From 4ae234ef247ff9e2d350ce3a3173d43420ed661a Mon Sep 17 00:00:00 2001 From: siivonek Date: Thu, 7 Sep 2023 21:02:26 +0300 Subject: [PATCH] Implement 16xN planar prediction. --- src/strategies/avx2/intra-avx2.c | 50 ++++++++++++++++++++++++++++++-- 1 file changed, 48 insertions(+), 2 deletions(-) diff --git a/src/strategies/avx2/intra-avx2.c b/src/strategies/avx2/intra-avx2.c index 5ae1440b..a71d9de3 100644 --- a/src/strategies/avx2/intra-avx2.c +++ b/src/strategies/avx2/intra-avx2.c @@ -665,7 +665,24 @@ static void intra_pred_planar_hor_w8(const uvg_pixel* ref, const int line, const dst[d] = _mm256_slli_epi16(v_tmp, shift); } } -static void intra_pred_planar_hor_w16(const uvg_pixel* ref, const int line, const int shift, __m256i* dst) {} +static void intra_pred_planar_hor_w16(const uvg_pixel* ref, const int line, const int shift, __m256i* dst) +{ + const __m256i v_last_ref = _mm256_set1_epi16(ref[16 + 1]); + + __m256i v_ref_coeff = _mm256_setr_epi16(15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0); + __m256i v_last_ref_coeff = _mm256_setr_epi16(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16); + + __m256i v_last_ref_mul = _mm256_mullo_epi16(v_last_ref, v_last_ref_coeff); + + for (int i = 0, d = 0; i < line; ++i, ++d) { + __m256i v_ref = _mm256_set1_epi16(ref[i + 1]); + + __m256i v_tmp = _mm256_mullo_epi16(v_ref, v_ref_coeff); // TODO: the result is needed immediately after this. This leads to NOPs, consider doing multiple lines at a time + + v_tmp = _mm256_add_epi16(v_last_ref_mul, v_tmp); + dst[d] = _mm256_slli_epi16(v_tmp, shift); + } +} static void intra_pred_planar_hor_w32(const uvg_pixel* ref, const int line, const int shift, __m256i* dst) {} static void intra_pred_planar_ver_w1(const uvg_pixel* ref, const int line, const int shift, __m256i* dst) {} @@ -737,7 +754,36 @@ static void intra_pred_planar_ver_w8(const uvg_pixel* ref, const int line, const dst[d + 1] = _mm256_permute4x64_epi64(v_tmp1, _MM_SHUFFLE(3, 1, 2, 0)); } } -static void intra_pred_planar_ver_w16(const uvg_pixel* ref, const int line, const int shift, __m256i* dst) {} +static void intra_pred_planar_ver_w16(const uvg_pixel* ref, const int line, const int shift, __m256i* dst) +{ + const __m256i v_last_ref = _mm256_set1_epi8(ref[line + 1]); + + // Got 16 8-bit samples, or 128 bits of data. Duplicate to fill a whole 256-bit vector. + const __m128i v_ref_raw = _mm_load_si128((const __m128i*) &ref[1]); + __m256i v_ref = _mm256_castsi128_si256(v_ref_raw); + v_ref = _mm256_inserti128_si256(v_ref, v_ref_raw, 1); + + // Handle 2 lines at a time + for (int y = 0; y < line; y += 2) { + const int a1 = line - 1 - (y + 0); + const int b1 = (y + 0) + 1; + const int a2 = line - 1 - (y + 1); + const int b2 = (y + 1) + 1; + __m256i v_ys = _mm256_setr_epi8(a1, b1, a1, b1, a1, b1, a1, b1, + a1, b1, a1, b1, a1, b1, a1, b1, + a2, b2, a2, b2, a2, b2, a2, b2, + a2, b2, a2, b2, a2, b2, a2, b2); // TODO: these could be loaded from a table + __m256i v_lo = _mm256_unpacklo_epi8(v_ref, v_last_ref); + __m256i v_hi = _mm256_unpackhi_epi8(v_ref, v_last_ref); + + __m256i v_madd_lo = _mm256_maddubs_epi16(v_lo, v_ys); + __m256i v_madd_hi = _mm256_maddubs_epi16(v_hi, v_ys); + v_madd_lo = _mm256_slli_epi16(v_madd_lo, shift); + v_madd_hi = _mm256_slli_epi16(v_madd_hi, shift); + dst[y + 0] = _mm256_permute2x128_si256(v_madd_lo, v_madd_hi, 0x20); + dst[y + 1] = _mm256_permute2x128_si256(v_madd_lo, v_madd_hi, 0x31); + } +} static void intra_pred_planar_ver_w32(const uvg_pixel* ref, const int line, const int shift, __m256i* dst) {}