Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 20 Feb 2003 05:53:17 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 20 Feb 2003 05:53:17 -0500 Received: from gw1.cosmosbay.com ([62.23.185.226]:27849 "EHLO gw1.cosmosbay.com") by vger.kernel.org with ESMTP id ; Thu, 20 Feb 2003 05:53:14 -0500 Message-ID: <008601c2d8cf$a60e4c60$760010ac@edumazet> From: "dada1" To: "Andi Kleen" Cc: "Andi Kleen" , "Simon Kirby" , , , References: <20030219174757.GA5373@netnation.com.suse.lists.linux.kernel> <20030220092043.GA25527@netnation.com> <20030220093422.GA16369@wotan.suse.de> <006301c2d8c8$921c1d10$760010ac@edumazet> <20030220105435.GD10374@wotan.suse.de> Subject: Re: Longstanding networking / SMP issue? (duplextest) Date: Thu, 20 Feb 2003 12:03:06 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 779 Lines: 27 > > Yes IP is best-effort. But this argument cant explain why IP on linux works > > better if we disable SMP on linux... > > It has nothing to do with SMP. The lazy locking dropping packets can happen > on UP kernels too in extreme cases. Also with preempt. > Well, I too noticed that binding NIC IRQS one one CPU echo 1 > /proc/irq/18/smp_affinity helps a lot in normal cases. Some of us want SMP machine because application needs a lot of CPU, but we also need to not drop frames to save the limited network bandwith. Thanks - 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/