From: Maxime Coquelin <maxime.coquelin@redhat.com>
To: dev@dpdk.org, Yuanhan Liu <yliu@fridaylinux.org>
Cc: mst@redhat.com, vkaplans@redhat.com, jasowang@redhat.com,
jfreiman@redhat.com
Subject: Re: [dpdk-dev] [RFC 00/19] Vhost-user: Implement device IOTLB support
Date: Thu, 31 Aug 2017 11:10:07 +0200 [thread overview]
Message-ID: <285244a0-9959-6401-6b61-cbd7b559530e@redhat.com> (raw)
In-Reply-To: <20170704094922.11405-1-maxime.coquelin@redhat.com>
Hi,
On 07/04/2017 11:49 AM, Maxime Coquelin wrote:
> This first RFC, which targets v17.11, adds support for
> VIRTIO_F_IOMMU_PLATFORM feature, by implementing device IOTLB in the
> vhost-user backend. It improves the guest safety by enabling the
> possibility to isolate the Virtio device.
>
> It makes possible to use Virtio PMD in guest with using VFIO driver
> without enable_unsafe_noiommu_mode parameter set, so that the DPDK
> application on guest can only access memory its has been allowed to,
> and preventing malicious/buggy DPDK application in guest to make
> vhost-user backend write random guest memory. Note that Virtio-net
> Kernel driver also support IOMMU.
>
> The series depends on Qemu's "vhost-user: Specify and implement
> device IOTLB support" [0], available upstream and which will be part
> of Qemu v2.10 release.
>
> Performance-wise, even if this RFC has still room for optimizations,
> no performance degradation is noticed with static mappings (i.e. DPDK
> on guest) with PVP benchmark:
> Traffic Generator: Moongen (lua-trafficgen)
> Acceptable Loss: 0.005%
> Validation run time: 1 min
> Guest DPDK version/commit: v17.05
> QEMU version/commit: master (6db174aed1fd)
> Virtio features: default
> CPU: Intel(R) Xeon(R) CPU E5-2667 v4 @ 3.20GHz
> NIC: 2 x X710
> Page size: 1G host/1G guest
> Results (bidirectional, total of the two flows):
> - base: 18.8Mpps
> - base + IOTLB series, IOMMU OFF: 18.8Mpps
> - base + IOTLB series, IOMMU ON: 18.8Mpps
It seems that I did a mistake when benchmarking with IOMMU on.
Actually, with this RFC, the result is 14.5Mpps, which is a noticeable
performance degradation.
Next revision fixing this issue is coming soon, performance is recovered
to 18.8Mpps.
Cheers,
Maxime
prev parent reply other threads:[~2017-08-31 9:10 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-07-04 9:49 Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 01/19] vhost: protect virtio_net device struct Maxime Coquelin
2017-07-05 10:07 ` Jens Freimann
2017-07-07 7:31 ` Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 02/19] Revert "vhost: workaround MQ fails to startup" Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 03/19] vhost: prepare send_vhost_message() to slave requests Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 04/19] vhost: add support to slave requests channel Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 05/19] vhost: declare missing IOMMU-related definitions for old kernels Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 06/19] vhost: add iotlb helper functions Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 07/19] vhost-user: add support to IOTLB miss slave requests Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 08/19] vhost: initialize vrings IOTLB caches Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 09/19] vhost: implement IOTLB events notification mechanism Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 10/19] vhost-user: handle IOTLB update and invalidate requests Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 11/19] vhost: introduce guest IOVA to backend VA helper Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 12/19] vhost: use the guest IOVA to host " Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 13/19] vhost: enable rings at the right time Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 14/19] vhost: don't dereference invalid dev pointer after its reallocation Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 15/19] vhost: postpone rings adresses translation Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 16/19] vhost-user: translate ring addresses when IOMMU enabled Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 17/19] vhost-user: iommu: postpone device creation until ring are mapped Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 18/19] vhost: iommu: Invalidate vring in case of matching IOTLB invalidate Maxime Coquelin
2017-07-04 9:49 ` [dpdk-dev] [RFC 19/19] vhost: enable IOMMU support Maxime Coquelin
2017-08-31 9:10 ` Maxime Coquelin [this message]
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=285244a0-9959-6401-6b61-cbd7b559530e@redhat.com \
--to=maxime.coquelin@redhat.com \
--cc=dev@dpdk.org \
--cc=jasowang@redhat.com \
--cc=jfreiman@redhat.com \
--cc=mst@redhat.com \
--cc=vkaplans@redhat.com \
--cc=yliu@fridaylinux.org \
/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).