From: Lee Revell <rlrevell@joe-job.com>
To: Andrea Arcangeli <andrea@novell.com>
Cc: Ingo Molnar <mingo@elte.hu>, Chris Wedgwood <cw@f00f.org>,
Arjan van de Ven <arjanv@redhat.com>,
Hugh Dickins <hugh@veritas.com>,
"Martin J. Bligh" <mbligh@aracnet.com>,
Alan Cox <alan@lxorguk.ukuu.org.uk>,
LKML <linux-kernel@vger.kernel.org>,
Christoph Hellwig <hch@infradead.org>
Subject: Re: [PATCH 1/3] Separate IRQ-stacks from 4K-stacks option
Date: Sun, 12 Sep 2004 18:25:51 -0400 [thread overview]
Message-ID: <1095027951.22893.69.camel@krustophenia.net> (raw)
In-Reply-To: <20040912220720.GC3049@dualathlon.random>
On Sun, 2004-09-12 at 18:07, Andrea Arcangeli wrote:
> On Sun, Sep 12, 2004 at 05:36:41PM -0400, Lee Revell wrote:
> > But in this case the hardirq handler can run for 2ms, which caused a
> > scheduler latency problem, because nothing could run but other IRQs.
> > The IRQ threading in Ingo's patches solves the problem, and seems to me
> > to be the correct solution.
>
> the irq threading must have a cost, doesn't it? I doubt you want to
> offload irqs to a kernel thread on a server, *that* would be slow (irq
> nesting is zerocost compared to scheduling a kernel thread).
Yes, on a server you would probably disable threading for the disk and
network IRQs (the VP patch lets you set this via /proc). This feature
effectively gives you IPLs on Linux, albeit only two of them. For
example to prevent heavy traffic on one network interface from impacting
the latency of the other, you could enable threading for the first and
disable it for the second.
I am still unsure why the IDE i/o completion is the one place that
breaks the assumption that hardirq handlers execute quickly.
Lee
next prev parent reply other threads:[~2004-09-12 22:25 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-09-09 23:25 [PATCH 1/3] Separate IRQ-stacks from 4K-stacks option Chris Wedgwood
2004-09-09 23:55 ` Brian Gerst
2004-09-09 23:59 ` Chris Wedgwood
2004-09-10 6:40 ` Arjan van de Ven
2004-09-10 6:45 ` Chris Wedgwood
2004-09-10 6:52 ` Arjan van de Ven
2004-09-10 7:15 ` Chris Wedgwood
2004-09-10 7:21 ` Arjan van de Ven
2004-09-10 7:23 ` Chris Wedgwood
2004-09-10 7:35 ` Arjan van de Ven
2004-09-10 9:12 ` Alan Cox
2004-09-10 9:14 ` Alan Cox
2004-09-10 10:27 ` Russell King
2004-09-10 14:30 ` Martin J. Bligh
2004-09-10 15:06 ` Chris Wedgwood
2004-09-10 15:07 ` Hugh Dickins
2004-09-10 15:12 ` Alan Cox
2004-09-10 15:15 ` Arjan van de Ven
2004-09-10 15:28 ` Andrea Arcangeli
2004-09-10 15:09 ` Alan Cox
2004-09-10 15:34 ` Arjan van de Ven
2004-09-12 14:17 ` Andrea Arcangeli
2004-09-12 15:03 ` Martin J. Bligh
2004-09-12 15:45 ` Zwane Mwaikambo
2004-09-12 16:00 ` Martin J. Bligh
2004-10-20 16:27 ` Lee Revell
2004-10-21 15:18 ` Zwane Mwaikambo
2004-09-12 19:18 ` Lee Revell
2004-09-12 19:25 ` Chris Wedgwood
2004-09-12 19:35 ` Ingo Molnar
2004-09-12 19:40 ` Chris Wedgwood
2004-09-12 19:44 ` Ingo Molnar
2004-09-12 20:33 ` Andrea Arcangeli
2004-09-12 21:36 ` Lee Revell
2004-09-12 22:07 ` Andrea Arcangeli
2004-09-12 22:25 ` Lee Revell [this message]
2004-09-13 6:16 ` Ingo Molnar
2004-09-13 7:27 ` Lee Revell
2004-09-15 19:43 ` Bill Davidsen
2004-09-15 19:51 ` Andrea Arcangeli
2004-09-15 20:00 ` Ingo Molnar
2004-09-15 20:39 ` Lee Revell
2004-10-20 17:50 ` Timothy Miller
2004-09-13 7:32 ` Jens Axboe
2004-09-13 10:31 ` Andrea Arcangeli
2004-09-13 10:34 ` Arjan van de Ven
2004-09-13 12:30 ` Jens Axboe
2004-10-20 17:46 ` Timothy Miller
2004-09-13 10:51 ` Alan Cox
2004-09-13 20:26 ` Bill Davidsen
2004-09-14 5:20 ` Lee Revell
2004-10-20 15:46 ` Timothy Miller
2004-10-20 15:35 ` Arjan van de Ven
2004-10-20 16:39 ` Lee Revell
2004-10-20 16:50 ` Andrea Arcangeli
2004-10-20 16:55 ` Lee Revell
2004-10-20 17:08 ` Andrea Arcangeli
2004-10-20 17:15 ` Lee Revell
2004-10-20 17:34 ` Andrea Arcangeli
2004-10-20 17:58 ` Timothy Miller
2004-10-20 17:55 ` Lee Revell
2004-10-20 18:25 ` Timothy Miller
2004-10-20 18:28 ` Lee Revell
2004-10-20 18:51 ` Alan Cox
2004-10-20 23:19 ` Lee Revell
2004-09-10 15:16 ` Martin J. Bligh
2004-10-20 15:43 ` Timothy Miller
2004-10-20 15:35 ` Arjan van de Ven
2004-09-10 15:58 ` Alexander E. Patrakov
2004-09-10 16:19 ` Nathan Bryant
2004-09-10 17:54 ` Martin J. Bligh
2004-09-10 17:54 ` Bill Davidsen
2004-09-10 9:27 ` Alan Cox
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=1095027951.22893.69.camel@krustophenia.net \
--to=rlrevell@joe-job.com \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=andrea@novell.com \
--cc=arjanv@redhat.com \
--cc=cw@f00f.org \
--cc=hch@infradead.org \
--cc=hugh@veritas.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mbligh@aracnet.com \
--cc=mingo@elte.hu \
/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.