From mboxrd@z Thu Jan 1 00:00:00 1970 From: John Fastabend Subject: Re: [RFC Patch net-next] net_sched: make classifying lockless on ingress Date: Fri, 20 Dec 2013 18:32:49 -0800 Message-ID: <52B4FDD1.10608@intel.com> References: <1387582105-1789-1-git-send-email-xiyou.wangcong@gmail.com> <1387583344.19078.475.camel@edumazet-glaptop2.roam.corp.google.com> <1387584529.19078.482.camel@edumazet-glaptop2.roam.corp.google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Cc: Eric Dumazet , Linux Kernel Network Developers , "David S. Miller" , Jamal Hadi Salim To: Cong Wang Return-path: Received: from mga01.intel.com ([192.55.52.88]:59229 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752114Ab3LUCdG (ORCPT ); Fri, 20 Dec 2013 21:33:06 -0500 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On 12/20/2013 4:24 PM, Cong Wang wrote: > On Fri, Dec 20, 2013 at 4:08 PM, Eric Dumazet wrote: >> On Fri, 2013-12-20 at 15:57 -0800, Cong Wang wrote: >> >>> Just push the lock down to ->enqueue() ? Since ingress qdisc is classless >>> and its ->dequeue() is nop. >> >> >> https://github.com/jrfastab/Linux-Kernel-QOS >> >> https://github.com/jrfastab/Linux-Kernel-QOS/commit/67746f95acd77cf15d7ce34f644b76058ce19813 >> > > It's a nice idea to introduce a ->prequeue(). > > And so brave to play with RCU on singly linked list... maybe I can > just steal the ->prequeue() piece. John? :-D If you only steal the prequeue piece then you don't solve the lock contention part so I don't think it helps. At which point I suspect you might as well use one of the existing qdiscs not designed for multiqueue nics. > > BTW, the last commit is about 5 months ago... Yeah well I imagined I would write a rate limiting qdisc to use this infrastructure. Jamal hinted at using a systolic processes for this. But I work on this when I have time and have been busy the last few months with other things unfortunately. > -- > To unsubscribe from this list: send the line "unsubscribe netdev" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html >