public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Bob <spam@homeurl.co.uk>
To: linux-kernel@vger.kernel.org
Subject: Re: Kernel 2.6 SMP very slow with ServerWorks LE Chipset
Date: Fri, 08 Dec 2006 13:34:58 +0800	[thread overview]
Message-ID: <4578F982.70506@homeurl.co.uk> (raw)
In-Reply-To: <20061207110737.6c506c98@localhost.localdomain>

Alan wrote:
>> As a test of raw CPU power I've been decompressing the kernel tree, with 
>> a UP 2.6 kernel this takes about 1m 15s, I don't know if bz2 is 
>> multithreaded but even if it's not I would expect a slight speed increase 
>> but in fact with a SMP 2.6 kernel it take 13 ~ 26m, with a SMP 2.4 
>> kernel it takes 1m 28s and with a 2.4 UP 1m 35s. 
> 
> The 2.4 numbers look correct (slightly slower), the 2.6 numbers do not.
> 
> 
> Nothing obviously wrong from the traces however. If you pin the bzip to a
> given processor do you get different results according to which CPU ?
> 
> (see man taskset for info on the commands)
> 
> If you get very different times on the two processors that will be very
> useful information.

Mmm CPU 0 in 1m 32s, but I was running things like ps and taskset in 
another terminal to verify affinity, CPU 1 still hadn't finished in 40m 
when I killed it. I've downloaded the LFDK mentioned in the other post 
and will post the results when I have them.

Thanks for replying,

This is the script I'm using to time decompression
nas:~# cat ./cputest.sh
date
taskset -c 1 nice -n -8 tar -xjf /root/linux-2.6.18.1.tar.bz2
date
uname -a
rm -r /root/linux-2.6.18.1/
nas:~#

both the tar and bzip2 processes have their affinity and priority set by 
the above command.

  reply	other threads:[~2006-12-08  5:34 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-12-07  5:43 Kernel 2.6 SMP very slow with ServerWorks LE Chipset Bob
2006-12-07 11:07 ` Alan
2006-12-08  5:34   ` Bob [this message]
2006-12-07 11:15 ` Arjan van de Ven
2006-12-08  5:50   ` Bob
2006-12-14  3:09 ` Bob
2006-12-14 11:03   ` Need to enable caches in SMP ? (was Kernel 2.6 SMP very slow with ServerWorks LE Chipset) Alan
2006-12-14 17:29     ` Jeff V. Merkey
2006-12-15 11:05       ` Bob

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=4578F982.70506@homeurl.co.uk \
    --to=spam@homeurl.co.uk \
    --cc=linux-kernel@vger.kernel.org \
    /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