Another hang...

[Johnny Billquist <bqt%softjar.se@localhost>]

I have managed to get a new variant of hang on my 4000/90.

When it hung before, I managed to break into the debugger. Stepping 
through it a little and then continuing made it wake up again. The build 
I was running was aborted, though.

However, without rebooting, I decided to just restart the build again.
So under /usr/src, I ran build.sh -u build, and started watching the 
progress.
It came to the same point in libm as before, and hung again. So it 
didn't come any further. I find that kindof interesting. It would 
suggest that there is some combination of things happening that cause 
the hang, and it's not related to how much time have passed, or how much 
work the machine have been doing before.

Another interesting thing is that this time I can't break into ddb. The 
machine is not responding to any input on the console, and it's not 
responding to my ssh sessions either.
However, it is responding to pings.

So this time I can't find out at all what have happened.

Compared to the first hang, this is in a way inverted. First hang it 
didn't respond to pings, but I could break into ddb. This time it 
responds to pings, but I can't break into ddb.

I wonder if that will help anyone? :-)

        Johnny