From owner-svn-src-all@FreeBSD.ORG Wed Apr 10 06:21:01 2013 Return-Path: Delivered-To: svn-src-all@freebsd.org Received: from mx1.freebsd.org (mx1.FreeBSD.org [8.8.178.115]) by hub.freebsd.org (Postfix) with ESMTP id A5618914; Wed, 10 Apr 2013 06:21:01 +0000 (UTC) (envelope-from glebius@FreeBSD.org) Received: from cell.glebius.int.ru (glebius.int.ru [81.19.69.10]) by mx1.freebsd.org (Postfix) with ESMTP id 3205F6C2; Wed, 10 Apr 2013 06:21:00 +0000 (UTC) Received: from cell.glebius.int.ru (localhost [127.0.0.1]) by cell.glebius.int.ru (8.14.6/8.14.6) with ESMTP id r3A6KxPe058989; Wed, 10 Apr 2013 10:20:59 +0400 (MSK) (envelope-from glebius@FreeBSD.org) Received: (from glebius@localhost) by cell.glebius.int.ru (8.14.6/8.14.6/Submit) id r3A6KxdH058988; Wed, 10 Apr 2013 10:20:59 +0400 (MSK) (envelope-from glebius@FreeBSD.org) X-Authentication-Warning: cell.glebius.int.ru: glebius set sender to glebius@FreeBSD.org using -f Date: Wed, 10 Apr 2013 10:20:59 +0400 From: Gleb Smirnoff To: Andre Oppermann Subject: Re: svn commit: r249318 - in head/sys: kern net netinet Message-ID: <20130410062059.GZ76816@FreeBSD.org> References: <201304092102.r39L2LrO047044@svn.freebsd.org> MIME-Version: 1.0 Content-Type: text/plain; charset=koi8-r Content-Disposition: inline In-Reply-To: <201304092102.r39L2LrO047044@svn.freebsd.org> User-Agent: Mutt/1.5.21 (2010-09-15) Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org X-BeenThere: svn-src-all@freebsd.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: "SVN commit messages for the entire src tree \(except for " user" and " projects" \)" List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 10 Apr 2013 06:21:01 -0000 On Tue, Apr 09, 2013 at 09:02:21PM +0000, Andre Oppermann wrote: A> Author: andre A> Date: Tue Apr 9 21:02:20 2013 A> New Revision: 249318 A> URL: http://svnweb.freebsd.org/changeset/base/249318 A> A> Log: A> Change certain heavily used network related mutexes and rwlocks to A> reside on their own cache line to prevent false sharing with other A> nearby structures, especially for those in the .bss segment. A> A> NB: Those mutexes and rwlocks with variables next to them that get A> changed on every invocation do not benefit from their own cache line. A> Actually it may be net negative because two cache misses would be A> incurred in those cases. Although change is theoretically correct, IMO, such change should be supported by experimental performance benchmarking data. -- Totus tuus, Glebius.