Remove unused variables in Xtensa implmenetation of LSTM. (#2456)
We only started getting "unused variable" errors with latest versions of the Xtensa toolchain.
BUG=323856831
diff --git a/tensorflow/lite/micro/kernels/xtensa/lstm_eval.cc b/tensorflow/lite/micro/kernels/xtensa/lstm_eval.cc
index af5bad7..94e76a1 100644
--- a/tensorflow/lite/micro/kernels/xtensa/lstm_eval.cc
+++ b/tensorflow/lite/micro/kernels/xtensa/lstm_eval.cc
@@ -122,8 +122,7 @@
}
}
#else
- WORD32 err;
- err = xa_nn_elm_add_16x16_16(output, input_1, input_2, n_batch * n_input);
+ xa_nn_elm_add_16x16_16(output, input_1, input_2, n_batch * n_input);
#endif
}
@@ -227,8 +226,7 @@
}
#else // #if !(defined(HIFI3) || defined(HIFI4) || defined(HIFI5))
void Sigmoid(int16_t* data, int32_t data_size) {
- WORD32 err;
- err = xa_nn_vec_sigmoid_sym16s_sym16s(data, data, 0, 0, data_size);
+ xa_nn_vec_sigmoid_sym16s_sym16s(data, data, 0, 0, data_size);
}
void Sigmoid(float* data, int32_t data_size) {
@@ -251,9 +249,8 @@
input_multiplier = 3;
#endif
}
- WORD32 err;
- err = xa_nn_vec_tanh_sym16s_sym16s(output_data, input_data, input_multiplier,
- tanh_input_left_shift, data_size);
+ xa_nn_vec_tanh_sym16s_sym16s(output_data, input_data, input_multiplier,
+ tanh_input_left_shift, data_size);
}
void Tanh(int32_t cell_state_scale_power, float* input_data, float* output_data,
@@ -266,8 +263,7 @@
// Input and output have the same shape in LSTM
void Mul(const ArithmeticParams& params, const int16_t* input1_data,
const int16_t* input2_data, int8_t* output_data, int32_t data_size) {
- WORD32 err;
- err = xa_nn_elm_mul_sym16sxsym16s_asym8s(
+ xa_nn_elm_mul_sym16sxsym16s_asym8s(
output_data, params.output_offset, params.output_shift,
params.output_multiplier, params.quantized_activation_min,
params.quantized_activation_max, input1_data, input2_data, data_size);
@@ -277,8 +273,7 @@
void Mul(const ArithmeticParams& params, const int16_t* input1_data,
const int16_t* input2_data, int16_t* output_data, int32_t data_size) {
int dims_4D[4] = {1, 1, 1, data_size};
- WORD32 err;
- err = xa_nn_elm_mul_broadcast_4D_sym16sxsym16s_sym16s(
+ xa_nn_elm_mul_broadcast_4D_sym16sxsym16s_sym16s(
output_data, dims_4D, params.output_shift, params.output_multiplier,
params.quantized_activation_min, params.quantized_activation_max,
input1_data, dims_4D, input2_data, dims_4D);
@@ -299,10 +294,9 @@
const int32_t* bias_data, int16_t* output_data,
const int num_batches, const int output_depth,
const int accum_depth) {
- WORD32 err;
#pragma loop_count min = 1
for (int b = 0; b < num_batches; b++) {
- err = xa_nn_matXvec_out_stride_sym8sxasym8s_16(
+ xa_nn_matXvec_out_stride_sym8sxasym8s_16(
output_data + b * output_depth, filter_data,
input_data + b * accum_depth, bias_data, output_depth, accum_depth,
accum_depth, 1, params.input_offset, params.output_multiplier,
@@ -316,9 +310,7 @@
const int64_t* bias_data, int16_t* output_data,
const int num_batches, const int output_depth,
const int accum_depth) {
- WORD32 err;
-
- err = xa_nn_matmul_sym8sxsym16s_sym16s(
+ xa_nn_matmul_sym8sxsym16s_sym16s(
output_data, filter_data, input_data, bias_data, output_depth,
accum_depth, accum_depth, num_batches, accum_depth, output_depth, 1,
params.input_offset, params.output_multiplier, params.output_shift,
@@ -376,9 +368,8 @@
TFLITE_DCHECK_LE(step_info.CellStateOffset() + cell_state_shape.FlatSize(),
tflite::micro::GetTensorShape(cell_state).FlatSize());
- WORD32 err;
// Multiplier is equivalent to 0.5 here so adding 1 to shifts
- err = xa_nn_lstm_cell_state_update_16(
+ xa_nn_lstm_cell_state_update_16(
tflite::micro::GetTensorData<int16_t>(cell_state) +
step_info.CellStateOffset(),
forget_gate_output, cell_gate_output, input_gate_output,
