/external/XNNPACK/src/f32-argmaxpool/ |
D | 9p8x-sse2-c4.c | 83 vidx = _mm_or_si128(_mm_andnot_si128(vm1, vidx), _mm_and_si128(vm1, _mm_set1_epi32(1))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 87 vidx = _mm_or_si128(_mm_andnot_si128(vm2, vidx), _mm_and_si128(vm2, _mm_set1_epi32(2))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 91 vidx = _mm_or_si128(_mm_andnot_si128(vm3, vidx), _mm_and_si128(vm3, _mm_set1_epi32(3))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 95 vidx = _mm_or_si128(_mm_andnot_si128(vm4, vidx), _mm_and_si128(vm4, _mm_set1_epi32(4))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 99 vidx = _mm_or_si128(_mm_andnot_si128(vm5, vidx), _mm_and_si128(vm5, _mm_set1_epi32(5))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 103 vidx = _mm_or_si128(_mm_andnot_si128(vm6, vidx), _mm_and_si128(vm6, _mm_set1_epi32(6))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 107 vidx = _mm_or_si128(_mm_andnot_si128(vm7, vidx), _mm_and_si128(vm7, _mm_set1_epi32(7))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 111 vidx = _mm_or_si128(_mm_andnot_si128(vm8, vidx), _mm_and_si128(vm8, _mm_set1_epi32(8))); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 168 vidx = _mm_or_si128(_mm_andnot_si128(vm0, vidx), _mm_and_si128(vm0, vidx0)); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() 173 vidx = _mm_or_si128(_mm_andnot_si128(vm1, vidx), _mm_and_si128(vm1, vidx1)); in xnn_f32_argmaxpool_ukernel_9p8x__sse2_c4() [all …]
|
D | 9x-sse2-c4.c | 105 vidx = _mm_or_si128(_mm_andnot_si128(vm1, vidx), _mm_and_si128(vm1, _mm_set1_epi32(1))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 109 vidx = _mm_or_si128(_mm_andnot_si128(vm2, vidx), _mm_and_si128(vm2, _mm_set1_epi32(2))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 113 vidx = _mm_or_si128(_mm_andnot_si128(vm3, vidx), _mm_and_si128(vm3, _mm_set1_epi32(3))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 117 vidx = _mm_or_si128(_mm_andnot_si128(vm4, vidx), _mm_and_si128(vm4, _mm_set1_epi32(4))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 121 vidx = _mm_or_si128(_mm_andnot_si128(vm5, vidx), _mm_and_si128(vm5, _mm_set1_epi32(5))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 125 vidx = _mm_or_si128(_mm_andnot_si128(vm6, vidx), _mm_and_si128(vm6, _mm_set1_epi32(6))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 129 vidx = _mm_or_si128(_mm_andnot_si128(vm7, vidx), _mm_and_si128(vm7, _mm_set1_epi32(7))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 133 vidx = _mm_or_si128(_mm_andnot_si128(vm8, vidx), _mm_and_si128(vm8, _mm_set1_epi32(8))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 158 vidx = _mm_or_si128(_mm_andnot_si128(vm1, vidx), _mm_and_si128(vm1, _mm_set1_epi32(1))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() 162 vidx = _mm_or_si128(_mm_andnot_si128(vm2, vidx), _mm_and_si128(vm2, _mm_set1_epi32(2))); in xnn_f32_argmaxpool_ukernel_9x__sse2_c4() [all …]
