未验证 提交 78b30e97 编写于 作者: P Piotr Paturej 提交者: GitHub

[PHI] Migrate elementwise_div + all elementwise grad kernels (#48210)

* Migrate elementwise_div

* Migrate elementwise grad kernels
上级 27f49254
......@@ -25,11 +25,3 @@ REGISTER_OP_KERNEL(
dnnl::algorithm::binary_add>,
ops::EltwiseMKLDNNKernel<int8_t, dnnl::algorithm::binary_add>,
ops::EltwiseMKLDNNKernel<uint8_t, dnnl::algorithm::binary_add>)
REGISTER_OP_KERNEL(
elementwise_add_grad,
MKLDNN,
::paddle::platform::CPUPlace,
ops::EltwiseMKLDNNGradKernel<paddle::platform::bfloat16,
dnnl::algorithm::binary_add>,
ops::EltwiseMKLDNNGradKernel<float, dnnl::algorithm::binary_add>)
// 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/fluid/operators/elementwise/mkldnn/elementwise_mkldnn_op.h"
namespace ops = paddle::operators;
REGISTER_OP_KERNEL(elementwise_div,
MKLDNN,
paddle::platform::CPUPlace,
ops::EltwiseMKLDNNKernel<float, dnnl::algorithm::binary_div>,
ops::EltwiseMKLDNNKernel<paddle::platform::bfloat16,
dnnl::algorithm::binary_div>)
REGISTER_OP_KERNEL(
elementwise_div_grad,
MKLDNN,
paddle::platform::CPUPlace,
ops::EltwiseMKLDNNGradKernel<paddle::platform::bfloat16,
dnnl::algorithm::binary_div>,
ops::EltwiseMKLDNNGradKernel<float, dnnl::algorithm::binary_div>)
......@@ -25,11 +25,3 @@ REGISTER_OP_KERNEL(
dnnl::algorithm::binary_mul>,
ops::EltwiseMKLDNNKernel<int8_t, dnnl::algorithm::binary_mul>,
ops::EltwiseMKLDNNKernel<uint8_t, dnnl::algorithm::binary_mul>)
REGISTER_OP_KERNEL(
elementwise_mul_grad,
MKLDNN,
::paddle::platform::CPUPlace,
ops::EltwiseMKLDNNGradKernel<paddle::platform::bfloat16,
dnnl::algorithm::binary_mul>,
ops::EltwiseMKLDNNGradKernel<float, dnnl::algorithm::binary_mul>)
......@@ -25,11 +25,3 @@ REGISTER_OP_KERNEL(
dnnl::algorithm::binary_sub>,
ops::EltwiseMKLDNNKernel<int8_t, dnnl::algorithm::binary_sub>,
ops::EltwiseMKLDNNKernel<uint8_t, dnnl::algorithm::binary_sub>)
REGISTER_OP_KERNEL(
elementwise_sub_grad,
MKLDNN,
::paddle::platform::CPUPlace,
ops::EltwiseMKLDNNGradKernel<paddle::platform::bfloat16,
dnnl::algorithm::binary_sub>,
ops::EltwiseMKLDNNGradKernel<float, dnnl::algorithm::binary_sub>)
......@@ -414,3 +414,8 @@ PD_REGISTER_KERNEL(elementwise_pow,
float,
phi::dtype::float16) {}
#endif
#if defined PADDLE_WITH_MKLDNN
PD_REGISTER_KERNEL(
divide, OneDNN, ONEDNN, phi::DivideKernel, float, phi::dtype::bfloat16) {}
#endif
// 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/elementwise_add_grad_kernel.h"
#include "paddle/phi/kernels/elementwise_divide_grad_kernel.h"
#include "paddle/phi/kernels/elementwise_multiply_grad_kernel.h"
#include "paddle/phi/kernels/elementwise_subtract_grad_kernel.h"
#include "paddle/phi/backends/onednn/onednn_reuse.h"
#include "paddle/phi/core/kernel_registry.h"
namespace phi {
namespace funcs {
inline std::vector<int64_t> CalculateBroadcastedDims(
