From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH net-next-2.6] net: sk_dst_cache RCUification Date: Wed, 14 Apr 2010 01:04:05 +0200 Message-ID: <1271199845.16881.586.camel@edumazet-laptop> References: <1270803809.2623.69.camel@edumazet-laptop> <20100413.015232.67916764.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: netdev@vger.kernel.org, paulmck@linux.vnet.ibm.com To: David Miller Return-path: Received: from mail-bw0-f219.google.com ([209.85.218.219]:44600 "EHLO mail-bw0-f219.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752901Ab0DMXEL (ORCPT ); Tue, 13 Apr 2010 19:04:11 -0400 Received: by bwz19 with SMTP id 19so84611bwz.21 for ; Tue, 13 Apr 2010 16:04:09 -0700 (PDT) In-Reply-To: <20100413.015232.67916764.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: Le mardi 13 avril 2010 =C3=A0 01:52 -0700, David Miller a =C3=A9crit : > Applied, thanks for doing this work Eric. Thanks David :) I am now working on sk_callback_lock case, to speedup sock_def_readable(), sock_def_write_space() in typical cases (SOCK_FASYNC not set) Instead of using rcu on whole "struct socket", my plan is to use a smal= l structure : struct wait_queue_head_rcu { wait_queue_head_t wait; struct rcu_head rcu; } ____cacheline_aligned_in_smp; and make sk->sk_sleep points to this 'wait' field.