src/subgraph/leaky-relu.c

// Copyright 2020 Google LLC
//
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree.

#include <assert.h>
#include <math.h>
#include <stddef.h>
#include <stdint.h>

#include <xnnpack.h>
#include <xnnpack/log.h>
#include <xnnpack/operator.h>
#include <xnnpack/params.h>
#include <xnnpack/subgraph.h>
#include <xnnpack/subgraph-validation.h>


static enum xnn_status create_leaky_relu_operator(
  const struct xnn_node* node,
  const struct xnn_value* values,
  size_t num_values,
  struct xnn_operator_data* opdata,
  const struct xnn_caches* caches)
{
  assert(node->num_inputs == 1);
  const uint32_t input_id = node->inputs[0];
  assert(input_id != XNN_INVALID_VALUE_ID);
  assert(input_id < num_values);

  assert(node->num_outputs == 1);
  const uint32_t output_id = node->outputs[0];
  assert(output_id != XNN_INVALID_VALUE_ID);
  assert(output_id < num_values);

  const size_t num_input_dims = values[input_id].shape.num_dims;
  const size_t channel_dim = num_input_dims == 0 ? 1 : values[input_id].shape.dim[num_input_dims - 1];

  enum xnn_status status;
  switch (node->compute_type) {
#ifndef XNN_NO_F16_OPERATORS
    case xnn_compute_type_fp16:
      status = xnn_create_leaky_relu_nc_f16(
        channel_dim /* channels */, channel_dim /* input stride */, channel_dim /* output stride */,
        node->params.leaky_relu.negative_slope,
        node->flags,
        &opdata->operator_objects[0]);
      break;
#endif  // XNN_NO_F16_OPERATORS
    case xnn_compute_type_fp32:
      status = xnn_create_leaky_relu_nc_f32(
        channel_dim /* channels */, channel_dim /* input stride */, channel_dim /* output stride */,
        node->params.leaky_relu.negative_slope,
        node->flags,
        &opdata->operator_objects[0]);
      break;
#ifndef XNN_NO_QS8_OPERATORS
    case xnn_compute_type_qs8:
      status = xnn_create_leaky_relu_nc_qs8(
        channel_dim /* channels */, channel_dim /* input stride */, channel_dim /* output stride */,
        node->params.leaky_relu.negative_slope,
        (int8_t) values[input_id].quantization.zero_point, values[input_id].quantization.scale,
        (int8_t) values[output_id].quantization.zero_point, values[output_id].quantization.scale,
        node->flags,
        &opdata->operator_objects[0]);
      break;
#endif  // !defined(XNN_NO_QS8_OPERATORS)
#ifndef XNN_NO_QU8_OPERATORS
    case xnn_compute_type_qu8:
      status = xnn_create_leaky_relu_nc_qu8(
        channel_dim /* channels */, channel_dim /* input stride */, channel_dim /* output stride */,
        node->params.leaky_relu.negative_slope,
        (uint8_t) values[input_id].quantization.zero_point, values[input_id].quantization.scale,
        (uint8_t) values[output_id].quantization.zero_point, values[output_id].quantization.scale,
        node->flags,
        &opdata->operator_objects[0]);
      break;
#endif  // !defined(XNN_NO_QU8_OPERATORS)
    default:
      XNN_UNREACHABLE;
  }
  if (status == xnn_status_success) {
    opdata->batch_size = xnn_shape_multiply_non_channel_dims(&values[input_id].shape);
    opdata->inputs[0] = input_id;
    opdata->outputs[0] = output_id;
  }
  return status;
}

static enum xnn_status setup_leaky_relu_operator(
  const struct xnn_operator_data* opdata,
  const struct xnn_blob* blobs,
  size_t num_blobs,
  pthreadpool_t threadpool)
{
  const uint32_t input_id = opdata->inputs[0];
  assert(input_id != XNN_INVALID_VALUE_ID);
  assert(input_id < num_blobs);

