From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1762295AbYD2TTV (ORCPT ); Tue, 29 Apr 2008 15:19:21 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753155AbYD2TTN (ORCPT ); Tue, 29 Apr 2008 15:19:13 -0400 Received: from pasmtpb.tele.dk ([80.160.77.98]:43466 "EHLO pasmtpB.tele.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752462AbYD2TTN (ORCPT ); Tue, 29 Apr 2008 15:19:13 -0400 Date: Tue, 29 Apr 2008 21:19:46 +0200 From: Sam Ravnborg To: Andrew Morton Cc: riel@redhat.com, linux-kernel@vger.kernel.org, ajackson@redhat.com, airlied@redhat.com, benh@kernel.crashing.org Subject: Re: [PATCH 1/3] make access_process_vm work on device memory Message-ID: <20080429191946.GB32685@uranus.ravnborg.org> References: <20080429113237.614f89a1@cuia.boston.redhat.com> <20080429110919.dc7b8565.akpm@linux-foundation.org> <20080429190132.GA25793@uranus.ravnborg.org> <20080429121322.6da6b176.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080429121322.6da6b176.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 Tue, Apr 29, 2008 at 12:13:22PM -0700, Andrew Morton wrote: > On Tue, 29 Apr 2008 21:01:32 +0200 > Sam Ravnborg wrote: > > > > urgh. > > > > > > We have HAVE_ARCH* > > > We have __HAVE_ARCH* > > > We have ARCH_HAS* > > > We have __ARCH_HAS* > > > > > > what a mess. > > > > > > Probably the preferred (but still ugly) approach is to implement > > > CONFIG_ARCH_*. > > Only if it is a Kconfig symbol. > > > > We have recently started to use 'HAVE_*' for boolean symbols used > > to select a specific function for one architecture. > > See arch/Kconfig for a few samples. > > > > And Documentation/kbuild/kconfig-language.txt for usage hints. > > > > I think what you're telling us is to use CONFIG_HAVE_IOREMAP_PROT? I think so. But I just stumbled upon this mail and have not even tried to follow what you discuss. So yes - if it is a kconfig symbol. And if it is a bool. And if it has no dependencies, or we can create a helper with no dependencies - so yes. Sam