Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751883AbaAPGqc (ORCPT ); Thu, 16 Jan 2014 01:46:32 -0500 Received: from mx1.redhat.com ([209.132.183.28]:3630 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751584AbaAPGq2 (ORCPT ); Thu, 16 Jan 2014 01:46:28 -0500 Date: Thu, 16 Jan 2014 14:46:28 +0800 From: Dave Young To: "H. Peter Anvin" Cc: Arjan van de Ven , Borislav Petkov , Linux EFI , LKML , Borislav Petkov , Matt Fleming , Matthew Garrett , Toshi Kani Subject: Re: [PATCH 1/4] x86, ptdump: Add the functionality to dump an arbitrary pagetable Message-ID: <20140116064627.GC6356@dhcp-16-126.nay.redhat.com> References: <20140113122320.GA1900@darkstar.nay.redhat.com> <52D3FCAA.7070004@linux.intel.com> <20140114014003.GB4327@dhcp-16-126.nay.redhat.com> <52D57F75.7060308@zytor.com> <20140115011609.GD23767@dhcp-16-126.nay.redhat.com> <52D6970D.3050701@zytor.com> <20140116014457.GA3807@dhcp-16-126.nay.redhat.com> <52D746D9.7030206@linux.intel.com> <20140116030343.GF3807@dhcp-16-126.nay.redhat.com> <52D75A55.2080503@zytor.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <52D75A55.2080503@zytor.com> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 01/15/14 at 08:04pm, H. Peter Anvin wrote: > On 01/15/2014 07:03 PM, Dave Young wrote: > >> > >> making something harder to grep and less standardized is hardly cleaner > >> and these things compile to nothing for non-modules. > > > > It's not nothing, just very small increasement: > > text data bss dec hex filename > > 7636121 1391824 9355264 18383209 1188169 vmlinux > > > > text data bss dec hex filename > > 7636113 1391824 9355264 18383201 1188161 vmlinux > > > > I do not want to insist on this minor problem, if you want please keep it. > > > > Now *that* we can fix by making these macros expand to nothing if you > are compiling for a nonmodule, right? That will benefit code that > legitimately can compile as a module but is configured "y", too. I'm not sure expanding to nothing is a good fix. For built-in module we still can retrieve the module info in userspace via /usr/sbin/modinfo, if we expand the Macros to nothing then userspace will not see thus infomation anymore. Another advantage of built-in module is it has module specific parameters which should not necessary a generic kernel parameter. If the *module* is really a nonmodule, IMHO it should add the information to or add other macros macros instead of reuse the MODULE_*. BTW, for macros in this file, for a nonmodule "license GPL" is unnecessary at all. Thanks Dave -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/