Subject: Re: stop-a equiv?
To: Derek Peschel <email@example.com>
From: Christopher Tribo <firstname.lastname@example.org>
Date: 12/13/2002 21:55:35
There's also support for KGDB in kernel if you compile it in. If
you have the other serial port open you could wire that into the terminal
server, provided it has an open port, and KGDB works. Dunno if that buys
If you feel like rigging something up, It wouldn't be too big of a
stretch to come up with a circuit that for instance waits for DTR to go
high off the terminal server, and when it does, "flip" a bounceless switch
for like 5 seconds and wire it up to the soft power switch inside the
case. So when you connect to the port and DTR goes high, it turns the
computer off. Disconnect, and reconnect and turn it back on :) A few
transistors, a few gates and you never know hehe. Hell, you might even be
able to wire DTR directly to the power switch with just an isolation
transformer and a resistor to keep the voltage around 5v. Just a thought.
On Fri, 13 Dec 2002, Derek Peschel wrote:
> der Mouse explained the situation on Suns. On the Mac with NetBSD, OF
> stays in memory once the OS has booted, and the OS calls it for some
> hardware access. Keyboard access is handled by talking directly to
> the keyboard, so OF couldn't snoop for its "stop" keystroke.
I think there is an NMI that might be possible to program so that
it dumps you back into OF, I don't think this can be done from the
keyboard or serial console either though :(
> But der Mouse also mentioned ddb, the kernel debugger. Would that work
> for you? It can sync and reboot and do other things. On my machine,
> the keystroke is control-option-escape at the console. You might be
> able to set up a keystroke that will go through your terminal server.
> You should also make sure you can run ddb reliably (I discovered the
> 1.5 kernel needed some patches before it would work for me).