From mboxrd@z Thu Jan 1 00:00:00 1970 From: Shaohua Li Subject: Re: [PATCH V2 net] net: fix tcp reset packet flowlabel for ipv6 Date: Tue, 25 Jul 2017 14:32:15 -0700 Message-ID: <20170725213215.xfccqe4igfw3mui5@kernel.org> References: <25a0abdf61ab4698773d417af9e49fcd6e6559a4.1500393028.git.shli@fb.com> <20170724.133455.562898067272108474.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Miller , Linux Kernel Network Developers , kernel-team , Shaohua Li , Eric Dumazet , flo@fourcot.fr To: Cong Wang Return-path: Received: from mail.kernel.org ([198.145.29.99]:48240 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750880AbdGYVcR (ORCPT ); Tue, 25 Jul 2017 17:32:17 -0400 Content-Disposition: inline In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On Mon, Jul 24, 2017 at 04:12:53PM -0700, Cong Wang wrote: > On Mon, Jul 24, 2017 at 1:34 PM, David Miller wrote: > > From: Shaohua Li > > Date: Tue, 18 Jul 2017 12:03:37 -0700 > > > >> + /* Since this is being sent on the wire obfuscate hash a bit > >> + * to minimize possbility that any useful information to an > >> + * attacker is leaked. Only lower 20 bits are relevant. > >> + */ > >> + rol32(hash, 16); > > > > This doesn't help anything at all. > > I believe the above code is copy-n-pasted from ip6_make_flowlabel() > (with few adjustments). Don't know why we can't refactor that function > for reuse. There are just several lines of code, I really don't want to add adhoc if-else for fast path. Thanks, Shaohua