Subject: Re: ancient snapshot version of boot anyone?
To: Timm Wetzel <twetzel@gwdg.de>
From: Darrin B. Jewell <dbj@netbsd.org>
List: port-next68k
Date: 01/14/2000 10:28:36
I have the bootblocks as I originally received them from Rolf.
I have placed a copy at 
ftp://ftp.netbsd.org/pub/NetBSD/arch/next68k/archaic/sys.arch.next.tar.gz
It includes a binary, and is the oldest version I have.

I'm surprized that anything changed, but I should have bootblock
sources of all versions of boot since I received them from Rolf.
If the sources in public CVS aren't old enough, I have an archive
of my working CVS repository from before the initial NetBSD import.

Sometime in late '98, I did look at the turbo stuff a little.
Although I could get some code booted and executing off of the SCSI
disk, the kernel as loaded had large gaps of NULL bytes instead of
code and it would crash as soon as it hit the missing instructions.  I
put that project aside and haven't looked at turbo machines much
since.  Perhaps an older version of the bootloader wouldn't have had
this problem.

As far as getting a kernel loaded on the Turbo machines, it is
possible to link a Mach-O kernel using the native NeXT toolchain
and use the NeXT bootblocks to load it.  This is what I was doing
before Rolf's bootblocks conveniently solved that problem.

I hope this helps.  Please let us know how things go.

Darrin

Timm Wetzel  <twetzel@gwdg.de> writes:

> Does anyone happen to have a copy/backup of one of the very first next68k
> snapshots around?
> 
> I am looking for old versions of boot, since I used to be able to load a
> kernel from a SCSI disk on a turbo slab. Sadly I overwrote this bootblock
> during an upgrade.
> 
> The recent boot versions don't work on turbo hardware any more.
> I tried compiling with the obvious CVS-visible changes reverted, but this
> did not work, either (at least on my cross-compile setup).
> 
> So if anyone still has a _really_ ancient boot available, I'd appreciate a
> copy/pointer.
> 
> Pointers to the relevant code changes are also welcome. (I suspect there
> might have been revisions DBJ never committed to CVS.)
> 
> Timm
> PS: I was thinking of building a md-root kernel to experiment with color
> console and turbo hardware without the unsupported turbo DMA getting in
> the way. But this still needs to be loaded somehow ...
> 
> -- 
> Timm Wetzel     	<twetzel@gwdg.de>
> Biomed. NMR GmbH        Tel +49 551 201-1091    FAX +49 551 201-1307
> Am Fassberg 11, D-37077 Goettingen, Germany