From owner-freebsd-scsi Fri Dec 3 10:18: 0 1999 Delivered-To: freebsd-scsi@freebsd.org Received: from feral.com (feral.com [192.67.166.1]) by hub.freebsd.org (Postfix) with ESMTP id 881761507D; Fri, 3 Dec 1999 10:17:56 -0800 (PST) (envelope-from mjacob@feral.com) Received: from semuta.feral.com (semuta [192.67.166.70]) by feral.com (8.8.7/8.8.7) with ESMTP id KAA06637; Fri, 3 Dec 1999 10:19:06 -0800 Date: Fri, 3 Dec 1999 10:19:06 -0800 (PST) From: Matthew Jacob Reply-To: mjacob@feral.com To: "Justin T. Gibbs" Cc: scsi@FreeBSD.ORG Subject: bounce buffers (was Re: Tape driver problems) In-Reply-To: <199912031712.KAA00529@caspian.plutotech.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > >> The bounce buffering code doesn't perform coalessing of data yet because > >> I never found an algorithm I liked for it. The NetBSD approach of > >> making the bounce buffer 64K big and copying the whole thing if any > >> portion needs to be bounced doesn't appeal to me (it doesn't scale > >> well and consumes much too much memory). > > > >You pay for what you get. If you are really still using ISADMA mass > >storage, blow 128k or MAXPHYS already.. IMO... > > We do, but we allocate and dole it out as individual pages. NetBSD allocates > MAXPHYS per possible transaction on each card which turns out to be 1MB of > < 16MB memory per card. This doesn't scale. It's not important that ISADMA scales. If somebody wants to put 4 1542s in a system, they can pay a lot for the privilege of not buying 2940s. Still, it needn't be per card. It can be per-system and it's not for wired down usage (or you could add a flag for wired down usage). -matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message