How do I work around what seems to be integer overflow despite the type being large enough

Question

I am performing the following calculation:

unsigned long long int interval = (ui->spinBox_2->value() * 60 * 60 * 1000) + (ui->spinBox_3->value() * 60 * 1000) + (ui->spinBox_4->value() * 1000) + (ui->spinBox_5->value());

What it's basically doing is taking in the hours, minutes, seconds, and milliseconds and converting it all to milliseconds.

Example: 1 hour 0 minutes 2 seconds 100 milliseconds

= (1 * 60 * 60 1000) + (0 * 60 * 1000) + (2 * 1000) + (100)

If I put 999 hours for example, it outputs 18446744073010984420 milliseconds when it should output 3596400000 milliseconds.

The type is unsigned long long int, which should be able to store up to 18446744073709551615, which is significantly bigger than the calculation.

Is there a way to fix this, or an alternative way to save the time in milliseconds rather than calculating and storing it in an unsigned long long int variable?

Answer 1

The promotion to an unsigned long long happens on assignment. your calculation however happens as ints so the calculation overflows before it has a chance to be unsigned long long.

You can either cast your operands to unsigned long long or use a ULL suffix to make them unsigned long long : 60ULL