From mboxrd@z Thu Jan 1 00:00:00 1970 From: Max Kellermann Subject: Re: [conntrack-utils PATCH r7285 03/11] moved process function pointer to struct local_server Date: Wed, 23 Jan 2008 11:52:40 +0100 Message-ID: <20080123105240.GA9564@swift.blarg.de> References: <20080122141034.30077.12083.stgit@rabbit.intern.cm-ag> <20080122141051.30077.17154.stgit@rabbit.intern.cm-ag> <47971AAD.8080805@netfilter.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: netfilter-devel@vger.kernel.org To: Pablo Neira Ayuso Return-path: Received: from duempel.org ([78.31.71.42]:35802 "HELO duempel.org" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with SMTP id S1753516AbYAWKxR (ORCPT ); Wed, 23 Jan 2008 05:53:17 -0500 Content-Disposition: inline In-Reply-To: <47971AAD.8080805@netfilter.org> Sender: netfilter-devel-owner@vger.kernel.org List-ID: On 2008/01/23 11:45, Pablo Neira Ayuso wrote: > Max Kellermann wrote: > > include/local.h | 10 +++++++--- > > src/local.c | 11 +++++++---- > > src/run.c | 5 +++-- > > 3 files changed, 17 insertions(+), 9 deletions(-) > > This makes local.c less flexible since it sticks the local socket to a > given handler. With the current approach, you may call different > handlers in do_local_server_step with the same socket. I know, we aren't > using it in that way for now but perhaps in the near future. I have kept > this patch back. This is just the code which accepts connections from the unix socket. I assumed that the daemon will not change its behaviour during its lifetime, please explain your plan. In the process function, you can still choose what to do with the new client connection. You could also add a local_server_set_callback() or similar. Max