  const uint32_t output_id = opdata->outputs[0];
  assert(output_id != XNN_INVALID_VALUE_ID);
  assert(output_id < num_blobs);

  const struct xnn_blob* input_blob = blobs + input_id;
  const void* input_data = input_blob->data;
  assert(input_data != NULL);

  const struct xnn_blob* output_blob = blobs + output_id;
  void* output_data = output_blob->data;
  assert(output_data != NULL);

  switch (opdata->operator_objects[0]->type) {
#ifndef XNN_NO_F16_OPERATORS
    case xnn_operator_type_leaky_relu_nc_f16:
      return xnn_setup_leaky_relu_nc_f16(
        opdata->operator_objects[0],
        opdata->batch_size,
        input_data,
        output_data,
        threadpool);
#endif  // XNN_NO_F16_OPERATORS
    case xnn_operator_type_leaky_relu_nc_f32:
      return xnn_setup_leaky_relu_nc_f32(
        opdata->operator_objects[0],
        opdata->batch_size,
        input_data,
        output_data,
        threadpool);
#ifndef XNN_NO_QS8_OPERATORS
    case xnn_operator_type_leaky_relu_nc_qs8:
      return xnn_setup_leaky_relu_nc_qs8(
        opdata->operator_objects[0],
        opdata->batch_size,
        input_data,
        output_data,
        threadpool);
#endif  // !defined(XNN_NO_QS8_OPERATORS)
#ifndef XNN_NO_QU8_OPERATORS
    case xnn_operator_type_leaky_relu_nc_qu8:
      return xnn_setup_leaky_relu_nc_qu8(
        opdata->operator_objects[0],
        opdata->batch_size,
        input_data,
        output_data,
        threadpool);
#endif  // !defined(XNN_NO_QU8_OPERATORS)
    default:
      XNN_UNREACHABLE;
  }
}

enum xnn_status xnn_define_leaky_relu(
  xnn_subgraph_t subgraph,
  float negative_slope,
  uint32_t input_id,
  uint32_t output_id,
  uint32_t flags)
{
  enum xnn_status status;
  if ((status = xnn_subgraph_check_xnnpack_initialized(xnn_node_type_leaky_relu)) != xnn_status_success) {
    return status;
  }

  if (!isfinite(negative_slope)) {
    xnn_log_error(
      "failed to create %s operator with %f negative slope: finite number expected",
      xnn_node_type_to_string(xnn_node_type_leaky_relu),
      negative_slope);
    return xnn_status_invalid_parameter;
  }

  if ((status = xnn_subgraph_check_input_node_id(xnn_node_type_leaky_relu, input_id, subgraph->num_values)) !=
      xnn_status_success) {
    return status;
  }

  const struct xnn_value* input_value = &subgraph->values[input_id];
  status = xnn_subgraph_check_input_type_dense(xnn_node_type_leaky_relu, input_id, input_value);
  if (status != xnn_status_success) {
    return status;
  }

  switch (input_value->datatype) {
    case xnn_datatype_fp32:
#ifndef XNN_NO_QS8_OPERATORS
    case xnn_datatype_qint8:
#endif  // !defined(XNN_NO_QS8_OPERATORS)
#ifndef XNN_NO_QU8_OPERATORS
    case xnn_datatype_quint8:
#endif  // !defined(XNN_NO_QU8_OPERATORS)
      break;
    default:
      xnn_log_error(
        "failed to define %s operator with input ID #%" PRIu32 ": unsupported Value datatype %s (%d)",
        xnn_node_type_to_string(xnn_node_type_leaky_relu), input_id,
        xnn_datatype_to_string(input_value->datatype), input_value->datatype);
      return xnn_status_invalid_parameter;
  }

  status = xnn_subgraph_check_output_node_id(xnn_node_type_leaky_relu, output_id, subgraph->num_values);
  if (status != xnn_status_success) {
    return status;
  }

