mirror of
https://github.com/fmtlib/fmt.git
synced 2025-11-28 13:19:48 +01:00
Add initial support for 128-bit floats
This commit is contained in:
@@ -1252,10 +1252,10 @@ template <> struct float_info<double> {
|
||||
static const int shorter_interval_tie_upper_threshold = -77;
|
||||
};
|
||||
|
||||
// 80-bit extended precision long double.
|
||||
// An 80- or 128-bit floating point number.
|
||||
template <typename T>
|
||||
struct float_info<T, enable_if_t<std::is_same<T, long double>::value &&
|
||||
std::numeric_limits<T>::digits == 64>> {
|
||||
struct float_info<T, enable_if_t<std::numeric_limits<T>::digits == 64 ||
|
||||
std::numeric_limits<T>::digits == 113>> {
|
||||
using carrier_uint = detail::uint128_t;
|
||||
static const int exponent_bits = 15;
|
||||
};
|
||||
|
||||
Reference in New Issue
Block a user