From mboxrd@z Thu Jan 1 00:00:00 1970 From: "David S. Miller" Subject: Re: Initial benchmarks of some VJ ideas [mmap memcpy vs copy_to_user]. Date: Mon, 08 May 2006 13:15:47 -0700 (PDT) Message-ID: <20060508.131547.73910171.davem@davemloft.net> References: <20060508122418.GA22554@2ka.mipt.ru> <20060508195132.GB19091@2ka.mipt.ru> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, caitlinb@broadcom.com, kelly@au1.ibm.com, rusty@rustcorp.com.au Return-path: Received: from dsl027-180-168.sfo1.dsl.speakeasy.net ([216.27.180.168]:59272 "EHLO sunset.davemloft.net") by vger.kernel.org with ESMTP id S1750773AbWEHUR2 (ORCPT ); Mon, 8 May 2006 16:17:28 -0400 To: johnpol@2ka.mipt.ru In-Reply-To: <20060508195132.GB19091@2ka.mipt.ru> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org From: Evgeniy Polyakov Date: Mon, 8 May 2006 23:51:32 +0400 > Since nocopy is actually equal to dma into mapped buffer, > so we get something close to 6 times less CPU usage, and if it can be > lineary transferred into performance gain, we found where the most > significant part of VJ channels lives. Van's machines were cpu limited. And once cpu limit was removed, they became bus bandwidth limited. > Unfortunately it is not backward compatible with recv() system call, > and requires major changes in application to use this advantage. I have stopped believing that compatible API for getting top performance in networking receive is possible a very long time ago. ABI change is an absolutely requirement.