|
D | 4x-sse2-c4.c | 67 vidx = _mm_or_si128(_mm_andnot_si128(vm1, vidx), _mm_and_si128(vm1, _mm_set1_epi32(1))); in xnn_f32_argmaxpool_ukernel_4x__sse2_c4() 71 vidx = _mm_or_si128(_mm_andnot_si128(vm2, vidx), _mm_and_si128(vm2, _mm_set1_epi32(2))); in xnn_f32_argmaxpool_ukernel_4x__sse2_c4() 75 vidx = _mm_or_si128(_mm_andnot_si128(vm3, vidx), _mm_and_si128(vm3, _mm_set1_epi32(3))); in xnn_f32_argmaxpool_ukernel_4x__sse2_c4() 95 vidx = _mm_or_si128(_mm_andnot_si128(vm1, vidx), _mm_and_si128(vm1, _mm_set1_epi32(1))); in xnn_f32_argmaxpool_ukernel_4x__sse2_c4() 99 vidx = _mm_or_si128(_mm_andnot_si128(vm2, vidx), _mm_and_si128(vm2, _mm_set1_epi32(2))); in xnn_f32_argmaxpool_ukernel_4x__sse2_c4() 103 vidx = _mm_or_si128(_mm_andnot_si128(vm3, vidx), _mm_and_si128(vm3, _mm_set1_epi32(3))); in xnn_f32_argmaxpool_ukernel_4x__sse2_c4()
|
/external/libavc/common/x86/ |
D | ih264_iquant_itrans_recon_dc_ssse3.c | 153 temp4 = _mm_and_si128(temp4, sign_reg); in ih264_iquant_itrans_recon_4x4_dc_ssse3() 155 temp5 = _mm_and_si128(temp5, sign_reg); in ih264_iquant_itrans_recon_4x4_dc_ssse3() 306 temp1 = _mm_and_si128(temp1, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() 308 temp2 = _mm_and_si128(temp2, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() 310 temp3 = _mm_and_si128(temp3, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() 312 temp4 = _mm_and_si128(temp4, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() 314 temp5 = _mm_and_si128(temp5, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() 316 temp6 = _mm_and_si128(temp6, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() 318 temp7 = _mm_and_si128(temp7, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() 320 temp8 = _mm_and_si128(temp8, sign_reg); in ih264_iquant_itrans_recon_8x8_dc_ssse3() [all …]
|
D | ih264_deblk_luma_ssse3.c | 196 flag1_16x8 = _mm_and_si128(flag1_16x8, flag2_16x8); in ih264_deblk_luma_vert_bs4_ssse3() 212 flag1_16x8 = _mm_and_si128(flag1_16x8, flag2_16x8); in ih264_deblk_luma_vert_bs4_ssse3() 227 flag2_16x8 = _mm_and_si128(flag1_16x8, flag2_16x8); in ih264_deblk_luma_vert_bs4_ssse3() 240 flag3_16x8 = _mm_and_si128(flag3_16x8, flag2_16x8); in ih264_deblk_luma_vert_bs4_ssse3() 253 flag4_16x8 = _mm_and_si128(flag4_16x8, flag2_16x8); in ih264_deblk_luma_vert_bs4_ssse3() 384 p0_16x8 = _mm_and_si128(p0_16x8, in ih264_deblk_luma_vert_bs4_ssse3() 386 p0_16x8_1 = _mm_and_si128(p0_16x8_1, flag1_16x8); in ih264_deblk_luma_vert_bs4_ssse3() 388 q0_16x8 = _mm_and_si128(q0_16x8, in ih264_deblk_luma_vert_bs4_ssse3() 390 q0_16x8_1 = _mm_and_si128(q0_16x8_1, flag1_16x8); in ih264_deblk_luma_vert_bs4_ssse3() 394 p0_16x8 = _mm_and_si128(p0_16x8, in ih264_deblk_luma_vert_bs4_ssse3() [all …]
|
