44:57 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id DAA29129
         for <[email protected]>; Thu, 4 Jul 1996 03:44:54 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id af19730; 4 Jul 96 10:44 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac02082; 4 Jul 96 11:41 +0100
Received: from toad by oink with SMTP (PP) id <02917-0@oink>;
         Thu, 4 Jul 1996 10:51:51 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA06607;
         Thu, 4 Jul 1996 10:51:45 +0000
Date: Thu, 4 Jul 1996 10:51:45 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DB2 disconnect problem
X-Sun-Charset: US-ASCII
content-length: 664
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Ric Foster <[email protected]>
>
> I am in transition from an OS/2 & REXX environment to AIX & PERL.
> DB2/6000 access is critical.  I installed PERL 5.002, DBI 0.70, and
> DBD-DB2-0.59.  When I run the query.pl test program I get the following
> on $dbh->disconnect;
>
>    disconnect(DBI::db=HASH(0x20018b80)) invalidates 1 associated cursor(s)
>    at query.pl line 23.

That's a warning that you have disconnected without closing a cursor.

> followed by:
>
>    IOT/Abort trap(coredump)

That might be a stray abort() I left in some DBI error handling code (removed
in the next version) but it might also be something in the DBD::DB2 code.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <03170-1@oink>;
         Thu, 4 Jul 1996 12:01:40 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836477898:15647:4; Thu, 04 Jul 96 11:58:18 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15114; 4 Jul 96 11:57 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id DAA29148 for dbi-users-real; Thu, 4 Jul 1996 03:54:36 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id DAA29143
         for <[email protected]>; Thu, 4 Jul 1996 03:54:34 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ad03148; 4 Jul 96 11:53 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa04286; 4 Jul 96 11:51 +0100
Received: from toad by oink with SMTP (PP) id <03093-0@oink>;
         Thu, 4 Jul 1996 11:51:24 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA06769;
         Thu, 4 Jul 1996 11:51:14 +0000
Date: Thu, 4 Jul 1996 11:51:14 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: DBD-Oracle-0.32 test failed on HP/UX 10.10
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1814
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Gildas Perrot <[email protected]>
>
> Hi Tim and thanks a lot for your help.
>
> I tried Perl 5.003, DBI-0.70, and DBD-0racle-0.35 on HP-UX 10.10 as you
> suggested me. I am using the Ansi cc compiler provided by the Development
> option of HP/UX.
>
> No problem with Perl 5.003 (all tests passed successfully).
> Now, when I installed DBI (linked dynamically for HP/UX 10.10 which should
> support that as suggested by Tim Bunce, its author), I got this 'make test'
> output :

This is a slightly subtle problem that's likely to affect people who have
built static versions of some extensions and then upgrade to 5.003...


>   LD_RUN_PATH="" ld -o ./blib/arch/auto/DBI/DBI.sl -b -L/usr/local/lib DBI.o

Notice that it's building a DBI.sl (shared library).

> Then I installed DBD-Oracle, this time statically, the default.
> I got no problem with 'make' but 'make test' gives an error :

> Writing Makefile.aperl for DBD::Oracle

>         cc -L/usr/local/lib -Wl,-E -Wl,-B,deferred  -o perl -O ./perlmain.o
> /usr/local/lib/perl5/PA-RISC1.1/5.003/CORE/libperl.a
> /usr/local/lib/perl5/site_perl/PA-RISC1.1/auto/DBI/DBI.a
> /usr/local/lib/perl5/PA-RISC1.1/5.003/auto/DynaLoader/DynaLoader.a
> ./blib/arch/auto/DBD/Oracle/Oracle.a `cat ./blib/arch/auto/DBD/Oracle/extralibs
> .all` -lnet -lnsl_s -lndbm -lgdbm -ldld -lm -lc -lndir -lcrypt
> /usr/ccs/bin/ld: Unsatisfied symbols:
>    dowantarray (code)
>    warn (code)
>    gv_fetchmethod (code)

> Any idea about this problem ? Thanks in advance for your help.

Notice that here it's linking DBI.a (a static library) which was built for
Perl5.002. Since Perl5.003 uses EMBED mode by default old extension libraries
are no longer compatible.

The simple fix is to delete the old DBI.a file and rebuild DBD::Oracle.

>       Gildas.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <04952-2@oink>;
         Thu, 4 Jul 1996 18:41:14 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836499416:17901:2; Thu, 04 Jul 96 17:56:56 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa17454; 4 Jul 96 17:56 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA29436 for dbi-users-real; Thu, 4 Jul 1996 09:45:29 -0700
Precedence: junk
Sender: [email protected]
Received: from cais.cais.com (cais.com [199.0.216.4])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id JAA29432
         for <[email protected]>; Thu, 4 Jul 1996 09:45:28 -0700
Received: from cais.cais.com (cais.com [199.0.216.4])
         by cais.cais.com (8.6.10/8.6.5) with SMTP id MAA12004;
         Thu, 4 Jul 1996 12:45:23 -0400
Date: Thu, 4 Jul 1996 12:45:23 -0400 (EDT)
From: Ric Foster <[email protected]>
Reply-To: [email protected]
To: Tim Bunce <[email protected]>
cc: [email protected]
Subject: Re: DB2 disconnect problem
In-Reply-To: <[email protected]>
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


I assume that close is accomplished by $cursor->finish.  Here is my
test program:
=======================================================
use DBI;
$ENV{'DB2_HOME'} = '/usr/lpp/db2_02_01';
$drh = DBI->install_driver( 'DB2' );

$dbh = $drh->connect( 'gallery', 'user', 'pw' );
die unless $dbh;

$cursor = $dbh->prepare( "SELECT FILE_NAME FROM nga.web_register_tb" );
$cursor->execute;
$field = $cursor->fetchrow;
print "Field: $field\n";
$rc = $cursor->finish;
print "$rc\n";

$dbh->rollback;
$dbh->disconnect;
exit;
==============================================================

the return code from finish is '1' (tho it seems from looking at the DBD
code that it alwasy returns 1).




On Thu, 4 Jul 1996, Tim Bunce wrote:

>
> > From: Ric Foster <[email protected]>
> >
> > I am in transition from an OS/2 & REXX environment to AIX & PERL.
> > DB2/6000 access is critical.  I installed PERL 5.002, DBI 0.70, and
> > DBD-DB2-0.59.  When I run the query.pl test program I get the following
> > on $dbh->disconnect;
> >
> >    disconnect(DBI::db=HASH(0x20018b80)) invalidates 1 associated cursor(s)
> >    at query.pl line 23.
>
> That's a warning that you have disconnected without closing a cursor.
>
> > followed by:
> >
> >    IOT/Abort trap(coredump)
>
> That might be a stray abort() I left in some DBI error handling code (removed
> in the next version) but it might also be something in the DBD::DB2 code.
>
> Tim.
>

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <05687-0@oink>;
         Thu, 4 Jul 1996 21:56:19 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836513561:09076:2; Thu, 04 Jul 96 21:52:41 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa08265; 4 Jul 96 21:52 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA29647 for dbi-users-real; Thu, 4 Jul 1996 13:44:14 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id NAA29643
         for <[email protected]>; Thu, 4 Jul 1996 13:42:38 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ae16338; 4 Jul 96 20:35 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa05120; 4 Jul 96 21:34 +0100
Received: from toad by oink with SMTP (PP) id <05182-0@oink>;
         Thu, 4 Jul 1996 19:39:15 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA08216;
         Thu, 4 Jul 1996 19:39:08 +0000
Date: Thu, 4 Jul 1996 19:39:08 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DB2 disconnect problem
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 340
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Ric Foster <[email protected]>
>
> I assume that close is accomplished by $cursor->finish.

Yes. I suspect that the DBD::DB2 finish method is not using the
internal DBI DBIc_ACTIVE_off() macro.

CC'd to Mike Moran <[email protected]>, the author.

Problems with specific drivers should be CC'd to the author of the driver.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <06335-3@oink>;
         Fri, 5 Jul 1996 01:36:55 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836516463:22836:2; Thu, 04 Jul 96 22:41:03 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa22301; 4 Jul 96 22:40 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA29592 for dbi-users-real; Thu, 4 Jul 1996 12:30:54 -0700
Precedence: junk
Sender: [email protected]
Received: from Shalmaneser.Stanford.EDU (Shalmaneser.Stanford.EDU [171.64.75.107])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id MAA29588
         for <[email protected]>; Thu, 4 Jul 1996 12:30:53 -0700
Received: (from agupta@localhost) by Shalmaneser.Stanford.EDU (8.7.1/8.7.1)
         id MAA08762; Thu, 4 Jul 1996 12:30:52 -0700
Date: Thu, 4 Jul 1996 12:30:52 -0700
From: Ashish Gupta <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: question
Reply-to: Ashish Gupta <[email protected]>


Hi,
I want to call Oracle 7 from using perl5 and am considering
different options. One of my requirements is that I want to be able to
pass to Oracle SQL queries that use oracle specific syntax. Does DBI
allow SQL queries to be passed down to the RDBMS without any sanity
check? Even if it checks each SQL statemeent, is all of Oracle's
syntax incorporated into DBI?  Am I forced to use SQL suported by all
the DB vendons?
Thanks much.

Ashish
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <07492-0@oink>;
         Fri, 5 Jul 1996 07:06:31 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836531631:04150:2; Fri, 05 Jul 96 02:53:51 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa04095; 5 Jul 96 2:53 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id SAA00181 for dbi-users-real; Thu, 4 Jul 1996 18:48:44 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id SAA00177
         for <[email protected]>; Thu, 4 Jul 1996 18:48:42 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ac14131; 5 Jul 96 2:48 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab27897; 5 Jul 96 2:48 +0100
Received: from toad by oink with SMTP (PP) id <06558-0@oink>;
         Fri, 5 Jul 1996 01:44:17 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA09233;
         Fri, 5 Jul 1996 01:44:10 +0000
Date: Fri, 5 Jul 1996 01:44:10 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: question
X-Sun-Charset: US-ASCII
content-length: 387
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Ashish Gupta <[email protected]>
>
>
> Hi,
> I want to call Oracle 7 from using perl5 and am considering
> different options. One of my requirements is that I want to be able to
> pass to Oracle SQL queries that use oracle specific syntax. Does DBI
> allow SQL queries to be passed down to the RDBMS without any sanity
> check?

Yes. (It never does anything else.)

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <07492-7@oink>;
         Fri, 5 Jul 1996 07:07:19 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836536102:26461:4; Fri, 05 Jul 96 04:08:22 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa26418; 5 Jul 96 4:08 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id UAA00249 for dbi-users-real; Thu, 4 Jul 1996 20:04:16 -0700
Precedence: junk
Sender: [email protected]
Received: from gateway1.srs.gov (gateway1.srs.gov [192.33.240.10])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id UAA00244
         for <[email protected]>; Thu, 4 Jul 1996 20:03:16 -0700
Received: by gateway1.srs.gov id AA10015 (InterLock SMTP Gateway 3.0
         for [email protected]); Thu, 4 Jul 1996 23:00:31 -0400
Message-Id: <[email protected]>
Received: by gateway1.srs.gov (Internal Mail Agent-1);
         Thu, 4 Jul 1996 23:00:31 -0400
Date: Thu, 04 Jul 1996 22:56:39 -0400
From: James Taylor <[email protected]>
Subject: Re: Oracle Installation question
To: Tim Bunce <[email protected]>
Cc: [email protected], [email protected]
Mime-Version: 1.0
X-Mailer: Mozilla 2.0 (X11; I; SunOS 5.4 sun4c)
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
References: <[email protected]>

Tim Bunce wrote:
>
> Forwarded to dbi-users
>
> >From [email protected]  Wed Jul  3 17:03:48 1996

>
> What is the minimum installation one must do in order to be able to use
> DBD/DBI?
>
> Do you need to set up an Oracle Server?  (Or is SQL*Net enough?)
>
> Thanks,
>
> Jonathan

I think that this question has been asked before and I don't recall the
answer (if there was one).

My latest experience (trying to bring up a minimal Oracle installation
to test Oracle 7.3.2 and Solaris 2.5) suggests that you need SQL*net,
Pro*C, AND the rdbms to BUILD the DBD.

You would think that you would only need SQL*net and Pro*C, but the
build seems to need an include file from rdbms/demo.  I don't have my
notes, but I think that it was oratypes.h that the build couldn't find
if the rdbms tree was not installed.

After the DBD module is built, I would think that you would only need
SQL*net and the OCI libraries, but I HAVE NOT VERIFIED THIS.

It would be nice to see a good answer as to what you need as a minimum
to both BUILD and DEPLOY.

James
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <08893-3@oink>;
         Fri, 5 Jul 1996 13:54:52 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836562865:16504:2; Fri, 05 Jul 96 11:34:25 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa16187; 5 Jul 96 11:33 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id DAA00666 for dbi-users-real; Fri, 5 Jul 1996 03:25:01 -0700
Precedence: junk
Sender: [email protected]
Received: from thera.denkart.be (thera.denkart.be [193.74.137.253])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id DAA00662
         for <[email protected]>; Fri, 5 Jul 1996 03:24:51 -0700
From: Akkermans_Sven/[email protected]
Received: (from smap@localhost) by thera.denkart.be (8.6.12/8.6.9) id KAA25712
         for <[email protected]>; Fri, 5 Jul 1996 10:51:26 +0100
Received: from tyana.denkart.be(193.74.137.1) by thera.denkart.be
         via smap (V1.3) id sma025708; Fri Jul 5 10:50:55 1996
Received: from by tyana.denkart.be with SMTP (1.37.109.16/16.2) id AA166342095;
         Fri, 5 Jul 1996 12:21:35 +0200
X-Openmail-Hops: 1
Date: Fri, 5 Jul 96 12:20:46 +0200
Message-Id: <H000008b002ca1f1@MHS>
Subject: Compiling Oracle.c for DBD-Oracle-0.32
To: [email protected]

Hello,

I have a problem installing the DBD-Oracle-0.32 package.

System : HP 9000
OS : HP-UX 10.01
Perl : 5.003
DBI : DBI-0.69
Oracle : 7.1.6

after making the makefile through perl Makefile.PL, I ran make and
got the following result

cc -c -I/disc3/oracle/product/7.1.6/rdbms/demo
-I/users/sa/mrperl/lib/site_perl/9000/877-hpux/DBI -I/users/sa/mr
perl/lib/9000/877-hpux/5.003/DBI -I/usr/local/include -O
-DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" +z -I/users
/sa/mrperl/lib/9000/877-hpux/5.003/CORE  Oracle.c

cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 84:
warning 5: "signed" will become a keyword.
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 84: error
1000: Unexpected symbol: "int".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 112:
warning 5: "signed" will become a keyword.
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 112: error
1000: Unexpected symbol: "char".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 153:
warning 5: "signed" will become a keyword.
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 153: error
1000: Unexpected symbol: "short".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 201:
warning 5: "signed" will become a keyword.
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 201: error
1000: Unexpected symbol: "long".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 225:
warning 5: "signed" will become a keyword.
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/oratypes.h", line 225: error
1000: Unexpected symbol: "long".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/ocidfn.h", line 47: error
1000: Unexpected symbol: "sword".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/ocidfn.h", line 57: error
1000: Unexpected symbol: "sword".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/ocidfn.h", line 73: error
1000: Unexpected symbol: "sword".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/ocidfn.h", line 83: error
1000: Unexpected symbol: "sword".
cc: "/disc3/oracle/product/7.1.6/rdbms/demo/ocikpr.h", line 29: error
1000: Unexpected symbol: "obndra".
cc: error 2017: Cannot recover from earlier errors, terminating.
gmake: *** [Oracle.o] Error 1

I tried to add the -Aa option and this solved part of the problem, now I
got the following errors, after I did a 'make clean' and generated the
makefile again with the Aa option:

cp Oraperl.pm ./blib/lib/Oraperl.pm
cp Oracle.pm ./blib/lib/DBD/Oracle.pm
cp oraperl.ph ./blib/lib/oraperl.ph
./bin/perl5.003 -I/users/sa/mrperl/lib/9000/877-hpux/5.003
-I/users/sa/mrperl/lib /users/sa/mrperl/lib/ExtUtils
/xsubpp  -typemap /users/sa/mrperl/lib/ExtUtils/typemap Oracle.xs
>Oracle.tc && mv Oracle.tc Oracle.c
cc -Aa -c -I/disc3/oracle/product/7.1.6/rdbms/demo
-I/users/sa/mrperl/lib/site_perl/9000/877-hpux/DBI -I/users/s
a/mrperl/lib/9000/877-hpux/5.003/DBI -I/usr/local/include -O
-DVERSION=\"0.32\" -DXS_VERSION=\"0.32\" +z -I/u
sers/sa/mrperl/lib/9000/877-hpux/5.003/CORE  Oracle.c

cc: "/users/sa/mrperl/lib/9000/877-hpux/5.003/CORE/sv.h", line 250:
error 1000: Unexpected symbol: "DIR".
cc: "/users/sa/mrperl/lib/9000/877-hpux/5.003/CORE/av.h", line 12: error
1000: Unexpected symbol: "ssize_t".
cc: "/users/sa/mrperl/lib/9000/877-hpux/5.003/CORE/av.h", line 13: error
1000: Unexpected symbol: "xav_max".
cc: "/users/sa/mrperl/lib/9000/877-hpux/5.003/CORE/av.h", line 22: error
1000: Unexpected symbol: "}".
cc: "/users/sa/mrperl/lib/9000/877-hpux/5.003/CORE/hv.h", line 13: error
1000: Unexpected symbol: "HE".
cc: "/users/sa/mrperl/lib/9000/877-hpux/5.003/CORE/hv.h", line 30: error
1000: Unexpected symbol: "HE".
cc: "/users/sa/mrperl/lib/9000/877-hpux/5.003/CORE/perl.h", line 726:
error 1000: Unexpected symbol: "getuid".
cc: error 2017: Cannot recover from earlier errors, terminating.
gmake: *** [Oracle.o] Error 1

Any suggestions?

Thanks in advance
Sven Akkermans
[email protected]
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <09561-0@oink>;
         Fri, 5 Jul 1996 17:00:50 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836575931:14917:1; Fri, 05 Jul 96 15:12:11 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa14383; 5 Jul 96 15:11 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id HAA00856 for dbi-users-real; Fri, 5 Jul 1996 07:07:06 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id HAA00851
         for <[email protected]>; Fri, 5 Jul 1996 07:07:00 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id an10268; 5 Jul 96 15:06 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa13923; 5 Jul 96 14:40 +0100
Received: from toad by oink with SMTP (PP) id <09101-0@oink>;
         Fri, 5 Jul 1996 14:08:27 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA12206;
         Fri, 5 Jul 1996 14:08:20 +0000
Date: Fri, 5 Jul 1996 14:08:20 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: problem with compilation solved
X-Sun-Charset: US-ASCII
content-length: 532
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


----- Begin Included Message -----

From Akkermans_Sven/[email protected]  Fri Jul  5 13:56:03 1996
From: Akkermans_Sven/[email protected]
X-Openmail-Hops: 1
Date: Fri, 5 Jul 96 13:39:18 +0200
Subject: problem with compilation solved
To: [email protected]

Hello,

Yes, I should have read the README to the end.

The bug on HP-UX with oratypes.h in Oracle 7.1.3 applies also
to Oracle 7.1.6
The same sollution applies apparently.

Sorry for the trouble,

Sven Akkermans
[email protected]


----- End Included Message -----

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <09564-0@oink>;
         Fri, 5 Jul 1996 17:01:05 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836574915:15252:2; Fri, 05 Jul 96 14:55:15 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa14544; 5 Jul 96 14:54 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA00828 for dbi-users-real; Fri, 5 Jul 1996 06:47:58 -0700
Precedence: junk
Sender: [email protected]
Received: from pt.Cyanamid.COM (igate.cyanamid.com [141.173.63.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id GAA00824
         for <[email protected]>; Fri, 5 Jul 1996 06:47:57 -0700
Received: from jmsun by apbr1.pt.Cyanamid.COM (4.1/SMI-4.1) id AA17257;
         Fri, 5 Jul 96 09:47:22 EDT
Date: Fri, 5 Jul 1996 09:47:21 -0400 (EDT)
From: Jon Meek <[email protected]>
Reply-To: Jon Meek <[email protected]>
Subject: Re: Oracle Installation - What is required?
To: [email protected]
In-Reply-To: "Your message with ID" <[email protected]>
Message-Id: <Roam.3.0.1.836574441.25692.meekj@apbr1>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII

I was able to BUILD by installing just SQL*net and Pro*C, then getting the few
RDBMS files that were missing directly from the Oracle CD.

