[dpdk-dev] [PATCH] vhost: adaptively batch small guest memory copies

[Yao, Lei A]

> -----Original Message-----
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Jens Freimann
> Sent: Monday, August 28, 2017 2:31 PM
> To: Bie, Tiwei <tiwei.bie at intel.com>
> Cc: dev at dpdk.org; yliu at fridaylinux.org; maxime.coquelin at redhat.com;
> Wang, Zhihong <zhihong.wang at intel.com>; Yang, Zhiyong
> <zhiyong.yang at intel.com>
> Subject: Re: [dpdk-dev] [PATCH] vhost: adaptively batch small guest memory
> copies
> 
> Hi Tiwei,
> 
> On Thu, Aug 24, 2017 at 10:19:39AM +0800, Tiwei Bie wrote:
> >This patch adaptively batches the small guest memory copies.
> >By batching the small copies, the efficiency of executing the
> >memory LOAD instructions can be improved greatly, because the
> >memory LOAD latency can be effectively hidden by the pipeline.
> >We saw great performance boosts for small packets PVP test.
> 
> this sounds interesting. Do you have numbers showing how much
> performance improved for small packets?
> 
> regards,
> Jens
Hi, Jens

On my test bench, the performance gain as following for 
64B small packets
Mergeable:         19%
Vector        ：      21%
No-mergeable:  21%

CPU info: Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz
OS: Ubuntu 16.04