From owner-freebsd-hackers Wed Feb 11 15:20:58 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA12347 for hackers-outgoing; Wed, 11 Feb 1998 15:20:58 -0800 (PST) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from dingo.cdrom.com (dingo.cdrom.com [204.216.28.145]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA12331 for ; Wed, 11 Feb 1998 15:20:49 -0800 (PST) (envelope-from mike@dingo.cdrom.com) Received: from dingo.cdrom.com (localhost [127.0.0.1]) by dingo.cdrom.com (8.8.8/8.8.5) with ESMTP id PAA01611; Wed, 11 Feb 1998 15:20:05 -0800 (PST) Message-Id: <199802112320.PAA01611@dingo.cdrom.com> X-Mailer: exmh version 2.0zeta 7/24/97 To: dg@root.com cc: Brian McGovern , hackers@FreeBSD.ORG Subject: Re: Mapping phyical memory in to the PCI address range... In-reply-to: Your message of "Wed, 11 Feb 1998 11:20:28 PST." <199802111920.LAA21000@implode.root.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 11 Feb 1998 15:20:05 -0800 From: Mike Smith Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > The PCI devices have access to all of the PC's memory via DMA. The CPU can > also access the PCI device's memory if it is mapped. I'm familiar with the > PLX 9060 and it's a bit quirky, but not that difficult to setup DMA. See the > fxp driver for an example of a driver that does PCI DMA. > > >2 - Would it really be worthwhile pursuing this endeavor? After all, a 1-2% > >gain on moving a single character really isn't a big win. However 25+% > >very well might be. > > Maybe, maybe not. DMA will likely be slower when dealing with a small > number of characters since there is a significant amount of work to do > per DMA. I would guess that the fastest access would be to map the card's > RAM via the PCI space and access it directly via the CPU. Pardon my butting in; Brian, do you have any documentation on how the Cyclades card actually performs these memory accesses? In particular, does it just transfer single bytes, or does it only move in larger quantities? If the former, definitely stay with using the on-card memory, as that will be much more efficient. OTOH, if they are moving > 32 bits of data per DMA cycle you will start to pick up performance there. (Yes, I've worked with the PLX parts too, although my grounding with the 9060 is purely based on reading the datasheet.) -- \\ Sometimes you're ahead, \\ Mike Smith \\ sometimes you're behind. \\ mike@smith.net.au \\ The race is long, and in the \\ msmith@freebsd.org \\ end it's only with yourself. \\ msmith@cdrom.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe hackers" in the body of the message