On Wed, Oct 13, 2010 at 8:59 PM, Nick Holland
<nick@holland-consulting.net> wrote:
quoted text
>
> On 10/13/10 17:25, Robert wrote:
> > On Wed, 13 Oct 2010 16:55:18 -0400
> > Ted Unangst <ted.unangst@gmail.com> wrote:
> >> can be done about it, and 10 year old quirky PC hardware doesn't
> >> attract a of interest...
> >
> > As long as it's on [1] I hope it does?
> > I guess I'm not the only one who uses a Pentium 4 (or older stuff) for
> > firewalls and other systems, since they are very cheap to buy and
> > replace, and are more than sufficient (speed) for a lot of tasks.
> >
> > regards,
> > Robert
> >
> > [1] http://www.openbsd.org/i386.html
>
> There are big differences between well-designed hardware, poorly
> designed and implemented hardware, hardware is working properly and
> hardware that is malfunctioning.
>
> A lot of hardware out there was tested with Windows-of-the-Day (and
> maybe the day before) and that's it.  Anything else it works with,
> great, but it was by luck, not design.
>
> A lot of "early" AMD stuff was junk.  I'm not talking about the AMD
> chips themselves, I'm talking about the REST of the computer.  I've got
> a few AMD K6 systems, and NONE of them can build from source at the
> rated speed with OpenBSD.  They'll run the OS just fine, but they can't
> build, giving sig11's at random places during the process.  Replace the
> RAM with stuff that has worked well in 133MHz bus machines, same thing.
>  Slow down the bus speed, increase the multiplier, and suddenly they
> work fine.  I don't think that's an OpenBSD problem, and I really don't
> want developers fighting with that.  I have heard reports of these kinds
> of problems extending well into the Athlon days...
>
>
> In your case, though, yes, I'd look closely at your hardware.  Not sure
> why you have both a 150G disk and a 15G disk...double your chances of
> disk failure taking your system out...for 10% more storage.  I also see
> re2 is on irq12.  That's the PS/2 mouse IRQ.  Sure, you don't have a
> mouse on your machine, maybe you have the mouse port "off" in the
> BIOS...but I'd be completely unsurprised if your HW mfg screwed the
> pooch and didn't really disconnect the PS/2 hardware from IRQ
> controller, and that could be causing some of your issues there (twist
> knobs in the system BIOS, you can probably fix this).  And I'd not be
> surprised in the least if BOTH were problems for you...
>
> Nick.
>
Actually the same issue occurred previously when the only difference
was that I had configured re0 as the active interface.  I thought the
fact that it was using the same irq as pciide1 might be the source of
the issue.

Most recently, I have tried extracting the NIC in question.  The only
remaining NIC is now using irq 10 (along with pciide1.)  Similar
issues occurred.  I list here the output - similar sequences have been
listed many times, all values aside from c_bcount, c_skip, and
possibly cn vary.

pciide1:1:0: bus-master DMA error: missing interrupt, status=0x21
wd1a: device timeout writing fsbn 581104768 of 5811047680581104799
(wd1 bn 581184768; cn 576492 tn 13 sn 13), retrying
wd1(pciide1:1:0): timeout
        type: ata
        c_bcount: 16384
        c_skip: 0

I am going to see if I can eliminate the old PATA drive and associated
controller / driver from the picture with hardware I have available.

--
"Young man, in mathematics you don't understand things, you just get
used to them." - John von Neumann