From owner-freebsd-arch Sun Oct 13 20:11:28 2002 Delivered-To: freebsd-arch@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id B1E8A37B401 for ; Sun, 13 Oct 2002 20:11:27 -0700 (PDT) Received: from snipe.mail.pas.earthlink.net (snipe.mail.pas.earthlink.net [207.217.120.62]) by mx1.FreeBSD.org (Postfix) with ESMTP id 43D2A43EB1 for ; Sun, 13 Oct 2002 20:11:27 -0700 (PDT) (envelope-from tlambert2@mindspring.com) Received: from pool0015.cvx22-bradley.dialup.earthlink.net ([209.179.198.15] helo=mindspring.com) by snipe.mail.pas.earthlink.net with esmtp (Exim 3.33 #1) id 180vdG-0007g2-00; Sun, 13 Oct 2002 20:11:03 -0700 Message-ID: <3DAA357D.B0E51BB5@mindspring.com> Date: Sun, 13 Oct 2002 20:09:49 -0700 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: Wes Peters Cc: ticso@cicely.de, "M. Warner Losh" , hch@infradead.org, dillon@apollo.backplane.com, vova@sw.ru, nate@root.org, arch@FreeBSD.org Subject: Re: Database indexes and ram References: <3DA954CF.98B0891A@mindspring.com> <20021013.060851.113437955.imp@bsdimp.com> <3DA9B4A8.194A02FC@mindspring.com> <20021013.120847.31902907.imp@bsdimp.com> <20021013181633.GB34517@cicely8.cicely.de> <3DA9C3B9.E78BBFE6@mindspring.com> <3DAA2C4F.9E15CA75@softweyr.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-arch@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Wes Peters wrote: > No, and that's exactly why the Linux developers took the tack they did: > all of the DMA targets are allocated in the lower 4GB of physical address > space. It was quite an intelligent decision, one that made me grin when > I "got it." Did you ever have Dr. Tripp, e.g. for Analytical Mechanics? I think this falls into what he called IOTTMCO: Intuitively Obvious To The Most Casual Observer. 8-). It's reasonable to bounce the memory below 4G. It's the same deal for ISA drivers today. The Problem with the Alpha 2G limit is that it has the same type of issue -- basically, the code does not use the Bus Space code. In theory, the interface that's already there could handle all the necessary bouncing automatically (assuming you could tell if you could get away without the bouncing, without attributing the driver, like Linux does). I'm still not sure that, with the VM limitations, etc., that the ability to use the extra memory is worth all that much. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message