2 A bug in compile-time conversion of large floating-point constants, on host systems with a 32-bit long.
4 qual: const, volatile, const volatile,
11 // floating-point constant converted to unsigned integer.
12 {qual
} uint64_t f
=1e18
;
15 {qual
} uint64_t i
=1000000000000000000;
20 ASSERT (i
* 2 >= f
&& f
>= i
/ 2);