Gerd Hoffmann wrote:
Andrea Arcangeli wrote:
On Fri, Dec 12, 2008 at 12:17:58PM -0600, Anthony Liguori wrote:
You assume that anything using bdrv_aio_readv/writev will be going through
a DMA API. This isn't a safe assumption.
Well it's obviously a safe assumption right now... ;)
Well, the xen block backend in my xen patch queue calls bdrv_* directly
right now. Dunno whenever it is possible/useful to fit the grant table
stuff into the upcoming dma api somehow.
I don't know about grant table references b/c that's really foreign
memory. But this is a good argument against the DMA as it stands, b/c
you may be handing foreign memory to bdrv_aio_readv/writev.
A big reason for the map/unmap lock/unlock abstraction though would be
the qemu-dm map cache. I think you could use it to pretty reasonably
integrate the map cache which I know I've previously could never be done :-)
Regards,
Anthony Liguori
cheers,
Gerd
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
