FP flags

FP operations may set or test several bits that can be accessed by the programmer as pseudo-booleans:

FP_OVR        Floating point overflow; may be set on any FP operation (also if divide by zero). Read only.

FP_UND        Floating point underflow; may be set on any FP operation. Read only.

FP_IOP        Invalid floating point operation; may be set on any FP operation if argument does not match the function (e.g.: SQRT(-1)). Read only.

When FP_OVR is set, the offending function returns the REAL maximum value (≈6.8E38), adjusted according to the argument's sign.

When FP_UND is set, the offending function returns the REAL minimum value (≈1.1755E-38), adjusted according to the argument's sign.