To DEPLOY, I have found that only tnsnames.ora and sqlnet.ora are required
(for my installation anyway (Solaris 2.5, Oracle 7.2.x.x). It would be nice to
have the error message files available as well, but it is not required.

Jon

Related messages (abbreviated):

> >
> > What is the minimum installation one must do in order to be able to use
> > DBD/DBI?
> >
> > Do you need to set up an Oracle Server?  (Or is SQL*Net enough?)
> >

>
> My latest experience (trying to bring up a minimal Oracle installation
> to test Oracle 7.3.2 and Solaris 2.5) suggests that you need SQL*net,
> Pro*C, AND the rdbms to BUILD the DBD.
>
> You would think that you would only need SQL*net and Pro*C, but the
> build seems to need an include file from rdbms/demo.  I don't have my
> notes, but I think that it was oratypes.h that the build couldn't find
> if the rdbms tree was not installed.
>
> After the DBD module is built, I would think that you would only need
> SQL*net and the OCI libraries, but I HAVE NOT VERIFIED THIS.
>
> It would be nice to see a good answer as to what you need as a minimum
> to both BUILD and DEPLOY.
>
> James


---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <09561-6@oink>;
         Fri, 5 Jul 1996 17:01:20 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836581790:18645:4; Fri, 05 Jul 96 16:49:50 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa18325; 5 Jul 96 16:49 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA00999 for dbi-users-real; Fri, 5 Jul 1996 08:41:50 -0700
Precedence: junk
Sender: [email protected]
Received: from labs-n.bbn.com (LABS-N.BBN.COM [128.89.0.100])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id IAA00995
         for <[email protected]>; Fri, 5 Jul 1996 08:41:49 -0700
From: [email protected]
To: [email protected], [email protected]
cc: [email protected], [email protected], [email protected]
Subject: [email protected]: Re: DBD Module
Date: Fri, 05 Jul 96 11:41:23 -0400
Message-ID: <[email protected]>

We're stil awaiting some suggestions as to how to fix this problem/

One thing ocurred to me:  We compiled the DBI and DBD code (and Perl itself)
with gcc.  I know that we have had problems in the past in linking
gcc-compiled code with Oracle libraries; could this be the source of our
difficulties?  Might the problem go away if we used the Sunsoft compilers?

<j>
------- Forwarded Message

To: Amy McGovern <[email protected]>
cc: [email protected], [email protected], [email protected],
   [email protected]
Subject: Re: DBD Module
In-reply-to: Amy McGovern's message of Tue, 11 Jun 96 16:55:15 -0400.
            <[email protected]>
Reply-To: [email protected]
Date: Tue, 11 Jun 96 17:04:55 -0400
Message-ID: <[email protected]>
From: [email protected]

Oraperl test application $Revision: 1.20 $
Loaded DBI.pm at blib/lib/Oraperl.pm line 23
   DBI dispatch debug level set to 2
DBI->install_driver DBI Oracle at blib/lib/Oraperl.pm line 52
DBI->install_driver(Oracle) loaded
   New DBI::dr (for DBD::Oracle::dr, parent=, id=)
   dbih_setup_handle(DBI::dr=HASH(0x15208c)=>DBI::dr=HASH(0x176654), DBD::Oracle::dr, NULL)
   New DBI::dr => DBI::dr=HASH(0x15208c) (inner=DBI::dr=HASH(0x176654)) for DBD::Oracle::dr
DBI->install_driver(Oracle) = DBI::dr=HASH(0x15208c)
   >> debug       DISPATCH (DBI::dr=HASH(0x15208c) @2 g0 a1755a8 r2)
   -> debug in DBD::_::common for DBD::Oracle::dr (DBI::dr=HASH(0x15208c)~0x176654 0)
   <- debug= 0
   >> STORE       DISPATCH (DBI::dr=HASH(0x176654) @3 g0 a0 r2)
   -> STORE in DBD::_::common for DBD::Oracle::dr (DBI::dr=HASH(0x176654) 'CompatMode' 1)
   STORE DBI::dr=HASH(0x176654) 'CompatMode' => 1
   <- STORE= 1
   >> STORE       DISPATCH (DBI::dr=HASH(0x176654) @3 g0 a0 r2)
   -> STORE for DBD::Oracle::dr (DBI::dr=HASH(0x176654) 'Warn' 0)
   STORE DBI::dr=HASH(0x176654) 'Warn' => 0
   <- STORE= 1
   New DBI::dr (for DBD::Switch::dr, parent=, id=SCALAR(0x15a074))
   dbih_setup_handle(DBI::dr=HASH(0x1766c0)=>DBI::dr=HASH(0xc84bc), DBD::Switch::dr, SCALAR(0x15a074))
   New DBI::dr => DBI::dr=HASH(0x1766c0) (inner=DBI::dr=HASH(0xc84bc)) for DBD::Switch::dr

Oraperl Emulation Interface version 1.26
   >> FETCH       DISPATCH (DBI::dr=HASH(0x176654) @2 g0 a0 r1)
   <- FETCH= '0.32' QUICK
Oracle Driver 0.32
   >> FETCH       DISPATCH (DBI::dr=HASH(0xc84bc) @2 g0 a0 r1)
   -> FETCH for DBD::Switch::dr (DBI::dr=HASH(0xc84bc) 'Attribution')
   <- FETCH= 'DBI-0.69 Switch by Tim Bunce'
   >> FETCH       DISPATCH (DBI::dr=HASH(0xc84bc) @2 g0 a0 r1)
   <- FETCH= '0.69' QUICK
DBI-0.69 Switch by Tim Bunce, version 0.69


Connecting
to 'COR1' (from command line, else uses ORACLE_SID/TWO_TASK)
as 'argus/COR1FRESCA' (via ORACLE_USERID environment var or default)
   >> connect     DISPATCH (DBI::dr=HASH(0x15208c) @4 g1 a175688 r1)
   -> connect for DBD::Oracle::dr (DBI::dr=HASH(0x15208c)~0x176654 'COR1' 'argus/COR1FRESCA' '')
   >> debug       DISPATCH (DBI::dr=HASH(0x176654) @1 g0 a1755a8 r2)
   -> debug for DBD::Oracle::dr (DBI::dr=HASH(0x176654))
   <- debug= 0
libc internal error: _rmutex_unlock: rmutex not held.
Abort

------- End of Forwarded Message

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <09564-10@oink>;
         Fri, 5 Jul 1996 17:03:46 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836580573:14228:2; Fri, 05 Jul 96 16:29:33 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa13923; 5 Jul 96 16:28 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA00952 for dbi-users-real; Fri, 5 Jul 1996 08:22:56 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA00948
         for <[email protected]>; Fri, 5 Jul 1996 08:22:54 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id QAA19227 for [email protected]; Fri, 5 Jul 1996 16:15:04 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Hermetica page updates
To: [email protected]
Date: Fri, 5 Jul 1996 16:15:04 +0100 (BST)
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 785


The Hermetica pages have begun the process of getting revamped. All the
driver information has been updated to point correctly at Tom Christiansen's
groovy CPAN multiplexer instead of local copies.

The FAQ has started it's overhaul. Anything people want to see, let me know.
I'll post the FAQ to this mailing list on a semi-regular basis as well.

Also, a new perl module Devel::CoreStack has been released for generating
stack traces from core dumps. It's at:

       ftp://ftp.mcqueen.com/pub/research

Please take a copy of this if you're suffering from core dumping scripts.

Regards.

A.

--
Alligator Descartes     | "You fuckoffeh?" "Eh?" "Coffee? You for coffee?"
Anathematician          |     "Ohhh! Dearie me!"
[email protected]  |         -- The High Life
http://www.hermetica.com|
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <09564-12@oink>;
         Fri, 5 Jul 1996 17:04:03 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836580576:14239:2; Fri, 05 Jul 96 16:29:36 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa13867; 5 Jul 96 16:28 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA00958 for dbi-users-real; Fri, 5 Jul 1996 08:23:36 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA00954
         for <[email protected]>; Fri, 5 Jul 1996 08:23:29 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id QAA19234 for [email protected]; Fri, 5 Jul 1996 16:15:46 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: DBI FAQ v.0.1
To: [email protected]
Date: Fri, 5 Jul 1996 16:15:45 +0100 (BST)
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 10018


Original at:

       http://www.hermetica.com/technologia/DBI/doc/faq.html

  DBI Frequently Asked Questions v.0.1
  Last updated: 05/07/96

   1. Basic Information & Information Sources

        1. What is DBI, DBPerl, Oraperl and oraperl?
        2. Where can I get it from?
        3. Where can I get more information?


   2. Compilation problems or "It fails the tests!"

         + DBI
         + Oracle
         + mSQL
         + Informix


   3. Further Information


    _________________________________________________________________



 1.1 What is DBI, DBPerl, Oraperl and oraperl?



  To quote Tim Bunce, the architect and author of DBI:

    "DBI is a database access Application Programming Interface (API)
    for the Perl Language. The DBI API Specification defines a set of
    functions, variables and conventions that provide a consistent
    database interface independant of the actual database being used."



  In simple language, the DBI interface allows users to access multiple
  database types transparently. So, if you connecting to an Oracle,
  Informix, mSQL, Sybase or whatever database, you don't need to know
  the underlying mechanics of the 3GL layer. The API defined by DBI will
  work on all these database types.

  A similar benefit is gained by the ability to connect to two different
  databases of different vendor within the one perl script, ie, I want
  to read data from an Oracle database and insert it back into an
  Informix database all within one program. The DBI layer allows you to
  do this simply and powerfully.

  DBPerl is the old name for the interface specification. It's usually
  now used to denote perl4 modules on database interfacing, such as,
  oraperl, isqlperl, ingperl and so on. These interfaces didn't have a
  standard API and are generally not supported.

  However, some DBI modules have DBPerl emulation layers, so,
  DBD::Oracle comes with an Oraperl emulation layer, which allows you
  to run legacy oraperl scripts wihtout modification. The emulation
  layer translates the oraperl API calls into DBI calls and executes
  them through the DBI switch.


    _________________________________________________________________



 1.2. Where can I get it from?



  DBI is primarily distributed from:



    ftp://ftp.demon.co.uk/pub/perl/db

  Unfortunately, this site tends to be clogged most of the day. A mirror
  site has been set up at:

    ftp://ftp.mcqueen.com/pub/databases/DBI

  which currently is more accessible.

  The Comprehensive Perl Archive Network resources should be used for
  retrieving up-to-date versions of the drivers, since local mirror
  sites usually lag. For more specific version information and exact
  URLs of drivers, please see the drivers list and the DBI Switch pages.



    _________________________________________________________________



 1.3. Where can I get more information?



  There are a few information sources on DBI.

    * DBI Specification
      http://www.hermetica.com/technologia/DBI/doc/dbispec

      The DBI Specification lays out the DBI interface. It should be
      noted that some modules, notably DBD::mSQL and DBD::Informix, vary
      from this occasionally. This document should be regarded as being
      of historical interest only and should not serve as a programming
      manual, or authoratative in any sense.

    * Oraperl POD documentation

      For users of the Oraperl emulation layer bundled with DBD::Oracle,
      typing:

    perldoc Oraperl


      will produce an updated copy of the original oraperl man page
      written by Kevin Stock for perl4. The oraperl API is fully listed
      and described there.

    * Rambles, Tidbits and Observations
      http://www.hermetica.com/technologia/DBI/index.html#tidbits

      There are a serious of occasional rambles from various people on
      the DBI mailing lists who, in an attempt to clear up a simple
      point, end up drafting fairly comprehensive documents. These are
      quite often varying in quality, but do provide some insights into
      the workings of the interfaces.

    * The DBI driver

      If you locate the driver file DBI.pm in your perl library
      directory, at the end of that file, the DBI interface is listed
      with correct invocations.

    * Mailing Lists
      http://www.fugue.com/dbi

      There are three mailing lists for DBI run by Ted Lemon. These are:


         + dbi-announce

           This mailing list is for announcements only. Very low
           traffic. The announcements are usually posted on the main DBI
           WWW page.

           If you cannot successfully use the form on the above WWW
           page, please subscribe to the list in the following manner:

           Email: [email protected] with a message body of
           subscribe

         + dbi-dev

           This mailing list is intended for the use of developers
           discussing ideas and concepts for the DBI interface, API and
           driver mechanics. Only any use for developers, or interested
           parties. Low traffic.

           If you cannot successfully use the form on the above WWW
           page, please subscribe to the list in the following manner:

           Email: [email protected] with a message body of
           subscribe

         + dbi-users

           This mailing list is a general discussion list used for bug
           reporting, problem discussion and general enquiries. Medium
           traffic.

           If you cannot successfully use the form on the above WWW
           page, please subscribe to the list in the following manner:

           Email: [email protected] with a message body of
           subscribe



    * Mailing List Archives
      http://www.coe.missouri.edu/~faq/lists/dbi.html

      Searchable hypermail archives of the three mailing lists, and some
      of the much older traffic have been set up for users to browse.




    _________________________________________________________________





 2.1. Compilation problems or "It fails the test!"



  First off, consult the online information about the module, beit DBI
  itself, or a DBD, and see if it's a known compilation problem on your
  architecture. These documents can be found at:



    http://www.hermetica.com/technologia/DBI



  If it's a known problem, you'll probably have to wait till it gets
  fixed. If you're really needing it fixed, try the following:
    * Attempt to fix it yourself

      This technique is generally not recommended to the faint-hearted.
      If you do think you have managed to fix it, then, send a patch
      file ( context diff ) to the author with an explanation of:

         + What the problem was, and test cases, if possible.
         + What you needed to do to fix it. Please make sure you mention
           everything.
         + Platform information, database version, perl version, module
           version and DBI version.


    * Email the author Do NOT whinge!

      We tend to have real jobs to do, and we do read the mailing lists
      for problems. Besides, we may not have access to <insert your
      favourite brain-damaged platform here> and couldn't be of any
      assistance anyway! Apologies for sounding harsh, but that's the
      way of it!

      However, you might catch one of these creative genii at 3am when
      we're doing this sort of stuff anyway, and get a patch within 5
      minutes. The atmosphere in the DBI circle is that we do appreciate
      the users' problems, since we work in similar environments.

      If you are planning to email the author, please furnish as much
      information as possible, ie:

         + ALL the information off the README file in the problematic
           module. And we mean ALL of it. We don't put lines like that
           in documentation for the good of our health, or to meet
           obscure README file standards of length.
         + If you have a core dump, try the Devel::CoreStack module for
           generating a stack trace from the core dump. Send us that
           too. Devel::CoreStack can be found at
           ftp://ftp.mcqueen.com/pub/research/Devel-CoreStack-1.3.tar.g
           z
         + Module versions, perl version, test cases, operating system
           versions and any other pertinent information.


      Remember, the more information you send us, the quicker we can
      track problems down. If you send us nothing, expect nothing back.

    * Email the dbi-users Mailing List

      It's usually a fairly intelligent idea to cc the mailing list
      anyway with problems. The authors all read the lists, so you lost
      nothing by mailing there.





 2.2 DBI Compilation Problems



  Here are some of the most common DBI compilation and test failure
  problems.


    _________________________________________________________________




    _________________________________________________________________

  &copy 1995 Hermetica
  Alligator Descartes - Hermetica
--
Alligator Descartes     | "You fuckoffeh?" "Eh?" "Coffee? You for coffee?"
Anathematician          |     "Ohhh! Dearie me!"
[email protected]  |         -- The High Life
http://www.hermetica.com|
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <11229-4@oink>;
         Sat, 6 Jul 1996 00:45:46 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836587550:21606:2; Fri, 05 Jul 96 18:25:50 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa18792; 5 Jul 96 18:22 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA01111 for dbi-users-real; Fri, 5 Jul 1996 10:17:09 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id KAA01107
         for <[email protected]>; Fri, 5 Jul 1996 10:17:06 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id SAA20176; Fri, 5 Jul 1996 18:09:21 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: DBI FAQ v.0.1
To: [email protected] (Tim Bunce)
Date: Fri, 5 Jul 1996 18:09:20 +0100 (BST)
Cc: [email protected], [email protected]
In-Reply-To: <[email protected]> from "Tim Bunce" at Jul 5, 96 05:47:26 pm
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 2176


> >   1.1 What is DBI, DBPerl, Oraperl and oraperl?
>
> DBperl should be spelt rith a lowercase p.
>
> :-)

*bzzzrrppppttttt*

> >        The DBI Specification lays out the DBI interface. It should be
> +                                      ^^ and old version of

Corrected.

> >      * Oraperl POD documentation
>
> The DBD::Oracle README file also lists other sources of information for
> Oracle users.

OK, added that note.

> >      * Rambles, Tidbits and Observations
> >
> >        There are a serious of occasional rambles from various people on
>
> Serious? ;-)

Well, spellchecking? Pardon? 8-) Corrected.

> >   2.1. Compilation problems or "It fails the test!"
> >
> >      * Email the author Do NOT whinge!
>
> >             too. Devel::CoreStack can be found at
> >             ftp://ftp.mcqueen.com/pub/research/Devel-CoreStack-1.3.tar.g
> >             z
>
> Unfortunate letter wrapping.

Hmm, that's lynx's prettify function for you. I've twiddled the wording so it
now wraps back correctly......

> >      * Email the dbi-users Mailing List
> >
> >        It's usually a fairly intelligent idea to cc the mailing list
> >        anyway with problems. The authors all read the lists, so you lost
> >        nothing by mailing there.
>
> I specifically request people to CC the list. Personal mail puts too much
> presure on me and my time and I don't appreciate it for DBI/DBD issues.

OK, I've added this chunk to the "Email the author" section.

   Please email the address listed in the WWW pages for whichever driver you
   are having problems with. Do <B>not</B> directly email the author at a
   known address unless it corresponds with the one listed.

So, if any driver authors don't want directly contacted, change your email
adrresses in the README's and let me know for the pages. Tim, you're already
pointing at dbi-users.

I think this gets the point across.

Do we want regular posting of the FAQ to dbi-users?

> Tim.

--
Alligator Descartes     | "You fuckoffeh?" "Eh?" "Coffee? You for coffee?"
Anathematician          |     "Ohhh! Dearie me!"
[email protected]  |         -- The High Life
http://www.hermetica.com|
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11235-4@oink>;
         Sat, 6 Jul 1996 00:46:22 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836586651:16215:3; Fri, 05 Jul 96 18:10:51 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15697; 5 Jul 96 18:10 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA01083 for dbi-users-real; Fri, 5 Jul 1996 09:59:37 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA01079
         for <[email protected]>; Fri, 5 Jul 1996 09:59:32 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ac04697; 5 Jul 96 16:59 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ac02680; 5 Jul 96 17:49 +0100
Received: from toad by oink with SMTP (PP) id <10373-0@oink>;
         Fri, 5 Jul 1996 17:47:32 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA14581;
         Fri, 5 Jul 1996 17:47:26 +0000
Date: Fri, 5 Jul 1996 17:47:26 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DBI FAQ v.0.1
X-Sun-Charset: US-ASCII
content-length: 1308
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Alligator Descartes <[email protected]>
>
>
>   1.1 What is DBI, DBPerl, Oraperl and oraperl?

DBperl should be spelt with a lowercase p.

:-)

>   1.3. Where can I get more information?

>        The DBI Specification lays out the DBI interface. It should be
+                                      ^^ and old version of

>      * Oraperl POD documentation

The DBD::Oracle README file also lists other sources of information for
Oracle users.

>      * Rambles, Tidbits and Observations
>
>        There are a serious of occasional rambles from various people on

Serious? ;-)

>   2.1. Compilation problems or "It fails the test!"
>
>      * Email the author Do NOT whinge!

>             too. Devel::CoreStack can be found at
>             ftp://ftp.mcqueen.com/pub/research/Devel-CoreStack-1.3.tar.g
>             z

Unfortunate letter wrapping.

>      * Email the dbi-users Mailing List
>
>        It's usually a fairly intelligent idea to cc the mailing list
>        anyway with problems. The authors all read the lists, so you lost
>        nothing by mailing there.

I specifically request people to CC the list. Personal mail puts too much
presure on me and my time and I don't appreciate it for DBI/DBD issues.

