From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:49981) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dYbYK-00040L-CU for qemu-devel@nongnu.org; Fri, 21 Jul 2017 13:17:25 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dYbYH-0002DT-5A for qemu-devel@nongnu.org; Fri, 21 Jul 2017 13:17:24 -0400 Received: from zero.eik.bme.hu ([152.66.115.2]:22002) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1dYbYG-0002C1-PX for qemu-devel@nongnu.org; Fri, 21 Jul 2017 13:17:21 -0400 Date: Fri, 21 Jul 2017 19:17:17 +0200 (CEST) From: BALATON Zoltan In-Reply-To: Message-ID: References: <3E24AFDE-D25A-416B-A600-3C2221C3A9F1@gmail.com> <20170721092026.GD17693@redhat.com> <20170721114623.57b3be1d@nial.brq.redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII; format=flowed Subject: Re: [Qemu-devel] Commit 77af8a2b95b79699de650965d5228772743efe84 breaks Windows 2000 support List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Phil Dennis-Jordan Cc: Igor Mammedov , "Daniel P. Berrange" , Phil Dennis-Jordan , ehabkost@redhat.com, "qemu-devel@nongnu.org qemu-devel" , Programmingkid , Paolo Bonzini , Richard Henderson On Fri, 21 Jul 2017, Phil Dennis-Jordan wrote: > On Fri, Jul 21, 2017 at 12:50 PM, BALATON Zoltan wrote: >> I don't know if this helps but I've found that this same commit also broke >> booting OS X on q35 with OVMF and Clover (some old versions I had and worked >> before this commit). See here: >> http://lists.nongnu.org/archive/html/qemu-devel/2017-06/msg04306.html >> >> Maybe it's simpler to debug as these are open source but it's still not very >> clear why they fail because it starts to boot OS X but hangs during kernel >> init (and maybe Win2K fails to boot because of something else). I think I've >> seen this with OVMF 58f025afd and Clover 3354 (or around that, not >> completely sure about these versions). > > OVMF had 2 bugs that totally mucked up the ACPI tables if Qemu didn't > supply its loader commands "just so", and modified ACPI 2.0-5.0 FADTs > in such a way that they were no longer valid. These are fixed in > 198a46d and 072060a, which are about a year newer than 58f025afd, so > it's no surprise this combination doesn't work. I have no idea if > Clover on top makes any difference. I've tried with 072060a and with that version OS X boots with same Clover so my problem seems to have been too old OVMF version and Clover does not make a difference here. So this also likely has nothing do with the Win2k problem discussed in this thread so you can disregard my post. Thank you, BALATON Zoltan