Re: [Qemu-devel] bypassing pci_dma_read() and pci_dma_write() ?

[Paolo Bonzini]

Il 18/01/2013 17:33, Luigi Rizzo ha scritto:
>> > First, host memory mappings could change (though they rarely do on PC).
>> >  The result of address_space_map is not guaranteed to be stable.  To
>> > avoid problems with this, however, you could use something like
>> > hw/dataplane/hostmem.c and even avoid address_space_map altogether.
> I'll look into that. Hopefully there is something that i can
> use as a notification that the mapping has changed...

Yes, that's the MemoryListener interface that hw/dataplane/hostmem.c uses.

>> > Second, that pci_dma_*() could have the addresses translated by an
>> > IOMMU.  virtio is documented to have "real" physical memory addresses,
>> > but this does not apply to other devices.
> I see. I suppose the ability to have an iommu depends on the
> specific NIC ? I am only planning to use the above shortcut for
> e1000.

It depends on the bus, in this case PCI.

Paolo