All of lore.kernel.org
 help / color / mirror / Atom feed
From: Liu Yuan <namei.unix@gmail.com>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>,
	Rusty Russell <rusty@rustcorp.com.au>,
	Avi Kivity <avi@redhat.com>,
	kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
	Khoa Huynh <khoa@us.ibm.com>,
	Badari Pulavarty <pbadari@us.ibm.com>,
	Christoph Hellwig <hch@infradead.org>
Subject: Re: [RFC PATCH]vhost-blk: In-kernel accelerator for virtio block device
Date: Fri, 29 Jul 2011 22:45:46 +0800	[thread overview]
Message-ID: <4E32C79A.80606@gmail.com> (raw)
In-Reply-To: <CAJSP0QUbZ=RAkn_kbDupeA9QYYnzGd+u5M2duEg==e8ACjAqXg@mail.gmail.com>

On 07/29/2011 08:50 PM, Stefan Hajnoczi wrote:
> I hit a weirdness yesterday, just want to mention it in case you notice it too.
>
> When running vanilla qemu-kvm I forgot to use aio=native.  When I
> compared the results against virtio-blk-data-plane (which *always*
> uses Linux AIO) I was surprised to find average 4k read latency was
> lower and the standard deviation was also lower.
>
> So from now on I will run tests both with and without aio=native.
> aio=native should be faster and if I can reproduce the reverse I'll
> try to figure out why.
>
> Stefan
On my laptop, I don't meet this weirdo. the emulated POSIX AIO is much 
worse than the Linux AIO as expected. If iodepth goes deeper, the gap 
gets wider.

If not set aio=none, qemu uses emulated posix aio interface to do the 
IO. I peek at the posix-aio-compat.c,it uses thread pool and sync 
preadv/pwritev to emulate the AIO behaviour. The sync IO interface would 
even cause much poorer performance for random rw, since io-scheduler 
would possibly never get a chance to merge the requests stream. 
(blk_finish_plug->queue_unplugged->__blk_run_queue)

Yuan

  reply	other threads:[~2011-07-29 14:45 UTC|newest]

Thread overview: 54+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-28 14:29 [RFC PATCH]vhost-blk: In-kernel accelerator for virtio block device Liu Yuan
2011-07-28 14:29 ` [RFC PATCH] vhost-blk: An in-kernel accelerator for virtio-blk Liu Yuan
2011-07-28 14:47   ` Christoph Hellwig
2011-07-29 11:19     ` Liu Yuan
2011-07-28 15:18   ` Stefan Hajnoczi
2011-07-28 15:22   ` Michael S. Tsirkin
2011-07-29 15:09     ` Liu Yuan
2011-08-01  6:25     ` Liu Yuan
2011-08-01  8:12       ` Michael S. Tsirkin
2011-08-01  8:55         ` Liu Yuan
2011-08-01 10:26           ` Michael S. Tsirkin
2011-08-11 19:59     ` Dongsu Park
2011-08-12  8:56       ` Alan Cox
2011-07-28 14:29 ` [RFC PATCH] vhost: Enable vhost-blk support Liu Yuan
2011-07-28 15:44 ` [RFC PATCH]vhost-blk: In-kernel accelerator for virtio block device Stefan Hajnoczi
2011-07-29  4:48   ` Stefan Hajnoczi
2011-07-29  7:59     ` Liu Yuan
2011-07-29 10:55       ` Christoph Hellwig
2011-07-29  7:22   ` Liu Yuan
2011-07-29  9:06     ` Stefan Hajnoczi
2011-07-29 12:01       ` Liu Yuan
2011-07-29 12:29         ` Stefan Hajnoczi
2011-07-29 12:50           ` Stefan Hajnoczi
2011-07-29 14:45             ` Liu Yuan [this message]
2011-07-29 14:50               ` Liu Yuan
2011-07-29 15:25         ` Sasha Levin
2011-08-01  8:17           ` Avi Kivity
2011-08-01  9:18             ` Liu Yuan
2011-08-01  9:37               ` Avi Kivity
2011-07-29 18:12     ` Badari Pulavarty
2011-08-01  5:46       ` Liu Yuan
2011-08-01  8:12         ` Christoph Hellwig
2011-08-04 21:58         ` Badari Pulavarty
2011-08-05  7:56           ` Liu Yuan
2011-08-05 11:04           ` Liu Yuan
2011-08-05 18:02             ` Badari Pulavarty
2011-08-08  1:35               ` Liu Yuan
2011-08-08  5:04                 ` Badari Pulavarty
2011-08-08  7:31                   ` Liu Yuan
2011-08-08 17:16                     ` Badari Pulavarty
2011-08-10  2:19                       ` Liu Yuan
2011-08-10 20:37                         ` Badari Pulavarty
2011-08-11  3:01                           ` Liu Yuan
2011-08-11  3:19                             ` Liu Yuan
2011-08-11 23:51                               ` Badari Pulavarty
2011-08-12  4:50                               ` Badari Pulavarty
2011-08-12  6:46                                 ` Dongsu Park
2011-08-12  8:27                                 ` Liu Yuan
2011-08-12 11:40                                   ` Liu Yuan
2011-08-12 16:12                                     ` Badari Pulavarty
2011-08-15  3:20                                       ` Liu Yuan
2011-08-15  4:17                                         ` Badari Pulavarty
2011-08-16  5:44                                           ` Liu Yuan
2011-09-07 13:36                                           ` Liu Yuan

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=4E32C79A.80606@gmail.com \
    --to=namei.unix@gmail.com \
    --cc=avi@redhat.com \
    --cc=hch@infradead.org \
    --cc=khoa@us.ibm.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mst@redhat.com \
    --cc=pbadari@us.ibm.com \
    --cc=rusty@rustcorp.com.au \
    --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 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.