conv_add_kernel.cpp

/* Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License. */

#ifdef FUSION_CONVADD_OP

#include "operators/kernel/conv_add_kernel.h"
#ifdef PADDLE_MOBILE_MALI_GPU
#include "acl_operator.h"
#include "framework/operator.h"
#include "operators/op_param.h"

namespace paddle_mobile {
namespace operators {

template <typename DeviceType, typename T>
class AclConvAddOp : public acl::ACLOperator {
 public:
  AclConvAddOp() {
    this->force_bypass_acl_path_ =
        bypass_acl_class_layer & FLAGS_ENABLE_ACL_CONV;
  }
  ~AclConvAddOp() = default;
  AclConvAddOp(const AclConvAddOp&) = delete;
  AclConvAddOp& operator=(const AclConvAddOp&) = delete;
  AclConvAddOp(AclConvAddOp&&) = delete;
  AclConvAddOp& operator=(AclConvAddOp&&) = delete;

  acl::AclParameters& getargs() { return args; }
  void InitAclLayer(const FusionConvAddParam<DeviceType>& param) {
    setTargetHint(acl::TargetHint::OPENCL);
    arm_compute::TensorShape input_shape(args.in_cols, args.in_rows,
                                         args.in_depth, args.batch);
    arm_compute::TensorShape output_shape(args.out_cols, args.out_rows,
                                          args.out_depth, args.out_num);
    arm_compute::TensorShape weights_shape(args.filter_cols, args.filter_rows,
                                           args.in_depth / args.num_group,
                                           args.out_depth);
    arm_compute::TensorShape biases_shape(args.out_depth);
    arm_compute::PadStrideInfo conv_info(
        args.stride_cols, args.stride_rows, args.pad_cols, args.pad_rows,
        arm_compute::DimensionRoundingType::FLOOR);

    if (is_operator_init_done(input_shape)) return;
    set_operator_init_done();
    this->force_bypass_acl_path_ = false;

    // check_direct_conv();
    group() = args.num_group;
    //[kernel_x, kernel_y, IFM, OFM]
    new_tensor(weights(), weights_shape, args.weight_data);
    //[OFM]
    if (args.biases_data) {
      new_tensor(biases(), biases_shape, args.biases_data);
    }

    //[width, height, IFM]
    new_tensor(input(), input_shape, args.input_data);
    //[width, height, OFM]
    new_tensor(output(), output_shape, args.output_data);

    acl_configure(conv, this, conv_info);
  }

  void RunAcl(void* input, void* output) {
    acl::ACLOperator::acl_run(input, output);
  }
  bool Bypass_acl(const FusionConvAddParam<DeviceType>& param) {
    bool bypass_acl = false;
    AclParametersByContext(param);
    InitAclLayer(param);
    // for performance, more groups impact GPU performance
    if (this->force_bypass_acl_path_ || args.num_group >= 5) {
      bypass_acl = true;
    }
    if (args.dim > 2) {
      bypass_acl = true;
    }
    if (args.dilated) {
      bypass_acl = true;
    }
    return bypass_acl;
  }

 private:
  void check_direct_conv() {
    bool use_direct_conv = false;
    const char* pDirectConv;
    pDirectConv = getenv("DIRECTCONV");
    if (pDirectConv) {
      unsigned int bdirectconv;
      sscanf(pDirectConv, "%i", &bdirectconv);
      if (bdirectconv != use_direct_conv) {
        use_direct_conv = bdirectconv;
        printf("DIRECTCONV<%s>\n", pDirectConv);
        printf("DIRECTCONV: %x\n", use_direct_conv);
      }
    }
    int pad_data[2], kernel[2];
    pad_data[1] = args.pad_rows;
    pad_data[0] = args.pad_cols;
    kernel[1] = args.filter_rows;
    kernel[0] = args.filter_cols;
    if (use_direct_conv && ((kernel[0] == 1 && kernel[1] == 1 &&
                             pad_data[0] == 0 && pad_data[1] == 0) ||
                            (kernel[0] == 3 && kernel[1] == 3 &&
                             pad_data[0] <= 1 && pad_data[1] <= 1))) {
      setConvMethod();  // NEDirectConvolutionLayer only for 1x1 and 3x3
    }
  }

