This does not happen on any of my hp300-based boxen, which includes:

It does happen occasionally on my 425 with 32MB.  I've not had the
time or patience recently to look at it.

   However, on all of my machines, I have pmap problems if I have 
   MAPPEDCOPY in the kernel config.

I've never actually tried that.

   Also, all of 
   my 32 meg boxen (with the exception of the 433) are quirky...Like they 
   just `freeze', and occasionally panic with `out of address space', [...]

Never seen that.

   For my own personal edification (sp?), can someone please enlighten me on 
   just exactly what MAPPEDCOPY is supposed to accomplish?

It's supposed to improve copyin/copyout performance for large
transfers (like file system blocks), by avoiding the use of the `movs'
instruction.  Another way to accomplish this would be to move the
kernel and the user code into the same address space, which I plan to
experiment with when I have more time.