Generating an exception when NaN is assigned

[Paul Lalonde]

Does anybody know how to make this RS6000 generate a floating point exception
when I assign (inadvertently) NaN or Inf to a float or a double?  
I can't find anything on the FPU in the stack of manuals (God only knows
why IBM didn't stick to the traditional UNIX manual organization and
syntax - railroad tracks are hell).

	Thanks,
		Paul

 Paul A. Lalonde	 	Internet: lalonde at qucis.queensu.ca
 Home Phone: (613)546-4713	Work Phone: (613)545-7100

	"The only true law is that which leads to freedom"
			- Richard Bach, _Jonathan Livingston Seagull_