D | ih264_deblk_chroma_ssse3.c | 153 flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff)); in ih264_deblk_chroma_vert_bs4_ssse3() 157 flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff)); in ih264_deblk_chroma_vert_bs4_ssse3() 184 flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff)); in ih264_deblk_chroma_vert_bs4_ssse3() 188 flag2 = _mm_and_si128(flag2, _mm_cmpgt_epi16(beta_cbcr_16x8, diff)); in ih264_deblk_chroma_vert_bs4_ssse3() 207 p0_uv_8x16_1 = _mm_and_si128(p0_uv_16x8, in ih264_deblk_chroma_vert_bs4_ssse3() 209 p0_uv_8x16_2 = _mm_and_si128(p0_uv_8x16_2, flag1); in ih264_deblk_chroma_vert_bs4_ssse3() 212 q0_uv_8x16_1 = _mm_and_si128(q0_uv_16x8, in ih264_deblk_chroma_vert_bs4_ssse3() 214 q0_uv_8x16_2 = _mm_and_si128(q0_uv_8x16_2, flag1); in ih264_deblk_chroma_vert_bs4_ssse3() 322 flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff)); in ih264_deblk_chroma_horz_bs4_ssse3() 326 flag1 = _mm_and_si128(flag1, _mm_cmpgt_epi16(beta_cbcr_16x8, diff)); in ih264_deblk_chroma_horz_bs4_ssse3() [all …]
|
D | ih264_ihadamard_scaling_ssse3.c | 105 mult_val = _mm_and_si128(mult_val, mask); in ih264_ihadamard_scaling_4x4_ssse3() 175 src_r0 = _mm_and_si128(src_r0, mask); in ih264_ihadamard_scaling_4x4_ssse3() 176 src_r1 = _mm_and_si128(src_r1, mask); in ih264_ihadamard_scaling_4x4_ssse3() 177 src_r2 = _mm_and_si128(src_r2, mask); in ih264_ihadamard_scaling_4x4_ssse3() 178 src_r3 = _mm_and_si128(src_r3, mask); in ih264_ihadamard_scaling_4x4_ssse3()
|
D | ih264_iquant_itrans_recon_sse42.c | 283 temp0 = _mm_and_si128(temp0, sign_reg); in ih264_iquant_itrans_recon_4x4_sse42() 285 temp1 = _mm_and_si128(temp1, sign_reg); in ih264_iquant_itrans_recon_4x4_sse42() 485 pred_r0 = _mm_and_si128(pred_r0, chroma_mask); in ih264_iquant_itrans_recon_chroma_4x4_sse42() 486 pred_r1 = _mm_and_si128(pred_r1, chroma_mask); in ih264_iquant_itrans_recon_chroma_4x4_sse42() 487 pred_r2 = _mm_and_si128(pred_r2, chroma_mask); in ih264_iquant_itrans_recon_chroma_4x4_sse42() 488 pred_r3 = _mm_and_si128(pred_r3, chroma_mask); in ih264_iquant_itrans_recon_chroma_4x4_sse42() 538 temp0 = _mm_and_si128(temp0, sign_reg); in ih264_iquant_itrans_recon_chroma_4x4_sse42() 540 temp1 = _mm_and_si128(temp1, sign_reg); in ih264_iquant_itrans_recon_chroma_4x4_sse42() 558 out_r0 = _mm_and_si128(out_r0, chroma_mask); in ih264_iquant_itrans_recon_chroma_4x4_sse42() 559 out_r1 = _mm_and_si128(out_r1, chroma_mask); in ih264_iquant_itrans_recon_chroma_4x4_sse42() [all …]
|
D | ih264_resi_trans_quant_sse42.c | 301 cmp0 = _mm_and_si128(temp_1, cmp0); in ih264_resi_trans_quant_4x4_sse42() 314 cmp1 = _mm_and_si128(temp_1, cmp1); in ih264_resi_trans_quant_4x4_sse42() 409 src_r0 = _mm_and_si128(src_r0, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() 410 src_r1 = _mm_and_si128(src_r1, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() 411 src_r2 = _mm_and_si128(src_r2, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() 412 src_r3 = _mm_and_si128(src_r3, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() 423 pred_r0 = _mm_and_si128(pred_r0, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() 424 pred_r1 = _mm_and_si128(pred_r1, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() 425 pred_r2 = _mm_and_si128(pred_r2, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() 426 pred_r3 = _mm_and_si128(pred_r3, chroma_mask); in ih264_resi_trans_quant_chroma_4x4_sse42() [all …]
