From: Tejun Heo <htejun@gmail.com>
To: Jeff Garzik <jeff@garzik.org>
Cc: ric@emc.com, linux-ide@vger.kernel.org,
Mark Lord <mlord@pobox.com>, Jens Axboe <axboe@suse.de>
Subject: Re: [RFT] major libata update
Date: Wed, 17 May 2006 09:50:18 +0900 [thread overview]
Message-ID: <446A734A.6020504@gmail.com> (raw)
In-Reply-To: <446A6ECD.7080104@garzik.org>
Jeff Garzik wrote:
> Actually, looking even more at the code, I think the entire "spurious
> interrupt" code block is bogus.
>
> The device is free to send D2H or SDB FIS to clear Status bits, as well
> as SDB FIS's to update SActive. The ata_qc_complete_multiple() call
> will do the right thing, even if there is no work to do.
Yes, they are, but...
* Please note that SActive update via SDB FIS is edge-triggered. It
doesn't send the whole image of the register. The SActive bits in SDB
FIS indicate W1C (this is inevitable because controller and device
cannot be in full synchronization regarding SActive). So, they have can
have dangerous side effects.
* Those spurious FISes are actively setting the I bit requesting
interrupt explicitly. Those are not innocent status update FISes. And
the spec state machines don't allow D2H FISes during NCQ command phase.
I seem to recollect it's actually forbidden explicitly but currently
cannot find where. I'll keep looking.
> As another side note, command coalescing is worth experimenting with.
Might be, but I'm a bit skeptical. We might lose more to added latency.
--
tejun
next prev parent reply other threads:[~2006-05-17 0:50 UTC|newest]
Thread overview: 115+ 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
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 [this message]
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.-
-- strict thread matches above, loose matches on Subject: below --
2006-05-17 7:35 Matthieu CASTET
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=446A734A.6020504@gmail.com \
--to=htejun@gmail.com \
--cc=axboe@suse.de \
--cc=jeff@garzik.org \
--cc=linux-ide@vger.kernel.org \
--cc=mlord@pobox.com \
--cc=ric@emc.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;
as well as URLs for NNTP newsgroup(s).