From: "Chen, Kenneth W" <kenneth.w.chen@intel.com>
To: "'Andrew Morton'" <akpm@osdl.org>
Cc: <nickpiggin@yahoo.com.au>, <mingo@redhat.com>,
<linux-kernel@vger.kernel.org>
Subject: RE: Default cache_hot_time value back to 10ms
Date: Wed, 6 Oct 2004 16:14:06 -0700 [thread overview]
Message-ID: <200410062313.i96NDo609923@unix-os.sc.intel.com> (raw)
In-Reply-To: <20041006134317.03a22198.akpm@osdl.org>
Andrew Morton wrote on Wednesday, October 06, 2004 1:43 PM
> "Chen, Kenneth W" <kenneth.w.chen@intel.com> wrote:
> >
> > Secondly, let me ask the question again from the first mail thread: this value
> > *WAS* 10 ms for a long time, before the domain scheduler. What's so special
> > about domain scheduler that all the sudden this parameter get changed to 2.5?
>
> So why on earth was it switched from 10 to 2.5 in the first place?
>
> Please resend the final patch.
Here is a patch that revert default cache_hot_time value back to the equivalence
of pre-domain scheduler, which determins task's cache affinity via architecture
defined variable cache_decay_ticks.
This is a mere request that we go back to what *was* before, *NOT* as a new
scheduler tweak (Whatever tweak done for domain scheduler broke traditional/
industry recognized workload).
As a side note, I'd like to get involved on future scheduler tuning experiments,
we have fair amount of benchmark environments where we can validate things across
various kind of workload, i.e., db, java, cpu, etc. Thanks.
Signed-off-by: Ken Chen <kenneth.w.chen@intel.com>
patch against 2.6.9-rc3:
--- linux-2.6.9-rc3/kernel/sched.c.orig 2004-10-06 15:10:56.000000000 -0700
+++ linux-2.6.9-rc3/kernel/sched.c 2004-10-06 15:18:51.000000000 -0700
@@ -387,7 +387,7 @@ struct sched_domain {
.max_interval = 4, \
.busy_factor = 64, \
.imbalance_pct = 125, \
- .cache_hot_time = (5*1000000/2), \
+ .cache_hot_time = cache_decay_ticks*1000000,\
.cache_nice_tries = 1, \
.per_cpu_gain = 100, \
.flags = SD_BALANCE_NEWIDLE \
patch against 2.6.9-rc3-mm2:
--- linux-2.6.9-rc3/include/linux/topology.h.orig 2004-10-06 15:32:48.000000000 -0700
+++ linux-2.6.9-rc3/include/linux/topology.h 2004-10-06 15:33:25.000000000 -0700
@@ -113,7 +113,7 @@ static inline int __next_node_with_cpus(
.max_interval = 4, \
.busy_factor = 64, \
.imbalance_pct = 125, \
- .cache_hot_time = (5*1000/2), \
+ .cache_hot_time = (cache_decay_ticks*1000),\
.cache_nice_tries = 1, \
.per_cpu_gain = 100, \
.flags = SD_LOAD_BALANCE \
next prev parent reply other threads:[~2004-10-06 23:19 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-10-06 0:42 Default cache_hot_time value back to 10ms Chen, Kenneth W
2004-10-06 0:47 ` Con Kolivas
2004-10-06 1:02 ` Nick Piggin
2004-10-06 0:58 ` Nick Piggin
2004-10-06 3:55 ` Andrew Morton
2004-10-06 4:30 ` Nick Piggin
2004-10-06 4:51 ` Andrew Morton
2004-10-06 5:00 ` Nick Piggin
2004-10-06 5:09 ` Andrew Morton
2004-10-06 5:21 ` Nick Piggin
2004-10-06 5:33 ` Andrew Morton
2004-10-06 5:46 ` Nick Piggin
2004-10-06 6:19 ` new dev model (was Re: Default cache_hot_time value back to 10ms) Jeff Garzik
2004-10-06 6:39 ` Andrew Morton
2004-10-06 8:56 ` Paolo Ciarrocchi
2004-10-06 9:44 ` bert hubert
2004-10-06 14:00 ` Andries Brouwer
2004-10-06 19:40 ` Jeff Garzik
2004-10-06 19:48 ` Jeff Garzik
2004-10-06 19:58 ` Jeff Garzik
2004-10-06 20:37 ` Geert Uytterhoeven
2004-10-07 1:08 ` Jeff Garzik
2004-10-07 0:02 ` Matt Mackall
2004-10-06 9:23 ` Ingo Molnar
2004-10-06 9:57 ` Paolo Ciarrocchi
2004-10-06 19:33 ` Jeff Garzik
2004-10-06 22:23 ` Martin J. Bligh
2004-10-06 5:52 ` Default cache_hot_time value back to 10ms Chen, Kenneth W
2004-10-06 19:27 ` Chen, Kenneth W
2004-10-06 19:39 ` Andrew Morton
2004-10-06 20:38 ` Chen, Kenneth W
2004-10-06 20:43 ` Andrew Morton
2004-10-06 23:14 ` Chen, Kenneth W [this message]
2004-10-07 2:26 ` Nick Piggin
2004-10-07 6:29 ` Ingo Molnar
2004-10-07 7:08 ` Jeff Garzik
2004-10-07 7:26 ` Ingo Molnar
2004-10-06 20:50 ` Ingo Molnar
2004-10-06 21:03 ` Chen, Kenneth W
2004-10-06 7:48 ` Ingo Molnar
2004-10-06 17:18 ` Chen, Kenneth W
2004-10-06 19:55 ` Ingo Molnar
2004-10-06 22:46 ` Peter Williams
2004-10-06 13:29 ` [patch] sched: auto-tuning task-migration Ingo Molnar
2004-10-06 13:44 ` Nick Piggin
2004-10-06 17:49 ` Chen, Kenneth W
2004-10-06 20:04 ` Ingo Molnar
2004-10-06 21:18 ` Chen, Kenneth W
2004-10-07 6:10 ` Ingo Molnar
2005-02-21 5:08 ` Paul Jackson
[not found] <200410071028.01931.habanero@us.ibm.com>
2004-10-07 15:58 ` Default cache_hot_time value back to 10ms Andrew Theurer
2004-10-08 9:47 ` Nick Piggin
2004-10-08 14:11 ` Andrew Theurer
-- strict thread matches above, loose matches on Subject: below --
2004-10-07 18:44 Albert Cahalan
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=200410062313.i96NDo609923@unix-os.sc.intel.com \
--to=kenneth.w.chen@intel.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=nickpiggin@yahoo.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 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).