Date: Wed, 27 Apr 2016 18:09:31 +0300
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Joerg Roedel <joro@8bytes.org>
Cc: Andy Lutomirski <luto@amacapital.net>,
        David Woodhouse <dwmw2@infradead.org>, Kevin Wolf <kwolf@redhat.com>,
        Wei Liu <wei.liu2@citrix.com>, Andy Lutomirski <luto@kernel.org>,
        qemu-block@nongnu.org, Christian Borntraeger <borntraeger@de.ibm.com>,
        Jason Wang <jasowang@redhat.com>,
        Stefano Stabellini <stefano.stabellini@eu.citrix.com>,
        "qemu-devel@nongnu.org Developers" <qemu-devel@nongnu.org>,
        peterx@redhat.com,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Amit Shah <amit.shah@redhat.com>, iommu@lists.linux-foundation.org,
        Stefan Hajnoczi <stefanha@redhat.com>, kvm list <kvm@vger.kernel.org>,
        Cornelia Huck <cornelia.huck@de.ibm.com>,
        Paolo Bonzini <pbonzini@redhat.com>,
        Linux Virtualization <virtualization@lists.linux-foundation.org>,
        Anthony PERARD <anthony.perard@citrix.com>
Subject: Re: [PATCH V2 RFC] fixup! virtio: convert to use DMA api
Message-ID: <20160427180610-mutt-send-email-mst@redhat.com>
References: <1461245745-6710-1-git-send-email-mst@redhat.com>
 <20160421135416.GE11775@citrix.com>
 <1461759501.118304.149.camel@infradead.org>
 <20160427153345-mutt-send-email-mst@redhat.com>
 <20160427142331.GH17926@8bytes.org>
 <CALCETrVkSSJbjoK8i7pLsSYR0o=Wy1UP-mrmn2uxYUd81g18dg@mail.gmail.com>
 <20160427173603-mutt-send-email-mst@redhat.com>
 <CALCETrUnXSLApj--ypOHnLWLqVSHLpSa922F2C2OCrvA1B585A@mail.gmail.com>
 <20160427175031-mutt-send-email-mst@redhat.com>
 <20160427145851.GJ17926@8bytes.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20160427145851.GJ17926@8bytes.org>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1032
Lines: 27

On Wed, Apr 27, 2016 at 04:58:51PM +0200, Joerg Roedel wrote:
> On Wed, Apr 27, 2016 at 05:54:57PM +0300, Michael S. Tsirkin wrote:
> > Point is, QEMU is not the only virtio implementation out there.
> > So we can't know no virtio implementations have an IOMMU as long as
> > linux supports this IOMMU.
> > virtio always used physical addresses since it was born and if it
> > changes that it must do this in a way that does not break existing
> > users.
> 
> FWIW, virtio in qemu can continue to just use physical addresses. But
> qemu needs to advertise that fact correctly to the OS in the DMAR table.
> This way old kernels (where virtio does not use DMA-API) will also
> continue to work on the fixed qemu.
> 
> 
> 
> 	Joerg

It's not clear it can do this since DMAR tables seem to assume
a given slot is either bypassing IOMMU or going through it.
QEMU allows reusing same slot for virtio and non-virtio devices.

Besides, this patch is not about it, it's a flag for QEMU to
tell guest that it can trust DMAR tables.

-- 
MST