third_party/opencv
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Updated AVX2 implementation of v_popcount for u8.

Browse Source
This commit is contained in:
Vitaly Tuzov
2019-05-14 18:48:36 +03:00
parent 1220dd4877
commit 7a55f2af3b
4 changed files with 32 additions and 82 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/core/include/opencv2/core/hal/intrin_avx.hpp
+6 -16
View File
@@ -1188,14 +1188,11 @@ inline float v_reduce_sad(const v_float32x8& a, const v_float32x8& b)
/** Popcount **/
inline v_uint8x32 v_popcount(const v_uint8x32& a)
{
__m256i m1 = _mm256_set1_epi32(0x55555555);
__m256i m2 = _mm256_set1_epi32(0x33333333);
__m256i m4 = _mm256_set1_epi32(0x0f0f0f0f);
__m256i p = a.val;
p = _mm256_add_epi32(_mm256_and_si256(_mm256_srli_epi32(p, 1), m1), _mm256_and_si256(p, m1));
p = _mm256_add_epi32(_mm256_and_si256(_mm256_srli_epi32(p, 2), m2), _mm256_and_si256(p, m2));
p = _mm256_add_epi32(_mm256_and_si256(_mm256_srli_epi32(p, 4), m4), _mm256_and_si256(p, m4));
return v_uint8x32(p);
__m256i _popcnt_table = _mm256_setr_epi8(0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4,
0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4);
__m256i _popcnt_mask = _mm256_set1_epi8(0x0F);
return v_uint8x32(_mm256_add_epi8(_mm256_shuffle_epi8(_popcnt_table, _mm256_and_si256( a.val , _popcnt_mask)),
_mm256_shuffle_epi8(_popcnt_table, _mm256_and_si256(_mm256_srli_epi16(a.val, 4), _popcnt_mask))));
}
inline v_uint16x16 v_popcount(const v_uint16x16& a)
{
@@ -1212,14 +1209,7 @@ inline v_uint32x8 v_popcount(const v_uint32x8& a)
}
inline v_uint64x4 v_popcount(const v_uint64x4& a)
{
__m256i m1 = _mm256_set1_epi32(0x55555555);
__m256i m2 = _mm256_set1_epi32(0x33333333);
__m256i m4 = _mm256_set1_epi32(0x0f0f0f0f);
__m256i p = a.val;
p = _mm256_add_epi32(_mm256_and_si256(_mm256_srli_epi32(p, 1), m1), _mm256_and_si256(p, m1));
p = _mm256_add_epi32(_mm256_and_si256(_mm256_srli_epi32(p, 2), m2), _mm256_and_si256(p, m2));
p = _mm256_add_epi32(_mm256_and_si256(_mm256_srli_epi32(p, 4), m4), _mm256_and_si256(p, m4));
return v_uint64x4(_mm256_sad_epu8(p, _mm256_setzero_si256()));
return v_uint64x4(_mm256_sad_epu8(v_popcount(v_reinterpret_as_u8(a)).val, _mm256_setzero_si256()));
}
inline v_uint8x32 v_popcount(const v_int8x32& a)
{ return v_popcount(v_reinterpret_as_u8(a)); }
modules/core/include/opencv2/core/hal/intrin_sse.hpp
+1 -8
View File
@@ -1580,14 +1580,7 @@ inline v_uint32x4 v_popcount(const v_uint32x4& a)
}
inline v_uint64x2 v_popcount(const v_uint64x2& a)
{
__m128i m1 = _mm_set1_epi32(0x55555555);
__m128i m2 = _mm_set1_epi32(0x33333333);
__m128i m4 = _mm_set1_epi32(0x0f0f0f0f);
__m128i p = a.val;
p = _mm_add_epi32(_mm_and_si128(_mm_srli_epi32(p, 1), m1), _mm_and_si128(p, m1));
p = _mm_add_epi32(_mm_and_si128(_mm_srli_epi32(p, 2), m2), _mm_and_si128(p, m2));
p = _mm_add_epi32(_mm_and_si128(_mm_srli_epi32(p, 4), m4), _mm_and_si128(p, m4));
return v_uint64x2(_mm_sad_epu8(p, _mm_setzero_si128()));
return v_uint64x2(_mm_sad_epu8(v_popcount(v_reinterpret_as_u8(a)).val, _mm_setzero_si128()));
}
inline v_uint8x16 v_popcount(const v_int8x16& a)
{ return v_popcount(v_reinterpret_as_u8(a)); }
modules/core/include/opencv2/core/hal/intrin_vsx.hpp
+2 -2
View File
@@ -766,8 +766,8 @@ inline scalartype v_reduce_##suffix(const _Tpvec& a)
}
OPENCV_HAL_IMPL_VSX_REDUCE_OP_8(v_uint8x16, vec_uchar16, uchar, max, vec_max)
OPENCV_HAL_IMPL_VSX_REDUCE_OP_8(v_uint8x16, vec_uchar16, uchar, min, vec_min)
OPENCV_HAL_IMPL_VSX_REDUCE_OP_8(v_int8x16, vec_char8, schar, max, vec_max)
OPENCV_HAL_IMPL_VSX_REDUCE_OP_8(v_int8x16, vec_char8, schar, min, vec_min)
OPENCV_HAL_IMPL_VSX_REDUCE_OP_8(v_int8x16, vec_char16, schar, max, vec_max)
OPENCV_HAL_IMPL_VSX_REDUCE_OP_8(v_int8x16, vec_char16, schar, min, vec_min)
inline v_float32x4 v_reduce_sum4(const v_float32x4& a, const v_float32x4& b,
const v_float32x4& c, const v_float32x4& d)