From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: rps perfomance WAS(Re: rps: question Date: Fri, 16 Apr 2010 07:18:26 +0200 Message-ID: <1271395106.16881.3645.camel@edumazet-laptop> References: <1271268242.16881.1719.camel@edumazet-laptop> <1271271222.4567.51.camel@bigi> <20100415.014857.168270765.davem@davemloft.net> <1271332528.4567.150.camel@bigi> <4BC741AE.3000108@hp.com> <1271362581.23780.12.camel@bigi> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: hadi@cyberus.ca, Rick Jones , David Miller , therbert@google.com, netdev@vger.kernel.org, robert@herjulf.net, andi@firstfloor.org To: Changli Gao Return-path: Received: from mail-bw0-f225.google.com ([209.85.218.225]:50111 "EHLO mail-bw0-f225.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752629Ab0DPFSh (ORCPT ); Fri, 16 Apr 2010 01:18:37 -0400 Received: by bwz25 with SMTP id 25so2533879bwz.28 for ; Thu, 15 Apr 2010 22:18:35 -0700 (PDT) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: Le vendredi 16 avril 2010 =C3=A0 07:56 +0800, Changli Gao a =C3=A9crit = : > On Fri, Apr 16, 2010 at 4:16 AM, jamal wrote: > > > > Sounds interesting. > > Wikipedia information overload. Any arch description of the HP9000? > > Did your scheme use IPIs to message the other CPUs? > > >=20 > If you doubt the cost of smp_call_function_single(), how about having > a try with my another patch, which implements the similar of RPS, but > uses kernel threads instead, so no explicit IPI. >=20 > http://patchwork.ozlabs.org/patch/38319/ >=20 >=20 Come on Changli. How do you wake up a thread on a remote cpu ? To answer Jamal question, we need to answer to Jamal question, that is timing cost of IPIS. A kernel module might do this, this could be integrated in perf bench s= o that we can regression tests upcoming kernels.