Return to BSD News archive
Newsgroups: comp.unix.bsd.bsdi.misc
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!newshost.telstra.net!asstdc.scgt.oz.au!metro!metro!munnari.OZ.AU!news.ecn.uoknor.edu!news.eng.convex.com!newsrelay.netins.net!news.dacom.co.kr!newsfeed.internetmci.com!in2.uu.net!news.new-york.net!wlbr!sms
From: sms@wlv.iipo.gtegsc.com (Steven M. Schultz)
Subject: Re: multiple pci cards [shared IRQs]
Sender: news@wlbr.iipo.gtegsc.com (Steven M. Schultz)
Organization: GTE Government Systems, Thousand Oaks CA USA
Message-ID: <Do8ALF.FBy@wlbr.iipo.gtegsc.com>
References: <313E5821.6D78@execpc.com> <4i0002$qg@raid.us.dell.com> <3144A92A.31FC@execpc.com> <1996Mar11.202048.1@spcvxb.spc.edu>
X-Nntp-Posting-Host: wlv.iipo.gtegsc.com
Date: Wed, 13 Mar 1996 23:00:50 GMT
Lines: 73
In article <1996Mar11.202048.1@spcvxb.spc.edu>,
Terry Kennedy, Operations Mgr. <terry@spcvxb.spc.edu> wrote:
>
> I've had no problem using multiple PCI cards that share an interrupt,
>although in my case it's been multiple identical cards. BSDI added this
>support during the 2.0 (yes, 2.0, not 2.1) beta test period. Thus, they...
>
> Steve Schultz has an identical configuration which also works fine on
>BSD/OS 2.x, although he's had horrible problems using other free Unix-
>like systems. He may want to comment...
Just a brief comment (after an "invitation" like that I could
refuse? :-)).
I needed to run dual SCSI chains (because the mix of devices exceeds
7 fairly often) and there were no IRQs left that the motherboard BIOS
would/could hand out.
FreeBSD 2.1R (the CD from Nov/Dec) didn't support/recognize the 946C
as a PCI device - thus interrupt sharing didn't work. I did try
building using the updates available via FTP for the next 'stable'
version. That didn't work either. The first SCSI chain was useable
though. The problem has been fixed I am told but I kinda lost interest
after BSD/OS 2.1 shipped.
Linux (all 3 distributions from Nov/Dec, Slackware, Redhat and one
other whose name escapes me) didn't even get past the boot/config
phase before starting to spew error messages on the console.
At that point I waited a few weeks and BSD/OS 2.1 showed up. With the
advent of the new sound drivers there was no longer a need to have
the other operating systems present and the disk space (couple spindles
of around 3gb) was reclaimed.
Steven Schultz
sms@wlv.iipo.gtegsc.com
sms@moe.2bsd.com
---------------------
BSDI BSD/OS 2.1 Kernel #7: Wed Feb 28 19:58:21 PST 1996
sms@moe.2bsd.com:/usr/src/sys/compile/MOE
cpu = Pentium (about 90 MHz) model 2, stepping 4
delay multiplier 2303
real mem = 67108864
avail mem = 64671744
buffer cache = 6545408
isa0 (root)
pci0 at root: configuration mechanism 2
wdc0 at isa0 iobase 0x1f0 irq 14: disk controller
wd0 at wdc0 drive 0: sec/int=4 2467584*512
...
bha0 at isa0 iobase 0x330 irq 11: Buslogic EISA/PCI/VLB rev 42 (32-bit)
Delaying for SCSI bus reset and device self tests
bha0: synch negotiation, parity
tg0 at bha0 target 0
sd0 at tg0 unit 0: disk: QUANTUM XP34301 rev 1051 (SCSI-2) 8410200*512
tg0: synch transfer period 200 ns, offset 15 (10 MB/s)
tg1 at bha0 target 1
sd1 at tg1 unit 0: disk: SEAGATE ST42400N rev 0119 (SCSI-2) 4159462*512
tg1: synch transfer period 200 ns, offset 15 (10 MB/s)
tg2 at bha0 target 2
sr0 at tg2 unit 0: CD-ROM: TOSHIBA CD-ROM XM-3701TA rev 2965 (SCSI-2) removable
tg2: synch transfer period 200 ns, offset 15 (10 MB/s)
WARNING: conflict at irq 11
bha1 at isa0 iobase 0x334 irq 11: Buslogic EISA/PCI/VLB rev 42 (32-bit)
Delaying for SCSI bus reset and device self tests
bha1: synch negotiation, parity
tg3 at bha1 target 2
sr1 at tg3 unit 0: CD-ROM: NEC CD-ROM DRIVE:841 rev 1.0 (SCSI-2) removable
tg4 at bha1 target 4
st0 at tg4 unit 0: WANGTEK 5150ES SCSI ES41 rev C560 (SCSI-1)
tg5 at bha1 target 5
st1 at tg5 unit 0: SONY SDT-5000 rev 3.02 (SCSI-2) DDS cassette streaming (512) tape
tg5: synch transfer period 200 ns, offset 8 (10 MB/s)