  void AclParametersByContext(const FusionConvAddParam<DeviceType>& param) {
    const Tensor* input = param.Input();
    Tensor filter = *param.Filter();
    Tensor* output = param.Output();
    Tensor* bias;

    int groups = param.Groups();
    std::vector<int> strides = param.Strides();
    std::vector<int> paddings = param.Paddings();
    std::vector<int> dilations = param.Dilations();

    const T* input_data = input->data<T>();
    T* output_data = output->mutable_data<T>();
    const T* weight_data = filter.data<T>();

    args.input_data = (void*)input_data;
    args.output_data = (void*)output_data;
    args.weight_data = (void*)weight_data;
    args.biases_data = nullptr;

    try {
      bias = param.Bias();
    } catch (const std::exception& e) {
    }
    if (bias) {
      const T* biases_data = bias->data<T>();
      args.biases_data = (void*)biases_data;
    }

    args.num_group = groups;

    args.dilation_rows = dilations[0];
    args.dilation_cols = dilations[1];
    if (dilations[0] != 1 || dilations[1] != 1) {
      args.dilated = true;
    }

    // NCHW
    // std::cout << "In dims: " << (input->dims()).size() << std::endl;
    args.batch = input->dims()[0];
    args.in_depth = input->dims()[1];
    args.in_rows = input->dims()[2];
    args.in_cols = input->dims()[3];
    // std::cout <<"In N: " << args.batch << " C: " <<  args.in_depth
    //  << " H: " << args.in_rows << " W: " << args.in_cols << "\n";
    // NCHW
    // std::cout << "Out dims: " << (output->dims()).size() << std::endl;
    args.out_num = output->dims()[0];
    args.out_depth = output->dims()[1];
    args.out_rows = output->dims()[2];
    args.out_cols = output->dims()[3];
    // std::cout <<"Out N: " << static_cast<int>(output->dims()[0])
    //  << " C: " <<  args.out_depth
    //  << " H: " << args.out_rows << " W: " << args.out_cols << "\n";
    // MCHW = OIHW
    args.filter_rows = filter.dims()[2];
    args.filter_cols = filter.dims()[3];
    // std::cout <<"Filter O: " << static_cast<int>(filter.dims()[0])
    //  << " I: " <<  static_cast<int>(filter.dims()[1])
    //  << " H: " << args.filter_rows << " W: " << args.filter_cols << "\n";

    // strides(h_stride, w_stride)
    args.stride_rows = strides[0];
    args.stride_cols = strides[1];
    // std::cout <<"Stride H: " << args.stride_rows << " W: " <<
    // args.stride_cols << "\n";

    // paddings(h_pad, w_pad)
    args.pad_rows = paddings[0];
    args.pad_cols = paddings[1];
    // std::cout <<"Pad H: " << args.pad_rows << " W: " << args.pad_cols <<
    // "\n";
  }
  acl::AclParameters args;
};

template <>
bool ConvAddKernel<GPU_MALI, float>::Init(FusionConvAddParam<GPU_MALI>* param) {
  AclConvAddOp<GPU_MALI, float>* acl_op =
      reinterpret_cast<AclConvAddOp<GPU_MALI, float>*>(this->GetAclOp());
  if (acl_op == nullptr) {
    acl_op = new AclConvAddOp<GPU_MALI, float>();
    this->SetAclOp((void*)acl_op, (void*)this);
  }
  if (acl_op->Bypass_acl(*param)) {
    std::cout << "init acl failed" << std::endl;
    return false;
  }
  return true;
}

template <>
void ConvAddKernel<GPU_MALI, float>::Compute(
    const FusionConvAddParam<GPU_MALI>& param) const {
  std::cout << "init acl" << std::endl;
  AclConvAddOp<GPU_MALI, float>* acl_op =
      reinterpret_cast<AclConvAddOp<GPU_MALI, float>*>(this->GetAclOp());
  if (acl_op == nullptr) {
    return;
  }
  acl::AclParameters& args = acl_op->getargs();

  acl_op->RunAcl(args.input_data, args.output_data);
}

template class ConvAddKernel<GPU_MALI, float>;
}  // namespace operators
}  // namespace paddle_mobile

#endif
#endif