On Wed, 19 Nov 1997, Michael R Zucca wrote:

> > This is definitely the video mapping problem (since that's pretty much
> > part of what occurs next).  If it were an ADB problem, you'd hang much
> > later in the boot sequence (after the copyright message, during ADB
> > initialization).
> 
> Yep. I sent many an hour finding out exactly what was happening in that gap.
> 
> > Yes, he did.  Of course, figuring out the "right way" is the hard part :-)
> > There are actually a fair number of machines which will depend on this,
> > including all of the PowerBooks, the Classic Macs, and the Quadra 605, to
> > name a few.
> 
> (This is from ancient memories of mine)
> Well, it looked to me like we were screwing something up in pmap_bootstrap.c
> and machdep.c.
> 
> The 160/180's were particularly susceptible to a bug in pmap_bootstrap.
> If I remember correctly, when we boot with intvid pmap_bootstrap twiddles
> with the way memory is mapped. When we boot serial, all of that behavior
> is side-stepped and we just inherit the mappings from MacOS.
> 
> To fix the 160/180's I simply put in a hack that did the serial mapping
> if it was that particular model. However, what we really need to do is
> figure out how memory is *really* mapped on every machine and then go
> back and fix our memory mapping code.

Are there any ncrscsi kernels compiiled with this change made?  I'm
anxious to see if my 145 will still boot with current kernels (last tested
based on a kernel Allen compiled for me a few months ago).


David

-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS/CC/FA/H/L/MC/M/MU/PA/TW d-@ s:>- a-- C++++>$ UBLAS*++++>$
P+?>$ L+++>$ !E--- W+++>$ N++(+++)>+++$ !o? K-? !w--- !O
M++>$ !V-- PS+>$ !PE- Y+>$ PGP+>$ t+++>$ 5+>++++$ !X- !R tv+>$
b++>$ !DI !D- G++(+++)>$ e>++++ h--! r--- !y-
------END GEEK CODE BLOCK------