From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH] ipv4: Remove fib_hash. Date: Wed, 02 Feb 2011 01:48:13 +0100 Message-ID: <1296607693.2607.7.camel@edumazet-laptop> References: <20110201.151941.48509941.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: netdev@vger.kernel.org To: David Miller Return-path: Received: from mail-ww0-f44.google.com ([74.125.82.44]:45636 "EHLO mail-ww0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753867Ab1BBAsX (ORCPT ); Tue, 1 Feb 2011 19:48:23 -0500 Received: by wwa36 with SMTP id 36so7706175wwa.1 for ; Tue, 01 Feb 2011 16:48:22 -0800 (PST) In-Reply-To: <20110201.151941.48509941.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: Le mardi 01 f=C3=A9vrier 2011 =C3=A0 15:19 -0800, David Miller a =C3=A9= crit : > The time has finally come to remove the hash based routing table > implementation in ipv4. >=20 > FIB Trie is mature, well tested, and I've done an audit of it's code > to confirm that it implements insert, delete, and lookup with the sam= e > identical semantics as fib_hash did. >=20 > If there are any semantic differences found in fib_trie, we should > simply fix them. >=20 > I've placed the trie statistic config option under advanced router > configuration. >=20 > Signed-off-by: David S. Miller > --- Hmm... I know having to maintain two implementations is time consuming, but I know fib_trie is bigger : # size net/ipv4/fib_*.o text data bss dec hex filename 7252 120 0 7372 1ccc net/ipv4/fib_frontend.o 7279 16 4 7299 1c83 net/ipv4/fib_hash.o 1479 0 0 1479 5c7 net/ipv4/fib_rules.o 7885 0 2080 9965 26ed net/ipv4/fib_semantics.o 16222 16 16 16254 3f7e net/ipv4/fib_trie.o In my tests, I know that fib_trie is more expensive for typical routing tables for hosts (no more than a dozen or entries), in latencies results, mostly because of icache misses, but also dcache ones. Thanks