From: Ric Wheeler <ric@emc.com>
To: "Eric D. Mudama" <edmudama@gmail.com>
Cc: Jeff Garzik <jeff@garzik.org>, Tejun Heo <htejun@gmail.com>,
linux-ide@vger.kernel.org, Mark Lord <mlord@pobox.com>,
Jens Axboe <axboe@suse.de>
Subject: Re: [RFT] major libata update
Date: Tue, 16 May 2006 22:13:23 -0400 [thread overview]
Message-ID: <446A86C3.4090904@emc.com> (raw)
In-Reply-To: <311601c90605161611s28538eebu61f887609f6b32f4@mail.gmail.com>
Eric D. Mudama wrote:
> On 5/16/06, Jeff Garzik <jeff@garzik.org> wrote:
>
>> Ric Wheeler wrote:
>> > But when I test with my synchronous write load, I am going at 6 times
>> > the normal rate (i.e., the rate I see with barriers disabled) ;-)
>>
>> Well, NCQ read/write commands have a 'FUA' bit... though maybe the NCQ
>> code forgot to check the global libata_fua.
>>
>> Jeff
>
>
> Is the workload a sequential write? SATA drives should be able to
> keep up with the media if they use FUA for their consistency when
> using NCQ.
>
> Flush cache as a barrier mechanism will blow revs in sequential
> workloads, especially using a write/flush write/flush technique.
That would be a huge win for us, but this instance I mounted with
barrier=flush (and that is the logged message in messages). I suspect
that our boot with write cache disabled + flip of the write cache to on
during boot is the culprit...
Once we get past the basic working/not working barriers, I will retest
the FUA vs flush barriers and post some results ;-)
ric
next prev parent reply other threads:[~2006-05-17 1:13 UTC|newest]
Thread overview: 117+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-05-15 17:00 [RFT] major libata update Jeff Garzik
2006-05-15 17:18 ` Andrew Morton
2006-05-15 18:06 ` Jeff Garzik
2006-05-15 19:06 ` Arkadiusz Miskiewicz
2006-05-15 20:45 ` Jeff Garzik
2006-05-15 19:33 ` Mark Lord
2006-05-15 22:52 ` Tejun Heo
2006-05-15 18:15 ` Jeff Garzik
2006-05-15 18:27 ` Andrew Morton
2006-05-15 18:44 ` Jeff Garzik
2006-05-15 18:37 ` Alan Cox
2006-05-15 17:19 ` Alan Cox
2006-05-15 17:13 ` Jeff Garzik
2006-05-15 18:29 ` Tomasz Torcz
2006-05-15 18:43 ` Jeff Garzik
2006-05-15 23:32 ` Tejun Heo
2006-05-15 23:49 ` Jeff Garzik
2006-05-16 0:04 ` Tejun Heo
2006-05-16 2:15 ` Tejun Heo
2006-05-15 19:15 ` Jeff Garzik
2006-05-15 23:02 ` Wakko Warner
2006-05-15 23:00 ` Jeff Garzik
2006-05-15 23:13 ` Wakko Warner
2006-05-15 23:19 ` Jeff Garzik
2006-05-15 23:40 ` Alan Cox
2006-05-15 23:50 ` Wakko Warner
2006-05-15 23:38 ` Alan Cox
2006-05-15 23:47 ` Wakko Warner
2006-05-15 23:45 ` Jeff Garzik
2006-05-15 23:30 ` Avuton Olrich
2006-05-15 23:36 ` Tejun Heo
2006-05-15 23:54 ` Jeff Garzik
2006-05-16 0:08 ` Avuton Olrich
2006-05-16 3:36 ` Avuton Olrich
2006-05-16 3:51 ` Jeff Garzik
2006-05-16 4:33 ` Avuton Olrich
2006-05-16 14:57 ` Linus Torvalds
2006-05-17 15:25 ` OGAWA Hirofumi
2006-05-17 23:40 ` Linus Torvalds
2006-05-17 23:48 ` Jeff Garzik
2006-05-18 1:48 ` Alan Cox
2006-05-17 23:49 ` Linus Torvalds
2006-05-16 15:02 ` Jeff Garzik
