Anthony Liguori wrote:
While it has served us well, it is long overdue that we eliminate the virtio-net tap hack. It turns out that zero-copy has very little impact on performance. The tap hack was gaining such a significant performance boost not because of zero-copy, but because it avoided dropping packets on receive which is apparently a significant problem with the tap implementation in QEMU.
FWIW, attached is a pretty straight forward zero-copy patch. What's interesting is that I see no change in throughput using this patch. The CPU is pegged at 100% during the iperf run. Since we're still using small MTUs, this isn't surprising. Copying a 1500 byte packet that we have to bring into the cache anyway doesn't seem significant. I think zero-copy will be more important with GSO though.
Regards, Anthony Liguori
Patches 3 and 4 in this series address the packet dropping issue and the net result is a 25% boost in RX performance even in the absence of zero-copy. Also worth mentioning, is that this makes merging virtio into upstream QEMU significantly easier. Signed-off-by: Anthony Liguori <[EMAIL PROTECTED]>
net-tap-zero-copy.patch
Description: application/mbox
------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________ kvm-devel mailing list kvm-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/kvm-devel
