Subject: Re: lost interrupt, device timeout...
To: None <netbsd-help@netbsd.org>
From: Jukka Salmi <j+nbsd@2004.salmi.ch>
List: netbsd-help
Date: 10/26/2004 11:22:47
Manuel Bouyer --> netbsd-help (2004-10-25 23:28:27 +0200):
> On Mon, Oct 25, 2004 at 09:35:01PM +0200, Jukka Salmi wrote:
> > > > Assuming the same applies
> > > > to RAIDframe: is it possible these errors are because of how I conn=
ected
> > > > the devices? Or is there something wrong with wd0?
> > >=20
> > > It could be too. Some master/slave combination don't work too well.
> > > I remember having such a problem with seagate/IBM combinations, I end=
ed up
> > > adding another pciide controller, so that each would be on a separate=
 channel.
> >=20
> > That's what I just tried: I added a Promise Ultra/100 IDE controller
> > and connected CD reader and writer. I was able to boot a Debian and
> > MS Fen=EAtre CDs, but booting Open- and NetBSD CDs failed. Stange. When
> > connected to the onboard IDE controller booting the BSD CDs works as
> > well.
>=20
> ATAPI devices on promise controllers are known to not work too well.
> I would connect the drives to the promise, and the CDs to the motherboard
> controller.

I tried this, and couldn't reproduce the problem since. Fine! But,
what I don't understand: connecting both disks to the same channel
(as master & slave) on the mainboard seems to work fine; if they are
connected to separate channels (both as master) on the mb I get

viaide0:0: lost interrupt
        type: ata tc_bcount: 8192 tc_skip: 0
viaide0:0:0: bus-master DMA error: missing interrupt, status=3D0x20
viaide0:0:0: device timeout, c_bcount=3D8192, c_skip0
wd0a: device timeout writing fsbn 64852684 of 64852684-64852699 (wd0 bn 648=
52747; cn 64338 tn 0 sn 43), retrying
wd0: soft error (corrected)

Is this more likely a problem with the chipset or with the disk?

However, using the Promise seems to work (around the problem...).
Thanks!


Regards, Jukka

--=20
bashian roulette:
$ ((RANDOM%6)) || rm -rf ~