From: bert hubert <bert.hubert@netherlabs.nl>
To: linux-kernel@vger.kernel.org
Cc: george@mvista.com
Subject: gettimeofday order of magnitude slower with pmtimer, which is default
Date: Mon, 20 Mar 2006 13:24:49 +0100 [thread overview]
Message-ID: <20060320122449.GA29718@outpost.ds9a.nl> (raw)
Hi everybody,
For my open source nameserver (http://www.powerdns.com) I need to do quite a
number of gettimeofday calls. I've pared it down to almost the bare minimum
of 1 gettimeofday per packet sent and received. With these calls, I can make
stats like http://ds9a.nl/tmp/rrd/ which my users need so they can verify
the proper performance of the nameserver.
Yesterday, together with Zwane, I discovered each gettimeofday call costs me
4 usec on some boxes and almost nothing on others. We did a fruitless chase
for vsyscall/sysenter happening but the problem turned out to be
CONFIG_X86_PM_TIMER.
This problem has been discussed before
http://www.ussg.iu.edu/hypermail/linux/kernel/0411.1/2135.html
Not only is the pm timer slow by design, it also needs to be read multiple
times to work around a bug in certain hardware.
What is new is that this option is now dependent on CONFIG_EMBEDDED. Unless
you select this option, the PM Timer will always be used.
Would a patch removing the link to EMBEDDED and adding a warning that while
this timer is of high quality, it is slow, be welcome?
Thanks.
--
http://www.PowerDNS.com Open source, database driven DNS Software
http://netherlabs.nl Open and Closed source services
next reply other threads:[~2006-03-20 12:24 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-03-20 12:24 bert hubert [this message]
2006-03-20 14:50 ` Andreas Mohr
2006-03-20 15:24 ` Con Kolivas
2006-03-21 1:26 ` OGAWA Hirofumi
2006-03-21 0:40 ` kernel
2006-03-21 2:59 ` OGAWA Hirofumi
2006-03-21 3:09 ` Con Kolivas
2006-03-21 8:53 ` Andreas Mohr
2006-03-21 9:06 ` Arjan van de Ven
2006-03-21 11:58 ` Con Kolivas
2006-03-21 12:04 ` Arjan van de Ven
2006-03-21 12:07 ` Con Kolivas
2006-03-21 19:23 ` john stultz
2006-03-21 21:19 ` OGAWA Hirofumi
2006-03-22 0:21 ` Con Kolivas
2006-03-22 18:49 ` [PATCH] PM-Timer: doesn't use workaround if chipset is not buggy OGAWA Hirofumi
2006-03-22 21:46 ` Andrew Morton
2006-03-23 7:31 ` OGAWA Hirofumi
2006-03-23 7:49 ` Andrew Morton
2006-03-23 17:04 ` Andreas Mohr
2006-03-23 18:21 ` OGAWA Hirofumi
2006-03-30 11:53 ` Andreas Mohr
2006-03-30 15:37 ` OGAWA Hirofumi
2006-03-30 16:02 ` Andreas Mohr
2006-03-25 12:00 ` bert hubert
2006-03-22 19:12 ` gettimeofday order of magnitude slower with pmtimer, which is default Avi Kivity
2006-03-22 19:54 ` OGAWA Hirofumi
2006-03-22 20:05 ` john stultz
2006-03-21 19:34 ` john stultz
2006-03-21 5:33 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=20060320122449.GA29718@outpost.ds9a.nl \
--to=bert.hubert@netherlabs.nl \
--cc=george@mvista.com \
--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
Powered by JetHome