From mboxrd@z Thu Jan 1 00:00:00 1970 From: Konrad Rzeszutek Wilk Subject: Re: Is: PVH - how to solve maxmem != memory scenario? Was:Re: [PATCH] libxl: create PVH guests with max memory assigned Date: Tue, 5 Aug 2014 11:56:51 -0400 Message-ID: <20140805155651.GA20827@laptop.dumpdata.com> References: <1405594942-20760-1-git-send-email-roger.pau@citrix.com> <1407228940.2120.22.camel@kazak.uk.xensource.com> <53E0A538.608@citrix.com> <53E0BB26.6050507@citrix.com> <20140805141844.GG13057@laptop.dumpdata.com> <53E0F2B8.7040302@citrix.com> <20140805154005.GD19858@laptop.dumpdata.com> <53E1198602000078000298ED@mail.emea.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mail6.bemta5.messagelabs.com ([195.245.231.135]) by lists.xen.org with esmtp (Exim 4.72) (envelope-from ) id 1XEh6l-0001Dt-Bj for xen-devel@lists.xenproject.org; Tue, 05 Aug 2014 15:57:05 +0000 Content-Disposition: inline In-Reply-To: <53E1198602000078000298ED@mail.emea.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Jan Beulich Cc: Ian Campbell , george.dunlap@eu.citrix.com, tim@xen.org, Ian Jackson , David Vrabel , xen-devel@lists.xenproject.org, roger.pau@citrix.com List-Id: xen-devel@lists.xenproject.org On Tue, Aug 05, 2014 at 04:51:02PM +0100, Jan Beulich wrote: > >>> On 05.08.14 at 17:40, wrote: > > There is a tiny problem with PoD: PCI passthrough. > > > > Right now we don't allow PoD + PCI until the hardware allows it > > (if it ever will). > > But PVH doesn't support pass-through so far, does it? Dom0 is an case of where it does work. My understanding is that for domU cases it just needs the proper E820 hookup and it _should_ work. > > > Thought for PV we do allow ballooning and PCI passthrough as we can > > at least control that the pages that are being ballooned are not going > > to be used for DMA operations (unless the guest does something stupid > > and returns an page back to the allocated but still does DMA ops on > > the PFN). > > And it would seem possible to permit the combination for PVH if > there was a documented requirement for the ballooning down to > happen before any device I/O starts. The same requirement should be applied to a normal PV guest as well. > > Jan >