Better comment.
This commit is contained in:
parent
0164723a8e
commit
758dc511fb
|
@ -67,9 +67,10 @@ really_inline double compute_float_64(int64_t power, uint64_t i, bool negative,
|
||||||
// We are going to need to do some 64-bit arithmetic to get a more precise product.
|
// We are going to need to do some 64-bit arithmetic to get a more precise product.
|
||||||
// We use a table lookup approach.
|
// We use a table lookup approach.
|
||||||
components c =
|
components c =
|
||||||
power_of_ten_components[power - FASTFLOAT_SMALLEST_POWER]; // safe because
|
power_of_ten_components[power - FASTFLOAT_SMALLEST_POWER];
|
||||||
// power_index
|
// safe because
|
||||||
// <= 2*308
|
// power >= FASTFLOAT_SMALLEST_POWER
|
||||||
|
// and power <= FASTFLOAT_LARGEST_POWER
|
||||||
// we recover the mantissa of the power, it has a leading 1. It is always
|
// we recover the mantissa of the power, it has a leading 1. It is always
|
||||||
// rounded down.
|
// rounded down.
|
||||||
uint64_t factor_mantissa = c.mantissa;
|
uint64_t factor_mantissa = c.mantissa;
|
||||||
|
|
Loading…
Reference in New Issue