From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([140.186.70.92]:54496) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1RcnWI-0004VT-91 for qemu-devel@nongnu.org; Mon, 19 Dec 2011 19:25:27 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1RcnWF-0000qg-40 for qemu-devel@nongnu.org; Mon, 19 Dec 2011 19:25:26 -0500 Received: from ozlabs.org ([203.10.76.45]:56752) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1RcnWE-0000qR-OV for qemu-devel@nongnu.org; Mon, 19 Dec 2011 19:25:23 -0500 Date: Tue, 20 Dec 2011 11:25:15 +1100 From: David Gibson Message-ID: <20111220002515.GA5133@truffala.fritz.box> References: <1323930340-24055-1-git-send-email-david@gibson.dropbear.id.au> <1323972307.2437.19.camel@x201.home> <20111216145353.GA29877@amd.com> <20111219001125.GA30390@truffala.fritz.box> <1324309598.2132.12.camel@shinybook.infradead.org> <20111219223120.GB5207@truffala.fritz.box> <1324335400.2132.47.camel@shinybook.infradead.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <1324335400.2132.47.camel@shinybook.infradead.org> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [RFC] Device isolation infrastructure v2 List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: David Woodhouse Cc: chrisw@redhat.com, aik@ozlabs.ru, Joerg Roedel , joro@8bytes.org, rusty@rustcorp.com.au, agraf@suse.de, iommu@lists.linux-foundation.org, qemu-devel@nongnu.org, B08248@freescale.com, Alex Williamson , scottwood@freescale.com, linux-kernel@vger.kernel.org On Mon, Dec 19, 2011 at 10:56:40PM +0000, David Woodhouse wrote: > On Tue, 2011-12-20 at 09:31 +1100, David Gibson wrote: > > When we're running paravirtualized under pHyp, it's impossible to > > merge multiple PEs into one domain per se. We could fake it rather > > nastily by replicating all map/unmaps across mutiple PEs. When > > running bare metal, we could do so a bit more nicely by assigning > > multiple PEs the same TCE pointer, but we have no mechanism to do so > > at present.=20 >=20 > VT-d does share the page tables, as you could on bare metal. But it's a= n > implementation detail =E2=80=94 there's nothing *fundamentally* wrong w= ith > having to do the map/unmap for each PE, is there? It's only at VM setup > time, so it doesn't really matter if it's slow. >=20 > Surely that's the only way you're going to present the guest with the > illusion of having no IOMMU; so that DMA to any given guest physical > address "just works". >=20 > On the other hand, perhaps you don't want to do that at all. Perhaps > you're better off presenting a virtualised IOMMU to the guest and > *insisting* that it fully uses it in order to do any DMA at all? Not only do we want to, we more or less *have* to. Existing kernels, which are used to being paravirt under phyp expect and need a paravirt iommu. DMA without iommu setup just doesn't happen. And the map/unmap hypercalls are frequently a hot path, so slow does matter. The other problem is that each domain's IOVA window is often fairly small, a limitation that would get even worse if we try to put too many devices in there. --=20 David Gibson | I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other= _ | _way_ _around_! http://www.ozlabs.org/~dgibson