On Wed, 2012-06-20 at 16:25 -0500, Anthony Liguori wrote:
> > +static void dma_aio_cancel(BlockDriverAIOCB *acb)
> > +{
> > +DMAAIOCB *dbs = container_of(acb, DMAAIOCB, common);
> > +
> > +trace_dma_aio_cancel(dbs);
> > +
> > +if (dbs->acb) {
> > +BlockDriverAIOCB *acb = dbs->a
On Wed, 2012-06-20 at 16:25 -0500, Anthony Liguori wrote:
> So this cancellation stuff is hopelessly broken
>
> It's simply not possible to fully cancel pending DMA in a synchronous
> callback.
Well, at least for PAPR H_PUT_TCE, cancellation must be synchronous, ie
the hypercall must not return
On 06/19/2012 01:39 AM, Benjamin Herrenschmidt wrote:
From: David Gibson
One new complication raised by IOMMU support over only handling DMA
directly to physical addresses is handling dma_memory_map() case
(replacing cpu_physical_memory_map()) when the IOMMU translation the
IOVAs covered by such
From: David Gibson
One new complication raised by IOMMU support over only handling DMA
directly to physical addresses is handling dma_memory_map() case
(replacing cpu_physical_memory_map()) when the IOMMU translation the
IOVAs covered by such a map are invalidated or changed while the map
is acti
From: David Gibson
One new complication raised by IOMMU support over only handling DMA
directly to physical addresses is handling dma_memory_map() case
(replacing cpu_physical_memory_map()) when the IOMMU translation the
IOVAs covered by such a map are invalidated or changed while the map
is acti
