scatter.cu.h 6.3 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
/* Copyright (c) 2022 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. */

#pragma once
16 17 18 19 20
#include "paddle/phi/backends/gpu/gpu_context.h"
#include "paddle/phi/backends/gpu/gpu_launch_config.h"
#include "paddle/phi/kernels/funcs/aligned_vector.h"

#define VecBytes 16
21 22 23 24 25 26 27 28 29 30 31 32 33 34

namespace phi {
namespace funcs {
namespace sparse {

/**
 * brief: scatter add
 * input: the inputs
 * unique_value: refer to UpdateIndexKernel notes
 * out_index: the output feature index
 * non_zero_num: the number of output features
 * rulebook_len: the length of rulebook
 * channels: the output channel size
 * out: the outputs
35
 **/
36
template <typename T, int VecSize>
37 38 39 40 41 42
__global__ void ScatterKernel(const T* input,
                              const int* unique_value,
                              const int* out_index,
                              const int non_zero_num,
                              const int rulebook_len,
                              const int channels,
43
                              T* out) {
44
  int tid = threadIdx.x + blockIdx.x * blockDim.x;
45 46 47 48 49 50 51
  const int vec_channels = channels / VecSize;
  using LoadT = phi::AlignedVector<T, VecSize>;
  using StoreT = phi::AlignedVector<T, VecSize>;
  for (int i = tid; i < non_zero_num * vec_channels;
       i += gridDim.x * blockDim.x) {
    int indices_i = i / vec_channels;
    int channels_i = i - indices_i * vec_channels;
52 53 54 55 56

    int start = unique_value[indices_i];
    int end = indices_i == non_zero_num - 1 ? rulebook_len
                                            : unique_value[indices_i + 1];
    // max(end-start) = kernel_size
57
    StoreT sums = {static_cast<T>(0)};
58 59
    for (int j = start; j < end; j++) {
      const int out_feature_i = out_index[j];
60 61 62 63 64 65 66
      LoadT vec_in;
      phi::Load<T, VecSize>(
          input + out_feature_i * channels + channels_i * VecSize, &vec_in);
#pragma unroll
      for (int k = 0; k < VecSize; k++) {
        sums[k] += vec_in[k];
      }
67
    }
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
    phi::Store<T, VecSize>(sums,
                           out + indices_i * channels + channels_i * VecSize);
  }
}

// scatter's index has been grouped in advance
// index_counts record the count of each group
// index_groups save the index of each group
template <typename T, int VecSize>
__global__ void ScatterKernelV2(const T* input,
                                const int* index_counts,
                                const int* index_groups,
                                const int non_zero_num,
                                const int kernel_size,
                                const int channels,
                                const int buffer_counts,
                                T* out) {
  int tid = threadIdx.x + blockIdx.x * blockDim.x;
  const int vec_channels = channels / VecSize;
  using LoadT = phi::AlignedVector<T, VecSize>;
  using StoreT = phi::AlignedVector<T, VecSize>;
  for (int i = tid; i < non_zero_num * vec_channels;
       i += gridDim.x * blockDim.x) {
    int indices_i = i / vec_channels;
    int channels_i = i - indices_i * vec_channels;

    StoreT sums = {static_cast<T>(0)};
    phi::Load<T, VecSize>(out + indices_i * channels + channels_i * VecSize,
                          &sums);
    for (int it = 0; it < buffer_counts; it++) {
      int len = index_counts[indices_i + it * non_zero_num];
99
      const int group_offset = it * kernel_size * non_zero_num;
100 101
      for (int j = 0; j < len; j++) {
        const int out_feature_i =
102
            index_groups[indices_i * kernel_size + j + group_offset];
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
        LoadT vec_in;
        phi::Load<T, VecSize>(
            input + out_feature_i * channels + channels_i * VecSize, &vec_in);
#pragma unroll
        for (int k = 0; k < VecSize; k++) {
          sums[k] += vec_in[k];
        }
      }
    }
    phi::Store<T, VecSize>(sums,
                           out + indices_i * channels + channels_i * VecSize);
  }
}

template <typename T>
void ScatterV2(const GPUContext& dev_ctx,
               const T* input,
               const int* index_counts,
               const int* index_groups,
               const int non_zero_num,
               const int kernel_size,
               const int channels,
               const int buffer_counts,
               T* output) {
  const int VecSize = VecBytes / sizeof(T);
  if (channels % VecSize == 0) {
    auto config = phi::backends::gpu::GetGpuLaunchConfig1D(
        dev_ctx, non_zero_num * channels / VecSize, 1);
    ScatterKernelV2<T, VecSize><<<config.block_per_grid.x,
                                  config.thread_per_block.x,
                                  0,
                                  dev_ctx.stream()>>>(input,
                                                      index_counts,
                                                      index_groups,
                                                      non_zero_num,
                                                      kernel_size,
                                                      channels,
                                                      buffer_counts,
                                                      output);
  } else {
    auto config = phi::backends::gpu::GetGpuLaunchConfig1D(
        dev_ctx, non_zero_num * channels, 1);
    ScatterKernelV2<T, 1><<<config.block_per_grid.x,
                            config.thread_per_block.x,
                            0,
                            dev_ctx.stream()>>>(input,
                                                index_counts,
                                                index_groups,
                                                non_zero_num,
                                                kernel_size,
                                                channels,
                                                buffer_counts,
                                                output);
156 157 158 159 160 161
  }
}

}  // namespace sparse
}  // namespace funcs
}  // namespace phi