Re: [Qemu-devel] [PATCH 1/9] Add stub functions for PCI device models to do PCI DMA

[David Gibson]

On Sun, Oct 16, 2011 at 03:15:53PM +0200, Avi Kivity wrote:
> On 10/14/2011 04:14 AM, David Gibson wrote:
> > > Virtio is a very, very special case.  virtio requires coherent RAM access.
> >
> > Right.  Virtio's access to memory is *not* emulated PCI DMA, it's
> > god-like hypervisor access to guest system memory.  It should
> > correctly bypass any IOMMU, and so should remain as
> > cpu_physical_memory_rw() or the atomic accessors, rather than being
> > converted to this new API.
> 
> virtio should definitely not bypass an iommu.

So, I just had a chat with Rusty about this.  Perhaps it shouldn't,
but it does.  The spec is in terms of guest physical addresses, not
bus/DMA addresses, and more to the point the Linux driver does *not*
do the necessary dma_map() and unmap operations to treat this as a PCI
DMA.  So like it or not, god-like hypervisor access rather than
emulated PCI DMA is what it does.

>  A guest may assign a
> virtio device to nested guests, and would wish it confined by the
> emulated iommu.

Well, that would be nice, but it can't be done.  It could be fixed,
but it would be an incompatible change so it would need a new feature
bit corresponding changes in the Linux driver to do the dma map/unmap
if it accepts the "respect IOMMU" feature.

> More generally, a guest sees a virtio device as just another pci device,
> and has no way to tell that it bypasses the iommu.

Well, except the fact that the driver knows its a virtio device,
because it's a virtio driver.  It's not like you can write a driver
that uses PCI DMA without knowing the particulars of the device you're
using.


-- 
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