|
/external/libvpx/libvpx/vpx_dsp/x86/ |
D | loopfilter_avx2.c | 66 abs_p1q1 = _mm_srli_epi16(_mm_and_si128(abs_p1q1, fe), 1); in vpx_lpf_horizontal_16_avx2() 99 filt = _mm_and_si128(_mm_subs_epi8(qs1ps1, qs1), hev); in vpx_lpf_horizontal_16_avx2() 105 filt = _mm_and_si128(filt, mask); in vpx_lpf_horizontal_16_avx2() 137 flat = _mm_and_si128(flat, mask); in vpx_lpf_horizontal_16_avx2() 163 flat2 = _mm_and_si128(flat2, flat); // flat2 & flat & mask in vpx_lpf_horizontal_16_avx2() 310 flat_q2p2 = _mm_and_si128(flat, flat_q2p2); in vpx_lpf_horizontal_16_avx2() 314 flat_q1p1 = _mm_and_si128(flat, flat_q1p1); in vpx_lpf_horizontal_16_avx2() 318 flat_q0p0 = _mm_and_si128(flat, flat_q0p0); in vpx_lpf_horizontal_16_avx2() 322 flat2_q6p6 = _mm_and_si128(flat2, flat2_q6p6); in vpx_lpf_horizontal_16_avx2() 328 flat2_q5p5 = _mm_and_si128(flat2, flat2_q5p5); in vpx_lpf_horizontal_16_avx2() [all …]
|
D | highbd_loopfilter_sse2.c | 41 ubounded = _mm_and_si128(ubounded, max); in signed_char_clamp_bd_sse2() 42 lbounded = _mm_and_si128(lbounded, min); in signed_char_clamp_bd_sse2() 121 mask = _mm_and_si128(mask, _mm_adds_epu16(limit_v, one)); in vpx_highbd_lpf_horizontal_16_sse2() 156 filt = _mm_and_si128(signed_char_clamp_bd_sse2(_mm_subs_epi16(ps1, qs1), bd), in vpx_highbd_lpf_horizontal_16_sse2() 162 filt = _mm_and_si128(filt, mask); in vpx_highbd_lpf_horizontal_16_sse2() 208 flat = _mm_and_si128(flat, mask); in vpx_highbd_lpf_horizontal_16_sse2() 246 flat2 = _mm_and_si128(flat2, flat); // flat2 & flat & mask in vpx_highbd_lpf_horizontal_16_sse2() 355 flat_p2 = _mm_and_si128(flat, flat_p2); in vpx_highbd_lpf_horizontal_16_sse2() 359 flat_q2 = _mm_and_si128(flat, flat_q2); in vpx_highbd_lpf_horizontal_16_sse2() 364 flat_p1 = _mm_and_si128(flat, flat_p1); in vpx_highbd_lpf_horizontal_16_sse2() [all …]
|
D | loopfilter_sse2.c | 57 mask = _mm_and_si128(mask, _mm_srli_si128(mask, 8)); \ 72 filter = _mm_and_si128(_mm_srli_si128(work, 8), hev); \ 77 filter = _mm_and_si128(filter, mask); /* & mask */ \ 275 abs_p1q1 = _mm_srli_epi16(_mm_and_si128(abs_p1q1, fe), 1); in vpx_lpf_horizontal_16_sse2() 306 filt = _mm_and_si128(_mm_subs_epi8(qs1ps1, qs1), hev); in vpx_lpf_horizontal_16_sse2() 312 filt = _mm_and_si128(filt, mask); in vpx_lpf_horizontal_16_sse2() 342 flat = _mm_and_si128(flat, mask); in vpx_lpf_horizontal_16_sse2() 361 flat2 = _mm_and_si128(flat2, flat); // flat2 & flat & mask in vpx_lpf_horizontal_16_sse2() 508 flat_q2p2 = _mm_and_si128(flat, flat_q2p2); in vpx_lpf_horizontal_16_sse2() 512 flat_q1p1 = _mm_and_si128(flat, flat_q1p1); in vpx_lpf_horizontal_16_sse2() [all …]
