From: linas@austin.ibm.com (Linas Vepstas)
To: Jan-Bernd Themann <ossthema@de.ibm.com>
Cc: Thomas Klein <tklein@de.ibm.com>,
Jan-Bernd Themann <themann@de.ibm.com>,
netdev <netdev@vger.kernel.org>,
linux-kernel <linux-kernel@vger.kernel.org>,
linux-ppc <linuxppc-dev@ozlabs.org>, Bodo Eggert <7eggert@gmx.de>,
Christoph Raisch <raisch@de.ibm.com>,
Marcus Eder <meder@de.ibm.com>,
Stefan Roscher <stefan.roscher@de.ibm.com>
Subject: Re: RFC: issues concerning the next NAPI interface
Date: Fri, 24 Aug 2007 16:35:14 -0500 [thread overview]
Message-ID: <20070824213514.GL4282@austin.ibm.com> (raw)
In-Reply-To: <46CF499C.60009@de.ibm.com>
On Fri, Aug 24, 2007 at 11:11:56PM +0200, Jan-Bernd Themann wrote:
> (when they are available for
> POWER in our case).
hrtimer worked fine on the powerpc cell arch last summer.
I assume they work on p5 and p6 too, no ??
> I tried to implement something with "normal" timers, but the result
> was everything but great. The timers seem to be far too slow.
> I'm not sure if it helps to increase it from 1000HZ to 2500HZ
> or more.
Heh. Do the math. Even on 1gigabit cards, that's not enough:
(1gigabit/sec) x (byte/8 bits) x (packet/1500bytes) x (sec/1000 jiffy)
is 83 packets a jiffy (for big packets, even more for small packets,
and more again for 10 gigabit cards). So polling once per jiffy is a
latency disaster.
--linas
WARNING: multiple messages have this Message-ID (diff)
From: linas@austin.ibm.com (Linas Vepstas)
To: Jan-Bernd Themann <ossthema@de.ibm.com>
Cc: Bodo Eggert <7eggert@gmx.de>, netdev <netdev@vger.kernel.org>,
Thomas Klein <tklein@de.ibm.com>,
Jan-Bernd Themann <themann@de.ibm.com>,
linux-kernel <linux-kernel@vger.kernel.org>,
linux-ppc <linuxppc-dev@ozlabs.org>,
Christoph Raisch <raisch@de.ibm.com>,
Marcus Eder <meder@de.ibm.com>,
Stefan Roscher <stefan.roscher@de.ibm.com>
Subject: Re: RFC: issues concerning the next NAPI interface
Date: Fri, 24 Aug 2007 16:35:14 -0500 [thread overview]
Message-ID: <20070824213514.GL4282@austin.ibm.com> (raw)
In-Reply-To: <46CF499C.60009@de.ibm.com>
On Fri, Aug 24, 2007 at 11:11:56PM +0200, Jan-Bernd Themann wrote:
> (when they are available for
> POWER in our case).
hrtimer worked fine on the powerpc cell arch last summer.
I assume they work on p5 and p6 too, no ??
> I tried to implement something with "normal" timers, but the result
> was everything but great. The timers seem to be far too slow.
> I'm not sure if it helps to increase it from 1000HZ to 2500HZ
> or more.
Heh. Do the math. Even on 1gigabit cards, that's not enough:
(1gigabit/sec) x (byte/8 bits) x (packet/1500bytes) x (sec/1000 jiffy)
is 83 packets a jiffy (for big packets, even more for small packets,
and more again for 10 gigabit cards). So polling once per jiffy is a
latency disaster.
