From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:41029) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TuoDQ-0007zO-7n for qemu-devel@nongnu.org; Mon, 14 Jan 2013 12:53:00 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1TuoDJ-0000gP-4V for qemu-devel@nongnu.org; Mon, 14 Jan 2013 12:52:56 -0500 Received: from mx1.redhat.com ([209.132.183.28]:46547) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TuoDI-0000g7-Qh for qemu-devel@nongnu.org; Mon, 14 Jan 2013 12:52:48 -0500 Message-ID: <50F44660.7010904@redhat.com> Date: Mon, 14 Jan 2013 18:54:40 +0100 From: Laszlo Ersek MIME-Version: 1.0 References: <1357681452-24963-1-git-send-email-lersek@redhat.com> <50F00384.10907@redhat.com> <50F43AD3.5030604@redhat.com> In-Reply-To: <50F43AD3.5030604@redhat.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [RFC PATCH] PIIX: reset the VM when the Reset Control Register's RCPU bit gets set List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Blue Swirl Cc: Paolo Bonzini , akong@redhat.com, qemu-devel@nongnu.org, Stefan Hajnoczi On 01/14/13 18:05, Laszlo Ersek wrote: > I guess I could try to implement the subregion by open-coding the > memory_region_init_io() and sysbus_add_io() functions and overriding as > needed, but it seems like a mess. (Which is of course caused by this > impossible hack of PIIX4 that places the reset control register inside > the PCI config index register! The IO port 0xcf9 actually has double > purpose, and it doesn't depend on any kind of "mapping".) After looking again at the docs, I'm now trying to add a generic region first (a container), and then, as its children, pci-conf-idx and rcr as two sibling regions. The docs suggest siblings can overlap. Thanks Laszlo