Return to BSD News archive
Newsgroups: comp.unix.bsd.bsdi.misc,comp.os.linux.misc
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!news.mira.net.au!vic.news.telstra.net!act.news.telstra.net!psgrain!usenet.eel.ufl.edu!news.mathworks.com!howland.erols.net!feed1.news.erols.com!uunet!in1.uu.net!news.new-york.net!news.spc.edu!spcunb.spc.edu!not-for-mail
From: Terry Kennedy <terry@spcunb.spc.edu>
Subject: Re: BSDi + AHA2944W + Eclipse RAID
Content-Type: text/plain; charset=US-ASCII
X-Newsreader: TIN [UNIX 1.3 unoff BETA release 961018]
X-Nntp-Posting-User: terry
Sender: news@spcuna.spc.edu (USENET News)
Content-Transfer-Encoding: 8bit
Organization: St. Peter's College, US
Lines: 75
Message-ID: <E144wK.5qq@spcuna.spc.edu>
References: <55j0ci$8l8@gol1.gol.com> <55tuo6$r02@news1.anet-dfw.com> <56e1qh$4t8@oldman.steinkamm.com> <56glrn$c1k@olympus.nwnet.net> <E10H2w.BC0@spcuna.spc.edu> <56qp4d$os2@olympus.nwnet.net>
Mime-Version: 1.0
X-Trace: 848399059/6558
X-Nntp-Posting-Host: spcunb.spc.edu
Date: Tue, 19 Nov 1996 10:24:20 GMT
Xref: euryale.cc.adfa.oz.au comp.unix.bsd.bsdi.misc:5309 comp.os.linux.misc:142580
In comp.unix.bsd.bsdi.misc Anthony Talltree <aad@nwnet.net> wrote:
> > You're not still talking about the wide-mode support (the "wide control-
> >lers are only supported in narrow mode" paragraph), are you?
> Yes, partially. The 2.1 book says that about the Adaptec boards, but doesn't
> list the wide or differential Buslogic boards at all. support@bsdi.com told
> me explicitly that no wide and no differential Buslogic boards would at all.
> You did indeed give me the information that showed that they were wrong. Even
> now I can't get multiple BT946C's to stay up for more than a week (90MHz CPU)
> or more than a day (133 or 150 MHz).
Well, I have a pair of 946C's in a Tomcat I (133MHz, 96MB) that's been up for
35 days - see my prior post for some ideas. Chris posted some dmesg output that
showed his dual 956's working (shhh - Chris, I told you it was a secret 8-).
Even before I updated his box, he was running the 2 956C's on stock 2.1 drivers
for months without any problems.
As soon as you told me about the wide issues, I made sure the BSDI folks were
aware of the actual configuration guidelines, and changed the code so that the
next release would fully support the additional ID's on wide buses.
> >motherboard that arbitrated the bus properly (Neptune-based boards such as
> >the Intel Plato had a problem when more than 2 bus masters are installed).
> I bought a Tyan Tomcat II to make sure, though the supplied manual doesn't
> seem to directly address the issue.
If you need info like that, you have to call the Intel Literature Hotline
and order the chipset documentation. Vendors are building for Windows users
who don't know/care about this level of detail.
> > The only problem I can see is that the newer BusLogic cards only offer the
> >"primary" (330) and "alternate" (334) address settings for "ISA compatible
> >port" in their setup menu
> Are these settings really important for PCI cards on BSDI machine, though?
> The description in the manual seems to imply that it's only for software that
> doesn't know about the PCI interface. Is this something that would prevent
> the card from working reliably, or just prevent it from showing up? After
> adding the requisite kernel config lines for bha2, I haven't had any problems
> getting them to show up, modulo the usual absurdity that the IRQ has to be
> set in both the card and the motherboard.
The bha driver runs the card using the ISA compatible ports by default. This
still occurs over the PCI bus at full PCI speeds - the only issue is with find-
ing the cards at probe time, since the bha probe routine doesn't know to look
in the PCI space.
You can either configure the kernel with the PCI address, use the -dev boot
switch, or enable ISA compatible addressing. If the board doesn't have any ISA
appearance, it won't probe. Note that I think the BusLogic microcode is smart
enough to see that ISA port references are being done even when the setting
is set to disabled.
The disadvantage to changing the kernel config or using -dev is that dif-
ferent motherboards (and even the same motherboard when different cards are
installed) may configure the BusLogics at different parts of the PCI address
space.
Of course, the real solution is to probe the PCI space a la the de driver.
I'll probably add that and send it to BSDI for a follow-on release after 3.0.
> >so the third (and so on) cards would need a "-dev
> >bha2 port=6000" (or whereever your BIOS configures it)
> Would I then set it to 'disabled' in Auto-SCSI?
For the 3rd and subsequent card I don't think it matters.
> > You seem to doubt that 3.0 will ship
> No, I'm sure it'll ship some year to somebody. I'm not so convinced that it'll
> ship soon, or to us without us calling BSDI and asking for it.
BSDI has previously been overly optimistic about their ship dates. I think
the past releases have taught them something.
Terry Kennedy Operations Manager, Academic Computing
terry@spcvxa.spc.edu St. Peter's College, Jersey City, NJ USA
+1 201 915 9381 (voice) +1 201 435-3662 (FAX)