From mboxrd@z Thu Jan 1 00:00:00 1970 From: Shaohua Li Subject: Re: [RFC net-next] net ipv6: convert fib6_table rwlock to a percpu lock Date: Mon, 31 Jul 2017 19:57:04 -0700 Message-ID: <20170801025704.fuhxll23d3eenwrz@kernel.org> References: <1a5ad2f0585aa66496b27e123d1c38b75552df4c.1501520674.git.shli@fb.com> <20170731161007.61c2d958@xeon-e3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: netdev@vger.kernel.org, davem@davemloft.net, Kernel-team@fb.com, Shaohua Li , Wei Wang To: Stephen Hemminger Return-path: Received: from mail.kernel.org ([198.145.29.99]:46918 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751809AbdHAC5G (ORCPT ); Mon, 31 Jul 2017 22:57:06 -0400 Content-Disposition: inline In-Reply-To: <20170731161007.61c2d958@xeon-e3> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, Jul 31, 2017 at 04:10:07PM -0700, Stephen Hemminger wrote: > On Mon, 31 Jul 2017 10:18:57 -0700 > Shaohua Li wrote: > > > From: Shaohua Li > > > > In a syn flooding test, the fib6_table rwlock is a significant > > bottleneck. While converting the rwlock to rcu sounds straighforward, > > but is very challenging if it's possible. A percpu spinlock is quite > > trival for this problem since updating the routing table is a rare > > event. In my test, the server receives around 1.5 Mpps in syn flooding > > test without the patch in a dual sockets and 56-CPU system. With the > > patch, the server receives around 3.8Mpps, and perf report doesn't show > > the locking issue. > > > > Cc: Wei Wang > > You just reinvented brlock... you mean lglock? It has been removed from kernel. > RCU is not that hard, why not do it right? Maybe. But don't think it's the reason why we shouldn't do the percpu lock now, this is a simple change, if some smart guys find a way of RCU, we can easily remove this.