const phi::DenseTensor* x, const phi::DenseTensor* y) {
const auto src_tz = phi::vectorize(x->dims());
const auto dst_tz = phi::vectorize(y->dims());
std::vector<int64_t> dst_tz_ex(src_tz.size(), 1);
if (src_tz.size() == dst_tz.size()) {
for (size_t i = 0; i < src_tz.size(); i++) {
dst_tz_ex[i] = (src_tz[i] == dst_tz[i]) ? dst_tz[i] : 1;
}
} else {
size_t j = 0;
for (size_t i = 0; i < src_tz.size(); i++) {
dst_tz_ex[i] = (src_tz[i] != dst_tz[j]) ? 1 : dst_tz[j++];
if (j == dst_tz.size()) break;
}
}
return dst_tz_ex;
}
inline void AddSubNonBroadcast(ReorderOneDNNHandler* reorder_handler,
phi::DenseTensor* grad_tensor,
const std::shared_ptr<dnnl::memory>& src_memory,
const std::shared_ptr<dnnl::memory>& dst_memory,
const std::vector<float>& scales) {
dnnl::primitive_attr reorder_attr;
reorder_attr.set_output_scales(0, scales);
auto reorder_p =
reorder_handler->AcquireReorder(dst_memory, src_memory, reorder_attr);
paddle::platform::RecordEvent record_reorder(
"int_reorder",
paddle::platform::TracerEventType::UserDefined,
2,
paddle::platform::EventRole::kUniqueOp);
reorder_p->execute(
OneDNNContext::tls().get_stream(), *src_memory, *dst_memory);
}
template <typename T>
inline void BroadcastReduction(const Place& place,
const dnnl::engine& onednn_engine,
phi::DenseTensor* grad_tensor,
const phi::DenseTensor* dout,
const std::shared_ptr<dnnl::memory>& src_memory,
std::shared_ptr<dnnl::memory> dst_memory,
const std::vector<float>& scales,
const bool is_sub) {
dnnl::primitive_attr broadcast_reduction_attr;
// Broadcasting
if (is_sub) {
dnnl::post_ops po;
po.append_eltwise(1.0f, dnnl::algorithm::eltwise_linear, scales[0], 0);
broadcast_reduction_attr.set_post_ops(po);
}
ReductionOneDNNHandler<T> reduction_handler(
dnnl::algorithm::reduction_sum,
0.0f,
0.0f,
onednn_engine,
place,
dout,
grad_tensor,
CalculateBroadcastedDims(dout, grad_tensor),
broadcast_reduction_attr);
dst_memory = reduction_handler.AcquireDstMemory(grad_tensor);
auto reduction_p = reduction_handler.AcquireForwardPrimitive();
auto astream = OneDNNContext::tls().get_stream();
reduction_p->execute(astream,
{
{DNNL_ARG_SRC, *src_memory},
{DNNL_ARG_DST, *dst_memory},
});
astream.wait();
grad_tensor->set_mem_desc(dst_memory->get_desc().reshape(
phi::vectorize<int64_t>(grad_tensor->dims())));
}
} // namespace funcs
template <typename T, dnnl::algorithm BINARY_OP>
void ElementwiseGradKernel(const OneDNNContext& dev_ctx,
const DenseTensor& x,
const DenseTensor& y,
const DenseTensor* out,
const DenseTensor& dout,
int axis,
DenseTensor* dx,
DenseTensor* dy) {
const auto& onednn_engine = dev_ctx.GetEngine();
// oneDNN's binary is optimized for broadcasting y into x, so in other case
// we have to swap tensors to achieve optimal performance
bool swap_x_y = false;
auto* non_const_x = &x;
auto* non_const_y = &y;
if (x.numel() < y.numel()) {
