On 05/20/2010 01:33 AM, Michael S. Tsirkin wrote:
Virtio is already way too bouncy due to the indirection between the
avail/used rings and the descriptor pool. A device with out of order
completion (like virtio-blk) will quickly randomize the unused
descriptor indexes, so every descriptor fetch will require a bounce.
In contrast, if the rings hold the descriptors themselves instead of
pointers, we bounce (sizeof(descriptor)/cache_line_size) cache lines for
every descriptor, amortized.
On the other hand, consider that on fast path we are never using all
of the ring. With a good allocator we might be able to keep
reusing only small part of the ring, instead of wrapping around
all of it all of the time.
It's still suboptimal, we have to bounce both the avail/used rings and
the descriptor pool, compared to just the descriptor ring with a direct
design. Plus we don't need a fancy allocator.
When amortizing cachelines, simple data structures win.
--
Do not meddle in the internals of kernels, for they are subtle and quick to
panic.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
