hashirama/uvg266
1
0
Fork 0
mirror of https://github.com/ultravideo/uvg266.git synced 2024-12-02 21:24:07 +00:00
Code Issues Actions 1 Projects Releases Wiki Activity

Clerified some 6 pixel loads

Browse source
This commit is contained in:
Reima Hyvönen 2019-03-22 14:56:51 +02:00 committed by Pauli Oikkonen
parent 1a8b211e05
commit bc0a36830d
1 changed files with 24 additions and 22 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

46
src/strategies/avx2/sao-avx2.c
View file

@ -36,6 +36,12 @@
// Might be useful to check that if (when) this file // Might be useful to check that if (when) this file
// is difficult to understand. // is difficult to understand.
static INLINE __m128i load_6_pixels(const kvz_pixel* data)
{
return _mm_insert_epi16(_mm_cvtsi32_si128(*(int32_t*)&(data[0])), *(int16_t*)&(data[4]), 2);
}
// Mapping of edge_idx values to eo-classes. // Mapping of edge_idx values to eo-classes.
static int sao_calc_eo_cat(kvz_pixel a, kvz_pixel b, kvz_pixel c) static int sao_calc_eo_cat(kvz_pixel a, kvz_pixel b, kvz_pixel c)
{ {
@ -50,15 +56,15 @@ static int sao_calc_eo_cat(kvz_pixel a, kvz_pixel b, kvz_pixel c)
// Mapping of edge_idx values to eo-classes. // Mapping of edge_idx values to eo-classes.
static __m256i sao_calc_eo_cat_avx2(__m128i vector_a_epi8, __m128i vector_b_epi8, __m128i vector_c_epi8) static __m256i sao_calc_eo_cat_avx2(__m128i* vector_a_epi8, __m128i* vector_b_epi8, __m128i* vector_c_epi8)
{ {
// Mapping relationships between a, b and c to eo_idx. // Mapping relationships between a, b and c to eo_idx.
__m256i vector_sao_eo_idx_to_eo_category_epi32 = _mm256_setr_epi32(1, 2, 0, 3, 4, 0, 0, 0); __m256i vector_sao_eo_idx_to_eo_category_epi32 = _mm256_setr_epi32(1, 2, 0, 3, 4, 0, 0, 0);
__m256i eo_idx_epi32 = _mm256_set1_epi32(2); __m256i eo_idx_epi32 = _mm256_set1_epi32(2);
__m256i vector_a_epi32 = _mm256_cvtepu8_epi32(vector_a_epi8); __m256i vector_a_epi32 = _mm256_cvtepu8_epi32(*vector_a_epi8);
__m256i vector_b_epi32 = _mm256_cvtepu8_epi32(vector_b_epi8); __m256i vector_b_epi32 = _mm256_cvtepu8_epi32(*vector_b_epi8);
__m256i vector_c_epi32 = _mm256_cvtepu8_epi32(vector_c_epi8); __m256i vector_c_epi32 = _mm256_cvtepu8_epi32(*vector_c_epi8);
__m256i temp1_epi32 = _mm256_sign_epi32(_mm256_set1_epi32(1), _mm256_sub_epi32(vector_c_epi32, vector_a_epi32)); __m256i temp1_epi32 = _mm256_sign_epi32(_mm256_set1_epi32(1), _mm256_sub_epi32(vector_c_epi32, vector_a_epi32));
__m256i temp2_epi32 = _mm256_sign_epi32(_mm256_set1_epi32(1), _mm256_sub_epi32(vector_c_epi32, vector_b_epi32)); __m256i temp2_epi32 = _mm256_sign_epi32(_mm256_set1_epi32(1), _mm256_sub_epi32(vector_c_epi32, vector_b_epi32));
@ -83,7 +89,7 @@ static int sao_edge_ddistortion_avx2(const kvz_pixel *orig_data,
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];