--linas
next prev parent reply other threads:[~2007-08-24 21:35 UTC|newest]
Thread overview: 94+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <8VHRR-45R-17@gated-at.bofh.it>
[not found] ` <8VKwj-8ke-27@gated-at.bofh.it>
2007-08-24 19:04 ` RFC: issues concerning the next NAPI interface Bodo Eggert
2007-08-24 19:04 ` Bodo Eggert
2007-08-24 20:42 ` Linas Vepstas
2007-08-24 20:42 ` Linas Vepstas
2007-08-24 21:11 ` Jan-Bernd Themann
2007-08-24 21:11 ` Jan-Bernd Themann
2007-08-24 21:35 ` Linas Vepstas [this message]
2007-08-24 21:35 ` Linas Vepstas
[not found] ` <E1IOeSm-0000bm-Jo__24045.532072387$1187982363$gmane$org@be1.lrz>
2007-08-24 20:24 ` Stephen Hemminger
2007-08-25 2:10 Mitchell Erblich
-- strict thread matches above, loose matches on Subject: below --
2007-08-24 13:59 Jan-Bernd Themann
2007-08-24 13:59 ` Jan-Bernd Themann
2007-08-24 15:37 ` akepner
2007-08-24 15:37 ` akepner
2007-08-24 15:47 ` Jan-Bernd Themann
2007-08-24 15:47 ` Jan-Bernd Themann
2007-08-24 15:52 ` Stephen Hemminger
2007-08-24 15:52 ` Stephen Hemminger
2007-08-24 16:50 ` David Stevens
2007-08-24 16:50 ` David Stevens
2007-08-24 21:44 ` David Miller
2007-08-24 21:44 ` David Miller
2007-08-24 21:51 ` Linas Vepstas
2007-08-24 21:51 ` Linas Vepstas
2007-08-24 16:51 ` Linas Vepstas
2007-08-24 16:51 ` Linas Vepstas
2007-08-24 17:07 ` Rick Jones
2007-08-24 17:07 ` Rick Jones
2007-08-24 17:45 ` Shirley Ma
2007-08-24 17:45 ` Shirley Ma
2007-08-24 17:16 ` James Chapman
2007-08-24 17:16 ` James Chapman
2007-08-24 18:11 ` Jan-Bernd Themann
2007-08-24 18:11 ` Jan-Bernd Themann
2007-08-24 21:47 ` David Miller
2007-08-24 21:47 ` David Miller
2007-08-24 22:06 ` akepner
2007-08-24 22:06 ` akepner
2007-08-26 19:36 ` James Chapman
2007-08-26 19:36 ` James Chapman
2007-08-27 1:58 ` David Miller
2007-08-27 1:58 ` David Miller
2007-08-27 9:47 ` Jan-Bernd Themann
2007-08-27 9:47 ` Jan-Bernd Themann
2007-08-27 20:37 ` David Miller
2007-08-27 20:37 ` David Miller
2007-08-28 11:19 ` Jan-Bernd Themann
2007-08-28 11:19 ` Jan-Bernd Themann
2007-08-28 20:21 ` David Miller
2007-08-28 20:21 ` David Miller
2007-08-29 7:10 ` Jan-Bernd Themann
2007-08-29 7:10 ` Jan-Bernd Themann
2007-08-29 8:15 ` James Chapman
2007-08-29 8:15 ` James Chapman
2007-08-29 8:43 ` Jan-Bernd Themann
2007-08-29 8:43 ` Jan-Bernd Themann
2007-08-29 8:29 ` David Miller
2007-08-29 8:29 ` David Miller
2007-08-29 8:31 ` Jan-Bernd Themann
2007-08-29 8:31 ` Jan-Bernd Themann
2007-08-27 15:51 ` James Chapman
2007-08-27 15:51 ` James Chapman
2007-08-27 16:02 ` Jan-Bernd Themann
2007-08-27 16:02 ` Jan-Bernd Themann
2007-08-27 17:05 ` James Chapman
2007-08-27 17:05 ` James Chapman
2007-08-27 21:02 ` David Miller
2007-08-27 21:02 ` David Miller
2007-08-27 21:41 ` James Chapman
2007-08-27 21:41 ` James Chapman
2007-08-27 21:56 ` David Miller
2007-08-27 21:56 ` David Miller
2007-08-28 9:22 ` James Chapman
2007-08-28 9:22 ` James Chapman
2007-08-28 11:48 ` Jan-Bernd Themann
2007-08-28 11:48 ` Jan-Bernd Themann
2007-08-28 12:16 ` Evgeniy Polyakov
2007-08-28 12:16 ` Evgeniy Polyakov
2007-08-28 14:55 ` James Chapman
2007-08-28 14:55 ` James Chapman
2007-08-28 11:21 ` Jan-Bernd Themann
2007-08-28 11:21 ` Jan-Bernd Themann
2007-08-28 20:25 ` David Miller
2007-08-28 20:25 ` David Miller
2007-08-28 20:27 ` David Miller
2007-08-28 20:27 ` David Miller
2007-08-24 16:45 ` Linas Vepstas
2007-08-24 16:45 ` Linas Vepstas
2007-08-24 21:43 ` David Miller
2007-08-24 21:43 ` David Miller
2007-08-24 21:32 ` David Miller
2007-08-24 21:32 ` David Miller
2007-08-24 21:37 ` David Miller
2007-08-24 21:37 ` David Miller
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=20070824213514.GL4282@austin.ibm.com \
--to=linas@austin.ibm.com \
--cc=7eggert@gmx.de \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@ozlabs.org \
--cc=meder@de.ibm.com \
--cc=netdev@vger.kernel.org \
--cc=ossthema@de.ibm.com \
--cc=raisch@de.ibm.com \
--cc=stefan.roscher@de.ibm.com \
--cc=themann@de.ibm.com \
--cc=tklein@de.ibm.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.