mirror of
https://github.com/g-truc/glm.git
synced 2024-12-02 04:34:34 +00:00
69 lines
1.7 KiB
C++
69 lines
1.7 KiB
C++
/// @ref core
|
|
/// @file glm/detail/func_integer_simd.inl
|
|
|
|
#include "../simd/integer.h"
|
|
|
|
#if GLM_ARCH & GLM_ARCH_SSE2_BIT
|
|
|
|
namespace glm{
|
|
namespace detail
|
|
{
|
|
template<glm::precision P>
|
|
struct compute_bitfieldReverseStep<4, uint32, P, vec, true, true>
|
|
{
|
|
GLM_FUNC_QUALIFIER static vec<4, uint32, P> call(vec<4, uint32, P> const& v, uint32 Mask, uint32 Shift)
|
|
{
|
|
__m128i const set0 = v.data;
|
|
|
|
__m128i const set1 = _mm_set1_epi32(static_cast<int>(Mask));
|
|
__m128i const and1 = _mm_and_si128(set0, set1);
|
|
__m128i const sft1 = _mm_slli_epi32(and1, Shift);
|
|
|
|
__m128i const set2 = _mm_andnot_si128(set0, _mm_set1_epi32(-1));
|
|
__m128i const and2 = _mm_and_si128(set0, set2);
|
|
__m128i const sft2 = _mm_srai_epi32(and2, Shift);
|
|
|
|
__m128i const or0 = _mm_or_si128(sft1, sft2);
|
|
|
|
return or0;
|
|
}
|
|
};
|
|
|
|
template<glm::precision P>
|
|
struct compute_bitfieldBitCountStep<4, uint32, P, vec, true, true>
|
|
{
|
|
GLM_FUNC_QUALIFIER static vec<4, uint32, P> call(vec<4, uint32, P> const& v, uint32 Mask, uint32 Shift)
|
|
{
|
|
__m128i const set0 = v.data;
|
|
|
|
__m128i const set1 = _mm_set1_epi32(static_cast<int>(Mask));
|
|
__m128i const and0 = _mm_and_si128(set0, set1);
|
|
__m128i const sft0 = _mm_slli_epi32(set0, Shift);
|
|
__m128i const and1 = _mm_and_si128(sft0, set1);
|
|
__m128i const add0 = _mm_add_epi32(and0, and1);
|
|
|
|
return add0;
|
|
}
|
|
};
|
|
}//namespace detail
|
|
|
|
# if GLM_ARCH & GLM_ARCH_AVX_BIT
|
|
template<>
|
|
GLM_FUNC_QUALIFIER int bitCount(uint32 x)
|
|
{
|
|
return _mm_popcnt_u32(x);
|
|
}
|
|
|
|
# if(GLM_MODEL == GLM_MODEL_64)
|
|
template<>
|
|
GLM_FUNC_QUALIFIER int bitCount(uint64 x)
|
|
{
|
|
return static_cast<int>(_mm_popcnt_u64(x));
|
|
}
|
|
# endif//GLM_MODEL
|
|
# endif//GLM_ARCH
|
|
|
|
}//namespace glm
|
|
|
|
#endif//GLM_ARCH & GLM_ARCH_SSE2_BIT
|