| /external/libaom/libaom/aom_dsp/x86/ |
| D | blend_a64_vmask_sse4.c | 39 const __m128i v_m0_w = _mm_set1_epi16(*mask); in blend_a64_vmask_w4_sse4_1() local
40 const __m128i v_m1_w = _mm_sub_epi16(v_maxval_w, v_m0_w); in blend_a64_vmask_w4_sse4_1()
42 const __m128i v_res_w = blend_4(src0, src1, &v_m0_w, &v_m1_w); in blend_a64_vmask_w4_sse4_1()
64 const __m128i v_m0_w = _mm_set1_epi16(*mask); in blend_a64_vmask_w8_sse4_1() local
65 const __m128i v_m1_w = _mm_sub_epi16(v_maxval_w, v_m0_w); in blend_a64_vmask_w8_sse4_1()
67 const __m128i v_res_w = blend_8(src0, src1, &v_m0_w, &v_m1_w); in blend_a64_vmask_w8_sse4_1()
90 const __m128i v_m0_w = _mm_set1_epi16(*mask); in blend_a64_vmask_w16n_sse4_1() local
91 const __m128i v_m1_w = _mm_sub_epi16(v_maxval_w, v_m0_w); in blend_a64_vmask_w16n_sse4_1()
93 const __m128i v_resl_w = blend_8(src0 + c, src1 + c, &v_m0_w, &v_m1_w); in blend_a64_vmask_w16n_sse4_1()
95 blend_8(src0 + c + 8, src1 + c + 8, &v_m0_w, &v_m1_w); in blend_a64_vmask_w16n_sse4_1()
[all …]
|
| D | blend_sse4.h | 27 const __m128i *v_m0_w, const __m128i *v_m1_w) { in blend_4() argument
33 const __m128i v_p0_w = _mm_mullo_epi16(v_s0_w, *v_m0_w); in blend_4()
42 const __m128i *v_m0_w, const __m128i *v_m1_w) { in blend_8() argument
48 const __m128i v_p0_w = _mm_mullo_epi16(v_s0_w, *v_m0_w); in blend_8()
104 const __m128i v_m0_w, const __m128i v_m1_w);
107 const __m128i v_m0_w, const __m128i v_m1_w) { in blend_4_b10() argument
111 const __m128i v_p0_w = _mm_mullo_epi16(v_s0_w, v_m0_w); in blend_4_b10()
122 const __m128i v_m0_w, const __m128i v_m1_w) { in blend_8_b10() argument
126 const __m128i v_p0_w = _mm_mullo_epi16(v_s0_w, v_m0_w); in blend_8_b10()
137 const __m128i v_m0_w, const __m128i v_m1_w) { in blend_4_b12() argument
[all …]
|
| D | blend_a64_mask_sse4.c | 290 const __m128i v_m0_w = xx_roundn_epu16(v_rs_w, 2); in blend_a64_mask_sx_sy_w4_sse4_1() local
291 const __m128i v_m0_b = _mm_packus_epi16(v_m0_w, v_m0_w); in blend_a64_mask_sx_sy_w4_sse4_1()
323 const __m128i v_m0_w = xx_roundn_epu16(v_rs_w, 2); in blend_a64_mask_sx_sy_w8_sse4_1() local
324 const __m128i v_m0_b = _mm_packus_epi16(v_m0_w, v_m0_w); in blend_a64_mask_sx_sy_w8_sse4_1()
438 const __m128i v_m0_w = _mm_cvtepu8_epi16(v_m0_b); in blend_a64_mask_bn_w4_sse4_1() local
439 const __m128i v_m1_w = _mm_sub_epi16(v_maxval_w, v_m0_w); in blend_a64_mask_bn_w4_sse4_1()
441 const __m128i v_res_w = blend(src0, src1, v_m0_w, v_m1_w); in blend_a64_mask_bn_w4_sse4_1()
481 const __m128i v_m0_w = _mm_cvtepu8_epi16(v_m0_b); in blend_a64_mask_bn_w8n_sse4_1() local
482 const __m128i v_m1_w = _mm_sub_epi16(v_maxval_w, v_m0_w); in blend_a64_mask_bn_w8n_sse4_1()
484 const __m128i v_res_w = blend(src0 + c, src1 + c, v_m0_w, v_m1_w); in blend_a64_mask_bn_w8n_sse4_1()
[all …]
|
| D | blend_a64_mask_avx2.c | 458 const __m256i v_m0_w = yy_roundn_epu16(v_rsl_w, 2); in blend_a64_mask_sx_sy_w16_avx2() local
459 const __m256i v_m0_b = _mm256_packus_epi16(v_m0_w, v_m0_w); in blend_a64_mask_sx_sy_w16_avx2()
532 const __m128i v_m0_w = xx_roundn_epu16(v_rs_w, 2); in blend_a64_mask_sx_sy_avx2() local
533 const __m128i v_m0_b = _mm_packus_epi16(v_m0_w, v_m0_w); in blend_a64_mask_sx_sy_avx2()
555 const __m128i v_m0_w = xx_roundn_epu16(v_rs_w, 2); in blend_a64_mask_sx_sy_avx2() local
556 const __m128i v_m0_b = _mm_packus_epi16(v_m0_w, v_m0_w); in blend_a64_mask_sx_sy_avx2()
591 const __m256i v_m0_w = _mm256_and_si256(v_al_b, v_zmask_b); in blend_a64_mask_sx_w16_avx2() local
592 const __m256i v_m0_b = _mm256_packus_epi16(v_m0_w, _mm256_setzero_si256()); in blend_a64_mask_sx_w16_avx2()
|
| /external/libaom/libaom/av1/encoder/x86/ |
| D | wedge_utils_avx2.c | 51 const __m256i v_m0_w = _mm256_cvtepu8_epi16(v_m01_b); in av1_wedge_sse_from_residuals_avx2() local
53 const __m256i v_m0l_w = _mm256_unpacklo_epi16(v_m0_w, v_mask_max_w); in av1_wedge_sse_from_residuals_avx2()
54 const __m256i v_m0h_w = _mm256_unpackhi_epi16(v_m0_w, v_mask_max_w); in av1_wedge_sse_from_residuals_avx2()
107 const __m256i v_m0_w = in av1_wedge_sign_from_residuals_avx2() local
116 const __m256i v_p0_d = _mm256_madd_epi16(v_d0_w, v_m0_w); in av1_wedge_sign_from_residuals_avx2()
|
| D | wedge_utils_sse2.c | 55 const __m128i v_m0_w = _mm_unpacklo_epi8(v_m01_b, _mm_setzero_si128()); in av1_wedge_sse_from_residuals_sse2() local
58 const __m128i v_m0l_w = _mm_unpacklo_epi16(v_m0_w, v_mask_max_w); in av1_wedge_sse_from_residuals_sse2()
59 const __m128i v_m0h_w = _mm_unpackhi_epi16(v_m0_w, v_mask_max_w); in av1_wedge_sse_from_residuals_sse2()
130 const __m128i v_m0_w = _mm_unpacklo_epi8(v_m01_b, _mm_setzero_si128()); in av1_wedge_sign_from_residuals_sse2() local
139 const __m128i v_p0_d = _mm_madd_epi16(v_d0_w, v_m0_w); in av1_wedge_sign_from_residuals_sse2()
|