jp9000/obs-studio
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

libobs: Simplify util_mul_div64 for x64 on Windows

Browse Source 
Use intrinsics to perform 128-bit math directly.
This commit is contained in:
jpark37 2021-09-29 01:24:16 -07:00 committed by Jim
parent 9deb8fe292
commit 0c0ff1195f
1 changed files with 11 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
libobs/util/util_uint64.h
View File

@ -16,9 +16,19 @@
#pragma once
#if defined(_MSC_VER) && defined(_M_X64)
#include <intrin.h>
#endif
static inline uint64_t util_mul_div64(uint64_t num, uint64_t mul, uint64_t div)
{
#if defined(_MSC_VER) && defined(_M_X64)
unsigned __int64 high;
const unsigned __int64 low = _umul128(num, mul, &high);
unsigned __int64 rem;
return _udiv128(high, low, div, &rem);
#else
const uint64_t rem = num % div;
return (num / div) * mul + (rem * mul) / div;
#endif
}