Thanks for maintaining this Alligator.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11235-16@oink>;
         Sat, 6 Jul 1996 00:47:25 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836589719:19750:2; Fri, 05 Jul 96 19:01:59 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa19019; 5 Jul 96 19:00 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id KAA01143 for dbi-users-real; Fri, 5 Jul 1996 10:57:01 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id KAA01139
         for <[email protected]>; Fri, 5 Jul 1996 10:56:59 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id af12767; 5 Jul 96 17:56 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa02680; 5 Jul 96 17:49 +0100
Received: from toad by oink with SMTP (PP) id <10330-0@oink>;
         Fri, 5 Jul 1996 17:31:29 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA14538;
         Fri, 5 Jul 1996 17:31:19 +0000
Date: Fri, 5 Jul 1996 17:31:19 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: [email protected]: Re: DBD Module
Cc: [email protected], [email protected], [email protected]
X-Sun-Charset: US-ASCII
content-length: 599
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected]
>
> We're stil awaiting some suggestions as to how to fix this problem/
>
> One thing ocurred to me:  We compiled the DBI and DBD code (and Perl itself)
> with gcc.  I know that we have had problems in the past in linking
> gcc-compiled code with Oracle libraries; could this be the source of our
> difficulties?  Might the problem go away if we used the Sunsoft compilers?

> libc internal error: _rmutex_unlock: rmutex not held.

You're still waiting because you've not been paying enough attention :-)

It was fixed in DBD::Oracle 0.34 and we're now up to 0.35.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11235-21@oink>;
         Sat, 6 Jul 1996 00:47:46 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836591842:12954:4; Fri, 05 Jul 96 19:37:22 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa12873; 5 Jul 96 19:37 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA01207 for dbi-users-real; Fri, 5 Jul 1996 11:26:48 -0700
Precedence: junk
Sender: [email protected]
Received: from borg.mindspring.com (borg.mindspring.com [204.180.128.14])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id LAA01203
         for <[email protected]>; Fri, 5 Jul 1996 11:26:47 -0700
Received: from bean (max1-dyn5.mindspring.com [205.164.235.5])
         by borg.mindspring.com (8.6.12/8.6.12) with SMTP id OAA21010
         for <[email protected]>; Fri, 5 Jul 1996 14:26:43 -0400
Message-Id: <[email protected]>
X-Sender: [email protected]
X-Mailer: Windows Eudora Pro Version 2.2 (32)
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Fri, 05 Jul 1996 14:30:29 -0500
To: [email protected]
From: "Eric F. Palmer" <[email protected]>
Subject: DBI Oraperl Emulation on NT

Has anyone usccessfully gotten the Oraperl
emulation to work on NT or is this just a
pipe dream?

I use the original Oraperl with perl 4.036
and am finding myself unsatisfied with perl 4
now that I have tasted the delicious perl 5.

I did find a single email reference on one of
the modules list for a fellow named Fox that
supposed did something with Oraperl emulation
and NT but so far my email has gone unanswered.

TIA

--------------------------------------------
SERIS - SouthEast Regional Internet Society
Enhancing Your Time On-Line
http://www.seris.org/
--------------------------------------------
Eric F. Palmer
770.825.0535 Voice - 770.825.0796 Fax
http://www.etechinc.com/   *** Standard Disclaimer ***
--------------------------------------------

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11235-24@oink>;
         Sat, 6 Jul 1996 00:47:58 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836597378:11665:6; Fri, 05 Jul 96 21:09:38 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa11624; 5 Jul 96 21:09 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA01338 for dbi-users-real; Fri, 5 Jul 1996 13:04:26 -0700
Precedence: junk
Sender: [email protected]
Received: from pearl.mhtc.net ([email protected] [206.68.40.10])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA01334
         for <[email protected]>; Fri, 5 Jul 1996 13:04:23 -0700
Received: from [206.68.43.231] ([206.68.43.231])
         by pearl.mhtc.net (8.7.5/8.6.12) with SMTP id PAA15160
         for <[email protected]>; Fri, 5 Jul 1996 15:07:57 -0500 (CDT)
Message-Id: <v02140b02ae032455bf37@[206.68.43.231]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Fri, 5 Jul 1996 15:05:11 -0500
To: [email protected]
From: [email protected] (Stephen Dahmen)
Subject: Executing stored procedure wierdness?


DBD/DBI::Oracle/Perl versions 0.69/0.32/5.002.

Any ideas why the Perl statement

$dbh->do("EXECUTE PACKAGE.PROCEDURE('ARG1')");

is returning (figuratively speaking)

ERROR # 900: ORA-00900: invalid SQL statement (DBD: possibly parse error at char
acter 1 of 32 in 'EXECUTE PACKAGE.PROCEDURE('ARG1')')

(obviously I've swapped names to make it clear).

This statement functions in SQLPLUS without a hitch.

It must be something obvious, to be happening in col 1.

Stephen


PS. Please note my address has changed to [email protected].  The old
address will work also, but for fastest delivery use the new.

~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=
Stephen Dahmen            |            CLARITY COMMUNICATIONS:
[email protected]          |   Advanced Interface Solutions for the Web
__________________________|          http://www.clotho.com/clarity
Charter Member,           |_____________________________________________
Internet Consulting Alliance:  http://www.clotho.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <11235-28@oink>;
         Sat, 6 Jul 1996 00:48:15 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836600697:17196:5; Fri, 05 Jul 96 22:04:57 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa16911; 5 Jul 96 22:04 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA01401 for dbi-users-real; Fri, 5 Jul 1996 13:58:17 -0700
Precedence: junk
Sender: [email protected]
Received: from vivid.vivid.com (vivid.vivid.com [205.163.51.81])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA01397
         for <[email protected]>; Fri, 5 Jul 1996 13:58:16 -0700
Received: from [205.163.51.232] (greedo.vivid.com [205.163.51.232])
         by vivid.vivid.com (8.7.5/8.7.3) with SMTP id NAA11244;
         Fri, 5 Jul 1996 13:56:59 -0700 (PDT)
Date: Fri, 5 Jul 1996 13:56:59 -0700 (PDT)
X-Sender: [email protected]
Message-Id: <ae02cf3c440210040f78@[205.163.51.232]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: [email protected] (Stephen Dahmen)
From: [email protected] (Jake Donham)
Subject: Re: Executing stored procedure wierdness?
Cc: [email protected]

At 8:05 PM 7/5/96, Stephen Dahmen wrote:
>DBD/DBI::Oracle/Perl versions 0.69/0.32/5.002.
>
>Any ideas why the Perl statement
>
>$dbh->do("EXECUTE PACKAGE.PROCEDURE('ARG1')");

EXECUTE is a SQL*Plus extension. Try "BEGIN PACKAGE.PROCEDURE('ARG1');
END;" instead.

__________
J.A.K.E.
code droid
http://www.vivid.com/grey/jake.html


---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <12504-5@oink>;
         Sat, 6 Jul 1996 03:39:50 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836612944:27849:2; Sat, 06 Jul 96 01:29:04 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa27366; 6 Jul 96 1:28 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA01615 for dbi-users-real; Fri, 5 Jul 1996 17:23:17 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id RAA01611
         for <[email protected]>; Fri, 5 Jul 1996 17:23:14 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ah22485; 6 Jul 96 0:22 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab08328; 6 Jul 96 1:05 +0100
Received: from toad by oink with SMTP (PP) id <12165-0@oink>;
         Sat, 6 Jul 1996 01:00:44 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA16737;
         Sat, 6 Jul 1996 01:00:33 +0000
Date: Sat, 6 Jul 1996 01:00:33 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DBI Oraperl Emulation on NT
X-Sun-Charset: US-ASCII
content-length: 651
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: "Eric F. Palmer" <[email protected]>
>
> Has anyone usccessfully gotten the Oraperl
> emulation to work on NT or is this just a
> pipe dream?
>
> I use the original Oraperl with perl 4.036
> and am finding myself unsatisfied with perl 4
> now that I have tasted the delicious perl 5.
>
> I did find a single email reference on one of
> the modules list for a fellow named Fox that
> supposed did something with Oraperl emulation
> and NT but so far my email has gone unanswered.

I've lost track of any such work.

Most NT users seem to get by with the NT ODBC module.

You could probably knock up an Oraperl emulation layer for it.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-8@oink>;
         Mon, 8 Jul 1996 07:06:56 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836642759:06613:1; Sat, 06 Jul 96 09:45:59 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa06201; 6 Jul 96 9:45 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id BAA01968 for dbi-users-real; Sat, 6 Jul 1996 01:38:38 -0700
Precedence: junk
Sender: [email protected]
Received: from husc.harvard.edu ([email protected] [140.247.30.45])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id BAA01964
         for <[email protected]>; Sat, 6 Jul 1996 01:38:37 -0700
Original-Received: from
                  fas.HARVARD.EDU by husc.harvard.edu with ESMTP id EAA18849
PP-warning: Illegal Received field on preceding line
Original-Received: by
                  fas.HARVARD.EDU with id EAA27546
PP-warning: Illegal Received field on preceding line
Date: Sat, 6 Jul 1996 04:38:31 -0400 (EDT)
From: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***
To: [email protected]
cc: Kuan Wu <[email protected]>
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


-------------------------------------------------------------------------

                               Saturday, July 6, 1996

Dear DBI Users,

       Hi, I'm new to this e-mail group, and I've searched the Web for
info on using Perl with the Microsoft Access RDBMS, but so far have found
no info on this topic.  It's my hope that someone here would be kind
enough to help me with these 3 SPECIFIC QUESTIONS:

       1.      Is it possible for Perl to interface with Microsoft Access?

       2.      Has anyone ever tried it, if it's possible?

       3.      Are there any Perl code, modules, or libraries to do this?

       I would appreciate your assistance.  Thank you.

                               Sincerely,


                               Kuan-Chuen Wu
                               E-Mail:  [email protected]

---------------------------------------------------------------------------
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-10@oink>;
         Mon, 8 Jul 1996 07:07:12 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836645891:05124:1; Sat, 06 Jul 96 10:38:11 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa04681; 6 Jul 96 10:37 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA02006 for dbi-users-real; Sat, 6 Jul 1996 02:28:21 -0700
Precedence: junk
Sender: [email protected]
Received: from husc.harvard.edu ([email protected] [140.247.30.45])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id CAA02002;
         Sat, 6 Jul 1996 02:28:20 -0700
Original-Received: from fas.HARVARD.EDU by
                  husc.harvard.edu with ESMTP id FAA29191
PP-warning: Illegal Received field on preceding line
Original-Received: by fas.HARVARD.EDU
                  with id FAA28615
PP-warning: Illegal Received field on preceding line
Date: Sat, 6 Jul 1996 05:28:16 -0400 (EDT)
From: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***
        (fwd)
To: Perl DBI Mailing List <[email protected]>, [email protected],
   [email protected]
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII



---------- Forwarded message ----------
Date: Sat, 6 Jul 1996 04:38:31 -0400 (EDT)
From: Kuan Wu <[email protected]>
To: [email protected]
Cc: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***


-------------------------------------------------------------------------

                               Saturday, July 6, 1996

Dear DBI Users,

       Hi, I'm new to this e-mail group, and I've searched the Web for
info on using Perl with the Microsoft Access RDBMS, but so far have found
no info on this topic.  It's my hope that someone here would be kind
enough to help me with these 3 SPECIFIC QUESTIONS:

       1.      Is it possible for Perl to interface with Microsoft Access?

       2.      Has anyone ever tried it, if it's possible?

       3.      Are there any Perl code, modules, or libraries to do this?

       I would appreciate your assistance.  Thank you.

                               Sincerely,


                               Kuan-Chuen Wu
                               E-Mail:  [email protected]

---------------------------------------------------------------------------

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-12@oink>;
         Mon, 8 Jul 1996 07:07:32 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836645893:05107:3; Sat, 06 Jul 96 10:38:13 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa05071; 6 Jul 96 10:38 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA02015 for dbi-users-real; Sat, 6 Jul 1996 02:30:05 -0700
Precedence: junk
Sender: [email protected]
Received: from husc.harvard.edu ([email protected] [140.247.30.45])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id CAA02011;
         Sat, 6 Jul 1996 02:30:03 -0700
Original-Received: from fas.HARVARD.EDU by
                  husc.harvard.edu with ESMTP id FAA30134
PP-warning: Illegal Received field on preceding line
Original-Received: by fas.HARVARD.EDU
                  with id FAA28638
PP-warning: Illegal Received field on preceding line
Date: Sat, 6 Jul 1996 05:29:59 -0400 (EDT)
From: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***
        (fwd)
To: Perl DBI Mailing List <[email protected]>, [email protected],
   [email protected]
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII



---------- Forwarded message ----------
Date: Sat, 6 Jul 1996 04:38:31 -0400 (EDT)
From: Kuan Wu <[email protected]>
To: [email protected]
Cc: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***


-------------------------------------------------------------------------

                               Saturday, July 6, 1996

Dear DBI Users,

       Hi, I'm new to this e-mail group, and I've searched the Web for
info on using Perl with the Microsoft Access RDBMS, but so far have found
no info on this topic.  It's my hope that someone here would be kind
enough to help me with these 3 SPECIFIC QUESTIONS:

       1.      Is it possible for Perl to interface with Microsoft Access?

       2.      Has anyone ever tried it, if it's possible?

       3.      Are there any Perl code, modules, or libraries to do this?

       I would appreciate your assistance.  Thank you.

                               Sincerely,


                               Kuan-Chuen Wu
                               E-Mail:  [email protected]

---------------------------------------------------------------------------

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-15@oink>;
         Mon, 8 Jul 1996 07:08:09 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836647136:16855:0; Sat, 06 Jul 96 10:58:56 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa16485; 6 Jul 96 10:58 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA02020 for dbi-dev-real; Sat, 6 Jul 1996 02:30:16 -0700
Precedence: junk
Sender: [email protected]
Received: from husc.harvard.edu ([email protected] [140.247.30.45])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id CAA02002;
         Sat, 6 Jul 1996 02:28:20 -0700
Original-Received: from fas.HARVARD.EDU by
                  husc.harvard.edu with ESMTP id FAA29191
PP-warning: Illegal Received field on preceding line
Original-Received: by fas.HARVARD.EDU
                  with id FAA28615
PP-warning: Illegal Received field on preceding line
Date: Sat, 6 Jul 1996 05:28:16 -0400 (EDT)
From: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***
        (fwd)
To: Perl DBI Mailing List <[email protected]>, [email protected],
   [email protected]
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII



---------- Forwarded message ----------
Date: Sat, 6 Jul 1996 04:38:31 -0400 (EDT)
From: Kuan Wu <[email protected]>
To: [email protected]
Cc: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***


-------------------------------------------------------------------------

                               Saturday, July 6, 1996

Dear DBI Users,

       Hi, I'm new to this e-mail group, and I've searched the Web for
info on using Perl with the Microsoft Access RDBMS, but so far have found
no info on this topic.  It's my hope that someone here would be kind
enough to help me with these 3 SPECIFIC QUESTIONS:

       1.      Is it possible for Perl to interface with Microsoft Access?

       2.      Has anyone ever tried it, if it's possible?

       3.      Are there any Perl code, modules, or libraries to do this?

       I would appreciate your assistance.  Thank you.

                               Sincerely,


                               Kuan-Chuen Wu
                               E-Mail:  [email protected]

---------------------------------------------------------------------------

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-17@oink>;
         Mon, 8 Jul 1996 07:09:48 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836647805:23542:0; Sat, 06 Jul 96 11:10:05 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa22988; 6 Jul 96 11:09 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA02032 for dbi-dev-real; Sat, 6 Jul 1996 02:31:48 -0700
Precedence: junk
Sender: [email protected]
Received: from husc.harvard.edu ([email protected] [140.247.30.45])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id CAA02011;
         Sat, 6 Jul 1996 02:30:03 -0700
Original-Received: from fas.HARVARD.EDU by
                  husc.harvard.edu with ESMTP id FAA30134
PP-warning: Illegal Received field on preceding line
Original-Received: by fas.HARVARD.EDU
                  with id FAA28638
PP-warning: Illegal Received field on preceding line
Date: Sat, 6 Jul 1996 05:29:59 -0400 (EDT)
From: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***
        (fwd)
To: Perl DBI Mailing List <[email protected]>, [email protected],
   [email protected]
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII



---------- Forwarded message ----------
Date: Sat, 6 Jul 1996 04:38:31 -0400 (EDT)
From: Kuan Wu <[email protected]>
To: [email protected]
Cc: Kuan Wu <[email protected]>
Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***


-------------------------------------------------------------------------

                               Saturday, July 6, 1996

Dear DBI Users,

       Hi, I'm new to this e-mail group, and I've searched the Web for
info on using Perl with the Microsoft Access RDBMS, but so far have found
no info on this topic.  It's my hope that someone here would be kind
enough to help me with these 3 SPECIFIC QUESTIONS:

       1.      Is it possible for Perl to interface with Microsoft Access?

       2.      Has anyone ever tried it, if it's possible?

       3.      Are there any Perl code, modules, or libraries to do this?

       I would appreciate your assistance.  Thank you.

                               Sincerely,


                               Kuan-Chuen Wu
                               E-Mail:  [email protected]

---------------------------------------------------------------------------

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-43@oink>;
         Mon, 8 Jul 1996 07:16:30 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836782569:27777:2; Mon, 08 Jul 96 00:36:09 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa27379; 8 Jul 96 0:35 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id QAA04244 for dbi-users-real; Sun, 7 Jul 1996 16:21:52 -0700
Precedence: junk
Sender: [email protected]
Received: from pearl.aims.gov.AU (pearl.aims.gov.au [138.7.32.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id QAA04240
         for <[email protected]>; Sun, 7 Jul 1996 16:21:49 -0700
Received: from MONGIS by pearl.aims.gov.AU (SMI-8.6/SMI-SVR4) id JAA01020;
         Mon, 8 Jul 1996 09:21:41 +1000
Message-Id: <[email protected]>
Date: Mon, 08 Jul 96 09:21:46 -0700
From: Scott Bainbridge <[email protected]>
Organization: AIMS
X-Mailer: Mozilla 1.22 (Windows; I; 16bit)
MIME-Version: 1.0
Newsgroups: comp.lang.perl.modules
To: [email protected]
Subject: Perl-DBI and Oralce Web Browser
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=us-ascii

Help!

I am running Perl 5 and the DBI database connectivity on a Sun system to
talk to an Oracle Database, the Web server I am using is the Oracle Web
Server.

I have Perl 5 with the Oracle bits and pieces up and running and it works
well,from a terminal session (Sun Solaris 2.5) I can run Perl programs and
access the Oracle database with no problems. The problems start when I try
and do the same from the Web Browser.

A straight Perl progam works with the Web with no trouble but as soon as I
put in the line:

use DBI;

the Oracle web browser gives me an error message, the error message is:


The server had an internal error and could not process the request.


Which is not a lot of help. If I comment out the Oracle connectivity stuff
it works fine and the Oracle stuff works fine if I run the perl program
from a terminal session!


Any ideas??

I have tried setting the environemtnal variables in the Web session to be
the same as the terminal session using the $ENV variable (eg; $ENV{"HOME"}
= "/dir/dir/home";) but to no avail.


If anyone has any ideas could they please mail me direct to:

[email protected]

or reply to the group.


Cheers,



Scott Bainbridge
Australian Institute of Marine Science
Townsville, North Queensland
Australia.





---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-47@oink>;
         Mon, 8 Jul 1996 07:18:29 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836800953:15922:1; Mon, 08 Jul 96 05:42:33 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa15696; 8 Jul 96 5:41 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id VAA05049 for dbi-users-real; Sun, 7 Jul 1996 21:32:09 -0700
Precedence: junk
Sender: [email protected]
Received: from netmail1.austin.ibm.com (netmail1.austin.ibm.com [129.35.208.96])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id VAA05045
         for <[email protected]>; Sun, 7 Jul 1996 21:32:07 -0700
From: [email protected]
Received: from dax.austin.ibm.com (dax.austin.ibm.com [129.35.148.59])
         by netmail1.austin.ibm.com (8.6.12/8.6.11) with ESMTP id XAA38459;
         Sun, 7 Jul 1996 23:32:04 -0500
Received: (from mhm@localhost)
         by dax.austin.ibm.com (AIX 9610R-UP4.1/UCB 8.7/8.7-client1.0)
         id XAA21936; Sun, 7 Jul 1996 23:32:04 -0500 (CDT)
Message-Id: <[email protected]>
To: [email protected]
cc: [email protected]
Subject: Re: DB2 disconnect problem
In-reply-to: Your message of Wed, 03 Jul 96 20:04:47 -0400. <[email protected]>
Date: Sun, 07 Jul 96 23:32:04 -0500


Ric Foster <[email protected]>  writes:
|>
|> I am in transition from an OS/2 & REXX environment to AIX & PERL.
|> DB2/6000 access is critical.  I installed PERL 5.002, DBI 0.70, and
|> DBD-DB2-0.59.  When I run the query.pl test program I get the following
|> on $dbh->disconnect;
|>
|>    disconnect(DBI::db=HASH(0x20018b80)) invalidates 1 associated cursor(s)
|>    at query.pl line 23.
|>
|> followed by:
|>
|>    IOT/Abort trap(coredump)
|>
|> Any clue on what I did wrong? (the fetch worked fine).

