> 
>   And it got past the awk command again, so the patch gets around the
> problem - at least until something else feed "NaN" to awk.
> 
>   The real fix is to deal with the reserved operand fault better.
> 
The problem is that it thinks that it can use NaN as any other number,
which is illegal on vax (but legal with IEEE fp).  I assume that it
will give the same result under Ultrix.  The "reserved operand" fault
will beat you whenever any instruction touches an illegal number
so trying to do workarounds in the trap system won't help.

-- Ragge