Microsoft DirectX 8.1 (C++) |
The Int64x32Div32 function implements the formula ((a*b)+rnd)/c
where a is a 64-bit value and b, c, and rnd are 32-bit values.
Time stamps and seek times are 64-bit values, so this function is useful for performing conversions on 32-bit systems. For example, in MPEG-1 the system clock reference is 90-kHz, or 90,000 ticks per second. The formula to convert this to reference time (100-nanosecond units) is
(timestamp * 1000) / 9
which can be calculated as Int64x32Div32(
timestamp, 1000, 9, 0)
. Use the rnd parameter as a rounding factor.
Syntax
LONGLONG WINAPI Int64x32Div32(
LONGLONG a,
LONG b,
LONG c,
LONG rnd
);
Return Value
Returns either the (a * b + rnd)/c calculation or one of the following values.
0x7FFFFFFFFFFFFFFF | Overflow occurred because the result is too large (positive). |
0x8000000000000000 | Overflow occurred because the result is too large (negative). |
Remarks
Rounding on the division is toward zero. Division by zero is counted as an overflow condition.
See Also