hashirama/uvg266
1
0
Fork 0
mirror of https://github.com/ultravideo/uvg266.git synced 2024-11-28 03:34:06 +00:00
Code Issues Actions Projects Releases Wiki Activity

Vectorize the coeff group loop to some extent

Browse source
This commit is contained in:
Pauli Oikkonen 2018-12-10 23:47:07 +02:00
parent a01362e638
commit efd097f5a5
1 changed files with 47 additions and 18 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

65
src/strategies/avx2/encode_coding_tree-avx2.c
View file

@ -202,7 +202,7 @@ void kvz_encode_coeff_nxn_avx2(encoder_state_t * const state,
int8_t be_valid = encoder->cfg.signhide_enable; int8_t be_valid = encoder->cfg.signhide_enable;
int32_t scan_pos_sig; int32_t scan_pos_sig;
uint32_t go_rice_param = 0; uint32_t go_rice_param = 0;
uint32_t blk_pos, pos_y, pos_x, sig, ctx_sig; uint32_t ctx_sig;
// CONSTANTS // CONSTANTS
const uint32_t num_blk_side = width >> TR_MIN_LOG2_SIZE; const uint32_t num_blk_side = width >> TR_MIN_LOG2_SIZE;
@ -340,38 +340,67 @@ void kvz_encode_coeff_nxn_avx2(encoder_state_t * const state,
int32_t pattern_sig_ctx = kvz_context_calc_pattern_sig_ctx(sig_coeffgroup_flag, int32_t pattern_sig_ctx = kvz_context_calc_pattern_sig_ctx(sig_coeffgroup_flag,
cg_pos_x, cg_pos_y, width); cg_pos_x, cg_pos_y, width);
// TODO: reorder coeff and vectorize? ALIGNED(64) int16_t abs_coeff_buf[16];
const __m256i ns = _mm256_setr_epi16(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15); ALIGNED(32) int16_t pos_ys_buf[16];
ALIGNED(32) int16_t pos_xs_buf[16];
const __m256i coeff_pos_zero = _mm256_castsi128_si256(_mm_cvtsi32_si128(0xffff));
const __m128i log2_block_size_128 = _mm_cvtsi32_si128(log2_block_size);
__m256i coeffs = _mm256_load_si256((__m256i *)(coeff_reord + sub_pos)); __m256i coeffs = _mm256_load_si256((__m256i *)(coeff_reord + sub_pos));
__m256i sigs_inv = _mm256_cmpeq_epi16(coeffs, zero); __m256i sigs_inv = _mm256_cmpeq_epi16(coeffs, zero);
__m256i is = _mm256_set1_epi16(i); __m256i is = _mm256_set1_epi16(i);
__m256i is_zero = _mm256_cmpeq_epi16(is, zero); __m256i is_zero = _mm256_cmpeq_epi16(is, zero);
// TODO! __m256i coeffs_subzero = _mm256_cmpgt_epi16(zero, coeffs);
__m256i masked_coeffs = _mm256_andnot_si256(sigs_inv, coeffs);
__m256i abs_coeffs = _mm256_abs_epi16(masked_coeffs);
// TODO: obtain 16-bit block positions, maybe? :P
__m256i blk_poses_hi = _mm256_loadu_si256((__m256i *)(scan + sub_pos + 8));
__m256i blk_poses_lo = _mm256_loadu_si256((__m256i *)(scan + sub_pos + 0));
__m256i blk_poses_tmp = _mm256_packs_epi32(blk_poses_lo, blk_poses_hi);
__m256i blk_poses = _mm256_permute4x64_epi64(blk_poses_tmp, _MM_SHUFFLE(3, 1, 2, 0));
__m256i pos_ys = _mm256_srl_epi16(blk_poses, log2_block_size_128);
__m256i pos_xs = _mm256_sub_epi16(blk_poses, _mm256_sll_epi16(pos_ys, log2_block_size_128));
_mm256_store_si256((__m256i *)pos_ys_buf, pos_ys);
_mm256_store_si256((__m256i *)pos_xs_buf, pos_xs);
__m256i encode_sig_coeff_flags_inv = _mm256_andnot_si256(is_zero, coeff_pos_zero);
get_first_last_nz_int16(masked_coeffs, &first_nz_pos_in_cg, &last_nz_pos_in_cg);
_mm256_store_si256((__m256i *)abs_coeff_buf, abs_coeffs);
uint32_t esc_flags = ~(_mm256_movemask_epi8(encode_sig_coeff_flags_inv));
uint32_t sigs = ~(_mm256_movemask_epi8(sigs_inv));
uint32_t coeff_sign_buf = _mm256_movemask_epi8(coeffs_subzero);
// TODO: get first and last nz
for (; scan_pos_sig >= sub_pos; scan_pos_sig--) { for (; scan_pos_sig >= sub_pos; scan_pos_sig--) {
blk_pos = scan[scan_pos_sig]; uint32_t id = scan_pos_sig - sub_pos;
pos_y = blk_pos >> log2_block_size; uint32_t shamt = (id << 1) + 1;
pos_x = blk_pos - (pos_y << log2_block_size);
sig = (coeff_reord[scan_pos_sig] != 0) ? 1 : 0;
if (scan_pos_sig > sub_pos || i == 0 || num_non_zero != 0) { uint32_t curr_sig = (sigs >> shamt) & 1;
ctx_sig = kvz_context_get_sig_ctx_inc(pattern_sig_ctx, scan_mode, pos_x, pos_y, uint32_t curr_esc_flag = (esc_flags >> shamt) & 1;
uint32_t curr_coeff_sign = (coeff_sign_buf >> shamt) & 1;
uint32_t curr_pos_x = pos_xs_buf[id];
uint32_t curr_pos_y = pos_ys_buf[id];
if (curr_esc_flag | num_non_zero) {
ctx_sig = kvz_context_get_sig_ctx_inc(pattern_sig_ctx, scan_mode, curr_pos_x, curr_pos_y,
log2_block_size, type); log2_block_size, type);
cabac->cur_ctx = &baseCtx[ctx_sig]; cabac->cur_ctx = &baseCtx[ctx_sig];
CABAC_BIN(cabac, sig, "sig_coeff_flag"); CABAC_BIN(cabac, curr_sig, "sig_coeff_flag");
} }
if (sig) { if (curr_sig) {
abs_coeff[num_non_zero] = abs(coeff_reord[scan_pos_sig]); abs_coeff[num_non_zero] = abs_coeff_buf[id];
coeff_signs = 2 * coeff_signs + (coeff[blk_pos] < 0); coeff_signs = 2 * coeff_signs + curr_coeff_sign;
num_non_zero++; num_non_zero++;
} }
} }
__m256i masked_coeffs = _mm256_andnot_si256(sigs_inv, coeffs);
get_first_last_nz_int16(masked_coeffs, &first_nz_pos_in_cg, &last_nz_pos_in_cg);
} else { } else {
scan_pos_sig = sub_pos - 1; scan_pos_sig = sub_pos - 1;
} }