Return to BSD News archive
Xref: sserve comp.os.386bsd.bugs:2989 comp.os.386bsd.misc:5361 Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msunews!uwm.edu!news.alpha.net!news.mathworks.com!satisfied.elf.com!wizard.pn.com!Germany.EU.net!zib-berlin.de!rs1-hrz.uni-duisburg.de!RRZ.Uni-Koeln.DE!usenet From: se@parc03.MI.Uni-Koeln.DE (Stefan Esser) Newsgroups: comp.os.386bsd.bugs,comp.os.386bsd.misc Subject: Re: FreeBSD 2.0 and the ASUS PCI Motherboard Date: 11 Feb 1995 22:50:22 GMT Organization: Institute for Mathematics, University of Cologne, Germany Lines: 90 Distribution: world Message-ID: <3hjeve$2nd@news.rrz.uni-koeln.de> References: <3hhl6r$q66@deep.rsoft.bc.ca> NNTP-Posting-Host: parc03.mi.uni-koeln.de Thanks for the detailed report ! Though I think there are a few details mising, and one information doesn't seem consistent. } I went into the shop to pick up my new system today and discovered, } somewhat to my horror, that FreeBSD doesn't seem to boot on it. I'd } started with a boot disk made from the "January 1994" [sic] Walnut } Creek CD-ROM, but upon encountering the errors described below, went } and downloaded the boot disk from the 2.0-021092 snapshot and got the } exact same results. I then went and downloaded the NetBSD 1.0 boot } disks and found that the system booted just fine. (NetBSD seems to get } through its autoconfig about six or eight times as fast as FreeBSD, } BTW, though the messages aren't as informative.) } } Now this isn't that big a deal for me, as I was going to run NetBSD } anyway. (My NetBSD CD-ROM hasn't arrived yet, so I thought the FreeBSD } CD-ROM would do for testing.) However, I thought that the FreeBSD } folks would want to know about this. } } I've got a 486/66 on an ASUS PVI-486 SP3 motherboard, which is a } PCI-bus motherboard using an SiS chipset. One of the PCI slots has an Hmmm ... There is no ASUS 486 board with SiS chip set ... The ASUS PCI 486 AP4 is built around the Intel Aries, the PCI 486 SP3 uses rev. 2 of the Intel Saturn, and the PCI 486 SP3G Saturn rev. 4. (The PVI is the PCI+VLB+ISA version, PCI is PCI+ISA and the much better choice, IMHO). SiS chips are used in some ASUS Pentium motherboards. } ATI Graphics Pro, the other has the PCI-SC200 SCSI controller (NCR } 53C810-based) that optionally accompanies the system board. } } What happens is that FreeBSD configures the graphics card and the SCSI } controller just fine, and then goes on to find a second, identical } graphics card and SCSI controller on the same interrupts. After it } waits for the SCSI bus to settle on the "phantom" SCSI card (about } twenty seconds) it tries to probe the SCSI bus for devices at which } point the system hangs with the hard disk light on. FreeBSD claims to } find the graphics card and SCSI controler at pci0:11 and pci0:12 } respectively, and the phantom ones at pci0:23 and pci0:24, I believe. Doesn't happen on the PC3, for me (this was the development machine for the PCI and NCR drivers, BTW). So I assume you got the PVI 486 AP4 in fact ... } That's about all I have on the problem. If someone wants an exact copy } of those hardware messages, I can boot it up again, copy them out and } type them in. In fact the NetBSD PCI code was derived from an early version of that in FreeBSD at 1.1.5.1 times (written by Wolfgang Stanglmeier <wolf@kintaro.cologne.de> and me). The NetBSD version was then improved by Charles Hannum, independently from the FreeBSD version. We'll take at look at the code differences and will try correct the problem. (PCI is a bit of a moving target, since the specs aren't fixed yet, and we don't have docs of all the important chip sets. Those chip sets seem to diverge from the specs sometimes, and I REALLY would like to have hardware manuals for the Intel Aries, SiS 85c80x and Opti 82c546/597 and 82c822. So, if anybody knows how to contact SiS and Opti (best via the net or within Europe, but an US address will do ...), I'd be very grateful ... Could you please send the boot messages for your system ? We had only reports that the NCR plain won't work on the ASUS PVI (with nobody willing to find the reason), and you are the first person to report success ... I'm also interested in the chip set specific BIOS setup you use (i.e. cache options, PCI burst and porsting, ...). STefan -- Stefan Esser Internet: <se@ZPR.Uni-Koeln.DE> Zentrum fuer Paralleles Rechnen Tel: +49 221 4706019 Universitaet zu Koeln FAX: +49 221 4705160 Weyertal 80 50931 Koeln