From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH net-next] net dst: use a percpu_counter to track entries Date: Mon, 11 Oct 2010 23:29:33 +0200 Message-ID: <1286832573.30423.93.camel@edumazet-laptop> References: <1286555854.2959.605.camel@edumazet-laptop> <20101011.130758.70192030.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]:34383 "EHLO mail-ww0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756291Ab0JKV3y (ORCPT ); Mon, 11 Oct 2010 17:29:54 -0400 Received: by wwj40 with SMTP id 40so4229265wwj.1 for ; Mon, 11 Oct 2010 14:29:53 -0700 (PDT) In-Reply-To: <20101011.130758.70192030.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: Le lundi 11 octobre 2010 =C3=A0 13:07 -0700, David Miller a =C3=A9crit = : > When I first read the subject line for this patch, I was > scared, because I thought you were using a percpu counter > for dst entry refcounts :-) >=20 > Anyways this is fine, applied, thanks! I though about a crazy idea for loopback device (but not a percpu counter for dst refcounts) Add a cpu field somewhere so that each cpu manipulates a different dst. Or maybe just never cache such dst, so that each tcp socket uses its ow= n dst.