From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754616AbbI3Ie1 (ORCPT ); Wed, 30 Sep 2015 04:34:27 -0400 Received: from mail.skyhub.de ([78.46.96.112]:38124 "EHLO mail.skyhub.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753711AbbI3IeV (ORCPT ); Wed, 30 Sep 2015 04:34:21 -0400 Date: Wed, 30 Sep 2015 10:33:39 +0200 From: Borislav Petkov To: "H. Peter Anvin" Cc: Matt Fleming , Ingo Molnar , Ard Biesheuvel , Thomas Gleixner , "linux-kernel@vger.kernel.org" , "linux-efi@vger.kernel.org" , Leif Lindholm , Catalin Marinas , Will Deacon , "stable@vger.kernel.org" , Matt Fleming , Mark Rutland , Mark Salter , Linus Torvalds , Andrew Morton , Andy Lutomirski , Denys Vlasenko , Brian Gerst Subject: Re: [PATCH 2/2] arm64/efi: Don't pad between EFI_MEMORY_RUNTIME regions Message-ID: <20150930083339.GB6033@pd.tnic> References: <1443218539-7610-1-git-send-email-matt@codeblueprint.co.uk> <1443218539-7610-3-git-send-email-matt@codeblueprint.co.uk> <20150926060159.GB25877@gmail.com> <20150927070644.GC26125@gmail.com> <20150927104014.GA7631@pd.tnic> <20150929143612.GC4401@codeblueprint.co.uk> <560B3327.9020801@zytor.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <560B3327.9020801@zytor.com> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Sep 29, 2015 at 05:56:07PM -0700, H. Peter Anvin wrote: > On 09/29/2015 07:36 AM, Matt Fleming wrote: > > > > That's a pretty good summary for x86. I think specifically the reason > > we map the EFI memmap entries "backwards" (entry N has higher VA than > > entry N+1) is because the code was easier to write that way, but > > you'll know better than me ;-) > > > > There were two reasons: > > 1. The code was easier to write. > 2. Windows did it that way. > > Windows apparently broke and was changed due to this feature, too. So can we do the 1:1 thing again? I mean, we do create a special pagetable for EFI anyway, we can put in there whatever we want. I know, some apple boxes reportedly fail when 1:1 mapping is in use but we can do the VA mapping as a workaround for them. I.e., have the 1:1 mapping be the default. Apparently, there's not a single OS or tool which is used by fw writers to test their brain dumplings. Windoze breakage case-in-point. Because if there were, we'd simply do what that OS/tool does and be done with it. What really makes me climb the walls is when half-cooked, untested fw hits the wild and we have to support it indefinitely. -- Regards/Gruss, Boris. ECO tip #101: Trim your mails when you reply.