From: Avi Kivity <avi@qumranet.com>
To: qemu-devel@nongnu.org
Cc: kvm-devel@lists.sourceforge.net, Marcelo Tosatti <mtosatti@redhat.com>
Subject: Re: [kvm-devel] Re: [PATCH 1/3] Refactor AIO interface to allow other AIO implementations
Date: Tue, 22 Apr 2008 18:05:32 +0300 [thread overview]
Message-ID: <480DFEBC.1020504@qumranet.com> (raw)
In-Reply-To: <480DFBD9.4030802@codemonkey.ws>
Anthony Liguori wrote:
>>
>> If I submit sequential O_DIRECT reads with aio_read(), will they enter
>> the device read queue in the same order, and reach the disk in that
>> order (allowing for reordering when worthwhile by the elevator)?
>>
>
> There's no guarantee that any sort of order will be preserved by AIO
> requests. The same is true with writes. This is what fdsync is for,
> to guarantee ordering.
I believe he'd like a hint to get good scheduling, not a guarantee.
With a thread pool if the threads are scheduled out of order, so are
your requests. If the elevator doesn't plug the queue, the first few
requests may not be optimally sorted.
--
error compiling committee.c: too many arguments to function
WARNING: multiple messages have this Message-ID (diff)
From: Avi Kivity <avi@qumranet.com>
To: qemu-devel@nongnu.org
Cc: kvm-devel@lists.sourceforge.net, Marcelo Tosatti <mtosatti@redhat.com>
Subject: Re: [kvm-devel] [Qemu-devel] Re: [PATCH 1/3] Refactor AIO interface to allow other AIO implementations
Date: Tue, 22 Apr 2008 18:05:32 +0300 [thread overview]
Message-ID: <480DFEBC.1020504@qumranet.com> (raw)
In-Reply-To: <480DFBD9.4030802@codemonkey.ws>
Anthony Liguori wrote:
>>
>> If I submit sequential O_DIRECT reads with aio_read(), will they enter
>> the device read queue in the same order, and reach the disk in that
>> order (allowing for reordering when worthwhile by the elevator)?
>>
>
> There's no guarantee that any sort of order will be preserved by AIO
> requests. The same is true with writes. This is what fdsync is for,
> to guarantee ordering.
I believe he'd like a hint to get good scheduling, not a guarantee.
With a thread pool if the threads are scheduled out of order, so are
your requests. If the elevator doesn't plug the queue, the first few
requests may not be optimally sorted.
--
error compiling committee.c: too many arguments to function
next prev parent reply other threads:[~2008-04-22 15:05 UTC|newest]
Thread overview: 61+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-17 19:26 [PATCH 1/3] Refactor AIO interface to allow other AIO implementations Anthony Liguori
2008-04-17 19:26 ` [Qemu-devel] " Anthony Liguori
2008-04-17 19:26 ` [PATCH 2/3] Split out posix-aio code Anthony Liguori
2008-04-17 19:26 ` [Qemu-devel] " Anthony Liguori
2008-04-17 19:26 ` [PATCH 3/3] Implement linux-aio backend Anthony Liguori
2008-04-17 19:26 ` [Qemu-devel] " Anthony Liguori
2008-04-18 15:09 ` Marcelo Tosatti
2008-04-18 15:09 ` [Qemu-devel] " Marcelo Tosatti
2008-04-18 15:18 ` Anthony Liguori
2008-04-18 15:18 ` [Qemu-devel] " Anthony Liguori
2008-04-18 17:46 ` Marcelo Tosatti
2008-04-18 17:46 ` [Qemu-devel] " Marcelo Tosatti
2008-04-17 19:38 ` [PATCH 1/3] Refactor AIO interface to allow other AIO implementations Daniel P. Berrange
2008-04-17 19:38 ` [Qemu-devel] Re: [kvm-devel] " Daniel P. Berrange
2008-04-17 19:41 ` Anthony Liguori
2008-04-17 19:41 ` [Qemu-devel] " Anthony Liguori
2008-04-17 20:00 ` Daniel P. Berrange
2008-04-17 20:00 ` [Qemu-devel] " Daniel P. Berrange
2008-04-17 20:05 ` Anthony Liguori
2008-04-17 20:05 ` [Qemu-devel] Re: [kvm-devel] " Anthony Liguori
2008-04-18 12:43 ` Jamie Lokier
2008-04-18 12:43 ` [Qemu-devel] " Jamie Lokier
2008-04-18 15:23 ` Anthony Liguori
2008-04-18 15:23 ` [Qemu-devel] " Anthony Liguori
2008-04-18 16:22 ` [Qemu-devel] " Jamie Lokier
2008-04-18 16:22 ` [Qemu-devel] Re: [kvm-devel] " Jamie Lokier
2008-04-18 16:32 ` [Qemu-devel] " Avi Kivity
2008-04-18 16:32 ` [kvm-devel] " Avi Kivity
2008-04-20 15:49 ` Jamie Lokier
2008-04-20 15:49 ` [kvm-devel] " Jamie Lokier
2008-04-20 18:43 ` Avi Kivity
2008-04-20 18:43 ` [kvm-devel] " Avi Kivity
2008-04-20 23:39 ` Jamie Lokier
2008-04-20 23:39 ` [kvm-devel] " Jamie Lokier
2008-04-21 6:39 ` Avi Kivity
2008-04-21 6:39 ` [kvm-devel] " Avi Kivity
2008-04-21 12:10 ` Jamie Lokier
2008-04-21 12:10 ` [kvm-devel] " Jamie Lokier
2008-04-22 8:10 ` Avi Kivity
2008-04-22 8:10 ` [kvm-devel] " Avi Kivity
2008-04-22 14:28 ` [kvm-devel] " Jamie Lokier
2008-04-22 14:28 ` [kvm-devel] [Qemu-devel] " Jamie Lokier
2008-04-22 14:53 ` Anthony Liguori
2008-04-22 14:53 ` [kvm-devel] " Anthony Liguori
2008-04-22 15:05 ` Avi Kivity [this message]
2008-04-22 15:05 ` Avi Kivity
2008-04-22 15:23 ` [kvm-devel] " Jamie Lokier
2008-04-22 15:23 ` [kvm-devel] [Qemu-devel] " Jamie Lokier
2008-04-22 15:12 ` [kvm-devel] " Jamie Lokier
2008-04-22 15:12 ` [kvm-devel] [Qemu-devel] " Jamie Lokier
2008-04-22 15:03 ` Avi Kivity
2008-04-22 15:03 ` [kvm-devel] " Avi Kivity
2008-04-22 15:36 ` [kvm-devel] " Jamie Lokier
2008-04-22 15:36 ` [kvm-devel] [Qemu-devel] " Jamie Lokier
2008-05-02 16:37 ` Antonio Vargas
2008-05-02 17:18 ` Jamie Lokier
2008-05-02 17:52 ` Anthony Liguori
2008-05-02 18:24 ` Jamie Lokier
2008-04-22 15:47 ` Javier Guerra
2008-04-21 0:31 ` Javier Guerra Giraldez
2008-04-21 6:41 ` Avi Kivity
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=480DFEBC.1020504@qumranet.com \
--to=avi@qumranet.com \
--cc=kvm-devel@lists.sourceforge.net \
--cc=mtosatti@redhat.com \
--cc=qemu-devel@nongnu.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 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.