🚀 go-pugleaf

This last weekend I finally started some serious work on my long-planned
video-game-to-end-all-video-games, and encountered a strange problem.  When
I reboot my A2500 after a system crash, there's about a 30% chance that my
harddisk will be accessable in 68030 mode.  Most of the time, I'm forced to
select 68000 mode to be able to boot from the harddisk.  This is not good.

I am using a relatively new (6 months?) A2500/30 with a 40 meg harddisk.  I'm
doing my development in assembly language, using Devpac Assembler.  My code
doesn't disable Intuition, but it does call OwnBlitter and then writes directly
to the hardware registers (followed, of course, by DisownBlitter).  It doesn't
seem to matter whether the crash is inside or outside of the section of code
that owns the blitter.  At first, I thought might be related to my code
setting the Blitter Nasty bit and not resetting it, but when I stopped playing
with that bit the problem didn't go away.  I don't use setcpu to change any of
the default 68030 operating parameters.

The typical scenario is for me to be running my program from within the Devpac
Debugger in 68030 mode.  I'll get to whatever section hasn't been debugged yet
and cause a crash, for example by freeing an unallocated raster.  I hit CNTRL-
AMIGA-AMIGA, the display flashes a while, then I get the prompt for insert
Workbench disk.  Resetting again at this point gets me to the same point.  So,
I reset and hold the mouse buttons, select 68000 mode, and it boots from the
harddisk with no problem.  The next time the machine crashes, it will generally
boot directly from the harddisk without my selecting a processor, but the mode
it chooses is random.  Sometimes it comes up in 68030 mode, sometimes 68000.

What is causing this?  Is there anything I can do to fix the problem?

-brett

-- 
|Brett Coon           |  uunet!smosjc!brett         |
|S-MOS Systems, Inc.  |  "You like 'em, anchovies?" |
|San Jose, CA         |            -Runaway Train   |