From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:57956) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WiMhd-0008FH-OW for qemu-devel@nongnu.org; Thu, 08 May 2014 07:41:34 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WiMhY-0007xJ-RK for qemu-devel@nongnu.org; Thu, 08 May 2014 07:41:29 -0400 Received: from mx1.redhat.com ([209.132.183.28]:58108) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WiMhY-0007xF-Ie for qemu-devel@nongnu.org; Thu, 08 May 2014 07:41:24 -0400 Date: Thu, 8 May 2014 14:39:47 +0300 From: "Michael S. Tsirkin" Message-ID: <20140508113947.GB16456@redhat.com> References: <1399195515-1124-1-git-send-email-arei.gonglei@huawei.com> <20140507184848.GB15679@phenom.dumpdata.com> <20140507201433.GA11883@redhat.com> <33183CC9F5247A488A2544077AF19020815E6C65@SZXEMA503-MBS.china.huawei.com> <1399548381.9513.66.camel@kazak.uk.xensource.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1399548381.9513.66.camel@kazak.uk.xensource.com> Subject: Re: [Qemu-devel] [Xen-devel] [PATCH v3] Hvmloader: Modify ACPI to only supply _EJ0 methods for PCIslots that support hotplug by runtime patching List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Ian Campbell Cc: "kevin@koconnor.net" , "seabios@seabios.org" , "Huangweidong (C)" , "stefano.stabellini@eu.citrix.com" , Konrad Rzeszutek Wilk , "qemu-devel@nongnu.org" , "xen-devel@lists.xen.org" , "fabio.fantoni@m2r.biz" , "johannes.krampf@googlemail.com" , "Gonglei (Arei)" , "JBeulich@suse.com" , "anthony.perard@citrix.com" , "Hanweidong (Randy)" , "Gaowei (UVP)" , Paul Durrant On Thu, May 08, 2014 at 12:26:21PM +0100, Ian Campbell wrote: > On Thu, 2014-05-08 at 07:29 +0000, Gonglei (Arei) wrote: > > > + rmvc_pcrm = inl(PCI_RMV_BASE); > > > > > > > > > > > > So .. what is at this special 0xae0c address? > > > > > > > > Is there some code in QEMU that needs this treatment as well? > > > > > > Yep, the 0xae0c is defined in upstream QEMU, which show the hotplugging ability > > of PCI devices. > > Is this considered to be a ABI by qemu? Or is it subject to change? > > How is this going to interact with > http://lists.xen.org/archives/html/xen-devel/2014-05/msg00026.html ? (At > leas then it is a Xen interface I suppose) > > Ian. This is the ABI guarantee: There's a file (exposed with fwcfg file interface to guests) named etc/table-loader. If present, this file includes a list of other fwcfg files including available acpi tables (can be multiple acpi tables per file) together with some extra information helpful for guests (checksum location and pointers from one table to another). -- MST