  const struct xnn_value* output_value = &subgraph->values[output_id];
  status = xnn_subgraph_check_output_type_dense(xnn_node_type_leaky_relu, output_id, output_value);
  if (status != xnn_status_success) {
    return status;
  }

  enum xnn_compute_type compute_type = xnn_compute_type_invalid;
  switch (output_value->datatype) {
    case xnn_datatype_fp32:
      compute_type = xnn_compute_type_fp32;
      break;
#ifndef XNN_NO_QS8_OPERATORS
    case xnn_datatype_qint8:
      compute_type = xnn_compute_type_qs8;
      break;
#endif  // !defined(XNN_NO_QS8_OPERATORS)
#ifndef XNN_NO_QU8_OPERATORS
    case xnn_datatype_quint8:
      compute_type = xnn_compute_type_qu8;
      break;
#endif  // !defined(XNN_NO_QU8_OPERATORS)
    default:
      xnn_log_error(
        "failed to define %s operator with output ID #%" PRIu32 ": unsupported Value datatype %s (%d)",
        xnn_node_type_to_string(xnn_node_type_leaky_relu), output_id,
        xnn_datatype_to_string(output_value->datatype), output_value->datatype);
      return xnn_status_invalid_parameter;
  }
  assert(compute_type != xnn_compute_type_invalid);

  status = xnn_subgraph_check_datatype_matches(xnn_node_type_leaky_relu, input_id, input_value, output_id, output_value);
  if (status != xnn_status_success) {
    return status;
  }

#if !defined(XNN_NO_U8_OPERATORS) || !defined(XNN_NO_S8_OPERATORS)
  if (compute_type == xnn_datatype_qint8 || compute_type == xnn_datatype_quint8) {
    const float positive_input_output_scale = input_value->quantization.scale / output_value->quantization.scale;
    if (positive_input_output_scale < 0x1.0p-8f || positive_input_output_scale > 0x1.0p+7f) {
      xnn_log_error(
        "failed to define %s operator with %.7g positive-input-to-output scale ratio: scale ratio must be in [2**-8, 2**7] range",
        xnn_node_type_to_string(xnn_node_type_leaky_relu), positive_input_output_scale);
      return xnn_status_invalid_parameter;
    }

    const float negative_input_output_scale = positive_input_output_scale * negative_slope;
    if (negative_input_output_scale < -0x1.FFFC00p+6f || negative_input_output_scale > 0x1.0p+7f) {
      xnn_log_error(
        "failed to define %s operator with %.7g negative-input-to-output scale ratio: scale ratio must be in (-2**7, 2**7] range and ",
        xnn_node_type_to_string(xnn_node_type_leaky_relu), negative_input_output_scale);
      return xnn_status_invalid_parameter;
    }

    if (fabsf(negative_input_output_scale) < 0x1.0p-8f) {
      xnn_log_error(
        "failed to define %s operator with %.7g negative-input-to-output scale ratio: scale ratio must be at least 2**-8 in absolute value",
        xnn_node_type_to_string(xnn_node_type_leaky_relu), negative_input_output_scale);
      return xnn_status_invalid_parameter;
    }
  }
#endif  // !defined(XNN_NO_U8_OPERATORS) || !defined(XNN_NO_S8_OPERATORS)

  struct xnn_node* node = xnn_subgraph_new_node(subgraph);
  if (node == NULL) {
    return xnn_status_out_of_memory;
  }

  node->type = xnn_node_type_leaky_relu;
  node->compute_type = compute_type;
  node->params.leaky_relu.negative_slope = negative_slope;
  node->num_inputs = 1;
  node->inputs[0] = input_id;
  node->num_outputs = 1;
  node->outputs[0] = output_id;
  node->flags = flags;

  node->create = create_leaky_relu_operator;
  node->setup = setup_leaky_relu_operator;

  return xnn_status_success;
}