From owner-cvs-sys Fri Jul 28 02:05:39 1995 Return-Path: cvs-sys-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.11/8.6.6) id CAA02203 for cvs-sys-outgoing; Fri, 28 Jul 1995 02:05:39 -0700 Received: from FileServ1.MI.Uni-Koeln.DE (FileServ1.MI.Uni-Koeln.DE [134.95.212.1]) by freefall.cdrom.com (8.6.11/8.6.6) with SMTP id CAA02181 ; Fri, 28 Jul 1995 02:05:23 -0700 Received: by FileServ1.MI.Uni-Koeln.DE id AA13275 (5.67b/IDA-1.5); Fri, 28 Jul 1995 11:05:11 +0200 Message-Id: <199507280905.AA13275@FileServ1.MI.Uni-Koeln.DE> From: esser@zpr.uni-koeln.de (Stefan Esser) Date: Fri, 28 Jul 1995 11:05:10 +0200 In-Reply-To: "Justin T. Gibbs" "Re: cvs commit: src/sys/pci pcisupport.c" (Jul 27, 15:18) X-Mailer: Mail User's Shell (7.2.6 alpha(2) 7/9/95) To: "Justin T. Gibbs" Subject: Re: cvs commit: src/sys/pci pcisupport.c Cc: CVS-commiters@freefall.cdrom.com, cvs-sys@freefall.cdrom.com Sender: cvs-sys-owner@freebsd.org Precedence: bulk On Jul 27, 15:18, "Justin T. Gibbs" wrote: } Subject: Re: cvs commit: src/sys/pci pcisupport.c } >se 95/07/27 15:14:26 } > } > Modified: sys/pci pcisupport.c } > Log: } > Add a few chip set IDs. } } Do you have a policy on PCI ids? Would it be best to take the PCI } ids that are sprinkled throughout the drivers and place them in a } common place? I don't consider it a good idea to have all PCI IDs defined in one place. There are cases where the driver doesn't only need the PCI chip ID, but also the chip revision, and if you have to make all these available to the driver anyway, then I'd rather define all the symbolic PCI IDs in the driver. The current policy is to have VENDOR IDs in the generic PCI driver to aid in identifying unsupported PCI devices. The CHIP and VENDOR IDs are required in each driver to identify a particular chip. Making a driver recognise a new PCI chip requires having its CHIP ID defined, and then have the probe and attach code know how to deal with that particular chip. The CHIP IDs are only of local interest, and I rather would have short IDs ... One special case is the chip IDs for the PCI bus support chips (Neptun, Triton, ...) which are defined in "pcisupport.c" and which are only used to print the chip set configuration independently of the BIOS. Just to make sure that there isn't some SVGA card BIOS that switches off some PCI performance options in the chip set :). Of course it is possible to have all the PCI IDs in one header file, but we went to some effort to have adding a device require NO additions to generic PCI files. Just my opinion ... Regards, STefan -- Stefan Esser Internet: Zentrum fuer Paralleles Rechnen Tel: +49 221 4706021 Universitaet zu Koeln FAX: +49 221 4705160 Weyertal 80 50931 Koeln