Subject: Re: DMA and interrupt errors on DS10?
To: Manuel Bouyer <>
From: Eric Schnoebelen <>
List: port-alpha
Date: 02/09/2005 15:49:59
Manuel Bouyer writes:
- On Wed, Feb 09, 2005 at 02:46:41PM -0600, Eric Schnoebelen wrote:
- > 
- > "Johan A.van Zanten" writes:
- > - I've talked to a few friends who work with modern "PCish" hardware more
- > - than i, and i'm almost certain it's the cables which are garden-variety
- > - 40-pin, not the fancier 80-wire types meant for the newer, faster DMA
- > - drives.  I'll try some betetr cables.
- > 
- > 	Won't help, as the onboard IDE Controoller isn't able to
- > do more than PIO/ATA4.  
- Hum, no this controller should be able to do Ultra/33. Even if
- 80-wire cable are not required for Ultra/33, I found that it
- helped in some cases.
- > 	The reason you see the "degrade" is the driver tries
- > what the drives claim they can do, but the controller chip
- > fails..(or thats what I think happens..)
- The driver tries to do the min of what drives and controller are capable.

	My misunderstanding then.. Bummer..

	I've got a stack of DS10L's sitting about here, and they
all do this "degradation dance" during the first access..  Once
they've done that, they run just fine..

	And some of them have 80wire cables (round cables with
external shielding) and I still see the degradation.

Eric Schnoebelen
    Glossary:  compu-dinosaur: N, A computer guru stuck in time. (is
    happy with a TeleVideo terminal and a modem).  Most probably the
	      guy you learned everything you know from.