From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nikolay Aleksandrov Subject: Re: Race with ip=dhcp bootparameter in ip_rcv_finish on am335x Date: Wed, 23 Sep 2015 12:36:13 +0200 Message-ID: <5602809D.2020500@cumulusnetworks.com> References: <20150923102953.GA8918@omega> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit To: Alexander Aring , netdev@vger.kernel.org Return-path: Received: from mail-wi0-f176.google.com ([209.85.212.176]:32894 "EHLO mail-wi0-f176.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753687AbbIWKgQ (ORCPT ); Wed, 23 Sep 2015 06:36:16 -0400 Received: by wiclk2 with SMTP id lk2so232552269wic.0 for ; Wed, 23 Sep 2015 03:36:15 -0700 (PDT) In-Reply-To: <20150923102953.GA8918@omega> Sender: netdev-owner@vger.kernel.org List-ID: On 09/23/2015 12:29 PM, Alexander Aring wrote: > Hi, > > I am using am335x (beaglebone white) with "ip=dhcp" inside my > bootparameters. It seems there exists some race, rarely it boots fine > and often it crashed with following messages: > > First (Without successful dhcp ip assign and is very rarely): > > [ 6.249083] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready > [ 6.398943] . > [ 7.400822] Unable to handle kernel paging request at virtual address 7402a8c8 > [ 7.408419] pgd = c0004000 > [ 7.411351] [7402a8c8] *pgd=00000000 > [ 7.415121] Internal error: Oops: 5 [#1] SMP ARM > [ 7.419968] Modules linked in: > [ 7.423197] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W 4.2.0-11195-ge81be90 #124 > [ 7.432118] Hardware name: Generic AM33XX (Flattened Device Tree) > [ 7.438509] task: c0a996c8 ti: c0a92000 task.ti: c0a92000 > [ 7.444189] PC is at ip_route_input_noref+0x620/0xeb4 > [ 7.449491] LR is at trace_hardirqs_on_caller+0xb0/0x1e8 <<>> > [ 7.759169] [] (ip_route_input_noref) from [] (ip_rcv_finish+0xd4/0x67c) <<>> > > -------- > > I tried to bisect this issue without success, sometimes it's booting fine > and sometimes I got the above messages. Could anybody help me here to > fixing this issue? > > - Alex Hi Alex, This should be fixed by commit bde6f9ded1bd ("net: Initialize table in fib result") in net-next. Cheers, Nik