From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: [RFC] Die to NOTRACK/TRACE, long live MARK! Date: Sun, 23 May 2004 19:00:49 +0200 Sender: netfilter-devel-admin@lists.netfilter.org Message-ID: <40B0D8C1.6040209@trash.net> References: <40AF6A60.70406@trash.net> <1085302540.969.51.camel@tux.rsn.bth.se> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Cc: Jozsef Kadlecsik , Philip Craig , Jozsef Kadlecsik , Netfilter-devel Return-path: To: Martin Josefsson In-Reply-To: <1085302540.969.51.camel@tux.rsn.bth.se> Errors-To: netfilter-devel-admin@lists.netfilter.org List-Help: List-Post: List-Subscribe: , List-Unsubscribe: , List-Archive: List-Id: netfilter-devel.vger.kernel.org Martin Josefsson wrote: > Nice, I hope you succeed. We also need to rearrange struct ip_conntrack > so often updated fields aren't sharing the same cacheline as mostly read > fields. That and adding more test_bit() before we set_bit() (or > something similar) will bring down the cache-trashing we have right now, > especially on SMP. I will make a test_bit patch on top of my memsave-patches (which are on top of Joszef's conntrack patches). When we rearrange struct ip_conntrack we should also try to get rid of the holes in the structure, which are responsible for approximately 20bytes of wasted memory. Regards Patrick