Searched refs:vst1_s8 (Results 1 – 15 of 15) sorted by relevance
/external/tensorflow/tensorflow/lite/kernels/internal/optimized/integer_ops/ |
D | pooling.h | 117 vst1_s8(acc + channel, acc_reg); in MaxPool() 140 vst1_s8(output_ptr + channel, a); in MaxPool() 259 vst1_s8(output_ptr + channel, buf8); \ in AveragePool16() 276 vst1_s8(output_ptr + channel, buf8); in AveragePool16()
|
D | mul.h | 83 vst1_s8(output_data + i, clamped); in MulElementwise() 153 vst1_s8(output_data + i, clamped); in MulSimpleBroadcast()
|
D | add.h | 89 vst1_s8(output_data + i, clamped); in AddElementwise() 185 vst1_s8(output_data + i, clamped); in AddScalarBroadcast()
|
D | softmax.h | 206 vst1_s8(output_data_ptr + c, masked_output); in Softmax()
|
/external/tensorflow/tensorflow/lite/kernels/internal/optimized/ |
D | depthwiseconv_uint8_3x3_filter.h | 6690 vst1_s8(scratch_data + start_width + copy_done, half_work_reg); 6719 vst1_s8(scratch_data + start_width + copy_done, half_work_reg); 6727 vst1_s8(scratch_data + start_width + copy_done, half_work_reg); 6728 vst1_s8(scratch_data + start_width + copy_done + 8, half_work_reg); 6843 vst1_s8(scratch_data_base + scratch_data_offset, half_work_reg); 7019 vst1_s8(scratch_data + copy_done, half_work_reg); 7045 vst1_s8(scratch_data + copy_done, half_work_reg); 7052 vst1_s8(scratch_data + copy_done, half_work_reg); 7053 vst1_s8(scratch_data + copy_done + 8, half_work_reg);
|
D | depthwiseconv_uint8_transitional.h | 56 return vst1_s8(data_addr, reg); in util_vst1_x8() 2041 vst1_s8(scratch_data + start_width + copy_done, half_work_reg); 2069 vst1_s8(scratch_data + start_width + copy_done, half_work_reg); 2073 vst1_s8(scratch_data + start_width + copy_done, half_work_reg); 2074 vst1_s8(scratch_data + start_width + copy_done + 8, half_work_reg); 2184 vst1_s8(scratch_data_base + scratch_data_offset, half_work_reg); 2360 vst1_s8(scratch_data + copy_done, half_work_reg); 2386 vst1_s8(scratch_data + copy_done, half_work_reg); 2391 vst1_s8(scratch_data + copy_done, half_work_reg); 2392 vst1_s8(scratch_data + copy_done + 8, half_work_reg); [all …]
|
D | neon_tensor_utils.cc | 1645 vst1_s8(output + index, vmovn_s16(result)); in NeonCwiseMul() 2188 vst1_s8(&quantized_values[i], min_s8x8); in NeonSymmetricQuantizeFloats() 2268 vst1_s8(&quantized_values[i], min_s8x8); in NeonAsymmetricQuantizeFloats()
|
D | optimized_ops.h | 5721 vst1_s8(output + loc, res_s8); in Quantize() 6299 vst1_s8(dst, res8); in SaturateAndStore() 6715 vst1_s8(output_data + i, combined_val_narrowed); in AffineQuantize()
|
/external/gemmlowp/internal/ |
D | output_neon.h | 736 vst1_s8(dst_ptr, src.buf.reg[0]); 787 vst1_s8(dst_ptr + i * col_stride, src.buf.reg[i]); 851 vst1_s8(dst_ptr + i * stride, block.buf.reg[i]);
|
/external/libgav1/libgav1/src/dsp/arm/ |
D | common_neon.h | 165 vst1_s8(r.i8, val);
|
D | film_grain_neon.cc | 1007 vst1_s8(noise_image_row + x, vqrshrn_n_s16(grain, 5)); in WriteOverlapLine8bpp_NEON()
|
/external/libopus/silk/arm/ |
D | NSQ_del_dec_neon_intr.c | 143 vst1_s8( &pulses[ offset ], vmovn_s16( t_s16x8 ) ); in copy_winner_state_kernel()
|
/external/clang/test/CodeGen/ |
D | aarch64-neon-intrinsics.c | 11683 vst1_s8(a, b); in test_vst1_s8()
|
D | arm_neon_intrinsics.c | 18101 vst1_s8(a, b); in test_vst1_s8()
|
/external/neon_2_sse/ |
D | NEON_2_SSE.h | 1317 _NEON2SSESTORAGE void vst1_s8(__transfersize(8) int8_t * ptr, int8x8_t val); // VST1.8 {d0}, [r0] 9654 _NEON2SSESTORAGE void vst1_s8(__transfersize(8) int8_t * ptr, int8x8_t val); // VST1.8 {d0}, [r0] 9655 #define vst1_s8(ptr,val) vst1_u8((uint8_t*)ptr,val) macro
|