From mboxrd@z Thu Jan  1 00:00:00 1970
From: Avi Kivity <avi@qumranet.com>
Subject: Re: PCI passthrough with VT-d - native performance
Date: Wed, 16 Jul 2008 18:22:33 +0300
Message-ID: <487E1239.6080909@qumranet.com>
References: <1216214225-18030-1-git-send-email-benami@il.ibm.com>	 <487E076D.4050306@qumranet.com> <1216221530.31546.274.camel@cluwyn.haifa.ibm.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8;
	format=flowed
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: amit.shah@qumranet.com, kvm@vger.kernel.org,
	Muli Ben-Yehuda <muli@il.ibm.com>, weidong.han@intel.com,
	anthony@codemonkey.ws
To: Ben-Ami Yassour <benami@il.ibm.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from il.qumranet.com ([212.179.150.194]:30279 "EHLO il.qumranet.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1760448AbYGPPWf (ORCPT <rfc822;kvm@vger.kernel.org>);
	Wed, 16 Jul 2008 11:22:35 -0400
In-Reply-To: <1216221530.31546.274.camel@cluwyn.haifa.ibm.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

Ben-Ami Yassour wrote:
>>>  (with higher CPU utilization).
>>>  =20
>>>      =20
>> How much higher?
>>    =20
>
> Here are some numbers for running iperf -l 1M:
>
> e1000 NIC (behind a PCI bridge)
>                        Bandwidth (Mbit/sec)    CPU utilization
> Native OS                   771                      18%
> Native OS with VT-d         760                      18%=20
> KVM VT-d                    390                      95%=20
> =EF=BB=BFKVM VT-d with direct mmio   770                      84%
> KVM emulated                 57                     100% =20
>
> Comment: its not clear to me why the native linux can not get closer =
to 1G for this NIC,
> (I verified that its not external network issues). But clearly we sho=
uldn't hope to=20
> get more then the host does with a KVM guest (especially if the guest=
 and host are the=20
> same OS as in this case...).
>
> e1000e NIC (onboard)
>                        Bandwidth (Mbit/sec)    CPU utilization
> Native OS                   915                      18%
> Native OS with VT-d         915                      18%
> =EF=BB=BFKVM VT-d with direct mmio   914                      98%
>
> Clearly we need to try and improve the CPU utilization, but I think t=
hat this is good enough=20
> for the first phase.
>
>  =20

Agree;  part of the higher utilization is of course not the fault of th=
e=20
device assignment code, rather it is ordinary virtualization overhead. =
=20
We'll have to tune this.

--=20
error compiling committee.c: too many arguments to function