wehub-resource-sync
210 lines
8.4 KiB
Plaintext
210 lines
8.4 KiB
Plaintext
// 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.
|
|
|
|
#include "paddle/phi/kernels/send_u_recv_kernel.h"
|
|
|
|
#include <thrust/device_vector.h>
|
|
#include <thrust/execution_policy.h>
|
|
#include <thrust/fill.h>
|
|
|
|
#include <algorithm>
|
|
#include <vector>
|
|
|
|
#include "paddle/common/hostdevice.h"
|
|
#include "paddle/phi/backends/gpu/gpu_context.h"
|
|
#include "paddle/phi/core/kernel_registry.h"
|
|
#include "paddle/phi/kernels/full_kernel.h"
|
|
#include "paddle/phi/kernels/funcs/math_function.h"
|
|
#include "paddle/phi/kernels/gpu/graph_send_recv_funcs.h"
|
|
|
|
namespace phi {
|
|
|
|
template <typename Context, typename T, typename IndexT>
|
|
void GraphSendRecvOpCUDAKernelLaunchHelper(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& src_index,
|
|
const DenseTensor& dst_index,
|
|
const std::string& reduce_op,
|
|
int64_t out_size,
|
|
DenseTensor* out,
|
|
DenseTensor* dst_count = nullptr) {
|
|
// TODO(large-tensor): downstream functors may still use int; guard until
|
|
// upgraded.
|
|
const int64_t& index_size = src_index.dims()[0];
|
|
|
|
const auto& src_dims = x.dims();
|
|
int64_t memset_size = 1;
|
|
if (out_size <= 0) {
|
|
out->Resize(src_dims);
|
|
for (int i = 0; i < src_dims.size(); ++i) {
|
|
memset_size *= src_dims[i];
|
|
}
|
|
} else {
|
|
// Set out dim following out_size.
|
|
std::vector<int64_t> dims_ = vectorize(out->dims());
|
|
if (dims_.size() > 0) {
|
|
dims_[0] = out_size;
|
|
}
|
|
out->Resize(dims_);
|
|
memset_size = out_size;
|
|
for (int i = 1; i < src_dims.size(); ++i) {
|
|
memset_size *= src_dims[i];
|
|
}
|
|
}
|
|
dev_ctx.template Alloc<T>(out);
|
|
T* p_output = out->data<T>();
|
|
const size_t& memset_bytes = memset_size * sizeof(T);
|
|
funcs::SetConstant<Context, T> constant_functor;
|
|
if (reduce_op == "SUM" || reduce_op == "MEAN") {
|
|
constant_functor(dev_ctx, out, static_cast<T>(0));
|
|
} else if (reduce_op == "MAX") {
|
|
constant_functor(dev_ctx, out, std::numeric_limits<T>::lowest());
|
|
} else if (reduce_op == "MIN") {
|
|
constant_functor(dev_ctx, out, std::numeric_limits<T>::max());
|
|
}
|
|
|
|
if (index_size == 0) return;
|
|
|
|
int64_t slice_size = 1;
|
|
for (int i = 1; i < src_dims.size(); ++i) {
|
|
slice_size *= src_dims[i];
|
|
}
|
|
const T* p_src = x.data<T>();
|
|
const IndexT* s_index = src_index.data<IndexT>();
|
|
const IndexT* d_index = dst_index.data<IndexT>();
|
|
|
|
int block = 1024;
|
|
int64_t n = slice_size * index_size;
|
|
int64_t max_grid_dimx = dev_ctx.GetCUDAMaxGridDimSize()[0];
|
|
int64_t grid_tmp = (n + block - 1) / block;
|
|
int64_t grid = grid_tmp < max_grid_dimx ? grid_tmp : max_grid_dimx;
|
|
int64_t input_size = out_size <= 0 ? src_dims[0] : out_size;
|
|
if (reduce_op == "SUM") {
|
|
GraphSendRecvSumCUDAFunctor<T, IndexT> functor;
|
|
GraphSendRecvCUDAKernel<T, IndexT, GraphSendRecvSumCUDAFunctor<T, IndexT>>
|
|
<<<grid, block, 0, dev_ctx.stream()>>>(
|
|
p_src, s_index, d_index, p_output, index_size, slice_size, functor);
|
|
} else if (reduce_op == "MAX") {
|
|
GraphSendRecvMaxCUDAFunctor<T, IndexT> functor;
|
|
GraphSendRecvCUDAKernel<T, IndexT, GraphSendRecvMaxCUDAFunctor<T, IndexT>>
|
|
<<<grid, block, 0, dev_ctx.stream()>>>(
|
|
p_src, s_index, d_index, p_output, index_size, slice_size, functor);
|
|
|
|
int64_t grid_max_tmp = (input_size * slice_size + block - 1) / block;
