Re: [Xen-ia64-devel] [RFC] gnttab unmap_and_replace

[Alex Williamson <alex.williamson@xxxxxx>]

On Tue, 2007-06-05 at 23:27 +0900, Isaku Yamahata wrote:
> # HG changeset patch
> # User yamahata@xxxxxxxxxxxxx
> # Date 1181035451 -32400
> # Node ID cb024947fa0186c5e08a71ecd4c0550967c5509a
> # Parent  3d3656528c8fa62896b744d2441033ff7d72a122
> support dma tracking for swiotlb and xen_dma
> PATCHNAME: dma_tracking_swiotlb_xen_dma
> 

   My main concern on this one is the places where we call
gnttab_dma_map_page() immediately followed by gnttab_dma_unmap_page().
This isn't really all that much overhead, but it seems like a kludge.
Maybe we could call page_to_bus(), evaluate whether the address is
dma'able by the device, if it is, invent a new gnttab interface that
won't do the hypercall lookup again.  Thanks,

        Alex

-- 
Alex Williamson                             HP Open Source & Linux Org.


_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel