From owner-soc-status@FreeBSD.ORG Thu Jun 2 18:22:39 2011 Return-Path: Delivered-To: soc-status@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 3F3801065674; Thu, 2 Jun 2011 18:22:39 +0000 (UTC) (envelope-from syuu@dokukino.com) Received: from mail-wy0-f182.google.com (mail-wy0-f182.google.com [74.125.82.182]) by mx1.freebsd.org (Postfix) with ESMTP id 74B428FC1E; Thu, 2 Jun 2011 18:22:38 +0000 (UTC) Received: by wyf23 with SMTP id 23so1163720wyf.13 for ; Thu, 02 Jun 2011 11:22:37 -0700 (PDT) Received: by 10.216.62.195 with SMTP id y45mr1164000wec.15.1307038957298; Thu, 02 Jun 2011 11:22:37 -0700 (PDT) MIME-Version: 1.0 Received: by 10.216.6.196 with HTTP; Thu, 2 Jun 2011 11:22:17 -0700 (PDT) In-Reply-To: <5054184174934880962@unknownmsgid> References: <8259CBF7-B2E6-49C6-A7C4-6682ECBDBB9F@freebsd.org> <5054184174934880962@unknownmsgid> From: Takuya ASADA Date: Fri, 3 Jun 2011 03:22:17 +0900 Message-ID: To: George Neville-Neil Content-Type: text/plain; charset=UTF-8 Cc: "Robert N. M. Watson" , "soc-status@freebsd.org" , Kazuya Goda Subject: Re: Weekly status report (27th May) X-BeenThere: soc-status@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Summer of Code Status Reports and Discussion List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 02 Jun 2011 18:22:39 -0000 Hi, >> I see the test case isn't written yet, so >> how are you testing these changes? > > I modified libpcap/tcpdump just for the test - it can take extra > argument for filtering queues. > I'll send more detail of it when I get to home. I added three arguments on tcpdump, 'Q' for receive queue mask, 'g' for transmit queue mask, 'V' for other packets. tcpdump takes these args, tell mask params to (modified) libpcap. libpcap calls new ioctls which set queue mask on bpf descriptor. Example usage is something like this: tcpdump -Q 2 # only shows receive queue 2 tcpdump -g 3 # only shows transmit queue 3 Note that, current implementation only works on igb(4). On other devices, if_rxq_num/if_txq_num and m->m_pkthdr.rxqid/m->m_pkthdr.txqid may left uninitialized value, it may causes unexpected behavior on multiqueue bpf.