On Mon, Apr 07, 2003 at 01:25:28PM -0400, Gary Thorpe wrote:
> I tried a 1.6 install CD in the machine with the hard disk, cdrom, and
> cdrw drives: the only pause is just after the message "pciide channel 0
> wired to interrupt 15" or similar and it lasted less than 5 seconds.
> The primary channel does not in fact have any atapi devices on it: only
> the hard disk is on the channel and it is set as the master. The
> secondary channel is properly configured with master/slave settings
> also and experiences no pauses in the probe.
> 
> If atapi mandates a 30 second wait for the probe it apparently is not
> being applied in this case. Sounds like a bug, not a feature: either
> its not being applied when it should or there shouldn't be a pause at all.

it *up to* 30 seconds. If the devices comes ready before 30s there is no
reasons to wait that long. But if a device never comes ready,
we have to wait 30 seconds before declaring it dead.

A bus with a slave an no master can easily (depending on how things behaves)
appear as a dead master.

-- 
Manuel Bouyer <bouyer@antioche.eu.org>
     NetBSD: 24 ans d'experience feront toujours la difference
--