|
|
|
@ -58,11 +58,12 @@ public:
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
Float<M, E> operator*(const Float<M, E>& flt) const {
|
|
|
|
|
if ((this->value == 0.f && !std::isnan(flt.value)) ||
|
|
|
|
|
(flt.value == 0.f && !std::isnan(this->value)))
|
|
|
|
|
// PICA gives 0 instead of NaN when multiplying by inf
|
|
|
|
|
return Zero();
|
|
|
|
|
return Float<M, E>::FromFloat32(ToFloat32() * flt.ToFloat32());
|
|
|
|
|
float result = value * flt.ToFloat32();
|
|
|
|
|
// PICA gives 0 instead of NaN when multiplying by inf
|
|
|
|
|
if (!std::isnan(value) && !std::isnan(flt.ToFloat32()))
|
|
|
|
|
if (std::isnan(result))
|
|
|
|
|
result = 0.f;
|
|
|
|
|
return Float<M, E>::FromFloat32(result);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
Float<M, E> operator/(const Float<M, E>& flt) const {
|
|
|
|
@ -78,12 +79,7 @@ public:
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
Float<M, E>& operator*=(const Float<M, E>& flt) {
|
|
|
|
|
if ((this->value == 0.f && !std::isnan(flt.value)) ||
|
|
|
|
|
(flt.value == 0.f && !std::isnan(this->value)))
|
|
|
|
|
// PICA gives 0 instead of NaN when multiplying by inf
|
|
|
|
|
*this = Zero();
|
|
|
|
|
else
|
|
|
|
|
value *= flt.ToFloat32();
|
|
|
|
|
value = operator*(flt).value;
|
|
|
|
|
return *this;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|