From: "Michael S. Tsirkin" <mst@redhat.com>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: qemu-devel@nongnu.org
Subject: [Qemu-devel] Re: [PATCH v5 0/4] virtio: Use ioeventfd for virtqueue notify
Date: Sun, 12 Dec 2010 23:09:59 +0200 [thread overview]
Message-ID: <20101212210959.GA25136@redhat.com> (raw)
In-Reply-To: <20101212205634.GA24986@redhat.com>
On Sun, Dec 12, 2010 at 10:56:34PM +0200, Michael S. Tsirkin wrote:
> On Sun, Dec 12, 2010 at 10:42:28PM +0200, Michael S. Tsirkin wrote:
> > On Sun, Dec 12, 2010 at 10:41:28PM +0200, Michael S. Tsirkin wrote:
> > > On Sun, Dec 12, 2010 at 03:02:04PM +0000, Stefan Hajnoczi wrote:
> > > > See below for the v5 changelog.
> > > >
> > > > Due to lack of connectivity I am sending from GMail. Git should retain my
> > > > stefanha@linux.vnet.ibm.com From address.
> > > >
> > > > Virtqueue notify is currently handled synchronously in userspace virtio. This
> > > > prevents the vcpu from executing guest code while hardware emulation code
> > > > handles the notify.
> > > >
> > > > On systems that support KVM, the ioeventfd mechanism can be used to make
> > > > virtqueue notify a lightweight exit by deferring hardware emulation to the
> > > > iothread and allowing the VM to continue execution. This model is similar to
> > > > how vhost receives virtqueue notifies.
> > > >
> > > > The result of this change is improved performance for userspace virtio devices.
> > > > Virtio-blk throughput increases especially for multithreaded scenarios and
> > > > virtio-net transmit throughput increases substantially.
> > >
> > > Interestingly, I see decreased throughput for small message
> > > host to get netperf runs.
> > >
> > > The command that I used was:
> > > netperf -H $vguest -- -m 200
> > >
> > > And the results are:
> > > - with ioeventfd=off
> > > TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 11.0.0.104 (11.0.0.104) port 0 AF_INET : demo
> > > Recv Send Send Utilization Service Demand
> > > Socket Socket Message Elapsed Send Recv Send Recv
> > > Size Size Size Time Throughput local remote local remote
> > > bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB
> > >
> > > 87380 16384 200 10.00 3035.48 15.50 99.30 6.695 2.680
> > >
> > > - with ioeventfd=on
> > > TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 11.0.0.104 (11.0.0.104) port 0 AF_INET : demo
> > > Recv Send Send Utilization Service Demand
> > > Socket Socket Message Elapsed Send Recv Send Recv
> > > Size Size Size Time Throughput local remote local remote
> > > bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB
> > >
> > > 87380 16384 200 10.00 1770.95 18.16 51.65 13.442 2.389
> > >
> > >
> > > Do you see this behaviour too?
> >
> > Just a note: this is with the patchset ported to qemu-kvm.
>
> And just another note: the trend is reversed for larged messages,
> e.g. with 1.5k messages ioeventfd=on outputforms ioeventfd=off.
Another datapoint where I see a regression is with 4000 byte messages
for guest to host traffic.
