On Tue, 16 Feb 2016 12:13:12 +0000
"Stokes, Ian" <ian.sto...@intel.com> wrote:

> > -----Original Message-----
> > From: Flavio Leitner [mailto:f...@sysclose.org]
> > Sent: Friday, February 12, 2016 1:47 AM
> > To: Stokes, Ian
> > Cc: dev@openvswitch.org
> > Subject: Re: [ovs-dev] [PATCH] INSTALL.DPDK.md: Correct mergeable
> > buffers parameter.
> > 
> > On Wed, 10 Feb 2016 10:50:54 +0000
> > Ian Stokes <ian.sto...@intel.com> wrote:
> >   
> > > Update the mergeable buffers paramaters in performance tuning to the
> > > correct parameter mrg_rxbuf.
> > >
> > > Signed-off-by: Ian Stokes <ian.sto...@intel.com>
> > > ---
> > >  INSTALL.DPDK.md |    2 +-
> > >  1 files changed, 1 insertions(+), 1 deletions(-)
> > >
> > > diff --git a/INSTALL.DPDK.md b/INSTALL.DPDK.md index d892788..0f340e5
> > > 100644
> > > --- a/INSTALL.DPDK.md
> > > +++ b/INSTALL.DPDK.md
> > > @@ -392,7 +392,7 @@ Performance Tuning:
> > >   This behavior is typically supported and enabled by default,  
> > however  
> > >   in the case where the user knows that rx mergeable buffers are not  
> > needed  
> > >   i.e. jumbo frames are not needed, it can be forced off by adding
> > > - rx_mrgbuf=off to the QEMU command line options. By not reserving  
> > multiple  
> > > + mrg_rxbuf=off to the QEMU command line options. By not reserving
> > > +multiple
> > >   chains of descriptors it will make more individual virtio  
> > descriptors  
> > >   available for rx to the guest using dpdkvhost ports and this can  
> > improve  
> > >   performance.  
> > 
> > $ qemu-system-x86_64 -device virtio-net-pci,help 2>&1 | grep mrg
> > virtio-net-pci.mrg_rxbuf=bool (on/off)
> >   
> 
> Thanks Flavio, that command is very useful. 
> Do you think it should be added to the patch or was it meant as an FYI?

That isn't relevant to the documentation.  It's just how I verified that
the patch was doing the right thing.

Thanks,
-- 
fbl

_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev

Reply via email to