On Fri, Jan 27, 2023 at 6:29 PM Michael S. Tsirkin wrote:
>
> On Mon, Dec 26, 2022 at 12:12:42PM +0800, Jason Wang wrote:
> > > >@@ -682,6 +553,11 @@ static int vdpasim_dma_unmap(struct vdpa_device
> > > >*vdpa, unsigned int asid,
> > > > if (asid >= vdpasim->dev_attr.nas)
> > > >
On Mon, Dec 26, 2022 at 12:12:42PM +0800, Jason Wang wrote:
> > >@@ -682,6 +553,11 @@ static int vdpasim_dma_unmap(struct vdpa_device
> > >*vdpa, unsigned int asid,
> > > if (asid >= vdpasim->dev_attr.nas)
> > > return -EINVAL;
> > >
> > >+ if (vdpasim->iommu_pt[asid]) {
On Fri, Dec 23, 2022 at 5:27 PM Stefano Garzarella wrote:
>
> On Fri, Dec 23, 2022 at 02:00:21PM +0800, Jason Wang wrote:
> >We used to (ab)use the DMA ops for setting up identical mappings in
> >the IOTLB. This patch tries to get rid of the those unnecessary DMA
> >ops by maintaining a simple
On Fri, Dec 23, 2022 at 02:00:21PM +0800, Jason Wang wrote:
We used to (ab)use the DMA ops for setting up identical mappings in
the IOTLB. This patch tries to get rid of the those unnecessary DMA
ops by maintaining a simple identical/passthrough mappings by
default. When bound to virtio_vdpa
Looks good from the DMA subsysten POV:
Acked-by: Christoph Hellwig
Thanks for doing the work!
___
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization
We used to (ab)use the DMA ops for setting up identical mappings in
the IOTLB. This patch tries to get rid of the those unnecessary DMA
ops by maintaining a simple identical/passthrough mappings by
default. When bound to virtio_vdpa driver, DMA API will simply use PA
as the IOVA and we will be all