From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Tim Sander" Subject: Re: infinite spin in RT when booting with DHCP on Date: Thu, 2 Feb 2012 13:38:44 +0100 Message-ID: <201202021338.44950.tim.sander@hbm.com> References: <4F292FE0.7090302@digi.com> Mime-Version: 1.0 Content-Type: text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: , , , To: "Hector Palacios" Return-path: Received: from relay.medianet-world.de ([213.157.0.172]:43722 "HELO relay.medianet-world.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1751249Ab2BBMjm convert rfc822-to-8bit (ORCPT ); Thu, 2 Feb 2012 07:39:42 -0500 Content-Class: urn:content-classes:message In-reply-to: <4F292FE0.7090302@digi.com> Sender: linux-rt-users-owner@vger.kernel.org List-ID: Hi > When booting my platform with DHCP on, the DHCP request is sent by th= e > network driver before the PHY has even started the autonegotiation. > Since the PHY is not ready, the TX interrupt returns with NETDEV_TX_B= USY > but the softirq [sirq-net-tx] seems to have entered an infinite spin,= as > my system is practically hung and 'top' reveals [sirq-net-tx/0] is > consuming 95% of CPU. This is preventing the PHY autonegotiation (whi= ch is > scheduled as a delayed work) to start, so the PHY is never ready and = the > packet never reaches the network. Mh, i think i am hit by the same problem. I have a i.mx 35 and when dhc= p is enabled the ksoftirq is also running wild with 3.0-rt. This also happen= s when static ip is configured and the probably during network transfer the ne= twork interface is reconfigured. Basically it seems the sirq-net-tx thread tr= ies to send a packet on a non configured interface. But one thing that makes me thinking, is that this thing is only happen= ing on arm and not on x86. So Hector what platform are you working on. Are you= by chance using the same fec.c ethernet driver? I have verified that in my case the driver takes always the return stat= ement in line fec.c:247: return NETXDEV_TX_BUSY; =20 It never stops on a breakpoint set on line 250 which shows that the int= erface gets never configured. I have taken some screenshots of my hw debugger: trace:http://private.vlsi.informatik.tu-darmstadt.de/tstone/linux/fec_e= net_start_xmit.png stack:http://private.vlsi.informatik.tu-darmstadt.de/tstone/linux/fec_e= net_start_xmit_stacktrace.png locals:http://private.vlsi.informatik.tu-darmstadt.de/tstone/linux/fec_= enet_start_xmit_stack+locals.png Whats interesting to note is that phy_dev and mii_dev are both null poi= nters. Best regards Tim Hottinger Baldwin Messtechnik GmbH, Im Tiefen See 45, 64293 Darmstadt, = Germany | www.hbm.com=20 Registered as GmbH (German limited liability corporation) in the commer= cial register at the local court of Darmstadt, HRB 1147 =20 Company domiciled in Darmstadt | CEO: Andreas Huellhorst | Chairman of = the board: James Charles Webster Als Gesellschaft mit beschraenkter Haftung eingetragen im Handelsregist= er des Amtsgerichts Darmstadt unter HRB 1147=20 Sitz der Gesellschaft: Darmstadt | Geschaeftsfuehrung: Andreas Huellhor= st | Aufsichtsratsvorsitzender: James Charles Webster The information in this email is confidential. It is intended solely fo= r the addressee. If you are not the intended recipient, please let me k= now and delete this email. Die in dieser E-Mail enthaltene Information ist vertraulich und ledigli= ch f=FCr den Empfaenger bestimmt. Sollten Sie nicht der eigentliche Emp= faenger sein, informieren Sie mich bitte kurz und loeschen diese E-Mail= =2E -- To unsubscribe from this list: send the line "unsubscribe linux-rt-user= s" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html