Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 5 Jun 1995 14:53:25 -0700 (PDT)
From:      "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com>
To:        temp@temptation.interlog.com (Temptation)
Cc:        FreeBSD-hackers@FreeBSD.Org
Subject:   Re: NCR810 problem?
Message-ID:  <199506052153.OAA24419@gndrsh.aac.dev.com>
In-Reply-To: <Pine.3.89.9506051717.A21650-0100000@temptation.interlog.com> from "Temptation" at Jun 5, 95 05:40:39 pm

next in thread | previous in thread | raw e-mail | index | archive | help
> 
> 
> 
> On Mon, 5 Jun 1995, Rodney W. Grimes wrote:
> 
> > > > I too, am seeing the 1 boot out of 3 freaks out problem with the exact 
> > > > same symptoms as the above couple of messages.  This box (Asus SP3G, 
> > > > 486dx2/66, onboard NCR SCSI driving Quantum Empire 2100S and a HP 35480 
> > > > DAT drive) used to run 1.1.5.1 perfectly - never a problem on reboot.  
> > > > Now (2.0.5ish) freaks one out of three times with the NCR probing errors 
> > > > and then the river of errors.  Can fix it by hitting reset 9 times out of 10.
> > > > 
> > > Same motherboard, Asus (whatever) 486-66, NCR PCI, 2940, smc pci, 
> > > Seagate, Dat (on or off doesn't matter) 
> > > 
> > > reboots sometimes, Most of the not. Unmount correctly, but the 
> > > screen goes black, and sits there, Install program, Umounts, 
> > > but sits there.
> > > but then out of the blue it'll reboot correctly.
> > > Also *NO* disk problems. Nor any problems with anything else, 
> > > NT/WIN95/WARP/LINUX/SCO/UNIXWARE/DOS.
> 
> Don't know if thats good news for me or you, I personally hate this 
> motherboard, its one of the worse boards I've ever boughten.
> The reason is because I bought one for first ones, which never worked, 
> that was Rev1.0 and Rev1.1, then they gave me this one, Which is
>  rev1.2Feb1994, hmm better not get in that story. The Pent Trinton Asus 
> board is good tho. and works.

I will say it was very true that the Rev 1.0 and Rev 1.1 of this board
had serious problems (there where not ASUS's problem, it was buggy
Saturn I and II chip sets from Intel, they didn't work right until
Rev 1.2 of the board which uses the Rev 4 Saturn II chipset).  There
is still the problem that you *must* run this board with the external
cache in write back mode if you use the NCR controller (or any other
bus master device for that matter).

The nice thing about this board is that it does memory interleaving
(that is why you have to install simms in pairs) so that the memory
performance on it is very good.  (And with a DX4/100 can beat most
P5-60 machines, and even some P5-66 machines at make world!!).

> > For those of you with ASUS PCI/I-486SP3G boards I have kinda good news,
> > I will be getting in my backordered supply of them this week and once
> > I get done moving will be running at least 2 of them through the tests
> > right of the bat (due for customer shipment) and have a 3rd for me to
> > keep as a test board (sold too many of these now not to have one in
> > the test bench.)  The others go to stock and won't last a week :-)
> > 
> > I have seen this shutdown problem now, but only when using the 2.0.5A
> > install floppy, and that was on my known good test system.  Every kernel
> > I build here has worked just fine in this respect (and I do run GENERIC
> > on must of the test bench systems since hardware changes in there like
> > there is no tomarrow!!).
> > 
> > Have you folks supped the -current kernel sources and tried with it?
> > I am wondering if there is something funny going on with the build
> > environment Jordan is using on these kernels, as it seems to me the
> > folks installing the 2.0.5A are having problems, but those supping
> > -current (should be *IDENTICAL*) bits are not seeing many of the
> > strange problems reported :-(.
> 
> No. the 2.0.5A is the only version that would install, all other versions 
> crash or locked up before I could even get passed the Label program. 

I don't use the standard install tools, but I have been shipping systems
on this board since March, which means that FreeBSD does run on it with
out problem.  Sounds like you had the cache set to write back mode, which
is a known problem on this board (or at least known to me!!).

> Since I throw that cd in the garbage, I'm not going to try that those 
> other versions  again :) As for the supped thing, you need to be 
> connected to the "NET" for that don't you?

Yes.

> my FreeBSD is connected only 
> to my local network, and I have telnet or ftp to my linux server to get 
> out anyway. Guess I could try taring the directory? 

Yes.  It should be about 5MB tarred, 2.5MB as a .tgz or there abouts.

> it's not as big as the .packages or distfiles is it? I've been ftping 
> .packages for the last 25hours or so :) and it's still going :)

Your pulling the ports/distfiles directory most likely, which is HUGE!!!
Don't pull the distfiles, let make do that for you as it builds stuff,
unless you plan on building it ALL :-).

> I'll do what I can to test it out tho.
> (if any of my responses sounded a bit sarcast, there was no intent, I 
> thank you for your help, in helping solve MY problem)

Your welcome!


-- 
Rod Grimes                                      rgrimes@gndrsh.aac.dev.com
Accurate Automation Company                   Custom computers for FreeBSD



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199506052153.OAA24419>