[dpdk-dev] [PATCH v2] vhost: adaptively batch small guest memory copies

Tiwei Bie tiwei.bie at intel.com
Mon Sep 11 03:27:27 CEST 2017


On Sat, Sep 09, 2017 at 08:28:57PM +0530, santosh wrote:
> On Friday 08 September 2017 06:20 PM, 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 patch improves the performance for small packets, and has
> > distinguished the packets by size. So although the performance
> > for big packets doesn't change, it makes it relatively easy to
> > do some special optimizations for the big packets too.
> >
> > Signed-off-by: Tiwei Bie <tiwei.bie at intel.com>
> > Signed-off-by: Zhihong Wang <zhihong.wang at intel.com>
> > Signed-off-by: Zhiyong Yang <zhiyong.yang at intel.com>
> > Reviewed-by: Maxime Coquelin <maxime.coquelin at redhat.com>
> > ---
> > This optimization depends on the CPU internal pipeline design.
> > So further tests (e.g. ARM) from the community is appreciated.
> >
> > v2: s/burst_copy/batch_copy/g
> 
> In pvp setup for arm64 platform, no performance drop noticed with this patch but no significant performance gain either. Tested-by: Santosh Shukla <santosh.shukla at caviumnetworks.com>
> 

Thank you very much for testing this patch!

Best regards,
Tiwei Bie


More information about the dev mailing list