__m256i offsets_epi32 = _mm256_setr_epi32(offsets[0], offsets[1], offsets[2], offsets[3], offsets[4], 0, 0, 0); __m256i offsets_epi32 = _mm256_inserti128_si256(_mm256_castsi128_si256(_mm_loadu_si128((__m128i*) offsets)), _mm_insert_epi32(_mm_setzero_si128(), offsets[4], 0), 1);
__m256i tmp_diff_epi32; __m256i tmp_diff_epi32;
__m256i tmp_sum_epi32 = _mm256_setzero_si256(); __m256i tmp_sum_epi32 = _mm256_setzero_si256();
__m256i tmp_offset_epi32; __m256i tmp_offset_epi32;
@ -103,7 +109,7 @@ static int sao_edge_ddistortion_avx2(const kvz_pixel *orig_data,
__m128i vector_b_epi8 = _mm_loadl_epi64((__m128i*)&c_data[b_ofs.y * block_width + b_ofs.x]); __m128i vector_b_epi8 = _mm_loadl_epi64((__m128i*)&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); __m256i v_cat_epi32 = sao_calc_eo_cat_avx2(&vector_a_epi8, &vector_b_epi8, &vector_c_epi8);
tmp_diff_epi32 = _mm256_load_si256((__m256i*)&orig_data[y * block_width + x] - c); tmp_diff_epi32 = _mm256_load_si256((__m256i*)&orig_data[y * block_width + x] - c);
@ -132,21 +138,19 @@ static int sao_edge_ddistortion_avx2(const kvz_pixel *orig_data,
// Load the last 6 pixels to use // Load the last 6 pixels to use
const kvz_pixel *c_data = &rec_data[y * block_width + x]; const kvz_pixel *c_data = &rec_data[y * block_width + x];
const kvz_pixel *c_data2 = &rec_data[y * block_width + x + 2];
const kvz_pixel *c_data4 = &rec_data[y * block_width + x + 4];
kvz_pixel c = c_data[0]; kvz_pixel c = c_data[0];
__m128i vector_a_epi8 = _mm_setr_epi16(c_data[a_ofs.y * block_width + a_ofs.x], c_data2[a_ofs.y * block_width + a_ofs.x], c_data4[a_ofs.y * block_width + a_ofs.x], 0, 0, 0, 0, 0); __m128i vector_a_epi8 = load_6_pixels(&c_data[a_ofs.y * block_width + a_ofs.x]);
__m128i vector_c_epi8 = _mm_setr_epi16(c_data[0], c_data2[0], c_data4[0], 0, 0, 0, 0, 0); __m128i vector_c_epi8 = load_6_pixels(c_data);
__m128i vector_b_epi8 = _mm_setr_epi16(c_data[b_ofs.y * block_width + b_ofs.x], c_data2[b_ofs.y * block_width + b_ofs.x], c_data4[b_ofs.y * block_width + b_ofs.x], 0, 0, 0, 0, 0); __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); __m256i v_cat_epi32 = sao_calc_eo_cat_avx2(&vector_a_epi8, &vector_b_epi8, &vector_c_epi8);
tmp_diff_epi32 = _mm256_castsi128_si256(_mm_loadu_si128((__m128i*)&orig_data[y * block_width + x] - c)); tmp_diff_epi32 = _mm256_castsi128_si256(_mm_loadu_si128((__m128i*)&orig_data[y * block_width + x] - c));
__m128i diff_upper_epi32 = _mm_loadl_epi64((__m128i*)&orig_data[y * block_width + x + 4] - c); __m128i diff_upper_epi32 = _mm_loadl_epi64((__m128i*)&orig_data[y * block_width + x + 4] - c);
_mm256_insertf128_si256(tmp_diff_epi32, diff_upper_epi32, 0x1); _mm256_inserti128_si256(tmp_diff_epi32, diff_upper_epi32, 1);
tmp_offset_epi32 = _mm256_permutevar8x32_epi32(offsets_epi32, v_cat_epi32); tmp_offset_epi32 = _mm256_permutevar8x32_epi32(offsets_epi32, v_cat_epi32);
@ -245,7 +249,7 @@ static void calc_sao_edge_dir_avx2(const kvz_pixel *orig_data,
__m128i vector_b_epi8 = _mm_loadl_epi64((__m128i*)&c_data[b_ofs.y * block_width + b_ofs.x]); __m128i vector_b_epi8 = _mm_loadl_epi64((__m128i*)&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); __m256i v_cat_epi32 = sao_calc_eo_cat_avx2(&vector_a_epi8, &vector_b_epi8, &vector_c_epi8);
// Check wich values are right for specific cat amount. // Check wich values are right for specific cat amount.
