From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [RFC] sched: CHOKe packet scheduler (v0.2) Date: Fri, 07 Jan 2011 09:37:23 +0100 Message-ID: <1294389443.2704.101.camel@edumazet-laptop> References: <20110104162930.6fa672e3@nehalam> <1294208375.3420.46.camel@edumazet-laptop> <20110105091718.02f8a00f@nehalam> <1294248332.10633.25.camel@edumazet-laptop> <20110105112104.64ad3c86@nehalam> <1294286850.2723.65.camel@edumazet-laptop> <20110106205549.0de56de1@nehalam> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: David Miller , netdev@vger.kernel.org To: Stephen Hemminger Return-path: Received: from mail-ww0-f44.google.com ([74.125.82.44]:56149 "EHLO mail-ww0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753832Ab1AGIh2 (ORCPT ); Fri, 7 Jan 2011 03:37:28 -0500 Received: by wwa36 with SMTP id 36so18265697wwa.1 for ; Fri, 07 Jan 2011 00:37:27 -0800 (PST) In-Reply-To: <20110106205549.0de56de1@nehalam> Sender: netdev-owner@vger.kernel.org List-ID: Le jeudi 06 janvier 2011 =C3=A0 20:55 -0800, Stephen Hemminger a =C3=A9= crit : > The problem is that large tables of pointers in kernel require either > contiguous allocation or some indirect table algorithm. >=20 By large table, how many slots do you envision ? 1024 ? 8192 ? 65536 ? Even insane value like 65536 is OK most of the time. if kmalloc(GFP_KERNEL) fails, try vmalloc(). We are in process context and are allowed to sleep when qdisc is created. Anyway, accessing a random skb in a list of 65536 skbs is just crazy.