From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: Softirqs without captured packets Date: Sat, 09 Feb 2013 17:16:02 -0800 Message-ID: <1360458962.6696.38.camel@edumazet-glaptop> References: <20130208.145728.1110053781319882182.davem@davemloft.net> <20130208.151523.798634382725569230.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: David Miller , netdev@vger.kernel.org To: Javier Domingo Return-path: Received: from mail-da0-f53.google.com ([209.85.210.53]:36590 "EHLO mail-da0-f53.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760894Ab3BJBQH (ORCPT ); Sat, 9 Feb 2013 20:16:07 -0500 Received: by mail-da0-f53.google.com with SMTP id w3so22748dad.40 for ; Sat, 09 Feb 2013 17:16:05 -0800 (PST) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On Sat, 2013-02-09 at 21:27 +0100, Javier Domingo wrote: > Can you tell me something about the traces I introduced? > > I tried to insert the trace the cleanest way possible. I patched > dev.c, in the measurement places and tried to make the module the more > independent I could. It makes the measure with sched_clock() and in > each round, it sums up the work done. > > I have supposed that that softirq is just for capturing packets, so > that is why I asked my first question, Seems pretty easy to understand to me. A network device receives and transmits packets. n->poll() call done in net_rx_action() can both : - Receive packets - Perform the TX completion for most devices (some of them still use a hard IRQ driven TX completion path) The return value is only about receive part. (The TX completion is not limited by a per round quota)