Linmodems Mailing list Archives

Google
 
Web archives.linmodems.org

Return-Path: <robert.thornburrow@tsac.fsnet.co.uk>
Mailing-List: contact discuss-help@linmodems.org; run by ezmlm
Delivered-To: mailing list discuss@linmodems.org
Received: (qmail 17188 invoked from network); 16 Apr 2004 13:37:39 -0000
Received: from heisenberg.zen.co.uk (212.23.3.141)
  by www.linmodems.org with SMTP; 16 Apr 2004 13:37:38 -0000
Received: from [82.69.26.118] (helo=tsac.fsnet.co.uk)
	by heisenberg.zen.co.uk with esmtp (Exim 4.30)
	id 1BETXE-0005ao-H2; Fri, 16 Apr 2004 13:37:36 +0000
Message-ID: <407FE435.8020303@tsac.fsnet.co.uk>
Date: Fri, 16 Apr 2004 14:48:37 +0100
From: Robert Thornburrow <robert.thornburrow@tsac.fsnet.co.uk>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5) Gecko/20031031
X-Accept-Language: en-gb, en-us, en
MIME-Version: 1.0
CC: "Omar A. S. Guerrero" <osgcon@yahoo.com.mx>,  discuss@linmodems.org
Subject: Re: A problem compiling the driver...
References: <20040414205115.63047.qmail@web13125.mail.yahoo.com> <407DACA9.7040800@tsac.fsnet.co.uk>
In-Reply-To: <407DACA9.7040800@tsac.fsnet.co.uk>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-Originating-Heisenberg-IP: [82.69.26.118]

Hi again Omar,

This seems to be a problem perculiar to Fedora, as the vanilla kernel
2.4.22 does not seem to use atomic_t.  I will see if we can find some
convenient #define to make the code compile correctly under Fedora.
(I am assuming that they have applied some kernel patches, which are
not present in the vanilla 2.4.22 kernel - or any others for that
matter!).

In the meantime, if you change the offending line to read:

     if ((tty->count.counter == 1) && (state->count != 1))

instead of:

     if ((tty->count == 1) && (state->count != 1))

[Thanks to Alberto Monteiro for that one].

Best regards,

    Robert

Robert Thornburrow wrote:
> Hi Omar,
> 
> It looks like you've pretty much solved this one yourself ;-)
> 
> I've checked kernels 2.4.18-14 (RH8.), 2.4.22-rc2 and 2.4.23,
> and every single one of the defines "count" as type int.  I will
> check 2.4.26 when it finishes downloading!
> 
> Can you check what type "atomic_t" is defined as?  If it is a
> typedef to int, the a simple cast will fix the problem.  If it
> is a struct, then you need to pull out the appropriate member.
> 
> Not having Fedora installed, I can't check this myself.  It may
> be something specific to Fedora.  Anyway, we need to get to the
> bottom of it so that the sources can be fixed.
> 
> Best regards,
> 
>     Robert

Webmaster: Russell Nelson
Last modified: Wed Jul 30 11:02:43 EDT 2003