From: Jens Axboe <axboe@suse.de>
To: Jeff Garzik <jgarzik@pobox.com>
Cc: "Mudama, Eric" <eric_mudama@maxtor.com>,
Oleg Drokin <green@namesys.com>,
Bartlomiej Zolnierkiewicz <B.Zolnierkiewicz@elka.pw.edu.pl>,
Alan Cox <alan@lxorguk.ukuu.org.uk>,
Oliver Neukum <oliver@neukum.org>,
lkhelp@rekl.yi.org, linux-kernel@vger.kernel.org
Subject: Re: 2.5.69, IDE TCQ can't be enabled
Date: Tue, 13 May 2003 08:40:59 +0200 [thread overview]
Message-ID: <20030513064059.GL17033@suse.de> (raw)
In-Reply-To: <20030512195331.GD10089@gtf.org>
On Mon, May 12 2003, Jeff Garzik wrote:
> On Mon, May 12, 2003 at 09:42:45PM +0200, Jens Axboe wrote:
> > On Mon, May 12 2003, Jeff Garzik wrote:
> > > Most Linux people with TCQ drives seem to have Hitachi (nee IBM)
> > > ones AFAICS. These do not have a service interrupt (or at least,
> > > do not report such)
> >
> > Nonsense, it supports the service interrupt just fine. It will just
> > complain if you try to turn it off, iirc.
>
> Weird. Mine doesn't seem to assert it, nor does the identify page
> indicate it's supported. Maybe I have a broken drive firmware.
Then the linux code won't work on it, have you tried? I've tried a lot
of different IBM models, they all do service interrupts just fine.
> > > They do have the release interrupt.
> >
> > Which we don't use. To be interesting, you need to speculatively turn on
> > the dma engine for each command you want to start. If you don't do that,
> > then it's faster just to poll for release/no-release at command start
> > time.
>
> That's an annoying thing about ATA TCQ: the command _may_ execute
> immediately, or may be queued (even when queue is empty). At least
> that's how I read the code and specs...
That's correct, you can use the release interrupt to get around that...
> > I don't think the multiple pending _and_ active is that big a deal, and
> > besides _everybody_ uses write back caching on IDE which makes TCQ for
> > writes very uninteresting.
> [...]
> > I have to agree with Eric that the largest win is potentially not
> > getting hit by the rotational latency all the time. I don't think you'll
> > get much extra from actually having more than one active from the dma
> > POV.
>
> Yes and no. I am coming from a driver-complexity perspective:
> single-active is more annoying on the driver side.
>
> In terms of drive performance, multiple active probably doesn't make
> a huge difference. In terms of reduction in host CPU usage, there
> is a performance gain there with multiple active.
It should make a non-neglible difference in smart positioning in the
drive, some things just cannot be done in software for this stuff.
--
Jens Axboe
next prev parent reply other threads:[~2003-05-13 6:28 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2003-05-12 19:19 2.5.69, IDE TCQ can't be enabled Mudama, Eric
2003-05-12 19:35 ` Jeff Garzik
2003-05-12 19:42 ` Jens Axboe
2003-05-12 19:44 ` Jens Axboe
2003-05-12 19:53 ` Jeff Garzik
2003-05-13 6:40 ` Jens Axboe [this message]
2003-05-13 18:00 ` Dave Jones
2003-05-13 18:03 ` Jens Axboe
2003-05-13 18:05 ` Jeff Garzik
2003-05-13 18:06 ` Jens Axboe
2003-05-13 18:13 ` Jens Axboe
2003-05-13 18:42 ` Jeff Garzik
2003-05-13 20:29 ` Andre Hedrick
2003-05-14 7:04 ` Jens Axboe
2003-05-14 7:15 ` Jens Axboe
2003-05-13 20:03 ` Andre Hedrick
2003-05-13 22:55 ` Alan Cox
2003-05-13 18:03 ` Jeff Garzik
2003-05-14 13:30 ` Henning P. Schmiedehausen
2003-05-12 22:36 ` Christer Weinigel
2003-05-13 6:41 ` Jens Axboe
2003-05-13 20:04 ` Bill Davidsen
-- strict thread matches above, loose matches on Subject: below --
2003-05-13 20:43 Mudama, Eric
2003-05-13 20:31 Mudama, Eric
2003-05-13 20:34 ` Andre Hedrick
2003-05-13 20:28 Mudama, Eric
2003-05-12 17:58 Mudama, Eric
2003-05-12 18:59 ` Jens Axboe
2003-05-12 19:42 ` Jeff Garzik
2003-05-10 1:38 lkhelp
2003-05-12 12:46 ` Oleg Drokin
2003-05-12 12:55 ` Oliver Neukum
2003-05-12 13:16 ` Bartlomiej Zolnierkiewicz
2003-05-12 13:12 ` Alan Cox
2003-05-13 15:39 ` Jens Axboe
2003-05-12 13:22 ` Oleg Drokin
2003-05-12 13:23 ` Jens Axboe
2003-05-12 13:35 ` Jeff Garzik
2003-05-12 13:30 ` Bartlomiej Zolnierkiewicz
2003-05-12 13:22 ` Jens Axboe
2003-05-13 15:32 ` Jens Axboe
2003-05-13 17:25 ` Oliver Neukum
2003-05-13 17:28 ` Jens Axboe
2003-05-13 18:46 ` Oliver Neukum
2003-05-13 19:06 ` Jens Axboe
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=20030513064059.GL17033@suse.de \
--to=axboe@suse.de \
--cc=B.Zolnierkiewicz@elka.pw.edu.pl \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=eric_mudama@maxtor.com \
--cc=green@namesys.com \
--cc=jgarzik@pobox.com \
--cc=linux-kernel@vger.kernel.org \
--cc=lkhelp@rekl.yi.org \
--cc=oliver@neukum.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).