@ -327,20 +331,18 @@ static void calc_sao_edge_dir_avx2(const kvz_pixel *orig_data,
// Load the last 6 pixels to use // Load the last 6 pixels to use
const kvz_pixel *c_data = &rec_data[y * block_width + x]; const kvz_pixel *c_data = &rec_data[y * block_width + x];
const kvz_pixel *c_data2 = &rec_data[y * block_width + x + 2];
const kvz_pixel *c_data4 = &rec_data[y * block_width + x + 4];
kvz_pixel c = c_data[0]; kvz_pixel c = c_data[0];
__m128i vector_a_epi8 = _mm_setr_epi16(c_data[a_ofs.y * block_width + a_ofs.x], c_data2[a_ofs.y * block_width + a_ofs.x], c_data4[a_ofs.y * block_width + a_ofs.x], 0, 0, 0, 0, 0); __m128i vector_a_epi8 = load_6_pixels(&c_data[a_ofs.y * block_width + a_ofs.x]);
__m128i vector_c_epi8 = _mm_setr_epi16(c_data[0], c_data2[0], c_data4[0], 0, 0, 0, 0, 0); __m128i vector_c_epi8 = load_6_pixels(c_data);
__m128i vector_b_epi8 = _mm_setr_epi16(c_data[b_ofs.y * block_width + b_ofs.x], c_data2[b_ofs.y * block_width + b_ofs.x], c_data4[b_ofs.y * block_width + b_ofs.x], 0, 0, 0, 0, 0); __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); __m256i v_cat_epi32 = sao_calc_eo_cat_avx2(&vector_a_epi8, &vector_b_epi8, &vector_c_epi8);
__m256i temp_mem_epi32 = _mm256_castsi128_si256(_mm_loadu_si128((__m128i*)&orig_data[y * block_width + x] - c)); __m256i temp_mem_epi32 = _mm256_castsi128_si256(_mm_loadu_si128((__m128i*)&orig_data[y * block_width + x] - c));
__m128i temp_mem_upper_epi32 = _mm_loadl_epi64((__m128i*)&orig_data[y * block_width + x + 4] - c); __m128i temp_mem_upper_epi32 = _mm_loadl_epi64((__m128i*)&orig_data[y * block_width + x + 4] - c);
_mm256_insertf128_si256(temp_mem_epi32, temp_mem_upper_epi32, 0x1); _mm256_inserti128_si256(temp_mem_epi32, temp_mem_upper_epi32, 1);
// Check wich values are right for specific cat amount. // Check wich values are right for specific cat amount.
// It's done for every single value that cat could get {1, 2, 0, 3, 4} // It's done for every single value that cat could get {1, 2, 0, 3, 4}
@ -520,7 +522,7 @@ static void sao_reconstruct_color_avx2(const encoder_control_t * const encoder,
__m128i vector_b_epi8 = _mm_loadl_epi64((__m128i*)&c_data[b_ofs.y * stride + b_ofs.x]); __m128i vector_b_epi8 = _mm_loadl_epi64((__m128i*)&c_data[b_ofs.y * stride + b_ofs.x]);
__m256i v_cat_epi32 = sao_calc_eo_cat_avx2(vector_a_epi8, vector_b_epi8, vector_c_epi8); __m256i v_cat_epi32 = sao_calc_eo_cat_avx2(&vector_a_epi8, &vector_b_epi8, &vector_c_epi8);
v_cat_epi32 = _mm256_add_epi32(v_cat_epi32, offset_v_epi32); v_cat_epi32 = _mm256_add_epi32(v_cat_epi32, offset_v_epi32);