From mboxrd@z Thu Jan 1 00:00:00 1970 From: Roland McGrath Subject: Re: [PATCH V2] score: add regset support Date: Wed, 15 Jul 2009 13:18:17 -0700 (PDT) Message-ID: <20090715201817.9FF0840F7F@magilla.sf.frob.com> References: <200907131624.28286.arnd@arndb.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: Received: from mx1.redhat.com ([66.187.233.31]:50325 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932142AbZGOUS7 (ORCPT ); Wed, 15 Jul 2009 16:18:59 -0400 In-Reply-To: Arnd Bergmann's message of Monday, 13 July 2009 16:24:28 +0200 <200907131624.28286.arnd@arndb.de> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Arnd Bergmann Cc: liqin.chen@sunplusct.com, Christoph Hellwig , linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org, Linus Torvalds IMHO it is certainly a wise choice to have some struct (be it called user_regs_struct or whatever) that describes the exact userland layout. (I think it's also fine to just call that elf_gregset_t and have it be an array, with REG_* defines/enum somewhere to document its indices. It's just a matter of taste for the arch folks.) Given such a struct, the regset code can refer to it (with offsetof, e.g.) and that will make it far easier to read, understand, and review. Thanks, Roland