From owner-freebsd-net@FreeBSD.ORG Thu Mar 13 03:57:52 2008 Return-Path: Delivered-To: net@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 7863F1065672; Thu, 13 Mar 2008 03:57:52 +0000 (UTC) (envelope-from bms@FreeBSD.org) Received: from out4.smtp.messagingengine.com (out4.smtp.messagingengine.com [66.111.4.28]) by mx1.freebsd.org (Postfix) with ESMTP id 6426A8FC19; Thu, 13 Mar 2008 03:57:52 +0000 (UTC) (envelope-from bms@FreeBSD.org) Received: from compute1.internal (compute1.internal [10.202.2.41]) by out1.messagingengine.com (Postfix) with ESMTP id 8304DC3F38; Wed, 12 Mar 2008 23:57:51 -0400 (EDT) Received: from heartbeat1.messagingengine.com ([10.202.2.160]) by compute1.internal (MEProxy); Wed, 12 Mar 2008 23:57:51 -0400 X-Sasl-enc: UjIm2aq5dci79rNOVb9njsZ9/c26WuHgzO2X9opO5gam 1205380671 Received: from empiric.lon.incunabulum.net (82-35-112-254.cable.ubr07.dals.blueyonder.co.uk [82.35.112.254]) by mail.messagingengine.com (Postfix) with ESMTPSA id E2783D2A5; Wed, 12 Mar 2008 23:57:50 -0400 (EDT) Message-ID: <47D8A63D.3050903@FreeBSD.org> Date: Thu, 13 Mar 2008 03:57:49 +0000 From: "Bruce M. Simpson" User-Agent: Thunderbird 2.0.0.9 (X11/20080207) MIME-Version: 1.0 To: Robert Watson References: <20080312175151.V47697@fledge.watson.org> In-Reply-To: <20080312175151.V47697@fledge.watson.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: net@FreeBSD.org Subject: Re: FYI: inpcb/pcbinfo mutex -> rwlock at some point in the mid-distant future X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 13 Mar 2008 03:57:52 -0000 Robert Watson wrote: > One of those issues is that we need to demonstrate to ourselves that > exclusive access contention is managed as well with rwlocks as with > sleep mutexes, as these locks would continue to be fairly highly > contended in TCP. The other issue is that rwlocks don't support full > priority propagation for reader access, although Jeff Roberson has > recently improved fairness to writers with many readers. Don't forget that p4 bms_netdev contains a number of optimizations for the multicast paths -- there are lock acquisitions which are quite often unnecessary, or whose granularity is too high for the data structure(s) which need to be shared. BMS