From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752000Ab1KGVHi (ORCPT ); Mon, 7 Nov 2011 16:07:38 -0500 Received: from cavan.codon.org.uk ([93.93.128.6]:44936 "EHLO cavan.codon.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751246Ab1KGVHh (ORCPT ); Mon, 7 Nov 2011 16:07:37 -0500 Date: Mon, 7 Nov 2011 21:07:02 +0000 From: Matthew Garrett To: "H. Peter Anvin" Cc: Matt Fleming , Thomas Gleixner , Ingo Molnar , Zhang Rui , Huang Ying , linux-kernel@vger.kernel.org Subject: Re: [PATCH v3] x86, efi: Calling __pa() with an ioremap'd address is invalid Message-ID: <20111107210702.GA28781@srcf.ucam.org> References: <1320680088-2584-1-git-send-email-matt@console-pimps.org> <20111107202324.GA27515@srcf.ucam.org> <4EB8413D.6030500@zytor.com> <20111107203752.GA27875@srcf.ucam.org> <4EB8436B.20603@zytor.com> <20111107204839.GA28261@srcf.ucam.org> <4EB84644.8060102@zytor.com> <20111107205859.GA28681@srcf.ucam.org> <4EB84734.3070006@zytor.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <4EB84734.3070006@zytor.com> User-Agent: Mutt/1.5.20 (2009-06-14) X-SA-Exim-Connect-IP: X-SA-Exim-Mail-From: mjg59@cavan.codon.org.uk X-SA-Exim-Scanned: No (on cavan.codon.org.uk); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Nov 07, 2011 at 01:01:40PM -0800, H. Peter Anvin wrote: > On 11/07/2011 12:58 PM, Matthew Garrett wrote: > > > >> However, "EFI stub in the kernel" doesn't mean "can't use an external > >> bootloader." > > > > Agreed. It just means that we're still plausibly going to need some > > handshaking between them. Alternatively, as long as the bootloader > > passes us the memory map, we can just ignore any E820 map it gives us > > anyway. > > > > I know we need to be able to pass the initramfs in memory; anything else > we need other than the normal EFI executable entry conditions? If we're called before ExitBootServices(), no. If we're called after, we'll need the map from GetMemoryMap(). There's some other things that we may want to pass, such as option ROMs that we can get from firmware but which may not otherwise be mapped - I guess those could arguably be passed in the initramfs. -- Matthew Garrett | mjg59@srcf.ucam.org