From mboxrd@z Thu Jan  1 00:00:00 1970
From: Alexander Duyck <alexander.duyck@gmail.com>
Subject: Re: Having troubles binding an SR-IOV VF to
 uio_pci_generic on Amazon instance
Date: Thu, 1 Oct 2015 14:17:49 -0700
Message-ID: <560DA2FD.1050507@gmail.com>
References: <560BDE24.8000308@scylladb.com>
 <20150930165359-mutt-send-email-mst@redhat.com>
 <560BF782.4070308@scylladb.com>
 <20150930175848-mutt-send-email-mst@redhat.com>
 <560C0171.7080507@scylladb.com> <20150930204016.GA29975@redhat.com>
 <20151001113828-mutt-send-email-mst@redhat.com> <560CF44A.60102@scylladb.com>
 <20151001120027-mutt-send-email-mst@redhat.com>
 <560CFB66.5050904@scylladb.com>
 <20151001123335-mutt-send-email-mst@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
Cc: "dev@dpdk.org" <dev@dpdk.org>
To: "Michael S. Tsirkin" <mst@redhat.com>, Avi Kivity <avi@scylladb.com>
Return-path: <dev-bounces@dpdk.org>
Received: from mail-pa0-f51.google.com (mail-pa0-f51.google.com
 [209.85.220.51]) by dpdk.org (Postfix) with ESMTP id 827795A86
 for <dev@dpdk.org>; Thu,  1 Oct 2015 23:17:51 +0200 (CEST)
Received: by pacfv12 with SMTP id fv12so87158179pac.2
 for <dev@dpdk.org>; Thu, 01 Oct 2015 14:17:50 -0700 (PDT)
In-Reply-To: <20151001123335-mutt-send-email-mst@redhat.com>
List-Id: patches and discussions about DPDK <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>

On 10/01/2015 02:42 AM, Michael S. Tsirkin wrote:
> On Thu, Oct 01, 2015 at 12:22:46PM +0300, Avi Kivity wrote:
>> even when they are some users
>> prefer to avoid the performance penalty.
> I don't think there's a measureable penalty from passing through the
> IOMMU, as long as mappings are mostly static (i.e. iommu=pt).  I sure
> never saw any numbers that show such.

It depends on the IOMMU.  I believe Intel had a performance penalty on 
all CPUs prior to Ivy Bridge.  Since then things have improved to where 
they are comparable to bare metal.

The graph on page 5 of 
https://networkbuilders.intel.com/docs/Network_Builders_RA_vBRAS_Final.pdf 
shows the penalty clear as day.  Pretty much anything before Ivy Bridge 
w/ small packets is slowed to a crawl with an IOMMU enabled.

- Alex