ioeventfd=off
set_up_server could not establish a listen endpoint for port 12865 with family AF_UNSPEC
TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 11.0.0.4 (11.0.0.4) port 0 AF_INET : demo
Recv Send Send Utilization Service Demand
Socket Socket Message Elapsed Send Recv Send Recv
Size Size Size Time Throughput local remote local remote
bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB
87380 16384 4000 10.00 7717.56 98.80 15.11 1.049 2.566
ioeventfd=on
TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 11.0.0.4 (11.0.0.4) port 0 AF_INET : demo
Recv Send Send Utilization Service Demand
Socket Socket Message Elapsed Send Recv Send Recv
Size Size Size Time Throughput local remote local remote
bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB
87380 16384 4000 10.00 3965.86 87.69 15.29 1.811 5.055
--
MST
next prev parent reply other threads:[~2010-12-12 21:10 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-12-12 15:02 [Qemu-devel] [PATCH v5 0/4] virtio: Use ioeventfd for virtqueue notify Stefan Hajnoczi
2010-12-12 15:02 ` [Qemu-devel] [PATCH v5 1/4] virtio-pci: Rename bugs field to flags Stefan Hajnoczi
2010-12-12 15:02 ` [Qemu-devel] [PATCH v5 2/4] virtio-pci: Use ioeventfd for virtqueue notify Stefan Hajnoczi
2011-01-24 18:54 ` Kevin Wolf
2011-01-24 19:36 ` Michael S. Tsirkin
2011-01-24 19:48 ` Kevin Wolf
2011-01-24 19:47 ` Michael S. Tsirkin
2011-01-24 20:05 ` Kevin Wolf
2011-01-25 7:12 ` Stefan Hajnoczi
2011-01-25 9:49 ` Stefan Hajnoczi
2011-01-25 9:54 ` Stefan Hajnoczi
2011-01-25 11:27 ` Michael S. Tsirkin
2011-01-25 13:20 ` Stefan Hajnoczi
2011-01-25 14:07 ` Stefan Hajnoczi
2011-01-25 19:18 ` Anthony Liguori
2011-01-25 19:45 ` Stefan Hajnoczi
2011-01-25 19:51 ` Anthony Liguori
2011-01-25 19:59 ` Stefan Hajnoczi
2011-01-26 0:18 ` Anthony Liguori
2010-12-12 15:02 ` [Qemu-devel] [PATCH v5 3/4] virtio-pci: Don't use ioeventfd on old kernels Stefan Hajnoczi
2010-12-12 15:02 ` [Qemu-devel] [PATCH v5 4/4] docs: Document virtio PCI -device ioeventfd=on|off Stefan Hajnoczi
2010-12-12 15:14 ` [Qemu-devel] Re: [PATCH v5 0/4] virtio: Use ioeventfd for virtqueue notify Stefan Hajnoczi
2010-12-12 20:41 ` Michael S. Tsirkin
2010-12-12 20:42 ` Michael S. Tsirkin
2010-12-12 20:56 ` Michael S. Tsirkin
2010-12-12 21:09 ` Michael S. Tsirkin [this message]
2010-12-13 10:24 ` Stefan Hajnoczi
2010-12-13 10:38 ` Michael S. Tsirkin
2010-12-13 13:11 ` Stefan Hajnoczi
2010-12-13 13:35 ` Michael S. Tsirkin
2010-12-13 13:36 ` Michael S. Tsirkin
2010-12-13 14:06 ` Stefan Hajnoczi
2010-12-13 15:27 ` Stefan Hajnoczi
2010-12-13 16:00 ` Michael S. Tsirkin
2010-12-13 16:29 ` Stefan Hajnoczi
2010-12-13 16:30 ` Michael S. Tsirkin
2010-12-13 16:12 ` Michael S. Tsirkin
2010-12-13 16:28 ` Stefan Hajnoczi
2010-12-13 17:57 ` Stefan Hajnoczi
2010-12-13 18:52 ` Michael S. Tsirkin
2010-12-15 11:42 ` Stefan Hajnoczi
2010-12-15 11:48 ` Stefan Hajnoczi
2010-12-15 12:00 ` Michael S. Tsirkin
2010-12-15 12:14 ` Michael S. Tsirkin
2010-12-15 12:59 ` Stefan Hajnoczi
2010-12-16 16:40 ` Stefan Hajnoczi
2010-12-16 23:39 ` Michael S. Tsirkin
2010-12-19 14:49 ` Michael S. Tsirkin
2011-01-06 16:41 ` Stefan Hajnoczi
2011-01-06 17:04 ` Michael S. Tsirkin
2011-01-06 18:00 ` Michael S. Tsirkin
2011-01-07 8:56 ` Stefan Hajnoczi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20101212210959.GA25136@redhat.com \
--to=mst@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).