Not at the moment, as all of my code runs as expected.  I'll see
if I can force a core dump.

|> Also, there is a paragraph in the CAVEATS and HISTORY file that I
|> don't understand at all:
|>
|> > Immediately prior to calling $sth->execute() you must call
|> > $sth->bind_param(index, name, type_hash_ref).   Note, that a
|> > minimal set of type_hash_refs are provided by DB2.pm and they
|> > may be used to create new type_hash_refs
|> >    $new_type = { %$existing_type, 'Stype' => SQL_"NewTYPE" };
|>

Basically, if your statement to be executed contained any parameters
( select a, b, c from foo where a = ? and b = c) you need to execute
you need to call bind_param(index, name, type_hase_ref) once for
each question mark.  Index is the position indicator for the question
mark (1 in the contrived example), name is the name of the perl
variable, and type_hash_ref is a hash reference containing all the
information DB2 needs to know about your column but has forgotten
since the prepare() call.

Short example:

#!/usr/local/bin/perl
$ENV{"DBI_DRIVER"} = "DB2";  # set driver in the environment
$ENV{"DBI_NAME"} = "sample"; # set database name in the environment

use DBI;
use DBD::DB2::Constants;

# load the DB2 module import type information for
# int - integers, char - characters, float - floats
# date - date, ts - DB2 time_stamp
use DBD::DB2 qw($attrib_int $attrib_char $attrib_float
               $attrib_date $attrib_ts);

# set up a new hash ref for decimal types
$attrib_dec = { %$attrib_int, 'Stype' => SQL_DECIMAL, 'Scale' => 2,
                                                       'Prec' => 31 };
$DBI::dbi_debug=3; # increase the debug output

$dbh = DBI->connect('sample');  # connect to the DB2 sample database
if (!defined($dbh)) {           # check connection
# failed to connec to database
# dump the SQLerror number, the text of the error message
# and the SQLState, then exit
   print "err     ", $DBI::err, "\n";
   print "errstr  ", $DBI::errstr, "\n";
   print "state   ", $DBI::state, "\n";
       exit;
}

# select all employee which a gif format picture in the
# database
$stmt = "select empno, photo_format from inv.emp_photo where photo_format ='gif';";
$sth = $dbh->prepare($stmt);

$sth->execute();

while ((@row = $sth->fetchrow())) { # fetch empno and photo_format
# open output file with name being empno.photo_format
   open(OUTPUT,">$row[0].$row[1]") || die "Can't open $row[0].$row[1]";

#****************************************************************
#
# notice that the select statement contains a single parameter
#
   $stmt = "select picture from inv.emp_photo where empno = ? and photo_format = 'gif';";
   $pict_sth = $dbh->prepare($stmt);

#
#  After perparing the statement we need to tell DB2 where
#  to find all of the information about the variable (in this
#  this case row[0] (empno), that he (DB2) has forgotten
#
#  Note the only statements between bind_param and execute
#  are bind_param statements.
   $pict_sth->bind_param(1,$row[0],$attrib_char);
   $pict_sth->execute();
   $pict_sth->fetch();

   while ($buff = $pict_sth->readblob(1,0,4096)) {
       print OUTPUT $buff;
       $buff = "";
   }
   close(OUTPUT);
   $pict_sth->finish();
}

p
$pict_sth->finish();
$sth->finish();


Mike
----------------------------------------------------------------------
Michael H. Moran                  | Standard Disclaimer:  The content
[email protected]                | of this posting is independent of
Commercial Performance            | official IBM position.
IBM Corporation, Austin, Texas    |
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-49@oink>;
         Mon, 8 Jul 1996 07:19:15 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836802907:25331:2; Mon, 08 Jul 96 06:15:07 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa25068; 8 Jul 96 6:14 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id WAA05117 for dbi-users-real; Sun, 7 Jul 1996 22:03:10 -0700
Precedence: junk
Sender: [email protected]
Received: from netmail1.austin.ibm.com (netmail1.austin.ibm.com [129.35.208.96])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id WAA05113
         for <[email protected]>; Sun, 7 Jul 1996 22:03:09 -0700
From: [email protected]
Received: from dax.austin.ibm.com (dax.austin.ibm.com [129.35.148.59])
         by netmail1.austin.ibm.com (8.6.12/8.6.11) with ESMTP id AAA53412;
         Mon, 8 Jul 1996 00:03:03 -0500
Received: (from mhm@localhost)
         by dax.austin.ibm.com (AIX 9610R-UP4.1/UCB 8.7/8.7-client1.0)
         id AAA18096; Mon, 8 Jul 1996 00:03:01 -0500 (CDT)
Message-Id: <[email protected]>
To: Tim Bunce <[email protected]>
cc: [email protected], [email protected], [email protected]
Subject: Re: DB2 disconnect problem
In-reply-to: Your message of Thu, 04 Jul 96 19:39:08 +0000. <[email protected]>
Date: Mon, 08 Jul 96 00:03:00 -0500


Tim Bunce <[email protected]>  writes:
|>
|> > From: Ric Foster <[email protected]>
|> >
|> > I assume that close is accomplished by $cursor->finish.
|>
|> Yes. I suspect that the DBD::DB2 finish method is not using the
|> internal DBI DBIc_ACTIVE_off() macro.

Wrong, DBD::DB2 does use the DBIc_ACTIVE_off() macro.  The problem
actually looks like a symbole resolution problem in that
IOTrap is being caused by  dbhi_getcom() explicitly calling
abort.

Excerpt frm DBI-0.70:

   static imp_xxh_t *
   dbih_getcom(hrv)    /* Get com struct for handle. Must be fast.     */
       SV *hrv;
   {
       MAGIC *mg;
       SV *sv;

       /* important and quick sanity check (esp non-'safe' Oraperl)    */
       if (!SvROK(hrv)                 /* must at least be a ref */
           && hrv != DBI_LAST_HANDLE   /* special for var::FETCH */) {
           sv_dump(hrv);
           abort();
           croak("Invalid DBI handle %s", SvPV(hrv,na));
       }

[ rest intentionally deleted ]

I'll chase it some Monday morning.

Mike

----------------------------------------------------------------------
Michael H. Moran                  | Standard Disclaimer:  The content
[email protected]                | of this posting is independent of
Commercial Performance            | official IBM position.
IBM Corporation, Austin, Texas    |
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-50@oink>;
         Mon, 8 Jul 1996 07:19:35 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836803438:27859:1; Mon, 08 Jul 96 06:23:58 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa27628; 8 Jul 96 6:23 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id WAA05128 for dbi-users-real; Sun, 7 Jul 1996 22:10:18 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id WAA05124
         for <[email protected]>; Sun, 7 Jul 1996 22:10:16 -0700
Received: by gw.home.vix.com id WAA27444; Sun, 7 Jul 1996 22:10:13 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from wurrup.bha.oz.au (daemon@localhost)
         by online.tmx.com.au (8.6.5/8.6.5) with MHSnet id PAA19996
         for [email protected]; Mon, 8 Jul 1996 15:08:20 +1000
Received: by wurrup (AIX 3.2/UCB 5.64/4.03) id AA56231;
         Mon, 8 Jul 1996 15:04:21 +1000
From: [email protected] (adb)
Message-Id: <9607080504.AA56231@wurrup>
To: [email protected]
Subject: Problem inserting into tables with >50 columns
Date: Mon, 08 Jul 96 15:04:20 +1000


I'm using DBI (v0.70) and DBD-Oracle (v0.35) on IRIX 5.3 to load data
into tables using an SQL INSERT of the form:

 "INSERT INTO $table ($cols) VALUES (?, ?, ? ... )"

The insert works fine on about 30 different tables (largest is 35
columns), but perl/DBI/DBD dumps core (segmentation fault) on a couple
of tables that have >50 columns.

The core dump occurs when executing the insert (the `prepare' works
fine).  Note that there is no problem with the insert statement, and
none of the a table integrity constraints are violated by the data I'm
inserting.

I've attached the `perl -V' output and will follow up with a stack
trace (once I work out how to get one).  The current installation of
DBI and DBD passed all tests.  The problem occurs with our previous
installation as well (DBI v0.68, DBD v0.28).  I'm currently scouring
the DBI and DBD code to see if I can find the problem.

It appears that there is an array overflow happening somewhere.  I'd
appreciate any suggestions about where this overflow is occuring.

Ciao,

AndyB

------------------------------------------------------------------------
Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=irix, osver=5, archname=IP22-irix
   uname='irix kingfisher 5.3 11091812 ip22 mips '
   hint=recommended, useposix=true
 Compiler:
   cc='cc', optimize='-O', gccversion=
   cppflags='-D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -I/usr/local/include -DLANGUAGE_C'
   ccflags ='-D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -Olimit 3000 -I/usr/local/include -DLANGUAGE_C'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=1, d_casti32=define, d_castneg=
   intsize=4, alignbytes=8, usemymalloc=n, randbits=15
 Linker and Libraries:
   ld='ld', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /usr/lib /lib
   libs=-lmalloc -lsun -lm -lc -lcrypt -lbsd -lPW
   libc=/usr/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags=' ', lddlflags='-shared -L/usr/local/lib'

@INC: /usr/local/lib/perl5/IP22-irix/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/IP22-irix /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/IP22-irix .
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <18417-51@oink>;
         Mon, 8 Jul 1996 07:19:57 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836804452:03195:1; Mon, 08 Jul 96 06:40:52 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa02987; 8 Jul 96 6:39 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id WAA05153 for dbi-users-real; Sun, 7 Jul 1996 22:24:57 -0700
Precedence: junk
Sender: [email protected]
Received: from ny.engin.umich.edu ([email protected] [141.212.106.70])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id WAA05149
         for <[email protected]>; Sun, 7 Jul 1996 22:24:56 -0700
Received: from perfect (pm1-10.eecs.umich.edu [141.213.6.235])
         by ny.engin.umich.edu (8.6.12/8.6.4) with SMTP id BAA13256;
         Mon, 8 Jul 1996 01:24:37 -0400
Message-ID: <[email protected]>
Date: Mon, 08 Jul 1996 01:29:36 -0500
From: Dan DeMaggio <[email protected]>
X-Mailer: Mozilla 3.0b4 (WinNT; I)
MIME-Version: 1.0
To: [email protected], [email protected], [email protected]
CC: [email protected]
Subject: re: DBI on Windows NT (interm)
References: <[email protected]>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Windows NT Users:
-----------------

As far as I know, nobody has compiled DBI or oraperl for Windows NT.
(If they have, let me know). In the interm, you can use my ODBC perl
module (re-written by Dave Roth (Thanks Dave)).

Go to http://www.perl.com/CPAN/modules/01modules.index.html
and find Dave Roth's Win32ODBC module. With it, you will be able
to talk to any ODBC datasource (from Oracle to SQL Server to Excel
to Access, etc..)

Hope this helps!
                               -=Dan=- ([email protected])

P.S. Alligator, perhaps you could mention this in the DBI FAQ?
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <19219-0@oink>;
         Mon, 8 Jul 1996 08:10:09 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836809645:05323:2; Mon, 08 Jul 96 08:07:25 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa04916; 8 Jul 96 8:06 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id XAA05216 for dbi-users-real; Sun, 7 Jul 1996 23:51:15 -0700
Precedence: junk
Sender: [email protected]
Received: from gw.home.vix.com (gw.home.vix.com [192.5.5.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id XAA05212
         for <[email protected]>; Sun, 7 Jul 1996 23:51:14 -0700
Received: by gw.home.vix.com id XAA02438; Sun, 7 Jul 1996 23:51:01 -0700 (PDT)
X-btw: vix.com is also gw.home.vix.com and vixie.sf.ca.us
Received: from wurrup.bha.oz.au (daemon@localhost)
         by online.tmx.com.au (8.6.5/8.6.5) with MHSnet id QAA28748
         for [email protected]; Mon, 8 Jul 1996 16:49:36 +1000
Received: by wurrup (AIX 3.2/UCB 5.64/4.03) id AA25458;
         Mon, 8 Jul 1996 16:46:30 +1000
From: [email protected] (adb)
Message-Id: <9607080646.AA25458@wurrup>
To: [email protected]
Subject: Re: Problem inserting into tables with >50 columns
In-Reply-To: (Your message of Mon, 08 Jul 96 15:04:20 W.)
Date: Mon, 08 Jul 96 16:46:28 +1000



I wrote:

> I'm using DBI (v0.70) and DBD-Oracle (v0.35) on IRIX 5.3 to load data
> into tables using an SQL INSERT of the form:
>
>   "INSERT INTO $table ($cols) VALUES (?, ?, ? ... )"
>
> The insert works fine on about 30 different tables (largest is 35
> columns), but perl/DBI/DBD dumps core (segmentation fault) on a couple
> of tables that have >50 columns.

I found the problem (turns out that it *was* in the prepare).  The
'dbd_preparse' function of dbdimp.c in DBD-Oracle-0.35 was a little
optimistic in its memory allocation for a (modified) copy of the SQL
statement.  Context diffs for the change are attached:

Ciao,

AndyB

------------------------------------------------------------------------
diff -c -r1.1 dbdimp.c
*** dbdimp.c    1996/07/08 05:32:23     1.1
--- dbdimp.c    1996/07/08 05:34:51
*** 356,362 ****

     /* allocate room for copy of statement with spare capacity        */
     /* for editing ':1' into ':p1' so we can use obndrv.      */
!     imp_sth->statement = (char*)safemalloc(strlen(statement) + 100);

     /* initialise phs ready to be cloned per placeholder      */
     memset(&phs_tpl, 0, sizeof(phs_tpl));
--- 356,362 ----

     /* allocate room for copy of statement with spare capacity        */
     /* for editing ':1' into ':p1' so we can use obndrv.      */
!     imp_sth->statement = (char*)safemalloc(strlen(statement) * 3);

     /* initialise phs ready to be cloned per placeholder      */
     memset(&phs_tpl, 0, sizeof(phs_tpl));
------------------------------------------------------------------------
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <21425-1@oink>;
         Mon, 8 Jul 1996 10:09:26 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836812086:24800:1; Mon, 08 Jul 96 08:48:06 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa24415; 8 Jul 96 8:47 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id AAA05293 for dbi-users-real; Mon, 8 Jul 1996 00:38:48 -0700
Precedence: junk
Sender: [email protected]
Received: from cegelecproj.co.uk (ganymede.cegelecproj.co.uk [159.245.72.6])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id AAA05289
         for <[email protected]>; Mon, 8 Jul 1996 00:38:44 -0700
Received: from vampire.cegelecproj.co.uk (cerberus.cegelecproj.co.uk)
         by cegelecproj.co.uk (4.1/SMI-4.1) id AA05188;
         Mon, 8 Jul 96 08:14:16 BST
Received: from phantom.cegelecproj.co.uk
         by vampire.cegelecproj.co.uk (SMI-8.6/SMI-SVR4) id IAA11376;
         Mon, 8 Jul 1996 08:14:10 +0100
Received: by phantom.cegelecproj.co.uk (SMI-8.6/SMI-SVR4) id IAA04606;
         Mon, 8 Jul 1996 08:14:09 +0100
Date: Mon, 8 Jul 1996 08:14:09 +0100
From: [email protected]
Message-Id: <[email protected]>
To: Scott Bainbridge <[email protected]>, [email protected]
Subject: Re: Perl-DBI and Oralce Web Browser
X-Mailer: Wilymail 0.4

> A straight Perl progam works with the Web with no trouble but as soon as I
> put in the line:
>
> use DBI;
>
> the Oracle web browser gives me an error message, the error message is:
>
> The server had an internal error and could not process the request.

This means that the CGI program exited with non-zero status. Usually,
it's because of perl giving some compilation error, and this is particularly
likely if it's failing when you add a "use" line. You should be using perl -w,
and preferably "use strict" too, but it's often a good idea to check the program
with perl -c from the command line before invoking it with via a browser.

For what it's worth, here's Steve's Oracle Webserver Checklist. For the most
part, they assume that you're trying to invoke PL/SQL procedures directly
through Oracle's PL/SQL Agent, rather than just calling a Perl CGI script,
but I've found that I've had to use both Perl and PL/SQL, so perhaps you will
too. :-)

Symptom:
When attempting to access URLs such as:

http://host:8080/ows-bin/service/owa/pkg.proc

the server responds with "Request failed. We were unable to
process your request at this time. Please try again later."

Cause:
(i) The listener can't read the configuration file owa.cfg.
(ii) The listener is attempting to invoke the PL/SQL procedure named,
but can't, for some reason.

Fix:
(i) Make sure that /home/oracle/owa2/admin/owa.cfg is readable by the user
    that the listener runs as.
(ii) (a) Check that the Oracle user for the listener has the permissions to
        invoke such functions: do "pkg.proc" in sqlplus. There won't
        be any output, but it should say that it ran ok.
    (b) Make sure that the names of the input fields in the form
        match up exactly with the names of the parameters in the
        script. There must be a one-to-one mapping.
    (c) If you've got multiple-value fields, such as SELECTs,
        make sure they're of the right type (defined in the
        OWA_UTIL package), and that there's an extra value
        supplied by a hidden field to force at least one value
        to be selected.
(iii) If all else fails, the procedure is probably raising an
     exception. Put OTHERS exception handlers around everything
     and print out error messages in them.


Symptom:
When attempting to access URLs such as:

http://host:8080/ows-bin/service/owa/pkg.proc

the server responds with "Requested URL was not found on this
server".

Cause:
This is due to there not being a mapping in the service listener's
Virtual Directory Mappings configuration, of the form:

/home/oracle/ows2/bin/ CN /ows-bin/service/

so the "owa" program isn't found.

Fix:
Create the virtual directory mapping, halt the listener, and
restart it.

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <22111-1@oink>;
         Mon, 8 Jul 1996 12:13:22 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836817508:18513:1; Mon, 08 Jul 96 10:18:28 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa17723; 8 Jul 96 10:17 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA05413 for dbi-users-real; Mon, 8 Jul 1996 02:09:56 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-1.mail.demon.net (relay-1.mail.demon.net [158.152.1.140])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id CAA05409
         for <[email protected]>; Mon, 8 Jul 1996 02:09:54 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-1.mail.demon.net
         id ae21211; 8 Jul 96 10:09 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ab12574; 8 Jul 96 10:09 +0100
Received: from toad by oink with SMTP (PP) id <21401-0@oink>;
         Mon, 8 Jul 1996 10:08:26 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA21851;
         Mon, 8 Jul 1996 10:08:18 +0000
Date: Mon, 8 Jul 1996 10:08:18 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: DB2 disconnect problem
Cc: [email protected], [email protected]
X-Sun-Charset: US-ASCII
content-length: 767
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected]
>
> Tim Bunce <[email protected]>  writes:
> |>
> |> > From: Ric Foster <[email protected]>
> |> >
> |> > I assume that close is accomplished by $cursor->finish.
> |>
> |> Yes. I suspect that the DBD::DB2 finish method is not using the
> |> internal DBI DBIc_ACTIVE_off() macro.
>
> Wrong, DBD::DB2 does use the DBIc_ACTIVE_off() macro.  The problem
> actually looks like a symbole resolution problem in that
> IOTrap is being caused by  dbhi_getcom() explicitly calling
> abort.

The abort() got left in by accident (I was using it for some debugging).
I took it out awhile ago but haven't got around to making another release.

