hashirama/uvg266
1
0
Fork 0
mirror of https://github.com/ultravideo/uvg266.git synced 2024-11-30 12:44:07 +00:00
Code Issues Actions Projects Releases Wiki Activity

removed some variables to prevent memory losses

Browse source
This commit is contained in:
Reima Hyvönen 2019-03-29 15:32:29 +02:00 committed by Pauli Oikkonen
parent 247c3a7a71
commit f146de7acb
1 changed files with 21 additions and 54 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

75
src/strategies/avx2/sao-avx2.c
View file

@ -85,7 +85,6 @@ static int sao_edge_ddistortion_avx2(const kvz_pixel *orig_data,
int offsets[NUM_SAO_EDGE_CATEGORIES]) int offsets[NUM_SAO_EDGE_CATEGORIES])
{ {
int y, x; int y, x;
unsigned int sum = 0;
vector2d_t a_ofs = g_sao_edge_offsets[eo_class][0]; vector2d_t a_ofs = g_sao_edge_offsets[eo_class][0];
vector2d_t b_ofs = g_sao_edge_offsets[eo_class][1]; vector2d_t b_ofs = g_sao_edge_offsets[eo_class][1];
@ -123,73 +122,41 @@ static int sao_edge_ddistortion_avx2(const kvz_pixel *orig_data,
tmp_sum_epi32 = _mm256_add_epi32(tmp_sum_epi32, _mm256_sub_epi32(tmp1_vec_epi32, tmp2_vec_epi32)); tmp_sum_epi32 = _mm256_add_epi32(tmp_sum_epi32, _mm256_sub_epi32(tmp1_vec_epi32, tmp2_vec_epi32));
} }
bool use_6_elements = block_width - x - 1 == 6; // Load the last 6 pixels to use
switch (use_6_elements) const kvz_pixel *c_data = &rec_data[y * block_width + x];
{
case true:;
// Load the last 6 pixels to use
const kvz_pixel *c_data = &rec_data[y * block_width + x]; __m128i vector_a_epi8 = load_6_pixels(&c_data[a_ofs.y * block_width + a_ofs.x]);
__m128i vector_c_epi8 = load_6_pixels(c_data);
__m128i vector_b_epi8 = load_6_pixels(&c_data[b_ofs.y * block_width + b_ofs.x]);
__m128i vector_a_epi8 = load_6_pixels(&c_data[a_ofs.y * block_width + a_ofs.x]); __m256i v_cat_epi32 = sao_calc_eo_cat_avx2(&vector_a_epi8, &vector_b_epi8, &vector_c_epi8);
__m128i vector_c_epi8 = load_6_pixels(c_data);
__m128i vector_b_epi8 = load_6_pixels(&c_data[b_ofs.y * block_width + b_ofs.x]);
__m256i v_cat_epi32 = sao_calc_eo_cat_avx2(&vector_a_epi8, &vector_b_epi8, &vector_c_epi8); const kvz_pixel* orig_ptr = &(orig_data[y * block_width + x]);
const kvz_pixel* orig_ptr = &(orig_data[y * block_width + x]); tmp_diff_epi32 = _mm256_cvtepu8_epi32(load_6_pixels(orig_ptr));
tmp_diff_epi32 = _mm256_cvtepu8_epi32(load_6_pixels(orig_ptr)); tmp_diff_epi32 = _mm256_sub_epi32(tmp_diff_epi32, _mm256_cvtepu8_epi32(vector_c_epi8));
tmp_diff_epi32 = _mm256_sub_epi32(tmp_diff_epi32, _mm256_cvtepu8_epi32(vector_c_epi8)); tmp_offset_epi32 = _mm256_permutevar8x32_epi32(offsets_epi32, v_cat_epi32);
tmp_offset_epi32 = _mm256_permutevar8x32_epi32(offsets_epi32, v_cat_epi32); // (diff - offset) * (diff - offset)
tmp1_vec_epi32 = _mm256_mullo_epi32(_mm256_sub_epi32(tmp_diff_epi32, tmp_offset_epi32), _mm256_sub_epi32(tmp_diff_epi32, tmp_offset_epi32));
// (diff - offset) * (diff - offset) // diff * diff
tmp1_vec_epi32 = _mm256_mullo_epi32(_mm256_sub_epi32(tmp_diff_epi32, tmp_offset_epi32), _mm256_sub_epi32(tmp_diff_epi32, tmp_offset_epi32)); tmp2_vec_epi32 = _mm256_mullo_epi32(tmp_diff_epi32, tmp_diff_epi32);
// diff * diff // Offset is applied to reconstruction, so it is subtracted from diff.
tmp2_vec_epi32 = _mm256_mullo_epi32(tmp_diff_epi32, tmp_diff_epi32); // sum += (diff - offset) * (diff - offset) - diff * diff;
// Offset is applied to reconstruction, so it is subtracted from diff. tmp_sum_epi32 = _mm256_add_epi32(tmp_sum_epi32, _mm256_sub_epi32(tmp1_vec_epi32, tmp2_vec_epi32));
// sum += (diff - offset) * (diff - offset) - diff * diff;
tmp_sum_epi32 = _mm256_add_epi32(tmp_sum_epi32, _mm256_sub_epi32(tmp1_vec_epi32, tmp2_vec_epi32)); tmp_sum_epi32 = _mm256_hadd_epi32(tmp_sum_epi32, tmp_sum_epi32);
tmp_sum_epi32 = _mm256_hadd_epi32(tmp_sum_epi32, tmp_sum_epi32);
tmp_sum_epi32 = _mm256_hadd_epi32(tmp_sum_epi32, tmp_sum_epi32); tmp_sum_epi32 = _mm256_add_epi32(tmp_sum_epi32, _mm256_shuffle_epi32(tmp_sum_epi32, _MM_SHUFFLE(0, 1, 0, 1)));
tmp_sum_epi32 = _mm256_hadd_epi32(tmp_sum_epi32, tmp_sum_epi32);
tmp_sum_epi32 = _mm256_add_epi32(tmp_sum_epi32, _mm256_shuffle_epi32(tmp_sum_epi32, _MM_SHUFFLE(0, 1, 0, 1)));
sum += _mm_cvtsi128_si32(_mm256_castsi256_si128(tmp_sum_epi32));
break;
default:
// Only if theres odd number of pixels left
for (int i = x; i < block_width - 1; ++i) {
const kvz_pixel *c_data = &rec_data[y * block_width + i];
kvz_pixel a = c_data[a_ofs.y * block_width + a_ofs.x];
kvz_pixel c = c_data[0];
kvz_pixel b = c_data[b_ofs.y * block_width + b_ofs.x];
int offset = offsets[sao_calc_eo_cat(a, b, c)];
if (offset != 0) {
int diff = orig_data[y * block_width + i] - c;
// Offset is applied to reconstruction, so it is subtracted from diff.
sum += (diff - offset) * (diff - offset) - diff * diff;
}
}
break;
}
} }
return (_mm_cvtsi128_si32(_mm256_castsi256_si128(tmp_sum_epi32)));
return sum;
} }
/** /**