Make floating-point "NaN / 0" return NaN instead of raising an error.

Make floating-point "NaN / 0" return NaN instead of raising an error.

This is more consistent with the IEEE 754 spec and our treatment of
NaNs elsewhere; in particular, the case has always acted that way in
"numeric" arithmetic.

Noted by Dean Rasheed.

Discussion: http://postgr.es/m/3421746.1594927785@sss.pgh.pa.us