I'll make another release within a week or so. It'll require 5.003 so
people should upgrade now.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <22111-3@oink>;
         Mon, 8 Jul 1996 12:13:29 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836817511:18537:2; Mon, 08 Jul 96 10:18:31 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa18016; 8 Jul 96 10:17 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA05422 for dbi-users-real; Mon, 8 Jul 1996 02:10:05 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-1.mail.demon.net (relay-1.mail.demon.net [158.152.1.140])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id CAA05418
         for <[email protected]>; Mon, 8 Jul 1996 02:10:04 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-1.mail.demon.net
         id ad21220; 8 Jul 96 10:09 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa12574; 8 Jul 96 10:09 +0100
Received: from toad by oink with SMTP (PP) id <21385-0@oink>;
         Mon, 8 Jul 1996 10:05:17 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA21846;
         Mon, 8 Jul 1996 10:05:06 +0000
Date: Mon, 8 Jul 1996 10:05:06 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Problem inserting into tables with >50 columns
X-Sun-Charset: US-ASCII
content-length: 782
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (adb)
>
> I wrote:
>
> > I'm using DBI (v0.70) and DBD-Oracle (v0.35) on IRIX 5.3 to load data
> > into tables using an SQL INSERT of the form:
> >
> >   "INSERT INTO $table ($cols) VALUES (?, ?, ? ... )"
> >
> > The insert works fine on about 30 different tables (largest is 35
> > columns), but perl/DBI/DBD dumps core (segmentation fault) on a couple
> > of tables that have >50 columns.
>
> I found the problem (turns out that it *was* in the prepare).  The
> 'dbd_preparse' function of dbdimp.c in DBD-Oracle-0.35 was a little
> optimistic in its memory allocation for a (modified) copy of the SQL
> statement.  Context diffs for the change are attached:
>
> Ciao,
>
> AndyB

Thanks Andy. It'll be in the next release. Sorry for the hassle.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <22111-5@oink>;
         Mon, 8 Jul 1996 12:13:36 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836817578:19115:1; Mon, 08 Jul 96 10:19:38 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa18600; 8 Jul 96 10:18 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id CAA05431 for dbi-users-real; Mon, 8 Jul 1996 02:12:23 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-1.mail.demon.net (relay-1.mail.demon.net [158.152.1.140])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id CAA05427
         for <[email protected]>; Mon, 8 Jul 1996 02:12:22 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-1.mail.demon.net
         id ab21766; 8 Jul 96 10:11 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa12925; 8 Jul 96 10:11 +0100
Received: from toad by oink with SMTP (PP) id <21447-0@oink>;
         Mon, 8 Jul 1996 10:10:45 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA21868;
         Mon, 8 Jul 1996 10:10:38 +0000
Date: Mon, 8 Jul 1996 10:10:38 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Perl-DBI and Oralce Web Browser
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 678
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Scott Bainbridge <[email protected]>
>
> A straight Perl progam works with the Web with no trouble but as soon as I
> put in the line:
>
> use DBI;
>
> the Oracle web browser gives me an error message, the error message is:
> The server had an internal error and could not process the request.
>
> Which is not a lot of help. If I comment out the Oracle connectivity stuff
> it works fine and the Oracle stuff works fine if I run the perl program
> from a terminal session!
>
> Any ideas??

Sigh. Read the CGI programming FAQ's then fetch and use the CGI::ErrorWrap
module then you'll probably find that you've forgotten to set ORACLE_HOME.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <22111-9@oink>;
         Mon, 8 Jul 1996 12:13:49 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836820958:21507:2; Mon, 08 Jul 96 11:15:58 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa21297; 8 Jul 96 11:15 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id DAA05528 for dbi-users-real; Mon, 8 Jul 1996 03:06:55 -0700
Precedence: junk
Sender: [email protected]
Received: from nordwest.pop.de (mail.NordWest.POP.DE [193.100.96.7])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id DAA05524
         for <[email protected]>; Mon, 8 Jul 1996 03:06:32 -0700
Received: by nordwest.pop.de (Smail3.1.28.1 #6) id m0udDf3-0000rjC;
         Mon, 8 Jul 96 12:35 MES
Received: from hp01 (hp01 [194.15.215.78])
         by megatel.megatel.de (8.7.3/8.7.3/gfd951221) with SMTP id LAA02111;
         Mon, 8 Jul 1996 11:52:57 +0200 (MET DST)
Date: Mon, 8 Jul 1996 11:52:55 +0200 (METDST)
From: Sebastian Meyer <[email protected]>
X-Sender: meyer@hp01
To: Scott Bainbridge <[email protected]>
cc: [email protected]
Subject: Re: Perl-DBI and Oralce Web Browser
In-Reply-To: <[email protected]>
Message-ID: <Pine.HPP.3.91.960708114756.21391G-100000@hp01>
Organization: megatel Informations- und Kommunikationssysteme GmbH
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Mon, 8 Jul 1996, Scott Bainbridge wrote:

> Help!
>
> I am running Perl 5 and the DBI database connectivity on a Sun system to
> talk to an Oracle Database, the Web server I am using is the Oracle Web
> Server.
> [...]
> A straight Perl progam works with the Web with no trouble but as soon as I
> put in the line:
>
> use DBI;
>
> the Oracle web browser gives me an error message, the error message is:
> The server had an internal error and could not process the request.
> [...]
> Any ideas??

Shure, lots of. But you didn't supply sufficient background to decide
what is the right one.
First of all: You have a CGI problem, not a perl or dbi
problem. Following my sig is the advice Tom Christiansen collected for
people with this kind of probs. He called it "The Idiot's Guide to
Solving Perl CGI Problems".

You can find it at http://www.perl.com/perl/faq/idiots-guide.html, too.

