From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752821AbZA1O0r (ORCPT ); Wed, 28 Jan 2009 09:26:47 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751031AbZA1O0i (ORCPT ); Wed, 28 Jan 2009 09:26:38 -0500 Received: from relay2.sgi.com ([192.48.179.30]:52991 "EHLO relay.sgi.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1750994AbZA1O0h (ORCPT ); Wed, 28 Jan 2009 09:26:37 -0500 Date: Wed, 28 Jan 2009 08:26:35 -0600 From: Jack Steiner To: Andrew Morton Cc: linux-kernel@vger.kernel.org Subject: Re: [Patch 01/14] Add definitions of x86_64 GRU MMRs. Message-ID: <20090128142634.GA27540@sgi.com> References: <20090122174917.563627000@sgi.com> <20090122175025.475956000@sgi.com> <20090128013527.005648cc.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20090128013527.005648cc.akpm@linux-foundation.org> User-Agent: Mutt/1.4.2.1i Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Jan 28, 2009 at 01:35:27AM -0800, Andrew Morton wrote: > On Thu, 22 Jan 2009 11:49:18 -0600 steiner@sgi.com wrote: > > > Add definitions for x86_64 GRU MMRs. > > Breaks x86_32: > > arch/x86/include/asm/uv/uv_mmrs.h:409: error: width of 'real_time_cmpb' exceeds its type > arch/x86/include/asm/uv/uv_mmrs.h:425: error: width of 'real_time_cmpc' exceeds its type > arch/x86/include/asm/uv/uv_mmrs.h:441: error: width of 'real_time_cmpd' exceeds its type > arch/x86/include/asm/uv/uv_mmrs.h:491: error: width of 'address' exceeds its type > > > due to stuff flike > > unsigned long real_time_cmpb : 56; /* RW */ > > > this: > > --- a/arch/x86/kernel/io_apic.c~a > +++ a/arch/x86/kernel/io_apic.c > @@ -59,8 +59,11 @@ > #include > #include > #include > + > +#ifdef CONFIG_X86_64 > #include > #include > +#endif > > #include > #include > _ > > seems sufficient, but perhaps it'd be better if it was done within the > header? There is no reason that any of the UV definitions should be needed for 32-bit platforms. I'll find the best way to avoid them.... --- jack