🚀 go-pugleaf

Ed Hall (edhall@rand.org) wrote:

> In article <Bz3yB0.MC8@srgenprp.sr.hp.com> you write:
> >     I don't know if using cheap motherboards is the problem, but I
> >agree with the above quote.  I've been using a high-quality (read:
> >expensive) AMI motherboard -- where the ENTIRE motherboard, and not the
> >just BIOS, is made by AMI -- and I haven't had any DMA hardware problems
> >with my Adaptec 1542A.  I have had problems when multiple DMA bus
> >masters try to access the bus (causes memory refresh starvation,
> >resulting in parity errors), but this is a problem with the PC
> >architecture.
>
> You can increase the BUSOFF time on the 1542 to reduce the chance of
> refresh starvation when you have more than one of them.

     Yup, that's the fix that I used.  In my case, though, it wasn't
multiple Adaptec cards causing the problem -- it was my sound card while
I was recording sound at a sampling rate of 44110Hz.  The sound card DMA
and 1542A DMA starved memory refresh until I changed the bus on/off
times on the 1542A.

> I made pretty much the same choice you did--buying an AMI motherboard
> at a premium when I upgraded to a 486/33.  There were just too many
> reports of motherboards having problems with bus mastering cards,
> especially under UNIX, for me to take chances.  I've never regretted
> it in the year since then.

     I agree.

     -- Darryl Okahata
	Internet: darrylo@sr.hp.com

DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion or policy of Hewlett-Packard or of the
little green men that have been following him all day.