std::swap(non_const_x, non_const_y);
std::swap(dx, dy);
swap_x_y = true;
}
std::vector<float> scales{1.0};
if (swap_x_y) {
scales[0] = (BINARY_OP == dnnl::algorithm::binary_add) ? 1 : -1;
}
auto tz = phi::vectorize<int64_t>(dout.dims());
funcs::ReorderOneDNNHandler reorder_handler(
tz, dout.dtype(), funcs::ToOneDNNDataType(dout.dtype()), onednn_engine);
auto reorder_src_memory = reorder_handler.AcquireSrcMemory(
dout.mem_desc(), funcs::to_void_cast(dout.data<T>()));
std::shared_ptr<dnnl::memory> dst_memory;
std::shared_ptr<dnnl::memory> broadcast_src_memory = reorder_src_memory;
auto& astream = OneDNNContext::tls().get_stream();
if (dx) {
// elementwise_add & elementwise_sub
if (BINARY_OP == dnnl::algorithm::binary_add ||
BINARY_OP == dnnl::algorithm::binary_sub) {
if (dout.dims() == dx->dims()) {
dst_memory = reorder_handler.AcquireDstMemory(
dx, dout.mem_desc(), dev_ctx.GetPlace());
AddSubNonBroadcast(
&reorder_handler, dx, reorder_src_memory, dst_memory, scales);
}
} else { // elementwise_mul & elementwise_div
funcs::BinaryOneDNNHandler<T> binary_handler(BINARY_OP,
axis,
onednn_engine,
dev_ctx.GetPlace(),
&dout,
non_const_y,
dx,
1.0f,
1.0f,
1.0f,
false);
const auto src_dout_memory = binary_handler.AcquireSrcMemory(&dout);
const auto src_y_memory =
binary_handler.AcquireSecondSrcMemory(non_const_y);
dst_memory = binary_handler.AcquireDstMemory(dx);
const auto binary_prim = binary_handler.AcquireForwardPrimitive();
const std::unordered_map<int, dnnl::memory> args = {
{DNNL_ARG_SRC_0, *src_dout_memory},
{DNNL_ARG_SRC_1, *src_y_memory},
{DNNL_ARG_DST, *dst_memory}};
binary_prim->execute(astream, args);
}
astream.wait();
if (dout.dims() != dx->dims()) {
funcs::BroadcastReduction<T>(dev_ctx.GetPlace(),
onednn_engine,
dx,
&dout,
broadcast_src_memory,
dst_memory,
scales,
BINARY_OP == dnnl::algorithm::binary_sub);
} else {
dx->set_mem_desc(dst_memory->get_desc());
}
}
if (dy) {
// elementwise_add & elementwise_sub
if (BINARY_OP == dnnl::algorithm::binary_add ||
BINARY_OP == dnnl::algorithm::binary_sub) {
if (dout.dims() == dy->dims()) {
dst_memory = reorder_handler.AcquireDstMemory(
dy, dout.mem_desc(), dev_ctx.GetPlace());
AddSubNonBroadcast(
&reorder_handler, dy, reorder_src_memory, dst_memory, scales);
}
} else { // elementwise_mul & elementwise_div
std::unordered_map<int, dnnl::memory> args;
std::shared_ptr<dnnl::binary> binary_prim;
std::shared_ptr<dnnl::memory> post_op_memory;
std::shared_ptr<dnnl::memory> src_0_memory;
std::shared_ptr<dnnl::memory> src_1_memory;
funcs::BinaryOneDNNHandler<T> binary_handler(dnnl::algorithm::binary_mul,
axis,
onednn_engine,
dev_ctx.GetPlace(),
&dout,
non_const_x,
nullptr,
1.0f,
1.0f,
1.0f,
false);
src_1_memory = binary_handler.AcquireSecondSrcMemory(non_const_x);
if (BINARY_OP == dnnl::algorithm::binary_div) {
funcs::BinaryOneDNNHandler<T> post_op_binary_handler(
