From: Shirley Ma <mashirle@us.ibm.com>
To: xiaohui.xin@intel.com
Cc: netdev@vger.kernel.org, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, mst@redhat.com, mingo@elte.hu,
davem@davemloft.net, herbert@gondor.hengli.com.au,
jdike@linux.intel.com
Subject: Re: [RFC PATCH v9 00/16] Provide a zero-copy method on KVM virtio-net.
Date: Tue, 10 Aug 2010 18:23:24 -0700 [thread overview]
Message-ID: <1281489804.3391.23.camel@localhost.localdomain> (raw)
In-Reply-To: <1281086624-5765-1-git-send-email-xiaohui.xin@intel.com>
Hello Xiaohui,
On Fri, 2010-08-06 at 17:23 +0800, xiaohui.xin@intel.com wrote:
> Our goal is to improve the bandwidth and reduce the CPU usage.
> Exact performance data will be provided later.
Have you had any performance data to share here? I tested my
experimental macvtap zero copy for TX only. The performance I have seen
as below without any tuning, (default setting):
Before: netperf 16K message size results with 60 secs run is 7.5Gb/s
over ixgbe 10GbE card. perf top shows:
2103.00 12.9% copy_user_generic_string
1541.00 9.4% handle_tx
1490.00 9.1% _raw_spin_unlock_irqrestore
1361.00 8.3% _raw_spin_lock_irqsave
1288.00 7.9% _raw_spin_lock
924.00 5.7% vhost_worker
After: netperf results with 60 secs run is 8.1Gb/s, perf output:
1093.00 9.9% _raw_spin_unlock_irqrestore
1048.00 9.5% handle_tx
934.00 8.5% _raw_spin_lock_irqsave
864.00 7.9% _raw_spin_lock
644.00 5.9% vhost_worker
387.00 3.5% use_mm
I am still working on collecting more data (latency, cpu
utilization...). I will let you know once I get all data for macvtap TX
zero copy. Also I found some vhost performance regression on the new
kernel with tuning. I used to get 9.4Gb/s, now I couldn't get it.
Shirley
WARNING: multiple messages have this Message-ID (diff)
From: Shirley Ma <mashirle@us.ibm.com>
To: xiaohui.xin@intel.com
Cc: netdev@vger.kernel.org, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, mst@redhat.com, mingo@elte.hu,
davem@davemloft.net, herbert@gondor.apana.org.au,
jdike@linux.intel.com
Subject: Re: [RFC PATCH v9 00/16] Provide a zero-copy method on KVM virtio-net.
Date: Tue, 10 Aug 2010 18:23:24 -0700 [thread overview]
Message-ID: <1281489804.3391.23.camel@localhost.localdomain> (raw)
In-Reply-To: <1281086624-5765-1-git-send-email-xiaohui.xin@intel.com>
Hello Xiaohui,
On Fri, 2010-08-06 at 17:23 +0800, xiaohui.xin@intel.com wrote:
> Our goal is to improve the bandwidth and reduce the CPU usage.
> Exact performance data will be provided later.
Have you had any performance data to share here? I tested my
experimental macvtap zero copy for TX only. The performance I have seen
as below without any tuning, (default setting):
Before: netperf 16K message size results with 60 secs run is 7.5Gb/s
over ixgbe 10GbE card. perf top shows:
2103.00 12.9% copy_user_generic_string
1541.00 9.4% handle_tx
1490.00 9.1% _raw_spin_unlock_irqrestore
1361.00 8.3% _raw_spin_lock_irqsave
1288.00 7.9% _raw_spin_lock
924.00 5.7% vhost_worker
After: netperf results with 60 secs run is 8.1Gb/s, perf output:
1093.00 9.9% _raw_spin_unlock_irqrestore
1048.00 9.5% handle_tx
934.00 8.5% _raw_spin_lock_irqsave
864.00 7.9% _raw_spin_lock
644.00 5.9% vhost_worker
387.00 3.5% use_mm
I am still working on collecting more data (latency, cpu
utilization...). I will let you know once I get all data for macvtap TX
zero copy. Also I found some vhost performance regression on the new
kernel with tuning. I used to get 9.4Gb/s, now I couldn't get it.
Shirley
next prev parent reply other threads:[~2010-08-11 1:23 UTC|newest]
Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-08-06 9:23 [RFC PATCH v9 00/16] Provide a zero-copy method on KVM virtio-net xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 01/16] Add a new structure for skb buffer from external xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 02/16] Add a new struct for device to manipulate external buffer xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 03/16] Add a ndo_mp_port_prep func to net_device_ops xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 04/16] Add a function make external buffer owner to query capability xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 05/16] Add a function to indicate if device use external buffer xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 06/16] Use callback to deal with skb_release_data() specially xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 07/16] Modify netdev_alloc_page() to get external buffer xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 08/16] Modify netdev_free_page() to release " xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 09/16] Don't do skb recycle, if device use " xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 10/16] Add a hook to intercept external buffers from NIC driver xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 11/16] Add header file for mp device xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 13/16] Add a kconfig entry and make entry " xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 12/16] Add mp(mediate passthru) device xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 14/16] Provides multiple submits and asynchronous notifications xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 15/16] An example how to modifiy NIC driver to use napi_gro_frags() interface xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-08-06 9:23 ` [RFC PATCH v9 16/16] An example how to alloc user buffer based on " xiaohui.xin
2010-08-06 9:23 ` xiaohui.xin
2010-09-06 11:11 ` [RFC PATCH v9 12/16] Add mp(mediate passthru) device Michael S. Tsirkin
2010-09-10 13:40 ` Xin, Xiaohui
2010-09-11 7:41 ` Xin, Xiaohui
2010-09-12 13:37 ` Michael S. Tsirkin
2010-09-15 3:13 ` Xin, Xiaohui
2010-09-15 11:28 ` Michael S. Tsirkin
2010-09-17 3:16 ` Xin, Xiaohui
2010-09-20 8:08 ` xiaohui.xin
2010-09-20 11:36 ` Michael S. Tsirkin
2010-09-21 1:39 ` Xin, Xiaohui
2010-09-21 13:14 ` Michael S. Tsirkin
2010-09-22 11:41 ` Xin, Xiaohui
2010-09-22 11:55 ` Michael S. Tsirkin
2010-09-23 12:56 ` Xin, Xiaohui
2010-09-26 11:50 ` Michael S. Tsirkin
2010-09-27 0:42 ` Xin, Xiaohui
2010-09-11 9:42 ` Xin, Xiaohui
2010-08-11 1:23 ` Shirley Ma [this message]
2010-08-11 1:23 ` [RFC PATCH v9 00/16] Provide a zero-copy method on KVM virtio-net Shirley Ma
2010-08-11 1:43 ` Shirley Ma
2010-08-11 1:43 ` Shirley Ma
2010-08-11 6:01 ` Shirley Ma
2010-08-11 6:01 ` Shirley Ma
2010-08-11 6:55 ` Shirley Ma
2010-08-11 6:55 ` Shirley Ma
2010-09-03 10:52 ` Michael S. Tsirkin
2010-09-13 18:48 ` Shirley Ma
2010-09-13 21:35 ` Shirley Ma
2010-09-03 10:14 ` Michael S. Tsirkin
2010-09-03 20:29 ` Sridhar Samudrala
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=1281489804.3391.23.camel@localhost.localdomain \
--to=mashirle@us.ibm.com \
--cc=davem@davemloft.net \
--cc=herbert@gondor.hengli.com.au \
--cc=jdike@linux.intel.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=mst@redhat.com \
--cc=netdev@vger.kernel.org \
--cc=xiaohui.xin@intel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.