|
D | quantize_avx.c | 88 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_avx() 89 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_avx() 130 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_avx() 131 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_avx() 227 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_32x32_avx() 228 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_32x32_avx() 269 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_32x32_avx() 270 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_32x32_avx()
|
D | quantize_ssse3.c | 65 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_ssse3() 66 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_ssse3() 94 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_ssse3() 95 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_ssse3() 199 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_32x32_ssse3() 200 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_32x32_ssse3() 245 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in vpx_quantize_b_32x32_ssse3() 246 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in vpx_quantize_b_32x32_ssse3()
|
/external/OpenCL-CTS/test_common/harness/ |
D | mt19937.cpp | 140 … __m128i vy = _mm_or_si128( _mm_and_si128( _mm_load_si128( (__m128i*)(mt + kk) ), upper_mask.v ), in genrand_int32() 141 …_mm_and_si128( _mm_loadu_si128( (__m128i*)(mt + kk + 1) ), lower_mask.v )); // ((mt[kk]&UP… in genrand_int32() 143 …__m128i mask = _mm_cmpeq_epi32( _mm_and_si128( vy, one.v), one.v ); … in genrand_int32() 144 …__m128i vmag01 = _mm_and_si128( mask, matrix_a.v ); … in genrand_int32() 165 … __m128i vy = _mm_or_si128( _mm_and_si128( _mm_load_si128( (__m128i*)(mt + kk) ), upper_mask.v ), in genrand_int32() 166 …_mm_and_si128( _mm_loadu_si128( (__m128i*)(mt + kk + 1) ), lower_mask.v )); // ((mt[kk]&UP… in genrand_int32() 168 …__m128i mask = _mm_cmpeq_epi32( _mm_and_si128( vy, one.v), one.v ); … in genrand_int32() 169 …__m128i vmag01 = _mm_and_si128( mask, matrix_a.v ); … in genrand_int32() 189 …vy = _mm_xor_si128( vy, _mm_and_si128( _mm_slli_epi32( vy, 7 ), c0.v) ); // y ^= (y << 7) &… in genrand_int32() 190 …vy = _mm_xor_si128( vy, _mm_and_si128( _mm_slli_epi32( vy, 15 ), c1.v) ); // y ^= (y << 15) … in genrand_int32()
|
/external/XNNPACK/src/x8-zip/ |
D | x3-sse2.c | 39 const __m128i vxeye = _mm_or_si128(_mm_and_si128(vx, vmask0x00FF00FF), _mm_slli_epi16(vy, 8)); in xnn_x8_zip_x3_ukernel__sse2() 43 …const __m128i vzexo = _mm_or_si128(_mm_and_si128(vz, vmask0x00FF00FF), _mm_andnot_si128(vmask0x00F… in xnn_x8_zip_x3_ukernel__sse2() 46 …const __m128i vxeyezexo = _mm_or_si128(_mm_and_si128(vxeye, vmask0x0000FFFF), _mm_slli_epi32(vzexo… in xnn_x8_zip_x3_ukernel__sse2() 48 …const __m128i vyozoxeye = _mm_or_si128(_mm_and_si128(vyozo, vmask0x0000FFFF), _mm_andnot_si128(vma… in xnn_x8_zip_x3_ukernel__sse2() 88 const __m128i vxeye = _mm_or_si128(_mm_and_si128(vx, vmask0x00FF00FF), _mm_slli_epi16(vy, 8)); in xnn_x8_zip_x3_ukernel__sse2() 92 …const __m128i vzexo = _mm_or_si128(_mm_and_si128(vz, vmask0x00FF00FF), _mm_andnot_si128(vmask0x00F… in xnn_x8_zip_x3_ukernel__sse2() 95 …const __m128i vxeyezexo = _mm_or_si128(_mm_and_si128(vxeye, vmask0x0000FFFF), _mm_slli_epi32(vzexo… in xnn_x8_zip_x3_ukernel__sse2() 97 …const __m128i vyozoxeye = _mm_or_si128(_mm_and_si128(vyozo, vmask0x0000FFFF), _mm_andnot_si128(vma… in xnn_x8_zip_x3_ukernel__sse2()
|
/external/skqp/src/opts/ |
D | SkBlitRow_opts.h | 23 __m128i rb = _mm_and_si128(mask, c); in SkPMSrcOver_SSE2() 111 auto ANDed = _mm_and_si128(s3, _mm_and_si128(s2, _mm_and_si128(s1, s0))); in blit_row_s32a_opaque() 146 if (0xffff == _mm_movemask_epi8(_mm_cmpeq_epi8(_mm_and_si128(ORed, alphaMask), in blit_row_s32a_opaque() 160 auto ANDed = _mm_and_si128(s3, _mm_and_si128(s2, _mm_and_si128(s1, s0))); in blit_row_s32a_opaque() 161 if (0xffff == _mm_movemask_epi8(_mm_cmpeq_epi8(_mm_and_si128(ANDed, alphaMask), in blit_row_s32a_opaque()
|
/external/boringssl/src/crypto/poly1305/ |
D | poly1305_vec.c | 219 st->H[0] = _mm_and_si128(MMASK, T5); in poly1305_first_block() 220 st->H[1] = _mm_and_si128(MMASK, _mm_srli_epi64(T5, 26)); in poly1305_first_block() 222 st->H[2] = _mm_and_si128(MMASK, T5); in poly1305_first_block() 223 st->H[3] = _mm_and_si128(MMASK, _mm_srli_epi64(T5, 26)); in poly1305_first_block() 299 M0 = _mm_and_si128(MMASK, T5); in poly1305_blocks() 300 M1 = _mm_and_si128(MMASK, _mm_srli_epi64(T5, 26)); in poly1305_blocks() 302 M2 = _mm_and_si128(MMASK, T5); in poly1305_blocks() 303 M3 = _mm_and_si128(MMASK, _mm_srli_epi64(T5, 26)); in poly1305_blocks() 363 M0 = _mm_and_si128(MMASK, T5); in poly1305_blocks() 364 M1 = _mm_and_si128(MMASK, _mm_srli_epi64(T5, 26)); in poly1305_blocks() [all …]
|
/external/webp/src/dsp/ |
D | lossless_enc_sse41.c | 72 const __m128i gb0 = _mm_and_si128(in0, mask_gb); in CollectColorBlueTransforms_SSE41() 73 const __m128i gb1 = _mm_and_si128(in1, mask_gb); in CollectColorBlueTransforms_SSE41() 75 const __m128i g = _mm_and_si128(gb, mask_g); // g 0 in CollectColorBlueTransforms_SSE41() 80 const __m128i E = _mm_and_si128(D, mask_b); // 0 b'' in CollectColorBlueTransforms_SSE41() 110 const __m128i g0 = _mm_and_si128(in0, mask_g); // 0 0 | g 0 in CollectColorRedTransforms_SSE41() 111 const __m128i g1 = _mm_and_si128(in1, mask_g); in CollectColorRedTransforms_SSE41() 118 const __m128i D = _mm_and_si128(C, mask); // 0 r' in CollectColorRedTransforms_SSE41()
|
