From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Horman Subject: Re: [Patch 3/5] Network Drop Monitor: Adding kfree_skb_clean for non-drops and modifying end-of-line points for skbs Date: Wed, 11 Mar 2009 20:30:59 -0400 Message-ID: <20090312003059.GA12554@localhost.localdomain> References: <20090303170307.GD1480@hmsreliant.think-freely.org> <20090305210105.GC12835@hmsreliant.think-freely.org> <20090311.091830.152269823.davem@davemloft.net> <49B7E827.3070203@cosmosbay.com> <20090311170701.GA17005@hmsreliant.think-freely.org> <20090311213902.GA18878@ioremap.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Eric Dumazet , David Miller , netdev@vger.kernel.org, kuznet@ms2.inr.ac.ru, pekkas@netcore.fi, jmorris@namei.org, yoshfuji@linux-ipv6.org, kaber@trash.net To: Evgeniy Polyakov Return-path: Received: from charlotte.tuxdriver.com ([70.61.120.58]:46012 "EHLO smtp.tuxdriver.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753526AbZCLAbU (ORCPT ); Wed, 11 Mar 2009 20:31:20 -0400 Content-Disposition: inline In-Reply-To: <20090311213902.GA18878@ioremap.net> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, Mar 12, 2009 at 12:39:02AM +0300, Evgeniy Polyakov wrote: > On Wed, Mar 11, 2009 at 01:07:01PM -0400, Neil Horman (nhorman@tuxdriver.com) wrote: > > > I was wondering if packets droped at NIC level are handled by "Network Drop Monitor". > > > (See recent discussion about Multicast packet loss) > > > > > > I believe not, maybe I missed something ... > > > > > No, unfortunately, theres no really good way to detect drops at the driver level > > without polling them periodically, which I think is somewhat in opposition to the > > point of this utility. I'd welcome suggestions however. > > It could be possible to check softnet stats difference when napi is > scheduled or drop monitor is about to send a packet to the userspace > about different drop. > Those are both good ideas. I think I like the former suggestion, since the latter requires that a drop in the stack must occur to detect a stack closer to the hardware. Thank you, I'll keep that in mind for a subsequent enhancement. Regards Neil > -- > Evgeniy Polyakov >