Xen devel - Jul 2007 - question on c/s 15181

Herbert, are the changes below to use in_swiotlb_aperture() strictly necessary?
They get us further away from native code, so I wonder if, for the purpose of
being as close as possible to native, these could be reverted (in a larger patch
that would get us closer to native in many other respects).

Jan

********************************************************

@@ -605,10 +610,12 @@ swiotlb_unmap_sg(struct device *hwdev, s
 	BUG_ON(dir == DMA_NONE);
 
 	for (i = 0; i < nelems; i++, sg++)
-		if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
+		if (in_swiotlb_aperture(sg->dma_address))
 			unmap_single(hwdev, 
 				     (void *)bus_to_virt(sg->dma_address),
 				     sg->dma_length, dir);
+		else
+			gnttab_dma_unmap_page(sg->dma_address);
 }
 
 /*
@@ -627,7 +634,7 @@ swiotlb_sync_sg_for_cpu(struct device *h
 	BUG_ON(dir == DMA_NONE);
 
 	for (i = 0; i < nelems; i++, sg++)
-		if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
+		if (in_swiotlb_aperture(sg->dma_address))
 			sync_single(hwdev,
 				    (void *)bus_to_virt(sg->dma_address),
 				    sg->dma_length, dir);
@@ -642,7 +649,7 @@ swiotlb_sync_sg_for_device(struct device
 	BUG_ON(dir == DMA_NONE);
 
 	for (i = 0; i < nelems; i++, sg++)
-		if (sg->dma_address != SG_ENT_PHYS_ADDRESS(sg))
+		if (in_swiotlb_aperture(sg->dma_address))
 			sync_single(hwdev,
 				    (void *)bus_to_virt(sg->dma_address),
 				    sg->dma_length, dir);



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

On Thu, Jul 05, 2007 at 01:53:27PM +0100, Jan Beulich
wrote:
> Herbert, are the changes below to use in_swiotlb_aperture() strictly
necessary?
> They get us further away from native code, so I wonder if, for the purpose
of
> being as close as possible to native, these could be reverted (in a larger
patch
> that would get us closer to native in many other respects).
For the sg case technically we can go back to the original code
since the underlying grant pages should not get copied once it
has been mapped.

In fact, if we had the strcut page pointer for the unmap_single
case this could be greatly simplified.

Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

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