From owner-freebsd-current@FreeBSD.ORG Fri Jun 4 12:03:40 2004 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C55C716A4CE; Fri, 4 Jun 2004 12:03:40 -0700 (PDT) Received: from ebb.errno.com (ebb.errno.com [66.127.85.87]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9D9C843D2F; Fri, 4 Jun 2004 12:03:40 -0700 (PDT) (envelope-from sam@errno.com) Received: from [66.127.85.91] ([66.127.85.91]) (authenticated bits=0) by ebb.errno.com (8.12.9/8.12.6) with ESMTP id i54J3AWR046258 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO); Fri, 4 Jun 2004 12:03:10 -0700 (PDT) (envelope-from sam@errno.com) From: Sam Leffler Organization: Errno Consulting To: freebsd-hackers@freebsd.org Date: Fri, 4 Jun 2004 11:59:02 -0700 User-Agent: KMail/1.6.1 References: <20040531215101.GA60299@freefall.freebsd.org> <20040603151911.GB12727@Odin.AC.HMC.Edu> <20040604112733.GC98227@cell.sick.ru> In-Reply-To: <20040604112733.GC98227@cell.sick.ru> MIME-Version: 1.0 Content-Disposition: inline Content-Type: text/plain; charset="koi8-r" Content-Transfer-Encoding: 7bit Message-Id: <200406041159.02229.sam@errno.com> cc: Gleb Smirnoff cc: freebsd-current@freebsd.org cc: Bosko Milekic Subject: Re: [HEADS-UP] mbuma is in the tree X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 04 Jun 2004 19:03:40 -0000 On Friday 04 June 2004 04:27 am, Gleb Smirnoff wrote: > On Thu, Jun 03, 2004 at 08:19:11AM -0700, Brooks Davis wrote: > B> > On Wed, Jun 02, 2004 at 10:56:52AM -0700, Sam Leffler wrote: > B> > S> allocated using this mechanism. I did it once for vlan tags but > botched it B> > S> (didn't handle module references properly) so backed it. > But there's no B> > S> reason someone cannot redo it or convert other > heavily used fixed size tags B> > S> to use a zone. > B> > > B> > Have you saved your efforts? May I look at them? > B> > B> They are in the CVS history of sys/net/if_vlan.c. > > I see now, thanks. > > Question to Sam: have you performed any tests? Is this definitely > true, that UMAllocing in special zone is faster than general malloc()? Allocating from a zone was noticeable for gige interfaces, especially on my SMP configuration (which was running w/o Giant). For non-gige interfaces the overhead of using malloc is not noticeable (as I reported when I first converted vlan handling over to use tags). Regardless the point was that you can already use a zone for tags if you want. Sam