From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sowmini Varadhan Subject: Re: [RFC] Kernel unaligned access at __skb_flow_dissect Date: Fri, 29 Jan 2016 14:44:34 -0500 Message-ID: <20160129194434.GC17127@oracle.com> References: <20160129180651.GA17127@oracle.com> <1454092428.7627.52.camel@edumazet-glaptop2.roam.corp.google.com> <20160129.110634.1678121731419241040.davem@davemloft.net> <1454096250.7627.60.camel@edumazet-glaptop2.roam.corp.google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Miller , netdev@vger.kernel.org To: Eric Dumazet Return-path: Received: from aserp1040.oracle.com ([141.146.126.69]:23213 "EHLO aserp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756689AbcA2Ton (ORCPT ); Fri, 29 Jan 2016 14:44:43 -0500 Content-Disposition: inline In-Reply-To: <1454096250.7627.60.camel@edumazet-glaptop2.roam.corp.google.com> Sender: netdev-owner@vger.kernel.org List-ID: On (01/29/16 11:37), Eric Dumazet wrote: > > I have no idea why reading iph->saddr or iph->daddr would not hit the > problem, but accessing the 32bit ipv6 flow label would be an issue. > > Something is fishy. I was wondering about this myself. Even on sparc, I only first ran into the errors for ipv6. I dont know if the fact that the saddr is memcpy'ed masks the error (even though the problem is still there). But doing the check of: if (!IS_ALIGNED(..)) on the iph->saddr in the code does result in a positive. > But really adding unaligned() accesses in flow dissector would slow it > quite a lot on MIPS and others.