From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933808AbcKPF6B (ORCPT ); Wed, 16 Nov 2016 00:58:01 -0500 Received: from mga06.intel.com ([134.134.136.31]:42823 "EHLO mga06.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750939AbcKPF57 (ORCPT ); Wed, 16 Nov 2016 00:57:59 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.31,646,1473145200"; d="scan'208";a="786972198" Date: Tue, 15 Nov 2016 13:56:58 +0800 From: Ye Xiaolong To: "Duyck, Alexander H" Cc: "edumazet@google.com" , "torvalds@linux-foundation.org" , "linux-kernel@vger.kernel.org" , "davem@davemloft.net" , "willemb@google.com" , "lkp@01.org" , "ast@kernel.org" Subject: Re: [net] 34fad54c25: kernel BUG at include/linux/skbuff.h:1935! Message-ID: <20161115055658.GD2142@yexl-desktop> References: <582b7c30.nXQXP2V4/6pFiYwt%xiaolong.ye@intel.com> <1479249843.681.152.camel@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <1479249843.681.152.camel@intel.com> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 11/16, Duyck, Alexander H wrote: >On Wed, 2016-11-16 at 05:20 +0800, kernel test robot wrote: >>From what I can tell it looks like the size of the frame is 0x160 hex, >or 352.  For whatever reason we are only pulling 8 bytes into the >header which is giving us an skb->len of 352 (0x160), and a skb- >>data_len of 344 (0x158).  When we go to pull the 14 bytes for the >Ethernet header we end up at a skb->len of 338 (0x152) which is >resulting in the panic. > >The question is how are we coming up with 8 instead of 14 which is the >lowest limit supported by eth_get_headlen? My first thought was there >is an incorrect sizeof(eth) instead of the sizeof(*eth) somewhere in >the code but I can't find anything like that anywhere. > >Is there any way you can provide me with the net/ethernet/eth.o and >drivers/net/ethernet/igb/igb.o files?  With that I can look over the Is vmlinux ok for you? I've sent it to you separately. Btw: I've also tried gcc-5 (Debian 5.4.1-2) 5.4.1 20160904, and the result shows the same failures. Thanks, Xiaolong