Checks for all floating-point faulting conditions for the frcpa instruction. It propagates NaNs, NaTVals, and special IEEE results.