From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nikolay Aleksandrov Subject: Re: [RFC PATCH] netfilter: nf_tables: add new write expression Date: Sat, 15 Feb 2014 14:38:21 +0100 Message-ID: <52FF6DCD.6090509@redhat.com> References: <1392470242-4515-1-git-send-email-nikolay@redhat.com> <20140215133613.GB28751@breakpoint.cc> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: netfilter-devel@vger.kernel.org, pablo@netfilter.org, kaber@trash.net To: Florian Westphal Return-path: Received: from mx1.redhat.com ([209.132.183.28]:28903 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752928AbaBONjA (ORCPT ); Sat, 15 Feb 2014 08:39:00 -0500 In-Reply-To: <20140215133613.GB28751@breakpoint.cc> Sender: netfilter-devel-owner@vger.kernel.org List-ID: On 02/15/2014 02:36 PM, Florian Westphal wrote: > Nikolay Aleksandrov wrote: >> The new "write" expression can be used to manipulate packet data. > > Thanks. This is very useful feature (needed also for TCPMSS mangling > for instance). > > General question: How are/should packet checksums be handled after packet > content was altered? > hehe very good question, right now I do it with a hacked version of this patch, but I think we can add some more generic checksum op where you can choose which checksum to recompute. I've also thought about making the destination (where to write the checksum) variable for some corner cases, but this is a stretch. Cheers, Nik