| /external/XNNPACK/src/f32-gavgpool-cw/ |
| D | neon-x4.c | 82 const float32x4_t vsum = vcombine_f32(vpadd_f32(vget_low_f32(vsum01), vget_high_f32(vsum01)), in xnn_f32_gavgpool_cw_ukernel__neon_x4()
83 vpadd_f32(vget_low_f32(vsum23), vget_high_f32(vsum23))); in xnn_f32_gavgpool_cw_ukernel__neon_x4()
115 vsum = vpadd_f32(vsum, vsum); in xnn_f32_gavgpool_cw_ukernel__neon_x4()
|
| /external/tensorflow/tensorflow/lite/kernels/ |
| D | cpu_backend_gemm_custom_gemv.h | 752 vpadd_f32(vget_low_f32(acc0), vget_high_f32(acc0));
754 vpadd_f32(vget_low_f32(acc1), vget_high_f32(acc1));
756 vpadd_f32(vget_low_f32(acc2), vget_high_f32(acc2));
758 vpadd_f32(vget_low_f32(acc3), vget_high_f32(acc3));
760 vpadd_f32(pairwise_reduced_acc_0, pairwise_reduced_acc_1);
762 vpadd_f32(pairwise_reduced_acc_2, pairwise_reduced_acc_3);
|
| /external/libopus/celt/arm/ |
| D | pitch_neon_intr.c | 220 xy_f32x2 = vpadd_f32(xy_f32x2, xy_f32x2); in celt_inner_prod_neon()
268 xy01_f32x2 = vpadd_f32(xy01_f32x2, xy01_f32x2); in dual_inner_prod_neon()
269 xy02_f32x2 = vpadd_f32(xy02_f32x2, xy02_f32x2); in dual_inner_prod_neon()
|
| /external/webrtc/common_audio/resampler/ |
| D | sinc_resampler_neon.cc | 45 return vget_lane_f32(vpadd_f32(m_half, m_half), 0); in Convolve_NEON()
|
| /external/webrtc/common_audio/ |
| D | fir_filter_neon.cc | 66 out[i] = vget_lane_f32(vpadd_f32(m_half, m_half), 0); in Filter()
|
| /external/XNNPACK/src/f32-raddstoreexpminusmax/gen/ |
| D | neonfma-p5-x4.c | 174 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_p5_x4()
|
| D | neon-p5-x4.c | 175 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x4()
|
| D | neonfma-lut64-p2-x4.c | 212 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_lut64_p2_x4()
|
| D | neon-lut64-p2-x4.c | 213 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_lut64_p2_x4()
|
| D | neonfma-p5-x8.c | 247 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_p5_x8()
|
| D | neon-p5-x8.c | 248 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x8()
|
| D | neonfma-p5-x8-acc2.c | 250 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_p5_x8_acc2()
|
| D | neon-p5-x8-acc2.c | 251 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x8_acc2()
|
| D | neon-p5-x12.c | 264 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x12()
|
| D | neonfma-p5-x12-acc2.c | 266 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_p5_x12_acc2()
|
| D | neonfma-p5-x12.c | 263 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_p5_x12()
|
| D | neon-p5-x12-acc2.c | 267 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x12_acc2()
|
| D | neon-lut64-p2-x8.c | 311 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_lut64_p2_x8()
|
| D | neonfma-p5-x12-acc3.c | 268 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_p5_x12_acc3()
|
| D | neon-p5-x12-acc3.c | 269 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x12_acc3()
|
| D | neonfma-lut64-p2-x8-acc2.c | 313 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_lut64_p2_x8_acc2()
|
| D | neonfma-lut64-p2-x8.c | 310 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_lut64_p2_x8()
|
| D | neon-lut64-p2-x8-acc2.c | 314 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_lut64_p2_x8_acc2()
|
| D | neon-p5-x16-acc2.c | 283 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x16_acc2()
|
| D | neon-p5-x16.c | 280 vst1_lane_f32(sum, vpadd_f32(vacc_lo, vacc_lo), 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_p5_x16()
|