On Sun, 19 Jul 1998, Mike Hibler wrote:
> I haven't been following this very closely so I may be stating the
> obvious (or totally off base, take your pick :-) but it looks like you
> may be getting a serial line interrupt before the kernel has configured
> the dca (serial) device.  If the interrupt handler hasn't been
> registered, this could cause bad things to happen (though it should
> catch unexpected interrupts more gracefully I would think).  This might
> not happen with the graphics device as console since the boot ROM would
> then not be configuring the serial port (as it must when using the
> serial console). 

This would be consistent with observed behavior.

Can someone compile a kernel which probes and attaches the serial devices
before anything else?  I'm not even sure this would solve the problem but
its as good a shot in the dark as any.

/* 
   Matthew N. Dodd		| A memory retaining a love you had for life	
   winter@jurai.net		| As cruel as it seems nothing ever seems to
   http://www.jurai.net/~winter | go right - FLA M 3.1:53	
*/