Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:73767
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: error (pb1.pair.com: domain belski.net from 85.214.73.107 cause and error)
Message-ID: <ab107d6cfe977465fa86e21ed80c966c.squirrel@webmail.klapt.com>
In-Reply-To: <53545EA8.2020703@sugarcrm.com>
References: <53545EA8.2020703@sugarcrm.com>
Date: Tue, 22 Apr 2014 21:35:17 +0200
To: "Stas Malyshev" <smalyshev@sugarcrm.com>
Cc: "PHP Internals" <internals@lists.php.net>,
 "Pierre Joye" <pierre.php@gmail.com>,
 "Julien Pauli" <jpauli@php.net>
User-Agent: SquirrelMail/1.5.2 [SVN]
MIME-Version: 1.0
Content-Type: text/plain;charset=UTF-8
Content-Transfer-Encoding: 8bit
Subject: Re: [PHP-DEV] bug 50444 (PDO-ODBC on 64-bit)
From: anatol.php@belski.net ("Anatol Belski")

Hi Stas,

On Mon, April 21, 2014 01:56, Stas Malyshev wrote:
> Hi!
>
>
> There's a bug #50444 (https://bugs.php.net/bug.php?id=50444) regarding
> PDO-ODBC types with 64-bit builds. Looks like we are using wrong types
> in a couple of places. I've merged the patch for 5.6, but I'm not sure what
> to do with 5.5 and maybe 5.4. The thing is technically it's structure
> change which may break BC. OTOH, the old way didn't seem to work on 64-bit
> anyway. So, I wonder what is the status of our 64-bit support - e.g., I
> see we don't even have official windows build for 64-bit and 5.4, so it
> may not make sense to put it in 5.4, but what about 5.5? --
> Stanislav Malyshev, Software Architect
> SugarCRM: http://www.sugarcrm.com/
> (408)454-6900 ext. 227
>
>
I've just tested this patch - the compiler is fine with it, though I have
5 failing tests. Despite that I'd say it's fine to merge this patch in as
I see some places where the corresponding vars are passed by reference to
the SQL API. That is the 64 bit ussue which can lead to a crash without
this patch.

So for 5.4 it could only make sense if one does homemade 64 bit builds,
5.5 would make sense to be merged back though it's experimental, and into
str_size_and_int64 it'll slide automatically anyway. Otherwise it is just
harmless on 32 bit.

Shortly I'm going to investigate on the failing tests I've mentioned.

Regards

Anatol