From: Ming Lin <mlin@kernel.org>
To: linux-nvme@lists.infradead.org, qemu-devel@nongnu.org
Cc: fes@google.com, keith.busch@intel.com, tytso@mit.edu,
nab@linux-iscsi.org, virtualization@lists.linux-foundation.org,
axboe@fb.com, digitaleric@google.com,
Rob Nelson <rlnelson@google.com>, Christoph Hellwig <hch@lst.de>,
Mihai Rusu <dizzy@google.com>
Subject: [Qemu-devel] [RFC PATCH 0/2] Google extension to improve qemu-nvme performance
Date: Tue, 17 Nov 2015 21:47:02 -0800 [thread overview]
Message-ID: <1447825624-17011-1-git-send-email-mlin@kernel.org> (raw)
Hi Rob & Mihai,
I wrote vhost-nvme patches on top of Christoph's NVMe target.
vhost-nvme still uses mmio. So the guest OS can run unmodified NVMe
driver. But the tests I have done didn't show competitive performance
compared to virtio-blk/virtio-scsi. The bottleneck is in mmio. Your nvme
vendor extension patches reduces greatly the number of MMIO writes.
So I'd like to push it upstream.
I port these 2 patches to newer kernel and qemu.
I use ram disk as backend to compare performance.
qemu-nvme: 29MB/s
qemu-nvme+google-ext: 100MB/s
virtio-blk: 174MB/s
virtio-scsi: 118MB/s
I'll show you qemu-vhost-nvme+google-ext number later.
root@guest:~# cat test.job
[global]
bs=4k
ioengine=libaio
iodepth=64
direct=1
runtime=120
time_based
rw=randread
norandommap
group_reporting
gtod_reduce=1
numjobs=2
[job1]
filename=/dev/nvme0n1
#filename=/dev/vdb
#filename=/dev/sda
rw=read
Patches also available at:
kernel:
https://git.kernel.org/cgit/linux/kernel/git/mlin/linux.git/log/?h=nvme-google-ext
qemu:
http://www.minggr.net/cgit/cgit.cgi/qemu/log/?h=nvme-google-ext
Thanks,
Ming
next reply other threads:[~2015-11-18 5:47 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-11-18 5:47 Ming Lin [this message]
2015-11-18 5:47 ` [Qemu-devel] [PATCH -kernel] nvme: improve performance for virtual NVMe devices Ming Lin
2015-11-18 5:47 ` [Qemu-devel] [PATCH -qemu] nvme: support Google vendor extension Ming Lin
2015-11-19 10:37 ` Paolo Bonzini
2015-11-20 8:11 ` Ming Lin
2015-11-20 8:58 ` Paolo Bonzini
2015-11-20 23:05 ` Ming Lin
2015-11-21 12:56 ` Paolo Bonzini
2015-11-22 7:45 ` Ming Lin
2015-11-24 6:29 ` Ming Lin
2015-11-24 11:01 ` Paolo Bonzini
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=1447825624-17011-1-git-send-email-mlin@kernel.org \
--to=mlin@kernel.org \
--cc=axboe@fb.com \
--cc=digitaleric@google.com \
--cc=dizzy@google.com \
--cc=fes@google.com \
--cc=hch@lst.de \
--cc=keith.busch@intel.com \
--cc=linux-nvme@lists.infradead.org \
--cc=nab@linux-iscsi.org \
--cc=qemu-devel@nongnu.org \
--cc=rlnelson@google.com \
--cc=tytso@mit.edu \
--cc=virtualization@lists.linux-foundation.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).