2006-05-16 3:55 ` Tejun Heo
2006-05-16 4:37 ` Avuton Olrich
2006-05-16 11:36 ` Ric Wheeler
2006-05-16 14:25 ` Jeff Garzik
2006-05-16 15:24 ` Tejun Heo
2006-05-16 18:29 ` Ric Wheeler
2006-05-16 21:41 ` Ric Wheeler
2006-05-16 22:02 ` Jeff Garzik
2006-05-16 23:11 ` Eric D. Mudama
2006-05-17 2:13 ` Ric Wheeler [this message]
2006-05-16 23:23 ` Tejun Heo
2006-05-17 2:09 ` Ric Wheeler
2006-05-16 23:44 ` Tejun Heo
2006-05-16 23:53 ` Jeff Garzik
2006-05-17 0:00 ` Jeff Garzik
2006-05-17 0:29 ` Tejun Heo
2006-05-17 1:08 ` Jeff Garzik
2006-05-17 1:27 ` Tejun Heo
2006-05-17 2:26 ` Jeff Garzik
2006-05-17 3:05 ` Tejun Heo
2006-05-22 7:19 ` Jeff Garzik
2006-05-23 13:59 ` Tejun Heo
2006-05-17 0:31 ` Jeff Garzik
2006-05-17 0:50 ` Tejun Heo
2006-05-17 0:57 ` Tejun Heo
2006-05-17 2:22 ` Ric Wheeler
2006-05-17 1:37 ` Tejun Heo
2006-05-17 3:57 ` Ric Wheeler
2006-05-17 4:44 ` Tejun Heo
2006-05-17 11:30 ` Ric Wheeler
2006-05-17 20:45 ` Ric Wheeler
2006-05-17 21:01 ` Mark Lord
2006-05-17 21:04 ` Jeff Garzik
2006-05-17 21:50 ` Tejun Heo
2006-05-17 21:56 ` Mark Lord
2006-05-17 22:00 ` Jeff Garzik
2006-05-17 22:03 ` Mark Lord
2006-05-17 22:13 ` Jeff Garzik
2006-05-18 3:33 ` Ric Wheeler
2006-05-18 3:26 ` Tejun Heo
2006-05-18 11:58 ` Ric Wheeler
2006-05-18 12:52 ` Mark Lord
2006-05-18 13:22 ` Ric Wheeler
2006-05-18 13:37 ` Jens Axboe
2006-05-17 1:13 ` Jeff Garzik
2006-05-17 1:14 ` Jeff Garzik
2006-05-17 2:16 ` Ric Wheeler
2006-05-16 23:34 ` Jeff Garzik
2006-05-16 23:53 ` Tejun Heo
2006-05-17 2:05 ` Andrew Morton
2006-05-17 4:49 ` Tejun Heo
2006-05-17 4:56 ` Andrew Morton
2006-05-17 5:14 ` Tejun Heo
2006-05-17 6:35 ` Tejun Heo
2006-05-18 11:24 ` Albert Lee
2006-05-18 11:33 ` Tejun Heo
2006-05-19 10:37 ` Albert Lee
2006-05-19 11:03 ` Tejun Heo
2006-05-22 3:51 ` [PATCH 1/1] libata: use polling pio for identify device Albert Lee
2006-05-22 6:24 ` Jeff Garzik
2006-05-23 2:27 ` Albert Lee
2006-05-18 23:07 ` [RFT] major libata update Andrew Morton
2006-05-19 1:14 ` Tejun Heo
2006-05-19 2:06 ` Jeff Garzik
2006-05-19 2:16 ` Tejun Heo
2006-05-22 7:22 ` Jeff Garzik
2006-05-21 23:51 ` Michael Sterrett -Mr. Bones.-
2006-05-22 2:42 ` Tejun Heo
2006-05-22 3:42 ` Michael Sterrett -Mr. Bones.-
2006-05-22 6:23 ` Michael Sterrett -Mr. Bones.-
[not found] <6cOK9-5WZ-21@gated-at.bofh.it>
[not found] ` <6cUPu-6Dh-5@gated-at.bofh.it>
[not found] ` <6cV8V-71t-3@gated-at.bofh.it>
[not found] ` <6cYzI-3yr-3@gated-at.bofh.it>
[not found] ` <6cYT8-3Vy-7@gated-at.bofh.it>
[not found] ` <6cZvL-4Vb-1@gated-at.bofh.it>
2006-05-16 5:33 ` Robert Hancock
-- strict thread matches above, loose matches on Subject: below --
2006-05-17 7:35 Matthieu CASTET
2006-05-18 0:36 Brown, Len
2006-05-18 0:36 ` Brown, Len
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=446A86C3.4090904@emc.com \
--to=ric@emc.com \
--cc=axboe@suse.de \
--cc=edmudama@gmail.com \
--cc=htejun@gmail.com \
--cc=jeff@garzik.org \
--cc=linux-ide@vger.kernel.org \
--cc=mlord@pobox.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.