|
|
int64_t grid_max =
|
|
grid_max_tmp < max_grid_dimx ? grid_max_tmp : max_grid_dimx;
|
|
InputResetMaxCUDAKernel<T><<<grid_max, block, 0, dev_ctx.stream()>>>(
|
|
p_output, input_size, slice_size);
|
|
} else if (reduce_op == "MIN") {
|
|
GraphSendRecvMinCUDAFunctor<T, IndexT> functor;
|
|
GraphSendRecvCUDAKernel<T, IndexT, GraphSendRecvMinCUDAFunctor<T, IndexT>>
|
|
<<<grid, block, 0, dev_ctx.stream()>>>(
|
|
p_src, s_index, d_index, p_output, index_size, slice_size, functor);
|
|
|
|
int64_t grid_min_tmp = (input_size * slice_size + block - 1) / block;
|
|
int64_t grid_min =
|
|
grid_min_tmp < max_grid_dimx ? grid_min_tmp : max_grid_dimx;
|
|
InputResetMinCUDAKernel<T><<<grid_min, block, 0, dev_ctx.stream()>>>(
|
|
p_output, input_size, slice_size);
|
|
} else if (reduce_op == "MEAN") {
|
|
GraphSendRecvSumCUDAFunctor<T, IndexT> functor;
|
|
GraphSendRecvCUDAKernel<T, IndexT, GraphSendRecvSumCUDAFunctor<T, IndexT>>
|
|
<<<grid, block, 0, dev_ctx.stream()>>>(
|
|
p_src, s_index, d_index, p_output, index_size, slice_size, functor);
|
|
dst_count->Resize({input_size});
|
|
dev_ctx.template Alloc<int32_t>(dst_count);
|
|
int* p_dst_count = dst_count->data<int>();
|
|
|
|
#ifdef PADDLE_WITH_HIP
|
|
hipMemset(p_dst_count, 0, input_size * sizeof(int));
|
|
#else
|
|
cudaMemsetAsync(p_dst_count, 0, input_size * sizeof(int), dev_ctx.stream());
|
|
#endif
|
|
|
|
int64_t grid_count = (index_size + block - 1) / block;
|
|
ComputeCountCUDAKernel<T, IndexT>
|
|
<<<grid_count, block, 0, dev_ctx.stream()>>>(
|
|
p_dst_count, d_index, index_size);
|
|
|
|
int64_t grid_mean_tmp = (input_size * slice_size + block - 1) / block;
|
|
int64_t grid_mean =
|
|
grid_mean_tmp < max_grid_dimx ? grid_mean_tmp : max_grid_dimx;
|
|
ManipulateMeanCUDAKernel<T><<<grid_mean, block, 0, dev_ctx.stream()>>>(
|
|
p_output, p_dst_count, input_size, slice_size);
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void SendURecvKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& src_index,
|
|
const DenseTensor& dst_index,
|
|
const std::string& reduce_op,
|
|
const IntArray& out_size,
|
|
DenseTensor* out,
|
|
DenseTensor* dst_count) {
|
|
auto index_type = src_index.dtype();
|
|
auto& out_size_data = out_size.GetData();
|
|
|
|
if (x.numel() == 0 || src_index.numel() == 0 || dst_index.numel() == 0) {
|
|
if (out_size_data[0] <= 0) {
|
|
out->Resize(x.dims());
|
|
} else {
|
|
out->Resize(out_size_data);
|
|
}
|
|
if (reduce_op == "MEAN") {
|
|
int64_t input_size =
|
|
out_size_data[0] <= 0 ? x.dims()[0] : out_size_data[0];
|
|
dst_count->Resize({input_size});
|
|
}
|
|
Full<T, Context>(dev_ctx, out->dims(), 0, out);
|
|
Full<int32_t, Context>(dev_ctx, dst_count->dims(), 0, dst_count);
|
|
return;
|
|
}
|
|
|
|
if (index_type == DataType::INT32) {
|
|
GraphSendRecvOpCUDAKernelLaunchHelper<Context, T, int32_t>(dev_ctx,
|
|
x,
|
|
src_index,
|
|
dst_index,
|
|
reduce_op,
|
|
out_size_data[0],
|
|
out,
|
|
dst_count);
|
|
} else if (index_type == DataType::INT64) {
|
|
GraphSendRecvOpCUDAKernelLaunchHelper<Context, T, int64_t>(dev_ctx,
|
|
x,
|
|
src_index,
|
|
dst_index,
|
|
reduce_op,
|
|
out_size_data[0],
|
|
out,
|
|
dst_count);
|
|
}
|
|
}
|
|
|
|
} // namespace phi
|
|
|
|
PD_REGISTER_KERNEL(send_u_recv,
|
|
GPU,
|
|
ALL_LAYOUT,
|
|
phi::SendURecvKernel,
|
|
float,
|
|
double,
|
|
int,
|
|
int64_t,
|
|
phi::float16) {
|
|
kernel->OutputAt(1).SetDataType(phi::DataType::INT32);
|
|
}
|