D | alpha_processing_sse2.c | 50 const __m128i b1_lo = _mm_and_si128(b0_lo, rgb_mask); in DispatchAlpha_SSE2() 51 const __m128i b1_hi = _mm_and_si128(b0_hi, rgb_mask); in DispatchAlpha_SSE2() 59 all_alphas = _mm_and_si128(all_alphas, a0); in DispatchAlpha_SSE2() 123 const __m128i b0 = _mm_and_si128(a0, a_mask); in ExtractAlpha_SSE2() 124 const __m128i b1 = _mm_and_si128(a1, a_mask); in ExtractAlpha_SSE2() 130 all_alphas = _mm_and_si128(all_alphas, d0); in ExtractAlpha_SSE2() 242 const __m128i b0 = _mm_and_si128(a0, alpha_mask); in HasAlpha32b_SSE2() 243 const __m128i b1 = _mm_and_si128(a1, alpha_mask); in HasAlpha32b_SSE2() 244 const __m128i b2 = _mm_and_si128(a2, alpha_mask); in HasAlpha32b_SSE2() 245 const __m128i b3 = _mm_and_si128(a3, alpha_mask); in HasAlpha32b_SSE2() [all …]
|
D | common_sse2.h | 113 (OUT##0) = _mm_packus_epi16(_mm_and_si128((IN##0), v_mask), \ 114 _mm_and_si128((IN##1), v_mask)); \ 115 (OUT##1) = _mm_packus_epi16(_mm_and_si128((IN##2), v_mask), \ 116 _mm_and_si128((IN##3), v_mask)); \ 117 (OUT##2) = _mm_packus_epi16(_mm_and_si128((IN##4), v_mask), \ 118 _mm_and_si128((IN##5), v_mask)); \
|
/external/libvpx/libvpx/vp8/encoder/x86/ |
D | denoising_sse2.c | 83 __m128i adj2 = _mm_and_si128(mask2, l32); in vp8_denoiser_filter_sse2() 84 const __m128i adj1 = _mm_and_si128(mask1, l21); in vp8_denoiser_filter_sse2() 85 const __m128i adj0 = _mm_and_si128(mask0, clamped_absdiff); in vp8_denoiser_filter_sse2() 96 nadj = _mm_and_si128(diff_sign, adj); in vp8_denoiser_filter_sse2() 154 nadj = _mm_and_si128(diff_sign, adj); in vp8_denoiser_filter_sse2() 258 __m128i adj2 = _mm_and_si128(mask2, l32); in vp8_denoiser_filter_uv_sse2() 259 const __m128i adj1 = _mm_and_si128(mask1, l21); in vp8_denoiser_filter_uv_sse2() 260 const __m128i adj0 = _mm_and_si128(mask0, clamped_absdiff); in vp8_denoiser_filter_uv_sse2() 272 nadj = _mm_and_si128(diff_sign, adj); in vp8_denoiser_filter_uv_sse2() 341 nadj = _mm_and_si128(diff_sign, adj); in vp8_denoiser_filter_uv_sse2()
|
/external/skia/src/opts/ |
D | SkBlitRow_opts.h | 71 __m128i rb = _mm_and_si128(mask, c); in SkPMSrcOver_SSE2() 239 auto ANDed = _mm_and_si128(s3, _mm_and_si128(s2, _mm_and_si128(s1, s0))); in blit_row_s32a_opaque() 274 if (0xffff == _mm_movemask_epi8(_mm_cmpeq_epi8(_mm_and_si128(ORed, alphaMask), in blit_row_s32a_opaque() 288 auto ANDed = _mm_and_si128(s3, _mm_and_si128(s2, _mm_and_si128(s1, s0))); in blit_row_s32a_opaque() 289 if (0xffff == _mm_movemask_epi8(_mm_cmpeq_epi8(_mm_and_si128(ANDed, alphaMask), in blit_row_s32a_opaque()
|
/external/XNNPACK/src/requantization/ |
D | gemmlowp-sse.h | 35 const __m128i remainder = _mm_and_si128(x, mask); in gemmlowp_sse_rdivbypo2_s32() 67 mul_us_neg = _mm_and_si128(sign, mul_us_neg); in gemmlowp_sse_mul_s32() 77 _mm_and_si128(_mm_cmpeq_epi32(a, b), _mm_cmpeq_epi32(a, min)); in gemmlowp_sse_vqrdmulh_s32() 115 _mm_and_si128(saturation_mask, min), in gemmlowp_sse_vqrdmulh_s32()
|
/external/libaom/libaom/aom_dsp/x86/ |
D | quantize_sse2.c | 70 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in aom_quantize_b_sse2() 71 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in aom_quantize_b_sse2() 105 qcoeff0 = _mm_and_si128(qcoeff0, cmp_mask0); in aom_quantize_b_sse2() 106 qcoeff1 = _mm_and_si128(qcoeff1, cmp_mask1); in aom_quantize_b_sse2()
|