From: "Martin J. Bligh" <mbligh@google.com>
To: Peter Williams <pwil3058@bigpond.net.au>
Cc: Con Kolivas <kernel@kolivas.org>, Andrew Morton <akpm@osdl.org>,
linux-kernel@vger.kernel.org, Ingo Molnar <mingo@elte.hu>
Subject: Re: -mm seems significanty slower than mainline on kernbench
Date: Tue, 10 Jan 2006 22:21:29 -0800 [thread overview]
Message-ID: <43C4A3E9.1040301@google.com> (raw)
In-Reply-To: <43C4941D.6080302@bigpond.net.au>
Peter Williams wrote:
> Peter Williams wrote:
>
>> Con Kolivas wrote:
>>
>>> On Wed, 11 Jan 2006 01:38 pm, Peter Williams wrote:
>>>
>>>> Con Kolivas wrote:
>>>> > I guess we need to check whether reversing this patch helps.
>>>>
>>>> It would be interesting to see if it does.
>>>>
>>>> If it does we probably have to wear the cost (and try to reduce it) as
>>>> without this change smp nice support is fairly ineffective due to the
>>>> fact that it moves exactly the same tasks as would be moved without
>>>> it.
>>>> At the most it changes the frequency at which load balancing occurs.
>>>
>>>
>>>
>>>
>>> I disagree. I think the current implementation changes the balancing
>>> according to nice much more effectively than previously where by
>>> their very nature, low priority tasks were balanced more frequently
>>> and ended up getting their own cpu.
>>
>>
>>
>> I can't follow the logic here and I certainly don't see much
>> difference in practice.
>
>
> I think I've figured out why I'm not seeing much difference in
> practice. I'm only testing on 2 CPU systems and it seems to me that
> the main difference that the SMP nice patch will have is in selecting
> which CPU to steal tasks from (grabbing the one with the highest
> priority tasks) and this is a non issue on a 2 CPU system. :-(
>
> So I should revise my statement to say that it doesn't make much
> difference if there's only 2 CPUs.
>
If nothing's niced, why would it be affecting scheduling decisions at all?
That seems broken to me ?
M.
next prev parent reply other threads:[~2006-01-11 6:22 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-01-11 1:14 -mm seems significanty slower than mainline on kernbench Martin Bligh
2006-01-11 1:31 ` Andrew Morton
2006-01-11 1:41 ` Martin Bligh
2006-01-11 1:48 ` Andrew Morton
2006-01-11 1:49 ` Con Kolivas
2006-01-11 2:38 ` Peter Williams
2006-01-11 3:07 ` Con Kolivas
2006-01-11 3:12 ` Martin Bligh
2006-01-11 3:40 ` Peter Williams
2006-01-11 3:49 ` Con Kolivas
2006-01-11 4:33 ` Peter Williams
2006-01-11 5:14 ` Peter Williams
2006-01-11 6:21 ` Martin J. Bligh [this message]
2006-01-11 12:24 ` Peter Williams
2006-01-11 14:29 ` Con Kolivas
2006-01-11 22:05 ` Peter Williams
2006-01-12 0:54 ` Peter Williams
2006-01-12 1:18 ` Con Kolivas
2006-01-12 1:29 ` Peter Williams
2006-01-12 1:36 ` Con Kolivas
2006-01-12 2:23 ` Peter Williams
2006-01-12 2:26 ` Martin Bligh
2006-01-12 6:39 ` Con Kolivas
2006-01-23 19:28 ` Martin Bligh
2006-01-24 1:25 ` Peter Williams
2006-01-24 3:50 ` Peter Williams
2006-01-24 4:41 ` Martin J. Bligh
2006-01-24 6:22 ` Peter Williams
2006-01-24 6:42 ` Martin J. Bligh
2006-01-28 23:20 ` Peter Williams
2006-01-29 0:52 ` Martin J. Bligh
2006-01-12 2:27 ` Con Kolivas
2006-01-12 2:04 ` Martin Bligh
2006-01-12 6:35 ` Martin J. Bligh
2006-01-12 6:41 ` Con Kolivas
2006-01-12 6:54 ` Peter Williams
2006-01-12 18:39 ` Martin Bligh
2006-01-12 20:03 ` Peter Williams
2006-01-12 22:20 ` Peter Williams
2006-01-13 7:06 ` Peter Williams
2006-01-13 12:00 ` Peter Williams
2006-01-13 16:15 ` Martin J. Bligh
2006-01-13 16:26 ` Andy Whitcroft
2006-01-13 17:54 ` Andy Whitcroft
2006-01-13 20:41 ` Martin Bligh
2006-01-14 0:23 ` Peter Williams
2006-01-14 5:03 ` Nick Piggin
2006-01-14 5:40 ` Con Kolivas
2006-01-14 6:05 ` Nick Piggin
2006-01-14 5:53 ` Peter Williams
2006-01-14 6:13 ` Nick Piggin
2006-01-13 22:59 ` Peter Williams
2006-01-14 18:48 ` Martin J. Bligh
2006-01-15 0:05 ` Peter Williams
2006-01-15 2:04 ` Con Kolivas
2006-01-15 2:09 ` [PATCH] sched - remove unnecessary smpnice ifdefs Con Kolivas
2006-01-15 3:50 ` -mm seems significanty slower than mainline on kernbench Ingo Molnar
2006-01-12 1:25 ` Peter Williams
2006-01-11 1:52 ` Andrew Morton
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=43C4A3E9.1040301@google.com \
--to=mbligh@google.com \
--cc=akpm@osdl.org \
--cc=kernel@kolivas.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=pwil3058@bigpond.net.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).