sebastian
--
   __o     Sebastian Meyer  Tel: 0421/22095-32   | megatel GmbH
 _ \<,_    eMail: [email protected]               | Wienerstr. 3
(_)/ (_)   [http://muffin.physik.uni-bremen.de/] | D-28359 Bremen
~~~~~~~~~~ "To err is human, to really foul things up requires a computer"
===============================================================================
From: Tom Christiansen <[email protected]>
Newsgroups: comp.infosystems.www.authoring.cgi,comp.lang.perl.misc
Subject: The Idiot's Guide to Solving Perl CGI Problems (update)
Followup-To: comp.infosystems.www.authoring.cgi

If you're like most people on the newsgroup, you've presented us with
the clasiqumsimo CGI problem: "it doesn't work", without providing us
sufficient background to divine what the real problem is.  This is
infuriating, because is makes you come off like some drooling idiot
begging for free consulting help without having done any investigation
on your own part.  This drives people away who could otherwise help you
and impoverishes the Net by exacerbating the already annoyingly low
signal-to-noise ratio.

For a start, you need to look through the checklist below for a few simple
questions that need answering.  While some sample answers are provided,
you'll have to figure them out on your own.

=========================================================================

   Who owns the script?
       e.g. I do.

   What are the script's permissions?
       e.g. 0755

   Is your script in the right directory?
       e.g. No, I forgot to put it in /usr/local/etc/httpd/cgi-bin/.

   Is CGI execution enabled in your server for that directory
   and/or file suffix?
       e.g. No, my sysadmin forgot to configure it that way.

   Under what uid does the server runs its CGI programs?
       e.g. wwwuser (Oops, it can't write my files or directories.)

   Can the server's uid write any files you're trying to write?
       e.g. Nope -- I own the files, but it doesn't run as me
            and the permissions are 0600 instead of 0666.  I
            guess that's why it can't open my own files.

   What happens when you run it interactively?
       e.g. I didn't know you could run these interactively
            because I never bothered to read the documentation
            for the CGI.pm library.

   What's in the server error log?
       e.g. I never thought to look.  Oh, there it is -- never mind.

   Where's the server log?
       (No way to know -- it's system dependent.  Check with your
        admin if you can't find it in /usr/local/etc/httpd/logs/error_log)

   What's the Perl version and OS version?
           (try perl -v and uname -a)
       e.g. Perl version 5.002, SunOS version 4.1.3
           (if your Perl is below 5.002 in version, *UPGRADE NOW*)

   What's the library version?
       e.g. grep -i version in the library, or for CGI.pm, do this:
           $ perl -le 'use CGI; print $CGI::VERSION'
           2.21

   What's the path to your perl executable on the server?
       e.g. /contrib/bin/perl

   What's the path to your perl executable specified in your script?
       e.g. /usr/bin/perl  (oops, of course it can't find it)

   What's the http daemon server's version number so people who
   might help you have a better clue about your local environment?
       e.g. NCSA 1.5

   What happens when you add Perl's -w flag?
       e.g. It tells me about my silly mistakes that are listed in detail
            in the perldiag manpage where I diligently looked them up.

   What happens when you add Perl's -T flag?
       e.g. It tells me about security problems as described in the
            perlsec man page, which I've carefully read and understood.
            I even read the CGI Security faq in
            http://www-genome.wi.mit.edu/WWW/faqs/www-security-faq.html

   What happens when you add `use strict'?
       e.g. It makes me declare my variables and quote my strings
            and finds all these silly errors which I've carefully
            corrected using my() declarations and quotes.

   Did you remember to flush STDOUT via $| = 1 at the top of your script?
       e.g. Nope; gosh, no wonder it doesn't get out before it bombs.

   What happens when you check return values of EACH AND EVERY ONE
   of your syscalls, probably with something like
         open(FILE, ">some_file") || die("can't write some_file: $!");
       e.g. it show that $! contains "Permission denied" or
            "No such file or directory".

   Did you use the standard CGI.pm module to do this for you instead of
   parsing by hand (which would be a really idiotic idea) or use the
   more limited cgi-lib.pl library?
       e.g. No, I didn't know I didn't have to recreate the wheel
           (badly) and that I could get the neat CGI module from
               http://www.perl.com/cgi-bin/cgi_mod?modules=CGI

   If you used a library, did you type "make install" to place it in the
   proper system directory (somewhere in @INC, probably a lot like
   /usr/local/lib/perl/site_perl/CGI.pm) so it can be properly found and
   have its set permissions to 0644, or did you just blindly do a manual
   copy and screw up the path, permissions, or both?
       e.g. What's make?  Looks like I just misinstalled it myself by
            a silly copy to cgi-bin and then I made it mode 0700.
            I guess that doesn't work at all, does it?

   Did you remember to post to comp.infosystems.www.authoring.misc,
   instead of blasting the comp.lang.perl.* groups with questions
   that aren't related to Perl in the least?
       e.g. Nope -- is that why I don't get any useful answers but
            just a bunch of flames instead?

   If you're running some brain-dead, bondage-and-discipline pseudo-
   operating system from The Evil Empire, did you bother to read what
   the Windoze Perl FAQ says about the Web?
       e.g., No, I didn't know about http://www.perl.hip.com/PerlFaq.htm

   Did you avoid placing an interpreter in your cgi-bin the way CERT
   and http://www.perl.com//perl/news/latro-announce.html warn you about?
       e.g. No, what's wrong with that? (noise of disk grinding to
            dust after corporate secrets stolen heard in the background)

=========================================================================

Odds are that if you'd merely answer these questions and carry out the
specified steps, you yourself would solve your very own problem without
having to come begging for help and pissing off the Net.

Once you have thoroughly investigated *ALL* of these matters on your own
and perhaps even asking your buddy in the next cubicle whether he's seen
anything like it, should you find that it's still a problem, then by all
means please go ahead and post, but do make sure you give us plenty of
details, including your own answers to all (or at least most) of the
questions listed above.

On the other hand, if most of those questions don't even begin to make the
least bit of sense to you, you're already in way over your head.  If this
should be the case, you'd better hold off on CGI programming and go find
yourself a local guru and check out some of the many books listed at

   http://wwwiz.com/books/cgi-perl.html

IM!HO, the Addison-Wesley and O'Reilly books tend to be the most thorough,
but there are so many out there that you'd do well to look them over
yourself.  Remember that thickness needn't imply clarity, completeness,
or correctness.

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <24312-1@oink>;
         Mon, 8 Jul 1996 17:01:06 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836825381:06223:3; Mon, 08 Jul 96 12:29:41 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa06057; 8 Jul 96 12:29 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id EAA05662 for dbi-users-real; Mon, 8 Jul 1996 04:15:46 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-2.mail.demon.net (disperse.demon.co.uk [158.152.1.77])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id EAA05658
         for <[email protected]>; Mon, 8 Jul 1996 04:15:44 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-2.mail.demon.net
         id ae04146; 8 Jul 96 12:15 +0100
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa14070; 8 Jul 96 12:11 +0100
Received: from toad by oink with SMTP (PP) id <21736-0@oink>;
         Mon, 8 Jul 1996 10:31:27 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA22021;
         Mon, 8 Jul 1996 10:31:20 +0000
Date: Mon, 8 Jul 1996 10:31:20 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: *** Request for Info On Using Perl with Microsoft Access RDBMS ***
        (fwd)
X-Sun-Charset: US-ASCII
content-length: 1440
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> Original-Received: from fas.HARVARD.EDU by
>                    husc.harvard.edu with ESMTP id FAA30134
> Pp-Warning: Illegal Received field on preceding line
> Original-Received: by fas.HARVARD.EDU
>                    with id FAA28638
> Pp-Warning: Illegal Received field on preceding line

Tell your system admin people that their mail headers are invalid.

> Subject: *** Request for Info On Using Perl with Microsoft Access RDBMS ***

> To: Perl DBI Mailing List <[email protected]>, [email protected],
>         [email protected]

It's impolite to send a message to a mailing list without understanding
what the mailing list is for. Sending a message to all three of the dbi
groups (especially -announce) certainly does not encourage people to
want to spend time helping you.

Sending the message again to just dbi-users simply compounds the mistake
since many people would have already received three copies.

> Dear DBI Users,
>
>       Hi, I'm new to this e-mail group, and I've searched the Web for
> info on using Perl with the Microsoft Access RDBMS, but so far have found
> no info on this topic.  It's my hope that someone here would be kind
> enough to help me with these 3 SPECIFIC QUESTIONS:
>
>       1.      Is it possible for Perl to interface with Microsoft Access?

You might be able to use the ODBC module:
ftp://ftp.infowire.com/pub/ntperl/NTodbc_v0.5.zip
That module is not supported on this mailing list.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <24309-7@oink>;
         Mon, 8 Jul 1996 17:01:32 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836840611:16154:2; Mon, 08 Jul 96 16:43:31 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa15780; 8 Jul 96 16:42 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA06123 for dbi-users-real; Mon, 8 Jul 1996 08:30:55 -0700
Precedence: junk
Sender: [email protected]
Received: from ns.cdv.com (cdv.com [206.105.122.2])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA06119
         for <[email protected]>; Mon, 8 Jul 1996 08:30:52 -0700
Received: from nt351 by ns.cdv.com via SMTP (940816.SGI.8.6.9/940406.SGI.AUTO)
         for <[email protected]> id LAA11363; Mon, 8 Jul 1996 11:50:56 -0700
Message-ID: <[email protected]>
Date: Mon, 08 Jul 1996 11:27:01 -0400
From: Syme Kutz <[email protected]>
X-Mailer: Mozilla 3.0b4Gold (WinNT; I)
MIME-Version: 1.0
To: [email protected]
Subject: $dbh->do statments
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

#!/usr/bin/perl -w

use DBI;

%ENV{"ORACLE_HOME"}="oracle/home";
%ENV{"ORACLE_SID"}="database";

$drh = DBI->install_driver( 'Oracle' );
$dbh = $drh->connect( 'database','user','password');
die unless $dbh;

$dbh->do("delete cart;commit;");

# print "Content-type: text/html\n\n";
# $cursor = $dbh->prepare("select iname from item");
# $cursor->execute;
# while ($field = $cursor->fetchrow )
# { print "Field: $field";
# print "<br>"; }
# $cursor->finish;

$dbh->disconnect;
exit;

here is a snippet of code. If I unrem the cursor section it runs both
from command line and web browser. however no matter what i do with the
above do statment i doesn;t seem to work. I know this is an oversight on
my part but I give.... I'm runing oracle 7.1.6 with webserver 1.0.1,
perl 5.002, dbi .70, DBD::Oracle .35. I would like to be able to execute
a call similar to above via cron and I also need to execute stored
procedures which I see I should be able to do but.....

I have this test script in my cgi-bin dir owned by oracle who runs my
oracle daemons etc. The permissions are 777. I figure it's something in
my call as cursors run from everywhere just can't get the do to work.


Thanks in Advance.
 Syme Kutz
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <25486-0@oink>;
         Mon, 8 Jul 1996 19:52:39 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836843102:08330:4; Mon, 08 Jul 96 17:25:02 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa08063; 8 Jul 96 17:24 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id JAA06220 for dbi-users-real; Mon, 8 Jul 1996 09:14:39 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id JAA06216
         for <[email protected]>; Mon, 8 Jul 1996 09:14:35 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id aa15279; 8 Jul 96 16:14 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa27966; 8 Jul 96 17:05 +0100
Received: from toad by oink with SMTP (PP) id <24363-0@oink>;
         Mon, 8 Jul 1996 17:04:42 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA27100;
         Mon, 8 Jul 1996 17:04:34 +0000
Date: Mon, 8 Jul 1996 17:04:34 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Perl support for MS Access
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 461
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424

Forwarded to dbi-users mailing list.

----- Begin Included Message -----

From [email protected]  Mon Jul  8 17:02:53 1996
From: Satish <[email protected]>
Subject: Perl support for MS Access
To: [email protected]
Date: Mon, 8 Jul 96 14:17:25 IST


       For some reason, I want to interface my Perl script with
       a MS Access database. Any leads on the same would be appreciated.
       Best Regards,
       Satish Patil.


----- End Included Message -----

---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <25482-14@oink>;
         Mon, 8 Jul 1996 19:53:24 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836849522:04770:1; Mon, 08 Jul 96 19:12:02 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa03929; 8 Jul 96 19:11 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA06379 for dbi-users-real; Mon, 8 Jul 1996 11:02:13 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id LAA06375
         for <[email protected]>; Mon, 8 Jul 1996 11:02:11 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ah15134; 8 Jul 96 18:00 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa29619; 8 Jul 96 17:09 +0100
Received: from toad by oink with SMTP (PP) id <24451-0@oink>;
         Mon, 8 Jul 1996 17:09:08 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA27147;
         Mon, 8 Jul 1996 17:08:52 +0000
Date: Mon, 8 Jul 1996 17:08:52 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: $dbh->do statments
X-Sun-Charset: US-ASCII
content-length: 410
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Syme Kutz <[email protected]>
>
> If I unrem the cursor section it runs both
> from command line and web browser. however no matter what i do with the
> above do statment i doesn;t seem to work.

Anyone posting CGI related problems to this mailing list who has not
use the CGI::ErrorWrap module will be ignored by me. Sorry but I just
don't have the time. Perhaps others will feel more generous.

Tim.
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <26727-3@oink>;
         Tue, 9 Jul 1996 00:10:46 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836858801:27756:2; Mon, 08 Jul 96 21:46:41 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa27553; 8 Jul 96 21:46 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id NAA06629 for dbi-users-real; Mon, 8 Jul 1996 13:31:31 -0700
Precedence: junk
Sender: [email protected]
Received: from ws5.justintime.com (ws5.justintime.com [205.149.189.36])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id NAA06625
         for <[email protected]>; Mon, 8 Jul 1996 13:31:28 -0700
Received: (from ml@localhost) by ws5.justintime.com (8.7.3/8.7.3) id NAA28880;
         Mon, 8 Jul 1996 13:32:28 -0700 (PDT)
Date: Mon, 8 Jul 1996 13:32:28 -0700 (PDT)
From: Mike Linksvayer <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Problems with make test DBD-Oracle-0.35
Cc: [email protected], [email protected]


I'm also having problems with make test.  I tried the patch
Makefile.PL patch posted by Tim on July 1, but it doesn't
make any difference.  I know that I should be using proc.mk,
but I don't have it available.

Script of perl -V and attempted install follows.  I did a
make realclean first.

Any ideas?
Thanks!

Mike Linksvayer   [email protected]


redcomet:/export/home/ml/DBD-Oracle-0.35# perl -V
Summary of my perl5 (5.0 patchlevel 3 subversion 0) configuration:
 Platform:
   osname=solaris, osver=2.4, archname=sun4-solaris
   uname='sunos redcomet 5.4 generic_101945-39 sun4m sparc '
   hint=recommended, useposix=true, d_sigaction=define
 Compiler:
   cc='gcc', optimize='-O', gccversion=2.7.2
   cppflags='-I/usr/local/include'
   ccflags ='-I/usr/local/include'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='gcc', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /lib /usr/lib /usr/ccs/lib
   libs=-lsocket -lnsl -ldl -lm -lc -lcrypt
   libc=/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-fpic', lddlflags='-G -L/usr/local/lib'

@INC: /usr/local/lib/perl5/sun4-solaris/5.003 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/sun4-solaris /usr/local/lib/perl5/site_perl .
redcomet:/export/home/ml/DBD-Oracle-0.35# perl Makefile.PL -m /export/home/oracle/rdbms/lib/oracle.mk
Useless use of not in void context at Makefile.PL line 336.
Useless use of not in void context at Makefile.PL line 336.

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /export/home/oracle
Using /export/home/oracle/rdbms/lib/oracle.mk (version 1.14)
System: perl5.003 sunos redcomet 5.4 generic_101945-39 sun4m sparc
Compiler: gcc -O -I/usr/local/include
Linker:   /usr/ccs/bin/ld
Oracle proc.mk would have used these values but we override them:
 CC:       ld -dy $(COMPOBJS)
 LDFLAGS:  -L$(LIBHOME)
 LDSTRING: -R /opt/SUNWcluster/lib -Y P,$(LD_LIBRARY_PATH):/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib -Qy -lc $(COMPOBJ)/crtn.o
Linking with /export/home/oracle/lib/osntab.o /export/home/oracle/lib/crtn.o /export/home/oracle/lib/libocic.a  -lsqlnet -lora -lsqlnet -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lsocket -lnsl -lm -ldl -laio -lposix4 -R /opt/SUNWcluster/lib -YP,:/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib -Qy -lc  -lcore3

Checking if your kit is complete...
Looks good
Unrecognized argument in LIBS ignored: '/export/home/oracle/lib/libocic.a'
Unrecognized argument in LIBS ignored: '/opt/SUNWcluster/lib'
Unrecognized argument in LIBS ignored: '-YP,:/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib'
Unrecognized argument in LIBS ignored: '-Qy'

Warning: By default new modules are installed into your 'site_lib'
directories. Since site_lib directories come after the normal library
directories you must delete any old DBD::Oracle files and directories from
your 'privlib' and 'archlib' directories and their auto subdirectories.

Writing Makefile for DBD::Oracle
redcomet:/export/home/ml/DBD-Oracle-0.35# make
cp Oraperl.pm ./blib/lib/Oraperl.pm
cp Oracle.pm ./blib/lib/DBD/Oracle.pm
cp oraperl.ph ./blib/lib/oraperl.ph
//bin/perl -I/usr/local/lib/perl5/sun4-solaris/5.003 -I/usr/local/lib/perl5 /usr/local/lib/perl5/ExtUtils/xsubpp  -typemap /usr/local/lib/perl5/ExtUtils/typemap Oracle.xs >Oracle.tc && mv Oracle.tc Oracle.c
gcc -c -I/export/home/oracle/rdbms/demo  -I/export/home/oracle/rdbms/demo -I/usr/local/lib/perl5/site_perl/sun4-solaris/DBI -I/usr/local/lib/perl5/sun4-solaris/5.003/DBI -I/usr/local/include -O     -DVERSION=\"0.35\"  -DXS_VERSION=\"0.35\" -fpic -I/usr/local/lib/perl5/sun4-solaris/5.003/CORE -Wall -pedantic -Wno-comment -Wtraditional Oracle.c
Oracle.c: In function `XS_DBD__Oracle__db_disconnect':
Oracle.c:144: warning: ANSI C forbids braced-groups within expressions
Oracle.c: In function `XS_DBD__Oracle__st_execute':
Oracle.c:256: warning: ANSI C forbids braced-groups within expressions
Oracle.c:265: warning: ANSI C forbids braced-groups within expressions
Oracle.c: In function `XS_DBD__Oracle__st_fetchrow':
Oracle.c:305: warning: ANSI C forbids braced-groups within expressions
Oracle.c:311: warning: ANSI C forbids braced-groups within expressions
Oracle.c: In function `XS_DBD__Oracle__st_finish':
Oracle.c:400: warning: ANSI C forbids braced-groups within expressions
Oracle.c:404: warning: ANSI C forbids braced-groups within expressions
Oracle.c: In function `boot_DBD__Oracle':
Oracle.c:444: warning: ANSI C forbids braced-groups within expressions
gcc -c -I/export/home/oracle/rdbms/demo  -I/export/home/oracle/rdbms/demo -I/usr/local/lib/perl5/site_perl/sun4-solaris/DBI -I/usr/local/lib/perl5/sun4-solaris/5.003/DBI -I/usr/local/include -O     -DVERSION=\"0.35\"  -DXS_VERSION=\"0.35\" -fpic -I/usr/local/lib/perl5/sun4-solaris/5.003/CORE -Wall -pedantic -Wno-comment -Wtraditional dbdimp.c
dbdimp.c: In function `dbd_db_FETCH':
dbdimp.c:265: warning: value computed is not used
dbdimp.c: In function `dbd_st_blob_read':
dbdimp.c:756: warning: ANSI C forbids braced-groups within expressions
dbdimp.c: In function `dbd_st_FETCH':
dbdimp.c:910: warning: value computed is not used
dbdimp.c: At top level:
dbdimp.c:126: warning: `dump_error_status' defined but not used
Running Mkbootstrap for DBD::Oracle ()
chmod 644 Oracle.bs
LD_RUN_PATH="/export/home/oracle/lib:/lib" gcc -o blib/arch/auto/DBD/Oracle/Oracle.so -G -L/usr/local/lib Oracle.o  dbdimp.o /export/home/oracle/lib/osntab.o /export/home/oracle/lib/crtn.o /export/home/oracle/lib/__fstd.o   -L/export/home/oracle/lib -lsqlnet -lora -lsqlnet -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lsocket -lnsl -lm -ldl -laio -lposix4 -lc -lcore3
chmod 755 blib/arch/auto/DBD/Oracle/Oracle.so
cp Oracle.bs ./blib/arch/auto/DBD/Oracle/Oracle.bs
chmod 644 blib/arch/auto/DBD/Oracle/Oracle.bs
Manifying ./blib/man3/DBD::Oraperl.3
redcomet:/export/home/ml/DBD-Oracle-0.35# make test
PERL_DL_NONLAZY=1 //bin/perl -I./blib/arch -I./blib/lib -I/usr/local/lib/perl5/sun4-solaris/5.003 -I/usr/local/lib/perl5 -e 'use Test::Harness qw(&runtests $verbose); $verbose=0; runtests @ARGV;' t/*.t
t/base..............install_driver(Oracle) failed: Can't load './blib/arch/auto/DBD/Oracle/Oracle.so' for module DBD::Oracle: ld.so.1: //bin/perl: fatal: relocation error: symbol not found: oerhms: referenced in ./blib/arch/auto/DBD/Oracle/Oracle.so at /usr/local/lib/perl5/DynaLoader.pm line 140.

at blib/lib/DBD/Oracle.pm line 24
       DBI::install_driver called at t/base.t line 16
FAILED tests 4-5
       Failed 2/5 tests, 60.00% okay
       Test returned status 2 (wstat 512)
Failed 1 test script, 0.00% okay. 2/5 subtests failed, 60.00% okay.
*** Error code 29
make: Fatal error: Command failed for target `test_dynamic'
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <28230-2@oink>;
         Tue, 9 Jul 1996 03:06:17 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836874903:19836:3; Tue, 09 Jul 96 02:15:03 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa19755; 9 Jul 96 2:14 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id SAA07074 for dbi-users-real; Mon, 8 Jul 1996 18:03:54 -0700
Precedence: junk
Sender: [email protected]
Received: from cse.psu.edu (claven.cse.psu.edu [130.203.3.50])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id SAA07070
         for <[email protected]>; Mon, 8 Jul 1996 18:03:53 -0700
Received: from jovial.cse.psu.edu (jovial.cse.psu.edu [130.203.10.24])
         by cse.psu.edu (8.7.5/8.7.3) with ESMTP id VAA16975;
         Mon, 8 Jul 1996 21:04:36 -0400 (EDT)
Received: from localhost (groenvel@localhost)
         by jovial.cse.psu.edu (8.7.5/8.7.3) with SMTP id VAA12362;
         Mon, 8 Jul 1996 21:03:17 -0400 (EDT)
Message-Id: <[email protected]>
X-Authentication-Warning: jovial.cse.psu.edu: groenvel owned process doing -bs
X-Mailer: exmh version 1.6.5 12/11/95
X-uri: <URL:http://www.cse.psu.edu/~groenvel/>
X-work-address: Manufacturing Technology - Systems Development Group PO Box 30
               The Applied Research Lab State College, PA 16804
X-work-phone: +1 814 863 9896 (Voice) +1 814 863 1183 (Fax)
To: Mike Linksvayer <[email protected]>
cc: [email protected]
Subject: Re: Problems with make test DBD-Oracle-0.35
In-reply-to: Your message of "Mon, 08 Jul 1996 13:32:28 PDT." <[email protected]>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 08 Jul 1996 21:03:16 -0400
From: John D Groenveld <[email protected]>

What version of Oracle are you running? Is proc.mk available on your
installation CD?
John

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <28229-3@oink>;
         Tue, 9 Jul 1996 03:06:20 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836874904:19835:4; Tue, 09 Jul 96 02:15:04 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa19763; 9 Jul 96 2:14 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id RAA07053 for dbi-users-real; Mon, 8 Jul 1996 17:54:54 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id RAA07049
