From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vlad Yasevich Subject: Re: macvtap performance regression (bisected) between 3.13 and 3.14-rc1 Date: Fri, 28 Feb 2014 17:14:58 -0500 Message-ID: <53110A62.7070109@redhat.com> References: <530FA586.3010400@de.ibm.com> Reply-To: vyasevic@redhat.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: "David S. Miller" , Jason Wang , "Michael S. Tsirkin" , netdev@vger.kernel.org, KVM list , Linux Kernel Mailing List To: Christian Borntraeger Return-path: In-Reply-To: <530FA586.3010400@de.ibm.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On 02/27/2014 03:52 PM, Christian Borntraeger wrote: > Vlad, > > commit 6acf54f1cf0a6747bac9fea26f34cfc5a9029523 > macvtap: Add support of packet capture on macvtap device. > > causes a performance regression for iperf traffic between two KVM guests > on my s390 system. Both guests are connected via two macvtaps on the same OSA > network card. > Before that patch I get ~20 Gbit/sec between two guests, afterwards I get > ~4Gbit/sec > > Latency seems to be unchanges (uperf 1byte ping pong). > > According to ifconfig in the guest, I have ~ 1500 bytes per packet with this > patch and ~ 40000 bytes without. So for some reason this patch causes the > network stack to do segmentation. (the guest kernel stays the same, only host > kernel is changed). > > Any ideas? I am looking. It shouldn't cause addition segmentations and when I ran netperf on the code I didn't see any difference in the throughput. -vlad > > > Christian >