On Wed, Sep 14, 2016 at 04:17:26PM +0800, Peter Xu wrote:
> On Wed, Sep 14, 2016 at 05:15:03PM +1000, David Gibson wrote:
> 
> [...]
> 
> > Peter suggested commenting this next to the IOTLBEntry definition, and
> > I think that's probably ok for now.
> 
> Looks like we have something already (just not that obvious):
> 
> /**
>  * memory_region_notify_iommu: notify a change in an IOMMU translation entry.
>  *
>  * @mr: the memory region that was changed
>  * @entry: the new entry in the IOMMU translation table.  The entry
>  *         replaces all old entries for the same virtual I/O address range.
>  *         Deleted entries have .@perm == 0.
>  */
> void memory_region_notify_iommu(MemoryRegion *mr,
>                                 IOMMUTLBEntry entry);
> 
> Though it's quite simple, it did explain that perm==0 is for deleted
> entries.

That is definitely not sufficient.  It misses the crucial point is
that @perm != 0 is NOT ALLOWED if there was an existing mapping at
that address.

It's not that we really _need_ to support in-place change.  The point
is that we want to keep our interface contracts simple and clear.

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson