Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267880AbUIPJ3u (ORCPT ); Thu, 16 Sep 2004 05:29:50 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267882AbUIPJ3u (ORCPT ); Thu, 16 Sep 2004 05:29:50 -0400 Received: from sj-iport-1-in.cisco.com ([171.71.176.70]:31396 "EHLO sj-iport-1.cisco.com") by vger.kernel.org with ESMTP id S267880AbUIPJ3r (ORCPT ); Thu, 16 Sep 2004 05:29:47 -0400 X-BrightmailFiltered: true Message-Id: <5.1.0.14.2.20040916192213.04240008@171.71.163.14> X-Mailer: QUALCOMM Windows Eudora Version 5.1 Date: Thu, 16 Sep 2004 19:29:43 +1000 To: Jeff Garzik , "David S. Miller" From: Lincoln Dale Subject: Re: The ultimate TOE design Cc: alan@lxorguk.ukuu.org.uk, paul@clubi.ie, netdev@oss.sgi.com, leonid.grossman@s2io.com, linux-kernel@vger.kernel.org In-Reply-To: <4148B2E5.50106@pobox.com> References: <20040915141346.5c5e5377.davem@davemloft.net> <4148991B.9050200@pobox.com> <1095275660.20569.0.camel@localhost.localdomain> <4148A90F.80003@pobox.com> <20040915140123.14185ede.davem@davemloft.net> <20040915210818.GA22649@havoc.gtf.org> <20040915141346.5c5e5377.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1563 Lines: 42 not that i disagree with the general idea and rationale, but reality is what it is today for some reasons: At 07:23 AM 16/09/2004, Jeff Garzik wrote: > Jeff, who would love to have a bunch of Athlons > on PCI cards to play with. . . . this ignore the realities of power restrictions of PCI today . . . sure, one could create a PCI card that takes a power-connector, but that don't scale so well either . . . At 07:29 AM 16/09/2004, David S. Miller wrote: >I think a better goal is "offload 90+% of the net stack cost" which >is effectively what TSO does on the send side. > >This is why these discussions are so circular. TSO works on LAN-like environments (zero latency, minimal drop), it doesn't work so well across the internet . . . i believe that there are better alternatives than TSO, but it involves NICs having decent scatter-gather DMA engines and being able to be handled multiple transactions (packets/frames) at once. in theory, NICs like tg2/tg3 should be capable of implementing something like this -- if one could get to the ucode on the embedded cores. at least with PCI Express the general architecture of a PC starts to have a hope of keeping up with Moore's law. the same couldn't be said prior to DDR-SDRAM and higher front-side-bus frequencies. cheers, lincoln. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/