From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758393Ab1LNW6b (ORCPT ); Wed, 14 Dec 2011 17:58:31 -0500 Received: from terminus.zytor.com ([198.137.202.10]:34613 "EHLO mail.zytor.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758358Ab1LNW6a (ORCPT ); Wed, 14 Dec 2011 17:58:30 -0500 Message-ID: <4EE92A0C.2040207@zytor.com> Date: Wed, 14 Dec 2011 14:58:20 -0800 From: "H. Peter Anvin" User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:8.0) Gecko/20111115 Thunderbird/8.0 MIME-Version: 1.0 To: Matthew Garrett CC: Mike Waychison , linux-kernel@vger.kernel.org, x86@kernel.org Subject: Re: [PATCH 3/4] EFI: Add support for variables longer than 1024 bytes References: <1323896791-10858-1-git-send-email-mjg@redhat.com> <1323896791-10858-4-git-send-email-mjg@redhat.com> <20111214223950.GA29264@srcf.ucam.org> <4EE926D2.3030803@zytor.com> <20111214225728.GA29582@srcf.ucam.org> In-Reply-To: <20111214225728.GA29582@srcf.ucam.org> X-Enigmail-Version: 1.3.4 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 12/14/2011 02:57 PM, Matthew Garrett wrote: > On Wed, Dec 14, 2011 at 02:44:34PM -0800, H. Peter Anvin wrote: >> On 12/14/2011 02:39 PM, Matthew Garrett wrote: >>>> What are your thoughts on _not_ wrapping efi_variable with >>>> extended_efi_variable, and instead just using a >>>> "internal_efi_variable" structure that we copy stuff into/outof. I >>>> think that would make the memory management for dealing with the >>>> different sizes a lot easier to follow. >>> >>> Hm. I think that'd only work if we expose a new interface. Writes would >>> be easy enough to handle, but reads still need to work for old apps. >>> >> >> Would making the old ABI readonly and add a new write interface resolve >> the problem with compat mode? > > Yes, but we'd break existing userspace. > Obviously. However, you indicated above that that might be acceptable -- how many things use this ABI to set variables? -hpa