         for <[email protected]>; Mon, 8 Jul 1996 17:54:52 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id af20467; 9 Jul 96 0:54 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa15393; 9 Jul 96 1:33 +0100
Received: from toad by oink with SMTP (PP) id <27786-0@oink>;
         Tue, 9 Jul 1996 01:32:09 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA00957;
         Tue, 9 Jul 1996 01:31:54 +0000
Date: Tue, 9 Jul 1996 01:31:54 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Problems with make test DBD-Oracle-0.35
Cc: [email protected], [email protected]
X-Sun-Charset: US-ASCII
content-length: 1304
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Mike Linksvayer <[email protected]>
>
> I'm also having problems with make test.  I tried the patch
> Makefile.PL patch posted by Tim on July 1, but it doesn't
> make any difference.  I know that I should be using proc.mk,
> but I don't have it available.

> Any ideas?

> Linking with /export/home/oracle/lib/osntab.o /export/home/oracle/lib/crtn.o /export/home/oracle/lib/libocic.a  -lsqlnet -lora -lsqlnet -lora -lnlsrtl3 -lc3v6 -lcore3 -lnlsrtl3 -lcore3 -lsocket -lnsl -lm -ldl -laio -lposix4 -R /opt/SUNWcluster/lib -YP,:/opt/SUNWcluster/lib:/usr/ccs/lib:/usr/lib -Qy -lc  -lcore3
>
> Checking if your kit is complete...
> Looks good
> Unrecognized argument in LIBS ignored: '/export/home/oracle/lib/libocic.a'

Is Pro*C installed? If not then install it. It's required.

The real problem (apart from Oracle changing their makefiles with the
phases of the moon) is that MakeMaker's getting indigestion when given
a full library path. Here's a Makefile.PL workaround you can try:

 # get a cut down $linkwith to pass to MakeMaker liblist
 my $linkwith_s = expand_mkvars($linkwith, 1, 1);
+ # convert "/full/path/libFOO.a" into "-L/full/path -lFOO"
+ $linkwith_s =~ s!([/\w.]+)/lib(\w+).a!-L$1 -l$2!g;

Let me know if it works ASAP and I'll drop it into the next release ASAP.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <28991-2@oink>;
         Tue, 9 Jul 1996 07:06:00 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836879746:21836:4; Tue, 09 Jul 96 03:35:46 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa21580; 9 Jul 96 3:35 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id TAA07203 for dbi-users-real; Mon, 8 Jul 1996 19:20:32 -0700
Precedence: junk
Sender: [email protected]
Received: from online.tmx.com.au (online.tmx.com.au [192.150.129.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id TAA07199
         for <[email protected]>; Mon, 8 Jul 1996 19:20:29 -0700
Received: from wurrup.bha.oz.au (daemon@localhost)
         by online.tmx.com.au (8.6.5/8.6.5) with MHSnet id MAA19432
         for [email protected]; Tue, 9 Jul 1996 12:20:25 +1000
Received: by wurrup (AIX 3.2/UCB 5.64/4.03) id AA38037;
         Tue, 9 Jul 1996 12:16:26 +1000
From: [email protected] (adb)
Message-Id: <9607090216.AA38037@wurrup>
To: [email protected]
Cc: [email protected]
Subject: Inserting LONG RAW data using DBI/DBD-Oracle
Date: Tue, 09 Jul 96 12:16:25 +1000


I'm having problems inserting Oracle LONG RAW fields using the DBI.  I
noticed reports of success using the Oraperl interface on the mailing
list, but I'm constrained to use DBI.  When trying to insert a string
into a long raw column, I get the following Oracle error message from
the DBI `execute' function:

   ORA-01461: can bind a LONG value only for insert into a LONG column
              (DBD: oexec error)

I've tried the following variations with no success:

 1. $sth = $db->prepare ("INSERT INTO $table ($colname) VALUES (?)")
    (a) $sth->execute ($string)             # -- string contains raw data
    (b) $sth->execute ($hexidecimal_string) # -- hexidecimal for raw data

 2. $sth = $db->prepare ("INSERT INTO $table ($colname) VALUES (HEXTORAW(?))")
    $sth->execute ($hexidecimal_string)     # -- hexidecimal for raw data

From reading the interpretation of this error from the Oracle messages
and codes manual, it seems that either DBD-Oracle is forcing the type
to 'LONG' when binding the variable, or that Oracle is getting
confused about LONG and LONG RAW columns.  From a search through the
'dbdimp.c' source, I can't find any obvious problems.

Note that example 1(a) above works fine for LONG columns (and our
interim solution is to change the type of the columns to LONG).

Any comments or suggestions appreciated.  I'm happy to test code
changes if access to Oracle is a problem.

Ta.

AndyB
------------------------------------------------------------------------

Versions details:

 Perl:       5.002
 DBI:        0.70
 DBD-Oracle: 0.35

From `perl -V`:

Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=irix, osver=5, archname=IP22-irix
   uname='irix kingfisher 5.3 11091812 ip22 mips '
   hint=recommended, useposix=true
 Compiler:
   cc='cc', optimize='-O', gccversion=
   cppflags='-D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -I/usr/local/include -DLANGUAGE_C'
   ccflags ='-D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -Olimit 3000 -I/usr/local/include -DLANGUAGE_C'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=1, d_casti32=define, d_castneg=
   intsize=4, alignbytes=8, usemymalloc=n, randbits=15
 Linker and Libraries:
   ld='ld', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /usr/lib /lib
   libs=-lmalloc -lsun -lm -lc -lcrypt -lbsd -lPW
   libc=/usr/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags=' ', lddlflags='-shared -L/usr/local/lib'

@INC: /usr/local/lib/perl5/IP22-irix/5.002 /usr/local/lib/perl5 /usr/local/lib/perl5/site_perl/IP22-irix /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/IP22-irix .
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <01664-5@oink>;
         Tue, 9 Jul 1996 17:02:24 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836902784:11284:3; Tue, 09 Jul 96 09:59:44 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa10430; 9 Jul 96 9:59 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id BAA07537 for dbi-users-real; Tue, 9 Jul 1996 01:48:46 -0700
Precedence: junk
Sender: [email protected]
Received: from online.tmx.com.au (online.tmx.com.au [192.150.129.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id BAA07533
         for <[email protected]>; Tue, 9 Jul 1996 01:48:40 -0700
Received: from wurrup.bha.oz.au (daemon@localhost)
         by online.tmx.com.au (8.6.5/8.6.5) with MHSnet id SAA27382
         for [email protected]; Tue, 9 Jul 1996 18:48:22 +1000
Received: by wurrup (AIX 3.2/UCB 5.64/4.03) id AA44932;
         Tue, 9 Jul 1996 17:17:05 +1000
From: [email protected] (adb)
Message-Id: <9607090717.AA44932@wurrup>
To: [email protected]
Subject: Re: Inserting LONG RAW data using DBI/DBD-Oracle
Date: Tue, 09 Jul 96 17:17:05 +1000


I wrote:

> I'm having problems inserting Oracle LONG RAW fields using the DBI.  I
> noticed reports of success using the Oraperl interface on the mailing
> list, but I'm constrained to use DBI.  When trying to insert a string
> into a long raw column, I get the following Oracle error message from
> the DBI `execute' function:
>
>     ORA-01461: can bind a LONG value only for insert into a LONG column
>                (DBD: oexec error)
>
> I've tried the following variations with no success:
>
>   1. $sth = $db->prepare ("INSERT INTO $table ($colname) VALUES (?)")
>      (a) $sth->execute ($string)             # -- string contains raw data
>      (b) $sth->execute ($hexidecimal_string) # -- hexidecimal for raw data
>
>   2. $sth = $db->prepare ("INSERT INTO $table ($colname) VALUES (HEXTORAW(?))")
>      $sth->execute ($hexidecimal_string)     # -- hexidecimal for raw data


I've found a work-around.  The work around is, instead of relying on
`execute' to bind parameters (which gives them a VARCHAR2 type by
default), use the `bind_param' function to bind them manually with
explicit type attributes, e.g.

 $attrib{'ora_type'} = 24;                # -- 24 is LONG RAW type code
 $sth->bind_param(1, $string, \%attrib);  # -- must pass reference to attrib
 $sth->execute;                           # -- no parameters to execute

In order for this to work, a bug in the DBD-Oracle `dbdimp.c' code had
to be fixed.  Context diffs are attached.  From reading the code, it
seems that the `bind_param' function with non-null attributes is
untested (apart from my minimal testing).  Use it with caution ...

Ciao,

AndyB

------------------------------------------------------------------------
RCS file: RCS/dbdimp.c,v
retrieving revision 1.1
diff -c -r1.1 dbdimp.c
*** dbdimp.c    1996/07/08 05:32:23     1.1
--- dbdimp.c    1996/07/09 06:00:05
***************
*** 459,465 ****
       /* Setup / Clear attributes as defined by attribs.              */
       /* If attribs is EMPTY then reset attribs to default.           */
       ;       /* XXX */
!       if ( (svp=hv_fetch((HV*)SvRV(attribs), "ora_type",0, 0)) == NULL) {
           if (!dbtype_is_string(SvIV(*svp)))  /* mean but safe        */
               croak("bind_param %s ora_type %d not a simple string type",
                       name, (int)SvIV(*svp));
--- 459,465 ----
       /* Setup / Clear attributes as defined by attribs.              */
       /* If attribs is EMPTY then reset attribs to default.           */
       ;       /* XXX */
!       if ( (svp=hv_fetch((HV*)SvRV(attribs), "ora_type",8, 1)) != NULL) {
           if (!dbtype_is_string(SvIV(*svp)))  /* mean but safe        */
               croak("bind_param %s ora_type %d not a simple string type",
                       name, (int)SvIV(*svp));
------------------------------------------------------------------------
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <01664-19@oink>;
         Tue, 9 Jul 1996 17:04:37 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836920927:18482:1; Tue, 09 Jul 96 15:02:07 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa17689; 9 Jul 96 15:01 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id GAA07794 for dbi-users-real; Tue, 9 Jul 1996 06:49:27 -0700
Precedence: junk
Sender: [email protected]
Received: from dealer.cards.com ([128.126.162.2])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id GAA07790
         for <[email protected]>; Tue, 9 Jul 1996 06:49:23 -0700
Received: from cruel.cards.com by dealer.cards.com (4.1/mls/3.2) id AA17901;
         Tue, 9 Jul 96 09:49:38 EDT
Date: Tue, 9 Jul 96 09:49:37 EDT
From: [email protected] (Prasad Bammidi)
Message-Id: <[email protected]>
To: [email protected]
Subject: Setting ENV variables

Hi,

  We recently upgraded from 7.0.16 to 7.2.3 and seemd to
  be having problem with slow connection (upto 4 min).

  The problem is the same with oraperl 2.4 and Perl5 w/oraperl
  emulation (DBD-Oralce).  If we rebuild oraperl with the 7.0.16
  libs - it works fine with the new oracle - but we hate to keep
  the old libs around.

  Some people advised me that that The problem with the "slow connects"
  was traced to the use of a putenv() call in our application code;
  apparently part of Oracle's code was getting confused if the environment
  section of a process was getting changed at all. The work around
  suggested was setting the ENV variables seperately. This can be
  done by using a seperate shell script (setting env) and then running
  the ora-perl scripts. It does work but the problem we have is, we are
  using the ora-perl scripts to access the Oracle database thru the WEB.
  and running the shell scripts to set env variables every time an ora-perl
  script is called to be executed. Is there simple way to set ENV variables
  using the shell scripts once ?


Thanks in advance.

Regards,

Prasad Bammidi
Database Administrator
Fairmont, WV.
[email protected]
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <01664-26@oink>;
         Tue, 9 Jul 1996 17:05:43 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836926620:17152:3; Tue, 09 Jul 96 16:37:00 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa16796; 9 Jul 96 16:36 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa17492; 9 Jul 96 16:35 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id IAA08016 for dbi-users-real; Tue, 9 Jul 1996 08:22:31 -0700
Precedence: junk
Sender: [email protected]
Received: from river.it.gvsu.edu (river.it.gvsu.edu [148.61.1.16])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id IAA08012
         for <[email protected]>; Tue, 9 Jul 1996 08:22:28 -0700
Message-Id: <[email protected]>
Received: by river.it.gvsu.edu (1.37.109.16/16.2) id AA160115820;
         Tue, 9 Jul 1996 11:23:40 -0400
From: Joel Meulenberg <[email protected]>
Subject: Bug in 'Makefile.PL' for DBD::Oracle Version 0.35
To: [email protected]
Date: Tue, 9 Jul 96 11:23:40 EDT
Cc: [email protected]
Mailer: Elm [revision: 70.85.2.1]

I think I've found a bug in 'Makefile.PL' for DBD::Oracle version 0.35.
I was compiling it for Perl 5.003 under HP-UX version 10.01.
You can see from the diff output below that the 'Makefile.PL' that comes
with DBD::Oracle 0.35 refers to '$Config{osver}', but should refer to
'$Config{osvers}'.  The result is that 'Makefile.PL' thinks that the
OS version is undefined.  So the check to see if the version of HP-UX
is less than 10 will always evaluate to true.  This causes the static
linking of the Oracle lib's into the Perl interpreter even though
dynamic linking would work.  Here is the diff output:

# diff Makefile.PL Makefile.PL.broke
114,115c114,115
< if ($Config{osname} eq 'hpux' and $Config{osvers} < 10) {
<     print "Warning: forced to build static not dynamic on $Config{osname} $Config{osvers}\n";
---
> if ($Config{osname} eq 'hpux' and $Config{osver} < 10) {
>     print "Warning: forced to build static not dynamic on $Config{osname} $Config{osver}\n";

Hope this helps...

Joel Meulenberg
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <03846-3@oink>;
         Tue, 9 Jul 1996 22:49:22 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836942218:19469:1; Tue, 09 Jul 96 20:56:58 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa18647; 9 Jul 96 20:55 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA08608 for dbi-users-real; Tue, 9 Jul 1996 12:49:00 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id MAA08604
         for <[email protected]>; Tue, 9 Jul 1996 12:48:56 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id au09919; 9 Jul 96 18:58 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa13473; 9 Jul 96 19:56 +0100
Received: from toad by oink with SMTP (PP) id <03409-0@oink>;
         Tue, 9 Jul 1996 19:55:58 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA09665;
         Tue, 9 Jul 1996 19:55:51 +0000
Date: Tue, 9 Jul 1996 19:55:51 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: Re: Bug in 'Makefile.PL' for DBD::Oracle Version 0.35
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 344
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Joel Meulenberg <[email protected]>
>
> I think I've found a bug in 'Makefile.PL' for DBD::Oracle version 0.35.
> You can see from the diff output below that the 'Makefile.PL' that comes
> with DBD::Oracle 0.35 refers to '$Config{osver}', but should refer to
> '$Config{osvers}'.

Thanks. Fixed in next release (soon).

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <03850-6@oink>;
         Tue, 9 Jul 1996 22:50:13 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836940386:08061:2; Tue, 09 Jul 96 20:26:26 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa07553; 9 Jul 96 20:25 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id MAA08536 for dbi-users-real; Tue, 9 Jul 1996 12:14:58 -0700
Precedence: junk
Sender: [email protected]
Received: from cs.utah.edu (cs.utah.edu [128.110.4.21])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id MAA08532
         for <[email protected]>; Tue, 9 Jul 1996 12:14:52 -0700
Received: from lal.cs.utah.edu by cs.utah.edu (8.6.12/utah-2.21-cs) id NAA10605;
         Tue, 9 Jul 1996 13:12:39 -0600
Received: by lal.cs.utah.edu (8.6.12/utah-2.15sun-leaf) id NAA23450;
         Tue, 9 Jul 1996 13:12:37 -0600
Date: Tue, 9 Jul 1996 13:12:36 -0600 (MDT)
From: Jonathan Olson <[email protected]>
To: [email protected]
cc: [email protected]
Subject: DBD:Make Problems
Message-ID: <[email protected]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


I am having DBD:Oracle-0.35 make problems - PLEASE HELP!

1. A complete log of a all steps of the build :

# perl Makefile.PL

Configuring DBD::Oracle ...

>>>     Remember to actually *READ* the README file!

Using Oracle in /usr/oracle
Using /usr/oracle/proc/lib/proc.mk (version 1.9)
System: perl5.002 sunos orchid 5.5 generic sun4m sparc
sunw,sparcstation-20
Compiler: cc -O -I/usr/local/include
Linker:   /usr/ucb/ld
Oracle proc.mk would have used these values but we override them:
 CC:       /usr/5bin/cc
 CFLAGS:   -I. -O
 LDFLAGS:  -Bdynamic -L$(LIBHOME)
Linking with $(LIBHOME)/osntab.o -locic  -lsqlnet -lora -lsqlnet -lora
-lnlsrtl -lcv6 -lcore -lnlsrtl -lcv6 -lcore -lm

Checking if your kit is complete...
Looks good

Warning: By default new modules are installed into your 'site_lib'
directories. Since site_lib directories come after the normal library
directories you must delete any old DBD::Oracle files and directories from
your 'privlib' and 'archlib' directories and their auto subdirectories.

Writing Makefile for DBD::Oracle

# make
umask 0 && cp Oraperl.pm blib/lib/Oraperl.pm
umask 0 && cp oraperl.ph blib/lib/oraperl.ph
umask 0 && cp Oracle.pm blib/lib/DBD/Oracle.pm
/bin/perl -I/usr/local/lib/perl5/sun4-solaris/5.002
-I/usr/local/lib/perl5 /usr/local/lib/perl5/ExtUtils/xsubpp  -typemap
/usr/local/lib/perl5/ExtUtils/typemap Oracle.xs >Oracle.tc && mv
Oracle.tc Oracle.c
cc -c -I/usr/oracle/rdbms/demo
-I/usr/local/lib/perl5/site_perl/sun4-solaris/DBI
-I/usr/local/lib/perl5/sun4-solaris/5.002/DBI -I/usr/local/include -O
-DVERSION=\"0.35\" -DXS_VERSION=\"0.35\" -Kpic
-I/usr/local/lib/perl5/sun4-solaris/5.002/CORE  Oracle.c
"/usr/oracle/rdbms/demo/oratypes.h", line 168: warning: signed is a
keyword in ANSI C
"/usr/oracle/rdbms/demo/oratypes.h", line 168: undefined or not a type:
signed
"/usr/oracle/rdbms/demo/oratypes.h", line 168: warning: "typedef" valid
only for function declaration
"/usr/oracle/rdbms/demo/oratypes.h", line 168: parameter not in
identifier list: sbig_ora
"/usr/oracle/rdbms/demo/oratypes.h", line 208: warning: only "register"
valid as formal parameter storage class
"/usr/oracle/rdbms/demo/ocidfn.h", line 36: warning: dubious struct
declaration; use tag only: cda_head
"/usr/oracle/rdbms/demo/ocidfn.h", line 47: warning: dubious struct
declaration; use tag only: <unnamed>
"/usr/oracle/rdbms/demo/ocidfn.h", line 48: warning: dubious struct
declaration; use tag only: <unnamed>
"/usr/oracle/rdbms/demo/ocidfn.h", line 61: warning: dubious struct
declaration; use tag only: cda_def
"/usr/oracle/rdbms/demo/ocidfn.h", line 72: warning: dubious struct
declaration; use tag only: <unnamed>
"/usr/oracle/rdbms/demo/ocidfn.h", line 73: warning: dubious struct
declaration; use tag only: <unnamed>
"/usr/oracle/rdbms/demo/ocidfn.h", line 86: warning: only "register"
valid as formal parameter storage class
"/usr/oracle/rdbms/demo/ocidfn.h", line 90: warning: only "register"
valid as formal parameter storage class
"/usr/oracle/rdbms/demo/ociapr.h", line 32: parameter not in identifier
list: obndra
"/usr/oracle/rdbms/demo/ociapr.h", line 35: parameter not in identifier
list: obndrn
"/usr/oracle/rdbms/demo/ociapr.h", line 38: parameter not in identifier
list: obndrv
"/usr/oracle/rdbms/demo/ociapr.h", line 39: parameter not in identifier
list: obreak
"/usr/oracle/rdbms/demo/ociapr.h", line 40: parameter not in identifier
list: ocan
"/usr/oracle/rdbms/demo/ociapr.h", line 41: parameter not in identifier
list: oclose
"/usr/oracle/rdbms/demo/ociapr.h", line 42: parameter not in identifier
list: ocof
"/usr/oracle/rdbms/demo/ociapr.h", line 43: parameter not in identifier
list: ocom
"/usr/oracle/rdbms/demo/ociapr.h", line 44: parameter not in identifier
list: ocon
"/usr/oracle/rdbms/demo/ociapr.h", line 47: parameter not in identifier
list: odefin
"/usr/oracle/rdbms/demo/ociapr.h", line 53: parameter not in identifier
list: odessp
"/usr/oracle/rdbms/demo/ociapr.h", line 56: parameter not in identifier
list: odsc
"/usr/oracle/rdbms/demo/ociapr.h", line 59: parameter not in identifier
list: odescr
"/usr/oracle/rdbms/demo/ociapr.h", line 61: parameter not in identifier
list: oerhms
"/usr/oracle/rdbms/demo/ociapr.h", line 62: parameter not in identifier
list: oermsg
"/usr/oracle/rdbms/demo/ociapr.h", line 63: parameter not in identifier
list: oexec
"/usr/oracle/rdbms/demo/ociapr.h", line 65: parameter not in identifier
list: oexfet
"/usr/oracle/rdbms/demo/ociapr.h", line 66: parameter not in identifier
list: oexn
"/usr/oracle/rdbms/demo/ociapr.h", line 67: parameter not in identifier
list: ofen
"/usr/oracle/rdbms/demo/ociapr.h", line 68: parameter not in identifier
list: ofetch
"/usr/oracle/rdbms/demo/ociapr.h", line 70: parameter not in identifier
list: oflng
"/usr/oracle/rdbms/demo/ociapr.h", line 71: parameter not in identifier
list: ologof
"/usr/oracle/rdbms/demo/ociapr.h", line 73: parameter not in identifier
list: olon
"/usr/oracle/rdbms/demo/ociapr.h", line 76: parameter not in identifier
list: oopen
"/usr/oracle/rdbms/demo/ociapr.h", line 77: parameter not in identifier
list: oopt
"/usr/oracle/rdbms/demo/ociapr.h", line 79: parameter not in identifier
list: oname
"/usr/oracle/rdbms/demo/ociapr.h", line 81: parameter not in identifier
list: oparse
"/usr/oracle/rdbms/demo/ociapr.h", line 83: parameter not in identifier
list: orlon
"/usr/oracle/rdbms/demo/ociapr.h", line 84: parameter not in identifier
list: orol
"/usr/oracle/rdbms/demo/ociapr.h", line 85: parameter not in identifier
list: osql3
"/usr/oracle/rdbms/demo/ociapr.h", line 86: parameter not in identifier
list: sqlld2
"/usr/oracle/rdbms/demo/ociapr.h", line 87: parameter not in identifier
list: sqllda
"./dbdimp.h", line 19: warning: only "register" valid as formal parameter
storage class
"./dbdimp.h", line 19: warning: dubious tag declaration: struct imp_fbh_st
"./dbdimp.h", line 23: warning: useless declaration
"./dbdimp.h", line 31: warning: useless declaration
"./dbdimp.h", line 51: warning: useless declaration
"./dbdimp.h", line 77: warning: useless declaration
"./dbdimp.h", line 80: warning: only "register" valid as formal parameter
storage class
"./dbdimp.h", line 80: warning: dubious tag declaration: struct phs_st
"./dbdimp.h", line 86: warning: useless declaration
"./dbdimp.h", line 89: parameter not in identifier list: ora_error
"./dbdimp.h", line 90: parameter not in identifier list: fbh_dump
"./dbdimp.h", line 92: parameter not in identifier list: dbd_init
"./dbdimp.h", line 93: parameter not in identifier list: dbd_preparse
"./dbdimp.h", line 94: parameter not in identifier list: dbd_describe
"./dbdimp.h", line 96: parameter not in identifier list: dbd_st_blob_read
"./Oracle.h", line 45: parameter not in identifier list: dbd_init
"./Oracle.h", line 47: parameter not in identifier list: dbd_db_login
"./Oracle.h", line 48: parameter not in identifier list: dbd_db_do
"./Oracle.h", line 49: parameter not in identifier list: dbd_db_commit
"./Oracle.h", line 50: parameter not in identifier list: dbd_db_rollback
"./Oracle.h", line 51: parameter not in identifier list: dbd_db_disconnect
"./Oracle.h", line 52: parameter not in identifier list: dbd_db_destroy
"./Oracle.h", line 53: parameter not in identifier list: dbd_db_STORE
"./Oracle.h", line 54: parameter not in identifier list: dbd_db_FETCH
"./Oracle.h", line 57: parameter not in identifier list: dbd_st_prepare
"./Oracle.h", line 58: parameter not in identifier list: dbd_st_rows
"./Oracle.h", line 59: parameter not in identifier list: dbd_bind_ph
"./Oracle.h", line 60: parameter not in identifier list: dbd_st_execute
"./Oracle.h", line 61: parameter not in identifier list: dbd_st_fetch
"./Oracle.h", line 62: parameter not in identifier list: dbd_st_finish
"./Oracle.h", line 63: parameter not in identifier list: dbd_st_destroy
"./Oracle.h", line 65: parameter not in identifier list: dbd_st_readblob
"./Oracle.h", line 66: parameter not in identifier list: dbd_st_STORE
"./Oracle.h", line 67: parameter not in identifier list: dbd_st_FETCH
"Oracle.c", line 25: only "register" valid as formal parameter storage class
"Oracle.c", line 25: parameter not in identifier list: dbis
"Oracle.c", line 29: parameter not in identifier list: XS_DBD__Oracle_errstr
"Oracle.c", line 29: syntax error before or at: {
"Oracle.c", line 30: parameter not in identifier list: sp
"Oracle.c", line 30: cannot initialize parameter: sp
"Oracle.c", line 30: warning: statement not reached
"Oracle.c", line 30: parameter not in identifier list: mark
"Oracle.c", line 30: cannot initialize parameter: mark
"Oracle.c", line 30: parameter not in identifier list: ax
"Oracle.c", line 30: cannot initialize parameter: ax
"Oracle.c", line 30: parameter not in identifier list: items
"Oracle.c", line 30: cannot initialize parameter: items
"Oracle.c", line 31: syntax error before or at: if
"Oracle.c", line 32: parameter not in identifier list: croak
