downcasting a double/float to a uint8_t

Let’s say I have the following bit of code, which tries to ‘downcast’ a double value into a uint8_t:

uint8_t downcast(uint8_t *down1, uint8_t *down2)
{
    //the usual nullptr checks
    double fval = 100.0; //fits the uint8_t domain range
    *down1 = (uint8_t)fval;  //should work...

    fval = 1000.0; //does not fit uint8_t domain range
    *down2 = (uint8_t)fval;  //will this work??
}

I’ve seen a few posts that suggest doing this sort of "quantization-cast", but I’m not sure this is correct. I seem to recall reading that a double var whose value exceeds the output domain, will lead to undefined behavior from the cast (either C or static_cast<uint8_t>(fval)).

Thoughts?

>Solution :

The behaviour of the second cast is undefined.

C17 §6.3.1.4 ¶1 When a finite value of real floating type is converted to an integer type other than _Bool, the fractional part is discarded (i.e., the value is truncated toward zero). If the value of the integral part cannot be represented by the integer type, the behavior is undefined.