From: "John Hawkes" <hawkes@sgi.com>
To: "Nakajima, Jun" <jun.nakajima@intel.com>,
"Andi Kleen" <ak@suse.de>, "Ingo Molnar" <mingo@elte.hu>
Cc: <piggin@cyberone.com.au>, <linux-kernel@vger.kernel.org>,
<akpm@osdl.org>, <kernel@kolivas.org>, <rusty@rustcorp.com.au>,
<ricklind@us.ibm.com>, <anton@samba.org>,
<lse-tech@lists.sourceforge.net>, <mbligh@aracnet.com>
Subject: Re: [Lse-tech] [patch] sched-domain cleanups, sched-2.6.5-rc2-mm2-A3
Date: Thu, 25 Mar 2004 08:19:28 -0800 [thread overview]
Message-ID: <004101c41285$4e820270$6800a8c0@comcast.net> (raw)
In-Reply-To: 7F740D512C7C1046AB53446D3720017301119907@scsmsx402.sc.intel.com
From: "Nakajima, Jun" <jun.nakajima@intel.com>
> We have found some performance regressions (e.g. SPECjbb) with the
> scheduler on a large IA-64 NUMA machine, and we are debugging it. On SMP
> machines, we haven't seen performance regressions.
I've run a flavor of AIM7 and kernbench on a variety of ia64-NUMA CPU counts,
ranging up to 128p (and my 2.6.4 + Piggin-scheduler kernel hangs during boot
at >=192p, vs. 2.6.4 + vanilla-scheduler booting and running at 512p), and the
Piggin scheduler is 10-15% slower at 64p, and even worse at 128p. I did,
however, produce superior performance (superior to the vanilla scheduler) with
AIM7 by increasing the sched_domain->busy_factor by 4x. I tried a few other
things, but nothing improved performance on these big systems better than a
simple increase of busy_factor.
John Hawkes
next prev parent reply other threads:[~2004-03-25 16:24 UTC|newest]
Thread overview: 71+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-03-25 15:15 [Lse-tech] [patch] sched-domain cleanups, sched-2.6.5-rc2-mm2-A3 Nakajima, Jun
2004-03-25 16:19 ` John Hawkes [this message]
2004-03-25 16:53 ` Martin J. Bligh
-- strict thread matches above, loose matches on Subject: below --
2004-03-25 15:31 Nakajima, Jun
2004-03-25 15:40 ` Andi Kleen
2004-03-25 19:09 ` Ingo Molnar
2004-03-25 15:21 ` Andi Kleen
2004-03-25 19:39 ` Ingo Molnar
2004-03-25 20:30 ` Ingo Molnar
2004-03-29 8:45 ` Andi Kleen
2004-03-29 10:20 ` Rick Lindsley
2004-03-29 5:07 ` Andi Kleen
2004-03-29 11:28 ` Nick Piggin
2004-03-29 17:30 ` Rick Lindsley
2004-03-30 0:01 ` Nick Piggin
2004-03-30 1:26 ` Rick Lindsley
2004-03-29 11:20 ` Nick Piggin
2004-03-29 6:01 ` Andi Kleen
2004-03-29 11:46 ` Ingo Molnar
2004-03-29 7:03 ` Andi Kleen
2004-03-29 7:10 ` Andi Kleen
2004-03-29 20:14 ` Andi Kleen
2004-03-29 23:51 ` Nick Piggin
2004-03-30 6:34 ` Andi Kleen
2004-03-30 6:40 ` Ingo Molnar
2004-03-30 7:07 ` Andi Kleen
2004-03-30 7:14 ` Nick Piggin
2004-03-30 7:45 ` Ingo Molnar
2004-03-30 7:58 ` Nick Piggin
2004-03-30 7:15 ` Ingo Molnar
2004-03-30 7:18 ` Nick Piggin
2004-03-30 7:48 ` Andi Kleen
2004-03-30 8:18 ` Ingo Molnar
2004-03-30 9:36 ` Andi Kleen
2004-03-30 7:42 ` Ingo Molnar
2004-03-30 7:03 ` Nick Piggin
2004-03-30 7:13 ` Andi Kleen
2004-03-30 7:24 ` Nick Piggin
2004-03-30 7:38 ` Arjan van de Ven
2004-03-30 7:13 ` Martin J. Bligh
2004-03-30 7:31 ` Nick Piggin
2004-03-30 7:38 ` Martin J. Bligh
2004-03-30 8:05 ` Ingo Molnar
2004-03-30 8:19 ` Nick Piggin
2004-03-30 8:45 ` Ingo Molnar
2004-03-30 8:53 ` Nick Piggin
2004-03-30 15:27 ` Martin J. Bligh
2004-03-25 19:24 ` Martin J. Bligh
2004-03-25 21:48 ` Ingo Molnar
2004-03-25 22:28 ` Martin J. Bligh
2004-03-29 22:30 ` Erich Focht
2004-03-30 9:05 ` Nick Piggin
2004-03-30 10:04 ` Erich Focht
2004-03-30 10:58 ` Andi Kleen
2004-03-30 11:02 ` Andrew Morton
2004-03-31 18:59 ` Erich Focht
2004-03-31 2:08 ` Nick Piggin
2004-03-31 22:23 ` Erich Focht
2004-03-30 15:01 ` Martin J. Bligh
2004-03-31 21:23 ` Erich Focht
2004-03-31 21:33 ` Martin J. Bligh
2004-03-25 21:59 ` Ingo Molnar
2004-03-25 22:26 ` Rick Lindsley
2004-03-25 22:30 ` Andrew Theurer
2004-03-25 22:38 ` Martin J. Bligh
2004-03-26 1:29 ` Andi Kleen
2004-03-26 3:23 ` Nick Piggin
[not found] <1DF7H-22Y-11@gated-at.bofh.it>
[not found] ` <1DL3x-7iG-7@gated-at.bofh.it>
[not found] ` <1DLGd-7TS-17@gated-at.bofh.it>
[not found] ` <1FmNz-72J-73@gated-at.bofh.it>
[not found] ` <1FnzJ-7IW-15@gated-at.bofh.it>
2004-03-30 9:39 ` Andi Kleen
2004-03-31 1:56 ` Nick Piggin
2004-03-30 21:40 Nakajima, Jun
2004-03-30 22:15 ` Andrew Theurer
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='004101c41285$4e820270$6800a8c0@comcast.net' \
--to=hawkes@sgi.com \
--cc=ak@suse.de \
--cc=akpm@osdl.org \
--cc=anton@samba.org \
--cc=jun.nakajima@intel.com \
--cc=kernel@kolivas.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lse-tech@lists.sourceforge.net \
--cc=mbligh@aracnet.com \
--cc=mingo@elte.hu \
--cc=piggin@cyberone.com.au \
--cc=ricklind@us.ibm.com \
--cc=rusty@rustcorp.com.au \
/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.