/external/llvm/test/CodeGen/AMDGPU/ |
D | scratch-buffer.ll | 19 %scratch0 = alloca [8192 x i32] 22 %scratchptr0 = getelementptr [8192 x i32], [8192 x i32]* %scratch0, i32 0, i32 0 32 %if_ptr = getelementptr [8192 x i32], [8192 x i32]* %scratch0, i32 0, i32 %if_offset 57 %scratch0 = alloca [8192 x i32] 61 %scratchptr0 = getelementptr [8192 x i32], [8192 x i32]* %scratch0, i32 0, i32 %offset0 73 %if_ptr = getelementptr [8192 x i32], [8192 x i32]* %scratch0, i32 0, i32 %if_offset
|
/external/tensorflow/tensorflow/lite/kernels/ |
D | lstm_eval.cc | 331 float* scratch0, // size: n_batch in CalculateLstmGateHybrid() 362 input_to_gate_row_sums, compute_row_sums, scratch0, context); in CalculateLstmGateHybrid() 372 aux_input_to_gate_row_sums, compute_row_sums, scratch0, context); in CalculateLstmGateHybrid() 390 recurrent_to_gate_row_sums, compute_row_sums, scratch0, context); in CalculateLstmGateHybrid() 439 CpuBackendContext* context, float* scratch0, int8_t* scratch1, in CalculateLstmOutputHybrid() argument 442 activation, scratch0); in CalculateLstmOutputHybrid() 443 tensor_utils::VectorVectorCwiseProduct(output_gate, scratch0, in CalculateLstmOutputHybrid() 444 n_batch * n_cell, scratch0); in CalculateLstmOutputHybrid() 456 if (!tensor_utils::IsZeroVector(scratch0, n_batch * n_cell)) { in CalculateLstmOutputHybrid() 458 tensor_utils::BatchQuantizeFloats(scratch0, n_batch, n_cell, scratch1, in CalculateLstmOutputHybrid() [all …]
|
D | lstm_eval.h | 194 TfLiteTensor* scratch0, TfLiteTensor* scratch1, TfLiteTensor* scratch2, 220 TfLiteTensor* scratch0, TfLiteTensor* scratch1, TfLiteTensor* scratch2,
|
D | lstm.cc | 2077 TfLiteTensor* scratch0; in Eval() local 2079 GetTemporarySafe(context, node, 0, &scratch0)); in Eval() 2107 cell_state, output, scratch0, scratch1, scratch2, scratch3, in Eval() 2110 TfLiteTensor* scratch0; in Eval() local 2112 GetTemporarySafe(context, node, 0, &scratch0)); in Eval() 2145 &op_data->integer_lstm_param, scratch0, scratch1, scratch2, in Eval()
|
D | unidirectional_sequence_lstm.cc | 1398 TfLiteTensor* scratch0; in Eval() local 1400 GetTemporarySafe(context, node, 0, &scratch0)); in Eval() 1428 output, scratch0, scratch1, scratch2, scratch3, scratch4, scratch5, in Eval()
|
/external/llvm-project/llvm/test/CodeGen/AMDGPU/ |
D | scratch-buffer.ll | 18 %scratch0 = alloca [8192 x i32], addrspace(5) 21 %scratchptr0 = getelementptr [8192 x i32], [8192 x i32] addrspace(5)* %scratch0, i32 0, i32 0 31 %if_ptr = getelementptr [8192 x i32], [8192 x i32] addrspace(5)* %scratch0, i32 0, i32 %if_offset 58 %scratch0 = alloca [8192 x i32], addrspace(5) 62 …%scratchptr0 = getelementptr [8192 x i32], [8192 x i32] addrspace(5)* %scratch0, i32 0, i32 %offse… 74 %if_ptr = getelementptr [8192 x i32], [8192 x i32] addrspace(5)* %scratch0, i32 0, i32 %if_offset
|
D | memory-legalizer-multiple-mem-operands-nontemporal-2.mir | 7 %scratch0 = alloca [8192 x i32], addrspace(5) 9 %scratchptr01 = bitcast [8192 x i32] addrspace(5)* %scratch0 to i32 addrspace(5)* 17 …%if_ptr = getelementptr [8192 x i32], [8192 x i32] addrspace(5)* %scratch0, i32 0, i32 %if_offset,… 79 - { id: 0, name: scratch0, type: default, offset: 4, size: 32768, alignment: 4,
|
D | memory-legalizer-multiple-mem-operands-nontemporal-1.mir | 6 %scratch0 = alloca [8192 x i32], addrspace(5) 8 %scratchptr01 = bitcast [8192 x i32] addrspace(5)* %scratch0 to i32 addrspace(5)* 16 …%if_ptr = getelementptr [8192 x i32], [8192 x i32] addrspace(5)* %scratch0, i32 0, i32 %if_offset,… 99 - { id: 0, name: scratch0, type: default, offset: 4, size: 32768, alignment: 4,
|
/external/libopus/celt/ |
D | kiss_fft.c | 120 kiss_fft_cpx scratch0, scratch1; in kf_bfly4() local 122 C_SUB( scratch0 , *Fout, Fout[2] ); in kf_bfly4() 129 Fout[1].r = ADD32_ovflw(scratch0.r, scratch1.i); in kf_bfly4() 130 Fout[1].i = SUB32_ovflw(scratch0.i, scratch1.r); in kf_bfly4() 131 Fout[3].r = SUB32_ovflw(scratch0.r, scratch1.i); in kf_bfly4() 132 Fout[3].i = ADD32_ovflw(scratch0.i, scratch1.r); in kf_bfly4()
|
/external/rnnoise/src/ |
D | kiss_fft.c | 117 kiss_fft_cpx scratch0, scratch1; in kf_bfly4() local 119 C_SUB( scratch0 , *Fout, Fout[2] ); in kf_bfly4() 126 Fout[1].r = ADD32_ovflw(scratch0.r, scratch1.i); in kf_bfly4() 127 Fout[1].i = SUB32_ovflw(scratch0.i, scratch1.r); in kf_bfly4() 128 Fout[3].r = SUB32_ovflw(scratch0.r, scratch1.i); in kf_bfly4() 129 Fout[3].i = ADD32_ovflw(scratch0.i, scratch1.r); in kf_bfly4()
|
/external/tensorflow/tensorflow/lite/tools/optimize/calibration/builtin_logging_ops/ |
D | lstm.cc | 182 float* output_state_ptr, float* cell_state_ptr, float* scratch0, in LstmStepCalibration() 192 float* input_gate_scratch = scratch0; in LstmStepCalibration()
|