Mason Loring Bliss wrote:

>  SE/30, July 10th -current, GENERIC-36 kernel
> 
> The problem:
> 
>  I briefly tried using the GENERIC-35 kernel, and I had a lot of trouble
> with the adb routines. Here's a successful output from GENERIC-36:
> 
> adb: using II series hardware support
> adb: cleanup: nothing returned
> adb: ADBReInit complete
> adb: keyboard at 2
> adb: 100 dpi mouse at 3
> 
> When I was using GENERIC-35 (SBC, but that's irrelevant, I think) I'd get
> as far as the "adb: cleanup: nothing returned" line and the machine would
> hang. Alternately, sometimes it wouldn't hang, but the keyboard wouldn't
> talk to the rest of the system. The adb port into which the keyboard was
> plugged didn't really seem to matter.
> 
> I've read some of the send-pr docs, and I've submitted a request for a
> number, but I haven't received it just yet (it may be waiting for me to
> poll - I don't know) and I wanted folks to know about this. Incidentally,

Don't bother.  I don't think that NetBSD requires a specific send-pr
number.  You can just run send-pr and it will handle everything for you
(assuming you have a net connection which accepts mail, that is).

> when I queried the problem database, the only thing that came up was a
> report (on an alpha?) about the system freezing when the mouse was moved
> during an adb poll. The mouse isn't moving in this case, FWIW.

I'm not entirely sure that this is still a problem.  If I remember, I'll
try moving the mouse during boot when I go home tonight.

> It seems like there must be a bug in the adb manager code... I read in the
> readme file that lives with the recent kernels:
> 
> >           #     Source date
> >         ====    ===========
> >          30     30 Apr 1997 [1]
> >          32     01 Jun 1997
> >          33     30 Jun 1997 [2]
> >          34     10 Jul 1997 [2]
> >          35     23 Jul 1997 [1][2]
> >          36     23 Jul 1997 [2]
> >
> > [1] These kernels are compiled using the new ADB hardware driver from
> >     John P. Wittkoski, including support written by Takashi Hamada for
> >     many PowerBooks.
> 
> I had previously been using GENERIC-32 without problems, IIRC, and the
> problems only started when I slapped in GENERIC-35. Now that I'm using
> GENERIC-36, the problem doesn't seem to exist. This would seem to indicate
> a problem somewhere with John's code.

Hmmm...I had thought that John's code had been fairly well tested on most
machines that it was known to work on...can you compile your own kernels?
If so, you can compile a -current kernel with the line

options 	MRG_ADB

uncommented in it.

> If one hasn't been submitted by the time I get my send-pr ID, I'll send in
> a formal PR, but for now I figured it was worthwhile just to get the
> information out there somehow.

If no one comes up with any workable solutions, please do send a PR.

Later.

-- 
Colin Wood                                 cwood@ichips.intel.com
Component Design Engineer - MD6                 Intel Corporation
-----------------------------------------------------------------
I speak only on my own behalf, not for my employer.