28   // The long double representation, with the high and low portions of  in __fixunstfti()
32     double d[2];               // [0] is the high double, [1] is the low double.  in __fixunstfti()
33     unsigned long long ull[2]; // High and low doubles as 64-bit integers.  in __fixunstfti()
41   // Retrieve the 64-bit patterns of high and low doubles.  in __fixunstfti()
42   // Compute the unbiased exponent of both high and low doubles by  in __fixunstfti()
50   // CASE 1: High or low double fits in int64  in __fixunstfti()
53   // CASE 2: High or low double does not fit in int64  in __fixunstfti()
62   // from long double to 64-bit integer is needed. The low double can be either  in __fixunstfti()
64   // of the low double and to ensure it does not simply get converted to 0.  in __fixunstfti()
90   // CASE 1 - Low double fits in int64.  in __fixunstfti()
94     // CASE 2 - Low double does not fit in int64, scale and convert it.  in __fixunstfti()
102   // If the low double is negative, it may change the integer value of the  in __fixunstfti()
112   // Add the high and low doublewords together to form a 128 bit integer.  in __fixunstfti()