From owner-freebsd-bugs Fri Aug 11 10:23:36 1995 Return-Path: bugs-owner Received: (from majordom@localhost) by freefall.FreeBSD.org (8.6.11/8.6.6) id KAA28632 for bugs-outgoing; Fri, 11 Aug 1995 10:23:36 -0700 Received: from mail.barrnet.net (mail.barrnet.net [131.119.246.7]) by freefall.FreeBSD.org (8.6.11/8.6.6) with ESMTP id KAA28625 for ; Fri, 11 Aug 1995 10:23:35 -0700 Received: from gndrsh.aac.dev.com (gndrsh.aac.dev.com [198.145.92.241]) by mail.barrnet.net (8.6.10/MAIL-RELAY-LEN) with ESMTP id KAA09101 for ; Fri, 11 Aug 1995 10:23:25 -0700 Received: (from rgrimes@localhost) by gndrsh.aac.dev.com (8.6.11/8.6.9) id KAA04273; Fri, 11 Aug 1995 10:20:36 -0700 From: "Rodney W. Grimes" Message-Id: <199508111720.KAA04273@gndrsh.aac.dev.com> Subject: Re: Performance problem resolved. To: mrcpu@cdsnet.net (Jaye Mathisen) Date: Fri, 11 Aug 1995 10:20:36 -0700 (PDT) Cc: bugs@freebsd.org In-Reply-To: from "Jaye Mathisen" at Aug 11, 95 08:25:48 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1502 Sender: bugs-owner@freebsd.org Precedence: bulk > > > > Hmmm, well, I'm not sure why the VGA card on a PCI slot even gets an > interrupt, but moving everything to IRQ 15 seemed to solve the problem. Some VGA PCI bioses call the PCI bios and request an interrupt along with there block of I/O ports at PCI P-n-P configure time. This is for video cards that support vertical retrace interrupts (usefull for doing 3D split image display with shutter glasses). IMHO, it is a mis feature that should be disableable, but unfortenatly there is no way to turn it off (other than blowing a new PROM for the video card, removing the request for the interrupt, and disabling the logic that allows it to be created). Almost all ISA cards that could generate an IRQ2 had a way to turn it off so you could free that interrupt up for other use. Well, the PCI P-n-P folks forgot all about that nice little feature of jumpers and have taken all control of such things away from us users. P-n-P is creating headaches for fully loaded machines running out of interrupts due to this bit of braindamage :-(. Design computers that even an idiot can use, and only an idiot will use them :-(. > Transfers are now humming along at a nice 900+Kbps, which is more what I > expected. :-) > Assuming INN compiles fine, I'll be jammin'. You should be jamming buy now :-) > Thanks for the help everybody. > -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD