From: Jens Axboe <jens.axboe@oracle.com>
To: Zach Brown <zach.brown@oracle.com>
Cc: linux-kernel@vger.kernel.org, Ingo Molnar <mingo@elte.hu>,
Linus Torvalds <torvalds@linux-foundation.org>,
Arjan van de Ven <arjan@infradead.org>,
Christoph Hellwig <hch@infradead.org>,
Andrew Morton <akpm@zip.com.au>,
Alan Cox <alan@lxorguk.ukuu.org.uk>,
Ulrich Drepper <drepper@redhat.com>,
Evgeniy Polyakov <johnpol@2ka.mipt.ru>,
"David S. Miller" <davem@davemloft.net>,
Suparna Bhattacharya <suparna@in.ibm.com>,
Davide Libenzi <davidel@xmailserver.org>,
Thomas Gleixner <tglx@linutronix.de>
Subject: Re: Syslets, Threadlets, generic AIO support, v6
Date: Wed, 30 May 2007 09:40:32 +0200 [thread overview]
Message-ID: <20070530074031.GV15559@kernel.dk> (raw)
In-Reply-To: <20070529212718.GH7875@mami.zabbo.net>
On Tue, May 29 2007, Zach Brown wrote:
Thanks for picking this up, Zach!
> - cfq gets confused, share io_context amongst threads?
Yeah, it'll confuse CFQ a lot actually. The threads either need to share
an io context (clean approach, however will introduce locking for things
that were previously lockless), or CFQ needs to get better support for
cooperating processes. The problem is that CFQ will wait for a dependent
IO for a given process, which may arrive from a totally unrelated
process.
For the fio testing, we can make some improvements there. Right now you
don't get any concurrency of the io requests if you set eg iodepth=32,
as the 32 requests will be submitted as a linked chain of atoms. For io
saturation, that's not really what you want.
I'll take a stab at improving both of the above.
--
Jens Axboe
next prev parent reply other threads:[~2007-05-30 7:41 UTC|newest]
Thread overview: 71+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-05-29 21:27 Syslets, Threadlets, generic AIO support, v6 Zach Brown
2007-05-29 21:49 ` Linus Torvalds
2007-05-29 22:49 ` Zach Brown
2007-05-29 22:16 ` Jeff Garzik
2007-05-29 23:09 ` Zach Brown
2007-05-29 23:20 ` Ulrich Drepper
2007-05-30 1:11 ` Dave Jones
2007-05-30 17:08 ` Zach Brown
2007-05-30 7:26 ` Ingo Molnar
2007-05-30 7:20 ` Ingo Molnar
2007-05-30 7:31 ` Ulrich Drepper
2007-05-30 8:42 ` Ingo Molnar
2007-05-30 8:51 ` Evgeniy Polyakov
2007-05-30 9:05 ` Ingo Molnar
2007-05-30 15:16 ` Linus Torvalds
2007-05-30 15:39 ` Ulrich Drepper
2007-05-30 19:40 ` Davide Libenzi
2007-05-30 19:55 ` Ulrich Drepper
2007-05-30 20:00 ` Linus Torvalds
2007-05-30 20:21 ` Davide Libenzi
2007-05-30 20:31 ` Eric Dumazet
2007-05-30 20:44 ` Linus Torvalds
2007-05-30 21:53 ` Eric Dumazet
2007-05-30 21:31 ` Davide Libenzi
2007-05-30 21:16 ` Ulrich Drepper
2007-05-30 21:27 ` Linus Torvalds
2007-05-30 21:47 ` Ulrich Drepper
2007-05-30 22:06 ` Davide Libenzi
2007-05-30 21:48 ` Davide Libenzi
2007-05-30 22:01 ` Linus Torvalds
2007-05-31 6:13 ` Ingo Molnar
2007-05-31 7:35 ` Eric Dumazet
2007-05-31 9:26 ` Ingo Molnar
2007-05-31 9:02 ` Ingo Molnar
2007-05-31 10:41 ` Eric Dumazet
2007-05-31 10:50 ` Ingo Molnar
2007-05-31 9:32 ` Ingo Molnar
2007-05-31 9:34 ` Jens Axboe
2007-05-30 22:09 ` Eric Dumazet
2007-05-30 21:51 ` David M. Lloyd
2007-05-30 22:24 ` William Lee Irwin III
2007-05-30 21:38 ` Jeremy Fitzhardinge
2007-05-30 21:39 ` Davide Libenzi
2007-05-30 21:36 ` Jeremy Fitzhardinge
2007-05-30 21:44 ` Linus Torvalds
2007-05-30 21:48 ` Linus Torvalds
2007-05-30 21:54 ` Jeremy Fitzhardinge
2007-05-30 22:27 ` Matt Mackall
2007-05-30 22:38 ` William Lee Irwin III
2007-05-30 8:32 ` Evgeniy Polyakov
2007-05-30 8:54 ` Ingo Molnar
2007-05-30 9:30 ` Evgeniy Polyakov
2007-05-30 9:28 ` Jeff Garzik
2007-05-30 13:02 ` Ingo Molnar
2007-05-30 13:20 ` Ingo Molnar
2007-05-30 15:31 ` Linus Torvalds
2007-05-30 16:09 ` Ingo Molnar
2007-05-30 17:57 ` Jens Axboe
2007-05-30 19:05 ` Mark Lord
2007-05-30 19:10 ` Jens Axboe
2007-05-30 19:15 ` Linus Torvalds
2007-05-30 19:32 ` Jens Axboe
2007-05-30 20:07 ` Eric Dumazet
2007-05-30 20:31 ` Linus Torvalds
2007-05-30 20:46 ` Eric Dumazet
2007-05-30 19:52 ` Davide Libenzi
2007-05-30 7:40 ` Jens Axboe [this message]
2007-05-30 16:55 ` Zach Brown
2007-05-30 17:33 ` Jens Axboe
-- strict thread matches above, loose matches on Subject: below --
2007-05-31 8:15 Albert Cahalan
2007-05-31 9:50 ` Ingo Molnar
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=20070530074031.GV15559@kernel.dk \
--to=jens.axboe@oracle.com \
--cc=akpm@zip.com.au \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=arjan@infradead.org \
--cc=davem@davemloft.net \
--cc=davidel@xmailserver.org \
--cc=drepper@redhat.com \
--cc=hch@infradead.org \
--cc=johnpol@2ka.mipt.ru \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=suparna@in.ibm.com \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=zach.brown@oracle.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox