Hi Markus,

I read your announcement with interest. I wondered if you might have fixed
some things that were preventing MacBSD from booting multiuser on my
PB170 _without the serial console_. So I downloaded your kernel and
tried it. In fact, if I use ttye0 (instead of tty00, like I do when booting
with the serial console) then I do get a login: prompt with the multiuser
boot, though since the keyboard isn't working I can't login. I was never able
to get this far with the ADBTEST110 kernel or a few others I've tried.

I don't know if this info might be useful to those of you working or kernels
or not but I thought I'd throw it out.

Brian


 >Hi !
>
>Well, it might be the result of an lucky accident, but I have managed it to
>compile a kernel which is booting fully to the login prompt on a LCII. It boots
>in single user as well as in multi user mode, just the keyboard is missing. You
>should be able to login through the seriel ports, but I have not yet tested
>that.
>
>How did it happen? I prepared to play around somewhat with John Wittkowski's
>ADB code and sup'ed fresh sources (Yes, I have sup running on my Linux box,
>now). So I built a kernel from the current soure without and with John's
>patches. To make the code smaller and simpler I turned off a lot of the config
>options (all net support and FPSP). The result should be one kernel going as
>good as the ADBTEST.110 kernel. But in opposite to ADBTEST.110 this kernel
>seems to be fully operational (except ADB support). No 3 minute delay, no stop
>when looking for the disks as seen with ADBTEST.110.
>
>If someone wants a copy of this kernel, get it from
>ftp://ftp.trier.fh-rpl.de/pub/incoming/netbsd.lc5.gz
>
>Not tired I compiled a third kernel with some debugging messages (printf's)
>inside the new adb code without changing any functionality. And, surprise,
>surprise this third kernel is NOT running. It stops where ADBTEST.110 has it's
>3 minute delay.
>
>A look trough the code showed me that at the delay/stop time a lot of ROM
>addresses get moved. John commented out some stuff in this file (macrom.c) so
>the changing functionality may be a result of wrong/not initialized variables.
>Work is in progress.
>
>Question: In my source-code-travels I found some minor errors/typos in the
>sources. How can I put this changes into the master source tree? I've heard of
>a tool called 'cvs'. Is this for everyone's use? What is good behavior when
>changing source in the master tree?
>
>Good Night
>
>Markus
>
>
>--
>-----------------------------------------------------------------------
> Markus Hitter                 email: hitterm@trier.fh-rpl.de
>                               WWW: http://www.trier.fh-rpl.de/~hitterm
> Trier (Germany)               Voice: ++49/0651/828172
>-----------------------------------------------------------------------