Arithmetic overflow: Using operator 'operator' on a size-a byte value and then casting the result to a size-b byte value. Cast the value to the wider type before calling operator 'operator' to avoid overflow

This warning indicates incorrect behavior that results from integral promotion rules and types larger than the ones in which arithmetic is typically performed.

Code analysis detects when an integral value gets shifted left, multiplied, added, or subtracted, and the result gets cast to a wider integral type. If the operation overflows the narrower integral type, then data is lost. You can prevent this loss by casting the value to a wider type before the arithmetic operation.

The following code generates this warning:

void leftshift(int i) noexcept
{
  unsigned __int64 x;
  x = i << 31;  // C26451 reported here

  // code
}

To correct this warning, use the following code:

void leftshift(int i) noexcept
{
  unsigned __int64 x;
  x = static_cast<unsigned __int64>(i) << 31; // OK

  // code
}

void somefunc(__int64 /* y */) noexcept
{}

void callsomefunc(int x) noexcept
{
  somefunc(x * 2); // C26451 reported here
}

To correct this warning, use the following code:

void callsomefunc(int x) noexcept
{
  somefunc(static_cast<__int64>(x) * 2); // OK
}

__int64 add(int x) noexcept
{
  constexpr auto value = 2;
  return x += value; // C26451 reported here
}

To correct this warning, use the following code:

__int64 add(int x) noexcept
{
  constexpr auto value = 2;
  const __int64 y = static_cast<__int64>(x) + value; // OK
  return y;
}

• ES.103: Don't overflow