Subject: Re: Misc HWDIRECT fixes
To: None <port-mac68k@NetBSD.ORG>
From: Benoit MARTEL <magus@cs.mcgill.ca>
List: port-mac68k
Date: 02/02/1998 16:09:28
On Thu, 29 Jan 1998, Space Case wrote:
> On Jan 29, 1:46am, Benoit MARTEL wrote:
> >What source date is HWDIRECT.394 based on?
> >All I know is that it works fine with GENERIC56. I haven't found a newer
> >kernel and I haven't dowloaded the source to roll my own yet.
>
> There are newer kernels on:
>
> ftp://ftp.macbsd.com/pub/NetBSD/eskimo.copy/kernels/
>
> That's where I put my weekly snapshots. When I get my account back
> (lost in disk crash), I'll be putting them on netbsd.org as well.
>
> ~Steve
Thanks for the pointers Steve!
The problem I was getting was using HWDIRECT.394:
Boot hung at the very end (after the line "root on sd0a dumps on sd0b")
no matter what I do. If I move the mouse during serial initialisation,
the kernel panics and dumps its registers and stack.
Here is what I get with these new kernels from .../eskimo.copy/kernels:
HWDIRECTSBC57:
This kernel still crashes if I move the mouse during serial init.
However, if I dont move it, it doesn't hang and finishes booting fine and
works well.
GENERICSBC57:
This kernel works fine _except_ that I cannot use my cdrom anymore.
Note that the cdrom works fine using HWDIRECTSBC57. You would think if
one works the other would, but no. Here is what I get:
vex# mount -t cd9660 /dev/cd0c /cdrom
sbc0: parity error!
sbc0: parity error!
sbc0: parity error!
ncr5380_machine: unexpected disconnect.
sbc0: parity error!
Feb 2 12:49:34 vex /netbsd: ncr5380_machine: sbc0: parity error!
mount_cs9660: : Feb 2 12:49:35 vex last message repeated 2 times
Input/Output error
vex# Feb 2 12:439:35 vex /netbsd: ncr5380_machine: unexpected disconnect.
Feb 2 12:49:35 vex /netbsd: sbc0: parity error!
My system is a IIvx w/8Mb running 1.3 distrib binaries.
When I get the time, I'll get -current source and try it.
If I still get this cdrom problem with current, should I send-pr?
Also should I send-pr for the bootup crash with mouse movement?
Or is posting to this list the right thing to do?
Ben, for which only GENERICSBC56 seems to work perfectly.
-----------------------------------------------------------------
"Because user errors often produce unpredictable results, the user
should try to avoid them."
IBM MVS/XA System Programming Library.