From: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
To: Paul Jackson <pj@engr.sgi.com>
Cc: Jay Lan <jlan@engr.sgi.com>,
guillaume.thouvenin@bull.net, akpm@osdl.org, greg@kroah.com,
linux-kernel@vger.kernel.org, efocht@hpce.nec.com,
linuxram@us.ibm.com, gh@us.ibm.com,
elsa-devel@lists.sourceforge.net
Subject: Re: [patch 1/2] fork_connector: add a fork connector
Date: Wed, 30 Mar 2005 18:14:12 +0400 [thread overview]
Message-ID: <20050330181412.B13722@2ka.mipt.ru> (raw)
In-Reply-To: <20050329140106.2a9b8aa5.pj@engr.sgi.com>; from pj@engr.sgi.com on Tue, Mar 29, 2005 at 02:01:06PM -0800
Sorry for long delay - I was quite far from my test machines.
Here are results:
fork connector with turned off disk writes and direct connector's
methods calls.
pcix$ ./fork_test 100000
Average per process fork+exit time is 505 usecs [diff=50567251, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 512 usecs [diff=51248174, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 504 usecs [diff=50469379, max=100000].
fork connector with turned on disk writes and direct connector's
methods calls.
Each disk write has about 80 bytes which are:
time(&tm);
fprintf(out,
"%.24s : [%x.%x] [seq=%08x, ack=%08x] %s.\n",
ctime(&tm), data->id.idx, data->id.val,
data->seq, data->ack, (char *)data->data);
pcix$ ./fork_test 100000
Average per process fork+exit time is 539 usecs [diff=53944663, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 523 usecs [diff=52378314, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 540 usecs [diff=54078648, max=100000].
CBUS results.
Writing disabled:
pcix$ ./fork_test 100000
Average per process fork+exit time is 451 usecs [diff=45194377, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 454 usecs [diff=45416470, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 448 usecs [diff=44863153, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 453 usecs [diff=45312870, max=100000].
pcix$
Writing enabled like described above:
pcix$ ./fork_test 100000
Average per process fork+exit time is 456 usecs [diff=45680384, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 455 usecs [diff=45590682, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 453 usecs [diff=45376436, max=100000].
pcix$
fork connector is not compiled in:
Average per process fork+exit time is 452 usecs [diff=45280538, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 446 usecs [diff=44687388, max=100000].
pcix$ ./fork_test 100000
Average per process fork+exit time is 445 usecs [diff=44505999, max=100000].
pcix$ ./fork_test 100000
With 80 bytes write per fork with CBUS it takes from 0.5% to 2.5%.
So it still can be used for accounting :)
--
Evgeniy Polyakov
next prev parent reply other threads:[~2005-03-30 14:15 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-03-25 10:03 [patch 1/2] fork_connector: add a fork connector Guillaume Thouvenin
2005-03-25 22:45 ` dean gaudet
2005-03-28 21:42 ` Paul Jackson
2005-03-29 7:04 ` Evgeniy Polyakov
2005-03-29 7:02 ` Greg KH
2005-03-29 7:10 ` Evgeniy Polyakov
2005-03-29 8:49 ` Paul Jackson
2005-03-29 9:17 ` Guillaume Thouvenin
2005-03-29 15:23 ` Paul Jackson
2005-03-29 18:44 ` Jay Lan
2005-03-30 1:05 ` Paul Jackson
2005-03-30 5:39 ` Guillaume Thouvenin
2005-03-30 6:35 ` Paul Jackson
2005-03-30 10:25 ` Herbert Xu
2005-03-30 10:57 ` Evgeniy Polyakov
2005-03-30 11:01 ` Guillaume Thouvenin
2005-04-01 3:26 ` Drew Hess
2005-03-29 10:29 ` Evgeniy Polyakov
2005-03-29 17:03 ` Paul Jackson
2005-03-29 21:09 ` Jay Lan
2005-03-29 22:01 ` Paul Jackson
2005-03-30 14:14 ` Evgeniy Polyakov [this message]
2005-03-30 20:56 ` Paul Jackson
2005-03-30 6:06 ` dean gaudet
2005-03-30 6:25 ` Paul Jackson
2005-03-30 6:38 ` Guillaume Thouvenin
2005-03-30 18:11 ` Jay Lan
2005-03-29 8:05 ` Guillaume Thouvenin
2005-03-29 14:47 ` Paul Jackson
2005-03-29 12:51 ` Guillaume Thouvenin
2005-03-29 15:35 ` Paul Jackson
2005-03-30 5:52 ` Guillaume Thouvenin
2005-03-30 6:41 ` Paul Jackson
-- strict thread matches above, loose matches on Subject: below --
2005-03-17 9:04 Guillaume Thouvenin
2005-03-17 16:56 ` Jesse Barnes
2005-03-17 21:38 ` Evgeniy Polyakov
2005-03-17 22:05 ` Jesse Barnes
2005-03-21 8:23 ` Guillaume Thouvenin
2005-03-21 12:48 ` Guillaume Thouvenin
2005-03-21 20:52 ` Ram
2005-03-22 4:36 ` Evgeniy Polyakov
2005-03-22 18:40 ` Ram
2005-03-22 7:07 ` Guillaume Thouvenin
2005-03-22 18:15 ` Jay Lan
2005-03-23 8:15 ` Guillaume Thouvenin
2005-03-22 18:26 ` Ram
2005-03-22 19:22 ` Evgeniy Polyakov
2005-03-22 19:18 ` Ram
2005-03-22 20:25 ` Evgeniy Polyakov
2005-03-22 20:42 ` Ram
2005-03-23 4:52 ` Evgeniy Polyakov
2005-03-22 22:51 ` Jay Lan
2005-03-22 23:51 ` Jay Lan
2005-03-23 5:01 ` Evgeniy Polyakov
[not found] ` <1111557106.23532.65.camel@uganda>
2005-03-23 19:00 ` Ram
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=20050330181412.B13722@2ka.mipt.ru \
--to=johnpol@2ka.mipt.ru \
--cc=akpm@osdl.org \
--cc=efocht@hpce.nec.com \
--cc=elsa-devel@lists.sourceforge.net \
--cc=gh@us.ibm.com \
--cc=greg@kroah.com \
--cc=guillaume.thouvenin@bull.net \
--cc=jlan@engr.sgi.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxram@us.ibm.com \
--cc=pj@engr.sgi.com \
/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