Re: [PATCH] s390: protvirt: virtio: Refuse device without IOMMU

[Pierre Morel]

On 2020-06-15 12:37, Halil Pasic wrote:
> On Mon, 15 Jun 2020 11:01:55 +0800
> Jason Wang <jasowang@xxxxxxxxxx> wrote:
>
> > > hum, in between I found another way which seems to me much better:
> > >
> > > We already have the force_dma_unencrypted() function available which
> > > AFAIU is what we want for encrypted memory protection and is already
> > > used by power and x86 SEV/SME in a way that seems AFAIU compatible
> > > with our problem.
> > >
> > > Even DMA and IOMMU are different things, I think they should be used
> > > together in our case.
> > >
> > > What do you think?
> > >
> > > The patch would then be something like:
> > >
> > > diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c
> > > index a977e32a88f2..53476d5bbe35 100644
> > > --- a/drivers/virtio/virtio.c
> > > +++ b/drivers/virtio/virtio.c
> > > @@ -4,6 +4,7 @@
> > >  Â#include <linux/virtio_config.h>
> > >  Â#include <linux/module.h>
> > >  Â#include <linux/idr.h>
> > > +#include <linux/dma-direct.h>
> > >  Â#include <uapi/linux/virtio_ids.h>
> > >
> > >  Â/* Unique numbering for virtio devices. */
> > > @@ -179,6 +180,10 @@ int virtio_finalize_features(struct virtio_device
> > > *dev)
> > >  ÂÂÂÂÂÂÂ if (!virtio_has_feature(dev, VIRTIO_F_VERSION_1))
> > >  ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ return 0;
> > >
> > > +ÂÂÂÂÂÂ if (force_dma_unencrypted(&dev->dev) &&
> > > +ÂÂÂÂÂÂÂÂÂÂ !virtio_has_feature(dev, VIRTIO_F_IOMMU_PLATFORM))
> > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂ return -EIO;
> > > +
> > >  ÂÂÂÂÂÂÂ virtio_add_status(dev, VIRTIO_CONFIG_S_FEATURES_OK);
> > >  ÂÂÂÂÂÂÂ status = dev->config->get_status(dev);
> > >  ÂÂÂÂÂÂÂ if (!(status & VIRTIO_CONFIG_S_FEATURES_OK)) {
> >
> >
> > I think this can work but need to listen from Michael
>
> I don't think Christoph Hellwig will like force_dma_unencrypted()
> in virtio code:
> https://lkml.org/lkml/2020/2/20/630
>
> Regards,
> Halil

OK, then back to the first idea.
Thanks,

Pierre

--
Pierre Morel
IBM Lab Boeblingen