"Oracle.c", line 34: parameter not in identifier list: h
"Oracle.c", line 34: cannot initialize parameter: h
"Oracle.c", line 36: parameter not in identifier list: imp_xxh
"Oracle.c", line 36: cannot initialize parameter: imp_xxh
"Oracle.c", line 37: integral constant expression expected
"Oracle.c", line 37: parameter not in identifier list: stack_base
"Oracle.c", line 37: cannot initialize parameter: stack_base
"Oracle.c", line 37: warning: initialization type mismatch
"Oracle.c", line 37: non-constant initializer: op "SCONV"
"Oracle.c", line 38: syntax error before or at: }
"Oracle.c", line 39: parameter not in identifier list: stack_sp
"Oracle.c", line 39: cannot initialize parameter: stack_sp
"Oracle.c", line 39: warning: initialization type mismatch
"Oracle.c", line 39: non-constant initializer: op "SCONV"
"Oracle.c", line 39: syntax error before or at: return
"Oracle.c", line 39: warning: syntax error:  empty declaration
"Oracle.c", line 43: parameter not in identifier list:
XS_DBD__Oracle__dr_disconnect_all
"Oracle.c", line 44: parameter not in identifier list: sp
"Oracle.c", line 44: cannot initialize parameter: sp
"Oracle.c", line 44: warning: improper pointer/integer combination: op "="
"Oracle.c", line 44: parameter not in identifier list: mark
"Oracle.c", line 44: cannot initialize parameter: mark
"Oracle.c", line 44: warning: assignment type mismatch:
       pointer to pointer to struct sv {pointer to void sv_any, ulong
sv_refcnt, ulong sv_flags} "=" pointer to int
"Oracle.c", line 44: parameter not in identifier list: ax
"Oracle.c", line 44: cannot initialize parameter: ax
"Oracle.c", line 44: warning: improper pointer subtraction
"Oracle.c", line 44: parameter not in identifier list: items
"Oracle.c", line 44: cannot initialize parameter: items
"Oracle.c", line 48: parameter not in identifier list: drh
"Oracle.c", line 48: cannot initialize parameter: drh
"Oracle.c", line 48: warning: improper pointer/integer combination: op "="
"Oracle.c", line 50: parameter not in identifier list: imp_drh
"Oracle.c", line 50: cannot initialize parameter: imp_drh
"Oracle.c", line 56: warning: improper pointer/integer combination: op "="
"Oracle.c", line 58: warning: improper pointer/integer combination: op "="
"Oracle.c", line 60: warning: improper pointer/integer combination: op "="
"Oracle.c", line 61: cannot recover from previous errors
*** Error code 10
make: Fatal error: Command failed for target `Oracle.o'


2. Perl Version 5.002 Oracle7 Server Release 7.2.2.4.0 (but I don't think
this is the problem)

3. perl -V

Summary of my perl5 (5.0 patchlevel 2) configuration:
 Platform:
   osname=solaris, osver=2.5, archname=sun4-solaris
   uname='sunos orchid 5.5 generic sun4m sparc sunw,sparcstation-20 '
   hint=recommended, useposix=true
 Compiler:
   cc='cc', optimize='-O', gccversion=
   cppflags='-I/usr/local/include'
   ccflags ='-I/usr/local/include'
   stdchar='unsigned char', d_stdstdio=define, usevfork=false
   voidflags=15, castflags=0, d_casti32=define, d_castneg=define
   intsize=4, alignbytes=8, usemymalloc=y, randbits=15
 Linker and Libraries:
   ld='cc', ldflags =' -L/usr/local/lib'
   libpth=/usr/local/lib /lib /usr/lib /usr/ccs/lib
   libs=-lsocket -lnsl -ldl -lm -lc -lcrypt
   libc=/lib/libc.so, so=so
 Dynamic Linking:
   dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=, ccdlflags=' '
   cccdlflags='-Kpic', lddlflags='-G -L/usr/local/lib'

@INC: /usr/local/lib/perl5/sun4-solaris/5.002 /usr/local/lib/perl5
/usr/local/lib/perl5/site_perl/sun4-solaris
/usr/local/lib/perl5/site_perl .

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <03850-15@oink>;
         Tue, 9 Jul 1996 22:51:09 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836944996:26767:3; Tue, 09 Jul 96 21:43:16 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa25925; 9 Jul 96 21:42 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id LAA08356 for dbi-users-real; Tue, 9 Jul 1996 11:00:33 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id LAA08352
         for <[email protected]>; Tue, 9 Jul 1996 11:00:29 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id aj27525; 9 Jul 96 17:58 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id ad07040; 9 Jul 96 17:02 +0100
Received: from toad by oink with SMTP (PP) id <00637-0@oink>;
         Tue, 9 Jul 1996 12:13:03 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA02725;
         Tue, 9 Jul 1996 12:12:57 +0000
Date: Tue, 9 Jul 1996 12:12:57 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Inserting LONG RAW data using DBI/DBD-Oracle
Cc: [email protected]
X-Sun-Charset: US-ASCII
content-length: 1726
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (adb)
>
> I'm having problems inserting Oracle LONG RAW fields using the DBI.  I
> noticed reports of success using the Oraperl interface on the mailing
> list, but I'm constrained to use DBI.  When trying to insert a string
> into a long raw column, I get the following Oracle error message from
> the DBI `execute' function:
>
>     ORA-01461: can bind a LONG value only for insert into a LONG column
>                (DBD: oexec error)
>
> I've tried the following variations with no success:
>
>   1. $sth = $db->prepare ("INSERT INTO $table ($colname) VALUES (?)")
>      (a) $sth->execute ($string)             # -- string contains raw data
>      (b) $sth->execute ($hexidecimal_string) # -- hexidecimal for raw data
>
>   2. $sth = $db->prepare ("INSERT INTO $table ($colname) VALUES (HEXTORAW(?))")
>      $sth->execute ($hexidecimal_string)     # -- hexidecimal for raw data
>
> >From reading the interpretation of this error from the Oracle messages
> and codes manual, it seems that either DBD-Oracle is forcing the type
> to 'LONG' when binding the variable, or that Oracle is getting
> confused about LONG and LONG RAW columns.  From a search through the
> 'dbdimp.c' source, I can't find any obvious problems.
>
> Note that example 1(a) above works fine for LONG columns (and our
> interim solution is to change the type of the columns to LONG).

I've never had cause to use LONGs so I don't know for sure but something
like this might work:

       $sth = $db->prepare ("INSERT INTO $table ($colname) VALUES (?)")

       $sth->bind_param(1, $hexidecimal_string, { ora_type => 24 });

       $sth->execute();

Hopefully others who have actually played with longs can offer some
experience.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <03850-29@oink>;
         Tue, 9 Jul 1996 22:52:09 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836948892:08735:2; Tue, 09 Jul 96 22:48:12 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa08161; 9 Jul 96 22:47 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA08754 for dbi-users-real; Tue, 9 Jul 1996 14:38:26 -0700
Precedence: junk
Sender: [email protected]
Received: from iago.qualcomm.com (iago.qualcomm.com [129.46.50.28])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id OAA08750
         for <[email protected]>; Tue, 9 Jul 1996 14:38:25 -0700
Received: (from ckoder@localhost) by iago.qualcomm.com (8.7.5/0.7/8.7.2/1.9)
         id OAA11830 for [email protected];
         Tue, 9 Jul 1996 14:37:52 -0700 (PDT)
Date: Tue, 9 Jul 1996 14:37:52 -0700 (PDT)
From: Corinne Koder <[email protected]>
Message-Id: <[email protected]>
To: [email protected]
Subject: $cursor->fetchrow Fetches Nothing

Hello fellow dbi-users...

I am having a difficult time getting select statements out
of my Informix database using the cursor methods.  Any wise
words of wisdom would be appreciated!!

================
My code
================
$cursor = $dbh->prepare( 'select * from cori' );
$cursor->execute;
$cursor->debug(2);   # see below for results of debug
while ($row = $cursor->fetchrow) {
       print STDOUT "fetched row $row -- \n";
}
$cursor->finish;
$dbh->disconnect;
=================

I am using DBI 0.69, DBD-Informix, Perl 5.002.  Everything
connects, adds, and deletes fine, but 'select' does not
return any rows even though they exist!  Here is the debugging
result:

=================
$cursor->debug(2)
=================
Prepared cursor DBI::st=HASH(0x151cd4)
Cursor executed.
   DBI::st=HASH(0x151fec) debug level set to 2
   -> fetchrow for DBD::Informix::st (DBI::st=HASH(0x151cd4)~0x151fec)
   <- fetchrow= '................................................................'
fetched row  --
   -> fetchrow for DBD::Informix::st (DBI::st=HASH(0x151cd4)~0x151fec)
   <- fetchrow= '................................................................'
fetched row  --
   -> fetchrow for DBD::Informix::st (DBI::st=HASH(0x151cd4)~0x151fec)
   <- fetchrow= '................................................................'
fetched row  --
   -> fetchrow for DBD::Informix::st (DBI::st=HASH(0x151cd4)~0x151fec)
   <- fetchrow= '................................................................'
fetched row  --
   -> fetchrow for DBD::Informix::st (DBI::st=HASH(0x151cd4)~0x151fec)
   <- fetchrow= '................................................................'
fetched row  --
   -> fetchrow for DBD::Informix::st (DBI::st=HASH(0x151cd4)~0x151fec)
   <- fetchrow= undef
   -> finish for DBD::Informix::st (DBI::st=HASH(0x151cd4)~0x151fec)
   <- finish= 1
===================

What are all these dots?  The strange thing is, it knows that I have five rows in
the table.  But it won't read them for me!  Also, as a side note:

disconnect(DBI::db=HASH(0x151cd4)) invalidates 1 associated cursor(s) at test2.pl line 36.

Has anybody had a problem not getting 'select' statements to work?
Thank you and have a nice day...

Corinne Koder
[email protected]

---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <03912-0@oink>;
         Tue, 9 Jul 1996 22:54:28 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836949112:11077:0; Tue, 09 Jul 96 22:51:52 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa10409; 9 Jul 96 22:50 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa15146; 9 Jul 96 22:49 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA08764 for dbi-users-real; Tue, 9 Jul 1996 14:42:28 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id OAA08760
         for <[email protected]>; Tue, 9 Jul 1996 14:42:25 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id WAA15860; Tue, 9 Jul 1996 22:34:11 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: $cursor->fetchrow Fetches Nothing
To: Corinne Koder <[email protected]>
Date: Tue, 9 Jul 1996 22:34:10 +0100 (BST)
Cc: [email protected]
In-Reply-To: <[email protected]> from "Corinne Koder" at Jul 9, 96 02:38:39 pm
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1268


> I am having a difficult time getting select statements out
> of my Informix database using the cursor methods.  Any wise
> words of wisdom would be appreciated!!
>
> ================
> My code
> ================
> $cursor = $dbh->prepare( 'select * from cori' );
> $cursor->execute;
> $cursor->debug(2);   # see below for results of debug
> while ($row = $cursor->fetchrow) {

        ^^^^ should be ( $row ). fetchrow returns a LIST not a scalar.

>       print STDOUT "fetched row $row -- \n";
> }
> $cursor->finish;
> $dbh->disconnect;
> =================
>
> I am using DBI 0.69, DBD-Informix, Perl 5.002.  Everything
> connects, adds, and deletes fine, but 'select' does not
> return any rows even though they exist!  Here is the debugging
> result:

You should be checking return flags from each call as well!

> disconnect(DBI::db=HASH(0x151cd4)) invalidates 1 associated cursor(s) at test2.pl line 36.

Add:

       undef $cursor;

after the finish call.

> Has anybody had a problem not getting 'select' statements to work?
>
> Corinne Koder
> [email protected]

A.

--
Alligator Descartes     | "You fuckoffeh?" "Eh?" "Coffee? You for coffee?"
Anathematician          |     "Ohhh! Dearie me!"
[email protected]  |         -- The High Life
http://www.hermetica.com|
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <03912-2@oink>;
         Tue, 9 Jul 1996 22:54:35 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836948971:09510:1; Tue, 09 Jul 96 22:49:31 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa09274; 9 Jul 96 22:49 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA08764 for dbi-users-real; Tue, 9 Jul 1996 14:42:28 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id OAA08760
         for <[email protected]>; Tue, 9 Jul 1996 14:42:25 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id WAA15860; Tue, 9 Jul 1996 22:34:11 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: $cursor->fetchrow Fetches Nothing
To: [email protected] (Corinne Koder)
Date: Tue, 9 Jul 1996 22:34:10 +0100 (BST)
Cc: [email protected]
In-Reply-To: <[email protected]> from "Corinne Koder" at Jul 9, 96 02:38:39 pm
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1268


> I am having a difficult time getting select statements out
> of my Informix database using the cursor methods.  Any wise
> words of wisdom would be appreciated!!
>
> ================
> My code
> ================
> $cursor = $dbh->prepare( 'select * from cori' );
> $cursor->execute;
> $cursor->debug(2);   # see below for results of debug
> while ($row = $cursor->fetchrow) {

        ^^^^ should be ( $row ). fetchrow returns a LIST not a scalar.

>       print STDOUT "fetched row $row -- \n";
> }
> $cursor->finish;
> $dbh->disconnect;
> =================
>
> I am using DBI 0.69, DBD-Informix, Perl 5.002.  Everything
> connects, adds, and deletes fine, but 'select' does not
> return any rows even though they exist!  Here is the debugging
> result:

You should be checking return flags from each call as well!

> disconnect(DBI::db=HASH(0x151cd4)) invalidates 1 associated cursor(s) at test2.pl line 36.

Add:

       undef $cursor;

after the finish call.

> Has anybody had a problem not getting 'select' statements to work?
>
> Corinne Koder
> [email protected]

A.

--
Alligator Descartes     | "You fuckoffeh?" "Eh?" "Coffee? You for coffee?"
Anathematician          |     "Ohhh! Dearie me!"
[email protected]  |         -- The High Life
http://www.hermetica.com|
---+++---
Received: from punt2.demon.co.uk by oink with SMTP (PP) id <05376-0@oink>;
         Wed, 10 Jul 1996 02:59:45 +0100
Received: from punt-2.mail.demon.net by mailstore for [email protected]
         id 836950793:07148:1; Tue, 09 Jul 96 23:19:53 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-2.mail.demon.net
         id aa06484; 9 Jul 96 23:19 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id PAA08852 for dbi-users-real; Tue, 9 Jul 1996 15:11:03 -0700
Precedence: junk
Sender: [email protected]
Received: from fruitbat.mcqueen.com (pillshovel.mcqueen.com [194.72.106.246])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id PAA08848
         for <[email protected]>; Tue, 9 Jul 1996 15:10:57 -0700
Received: (from descarte@localhost) by fruitbat.mcqueen.com (8.6.12/8.6.12)
         id XAA16159; Tue, 9 Jul 1996 23:03:01 +0100
From: Alligator Descartes <[email protected]>
Message-Id: <[email protected]>
Subject: Re: $cursor->fetchrow Fetches Nothing
To: [email protected] (Corinne Koder)
Date: Tue, 9 Jul 1996 23:03:01 +0100 (BST)
Cc: [email protected]
In-Reply-To: <[email protected]> from "Corinne Koder" at Jul 9, 96 03:01:47 pm
X-Pants: ** The return address of this email should be: [email protected]
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1466


> > > ================
> > > My code
> > > ================
> > > $cursor = $dbh->prepare( 'select * from cori' );
> > > $cursor->execute;
> > > $cursor->debug(2);   # see below for results of debug
> > > while ($row = $cursor->fetchrow) {
> >
> >          ^^^^ should be ( $row ). fetchrow returns a LIST not a scalar.
>
>
> I don't understand what you mean by this.  I've tried capturing what
> fetchrow returns with an array, but it still comes out empty.  For example,
> if I use (while (@row = $cursor->fetchrow)...) and then print $row[0] or
> something, it returns nothing for the variable.  I can't think of a
> different way to read what "fetchrow" returns.

OK, what you want to do is explicitly name each column, for a start, eg, if
the table 'cori' contains 2 fields called 'name' and 'number', then:

$cursor = $dbh->prepare( "select name, number from cori" );
if ( !defined $cursor ) {
   die "prepare failed: $DBI::errstr\n";
 }
$cursor->execute;
while ( ( $name, $number ) = $cursor->fetchrow ) {
   print "Row: name = $name\tnumber = $number\n";
 }
$cursor->finish;
undef $cursor;

There is a way to do it groovily with a SELECT *, but I can't remember how
I did it. I'll post the source once I find it.

> Corinne Koder
> [email protected]

--
Alligator Descartes     | "You fuckoffeh?" "Eh?" "Coffee? You for coffee?"
Anathematician          |     "Ohhh! Dearie me!"
[email protected]  |         -- The High Life
http://www.hermetica.com|
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05375-4@oink>;
         Wed, 10 Jul 1996 03:00:06 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836953513:27460:2; Wed, 10 Jul 96 00:05:13 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa27013; 10 Jul 96 0:04 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id PAA08929 for dbi-users-real; Tue, 9 Jul 1996 15:57:30 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id PAA08925
         for <[email protected]>; Tue, 9 Jul 1996 15:57:28 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id bb28028; 9 Jul 96 22:56 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa14808; 9 Jul 96 23:13 +0100
Received: from toad by oink with SMTP (PP) id <04558-0@oink>;
         Tue, 9 Jul 1996 23:01:16 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA10282;
         Tue, 9 Jul 1996 23:00:43 +0000
Date: Tue, 9 Jul 1996 23:00:43 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: DBD:Make Problems
X-Sun-Charset: US-ASCII
content-length: 344
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: Jonathan Olson <[email protected]>
>
> I am having DBD:Oracle-0.35 make problems - PLEASE HELP!

Shouting PLEASE HELP won't endear you to those who are best placed to help.

> "/usr/oracle/rdbms/demo/oratypes.h", line 168: warning: signed is a
> keyword in ANSI C

Read the README for your answer and your enlightenment.

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05375-16@oink>;
         Wed, 10 Jul 1996 03:02:04 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836956861:02026:0; Wed, 10 Jul 96 01:01:01 BST
Received: from relay-1.mail.demon.net ([158.152.1.140])
         by punt-1.mail.demon.net id aa01562; 10 Jul 96 1:00 +0100
Received: from toccata.fugue.com ([204.254.239.2]) by relay-1.mail.demon.net
         id aa27494; 10 Jul 96 0:59 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id OAA08787 for dbi-users-real; Tue, 9 Jul 1996 14:46:47 -0700
Precedence: junk
Sender: [email protected]
Received: from mercury.Sun.COM (mercury.Sun.COM [192.9.25.1])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id OAA08783
         for <[email protected]>; Tue, 9 Jul 1996 14:46:46 -0700
Received: by mercury.Sun.COM (Sun.COM) id OAA05920;
         Tue, 9 Jul 1996 14:46:13 -0700
Received: from suntri.East.Sun.COM by East.Sun.COM (5.x/SMI-5.3) id AB21337;
         Tue, 9 Jul 1996 17:46:09 -0400
Received: from thebirds by suntri.East.Sun.COM (SMI-8.6/SMI-SVR4) id RAA17779;
         Tue, 9 Jul 1996 17:46:06 -0400
Message-Id: <[email protected]>
Date: Tue, 09 Jul 1996 17:46:07 -0400
From: Dave Anderson - SMCC Reseller SE - Cary NC <[email protected]>
Organization: SMCC
X-Mailer: Mozilla 2.01 (X11; I; SunOS 5.5.1 sun4m)
Mime-Version: 1.0
To: [email protected]
Cc: [email protected]
Subject: LD_LIBRARY_PATH problem
X-Url: http://www.hermetica.com/technologia/DBI/DBI/index.html
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Greetings,

I'm writing a small web application the uses the Oracle DB driver but
I've been having some problems.  Apparently, the script I've written
only works if I have LD_LIBRARY_PATH set to the $ORACLE_HOME/lib
directory.  Otherwise, it give me the following error:

install_driver(Oracle) failed: Can't load '/opt/local/perl5/lib/site_perl/sun4-solaris/auto/DBD/Oracle/Oracle.so'
for module DBD::Oracle: ld.so.1: /opt/local/bin/perl: fatal:
libclntsh.so.1.0: can't open file: errno=2 at
/opt/local/perl5/lib/DynaLoader.pm line 140.

at /opt/local/perl5/lib/site_perl/DBD/Oracle.pm line 24
       DBI::install_driver called at xx.pl line 7


This is not a problem when I ran my perl scripts by hand because I can
set the LD_LIBRARY_PATH variable but I don't see a way to do this with
CGI/Perl.  BTW, I have tried to use
$ENV{'LD_LIBRARY_PATH'}=$ORACLE_HOME/lib but I haven't had any luck
getting it to work.  Is there a way around this error or procedure so
that I don't need to use the LD_LIBRARY_PATH variable? Are there any
compiler flags such as -R that will help this problem?

Other information:
OS/Version: Solaris 2.5
Perl Version: 5.002
Oracle: 7.3.2


Any pointers or comments are welcome.

TIA,
Dave
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05375-24@oink>;
         Wed, 10 Jul 1996 03:03:02 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836961294:08512:2; Wed, 10 Jul 96 02:14:54 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa08396; 10 Jul 96 2:14 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id SAA09216 for dbi-users-real; Tue, 9 Jul 1996 18:04:21 -0700
Precedence: junk
Sender: [email protected]
Received: from relay-4.mail.demon.net (relay-4.mail.demon.net [158.152.1.108])
         by toccata.fugue.com (8.6.12/8.6.11) with SMTP id SAA09212
         for <[email protected]>; Tue, 9 Jul 1996 18:04:19 -0700
Received: from post.demon.co.uk ([158.152.1.72]) by relay-4.mail.demon.net
         id ae14029; 10 Jul 96 1:04 GMT
Received: from ignite.demon.co.uk ([158.152.8.99]) by relay-3.mail.demon.net
         id aa07166; 9 Jul 96 22:48 +0100
Received: from toad by oink with SMTP (PP) id <03541-0@oink>;
         Tue, 9 Jul 1996 20:08:20 +0100
Received: by toad.ig.co.uk (5.0/SMI-SVR4) id AA09723;
         Tue, 9 Jul 1996 20:08:13 +0000
Date: Tue, 9 Jul 1996 20:08:13 +0000
From: Tim Bunce <[email protected]>
Message-Id: <[email protected]>
To: [email protected], [email protected]
Subject: Re: Inserting LONG RAW data using DBI/DBD-Oracle
X-Sun-Charset: US-ASCII
content-length: 1111
Organization: Paul Ingram Group, Software Systems, +44 1 483 424424


> From: [email protected] (adb)
>
> I've found a work-around.  The work around is, instead of relying on
> `execute' to bind parameters (which gives them a VARCHAR2 type by
> default), use the `bind_param' function to bind them manually with
> explicit type attributes, e.g.
>
>   $attrib{'ora_type'} = 24;                # -- 24 is LONG RAW type code
>   $sth->bind_param(1, $string, \%attrib);  # -- must pass reference to attrib
>   $sth->execute;                           # -- no parameters to execute

It's not a workaround. It's the right way to do it!

> In order for this to work, a bug in the DBD-Oracle `dbdimp.c' code had
> to be fixed.  Context diffs are attached.  From reading the code, it
> seems that the `bind_param' function with non-null attributes is
> untested (apart from my minimal testing).  Use it with caution ...

Thanks. Your patch is just fine (I forget from time to time that hv_fetch
needs an actual string length and won't work it out itself like many other
perl functions). It'll be in the next release.

And what a joy it is to receive a patch rather than bug report :-)

Tim.
---+++---
Received: from punt.demon.co.uk by oink with SMTP (PP) id <05375-27@oink>;
         Wed, 10 Jul 1996 03:03:11 +0100
Received: from punt-1.mail.demon.net by mailstore for [email protected]
         id 836963083:20654:2; Wed, 10 Jul 96 02:44:43 BST
Received: from toccata.fugue.com ([204.254.239.2]) by punt-1.mail.demon.net
         id aa20234; 10 Jul 96 2:43 +0100
Received: (from daemon@localhost) by toccata.fugue.com (8.6.12/8.6.11)
         id SAA09263 for dbi-users-real; Tue, 9 Jul 1996 18:35:48 -0700
Precedence: junk
Sender: [email protected]
Received: from cs.pdx.edu ([email protected] [204.203.64.22])
         by toccata.fugue.com (8.6.12/8.6.11) with ESMTP id SAA09259
         for <[email protected]>; Tue, 9 Jul 1996 18:35:47 -0700
Received: (xuhao@localhost) by cs.pdx.edu (8.7.3/CATastrophe-2/10/96-P)
         id SAA28910; Tue, 9 Jul 1996 18:35:45 -0700 (PDT) for
From: [email protected] (Hao Xu)
Message-Id: <[email protected]>
Subject: Question
To: [email protected]
Date: Tue, 9 Jul 1996 18:35:44 -0700 (PDT)
Cc: [email protected] (Hao Xu)
X-Mailer: ELM [version 2.4 PL13]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

Hi, there,

Thanks for all the help from you.

Now, I have nearly finished compiling, but
whenever I run a DBperl script, there is
'Bus Error(core dump)' *after* the script
is executed.

I.E., things work now, but the last message
of core dump frustrates me a lot, though it
seems to be not prevent me from working:->

I found that the core dump is due to
"eval 'use Oraperl; 1' || die $@ if ] >=5;"

Could you give me any hint on this? This is my
problem of installation or your bug?

Thanks a lot,

-Hao

p.s.: besides, i didn't find the core file after all.
---+++---