dnnl::algorithm::binary_div,
axis,
onednn_engine,
dev_ctx.GetPlace(),
non_const_y,
non_const_y,
nullptr,
1.0f,
1.0f,
1.0f,
false);
post_op_memory = post_op_binary_handler.AcquireSrcMemory(non_const_y);
dnnl::post_ops po;
po.append_binary(dnnl::algorithm::binary_div,
post_op_memory->get_desc());
binary_handler =
funcs::BinaryOneDNNHandler<T>(dnnl::algorithm::binary_mul,
axis,
onednn_engine,
dev_ctx.GetPlace(),
&dout,
out,
nullptr,
-1.0f,
1.0f,
1.0f,
false,
po);
src_1_memory = binary_handler.AcquireSecondSrcMemory(out);
}
src_0_memory = binary_handler.AcquireSrcMemory(&dout);
const auto dst_dy_memory = (dout.dims() == dy->dims())
? binary_handler.AcquireDstMemory(dy)
: binary_handler.AcquireDstMemory();
binary_prim = binary_handler.AcquireForwardPrimitive();
args = {{DNNL_ARG_SRC_0, *src_0_memory},
{DNNL_ARG_SRC_1, *src_1_memory},
{DNNL_ARG_DST, *dst_dy_memory}};
if (BINARY_OP == dnnl::algorithm::binary_div)
args.insert({DNNL_ARG_ATTR_MULTIPLE_POST_OP(0) | DNNL_ARG_SRC_1,
*post_op_memory});
binary_prim->execute(astream, args);
broadcast_src_memory = dst_dy_memory;
dst_memory = dst_dy_memory;
}
astream.wait();
if (dout.dims() != dy->dims()) {
funcs::BroadcastReduction<T>(dev_ctx.GetPlace(),
onednn_engine,
dy,
&dout,
broadcast_src_memory,
dst_memory,
scales,
BINARY_OP == dnnl::algorithm::binary_sub);
} else {
dy->set_mem_desc(dst_memory->get_desc());
}
}
}
#define DEFINE_ONEDNN_ELEMENTWISE_GRAD_KERNEL(name, algorithm) \
template <typename T, typename Context> \
void name##GradKernel(const Context& dev_ctx, \
const DenseTensor& x, \
const DenseTensor& y, \
const DenseTensor& dout, \
int axis, \
DenseTensor* dx, \
DenseTensor* dy) { \
ElementwiseGradKernel<T, algorithm>( \
dev_ctx, x, y, nullptr, dout, axis, dx, dy); \
}
DEFINE_ONEDNN_ELEMENTWISE_GRAD_KERNEL(Add, dnnl::algorithm::binary_add)
DEFINE_ONEDNN_ELEMENTWISE_GRAD_KERNEL(Subtract, dnnl::algorithm::binary_sub)
DEFINE_ONEDNN_ELEMENTWISE_GRAD_KERNEL(Multiply, dnnl::algorithm::binary_mul)
template <typename T, typename Context>
void DivideGradKernel(const Context& dev_ctx,
const DenseTensor& x,
const DenseTensor& y,
const DenseTensor& out,
const DenseTensor& dout,
int axis,
DenseTensor* dx,
DenseTensor* dy) {
ElementwiseGradKernel<T, dnnl::algorithm::binary_div>(
dev_ctx, x, y, &out, dout, axis, dx, dy);
}
} // namespace phi
PD_REGISTER_KERNEL(
add_grad, OneDNN, ONEDNN, phi::AddGradKernel, float, phi::dtype::bfloat16) {
}
PD_REGISTER_KERNEL(subtract_grad,
OneDNN,
ONEDNN,
phi::SubtractGradKernel,
float,
phi::dtype::bfloat16) {}
PD_REGISTER_KERNEL(multiply_grad,
OneDNN,
ONEDNN,
phi::MultiplyGradKernel,
float,
phi::dtype::bfloat16) {}
PD_REGISTER_KERNEL(divide_grad,
OneDNN,
ONEDNN,
phi::DivideGradKernel,
float,
phi::dtype::bfloat16) {}
// 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/elementwise_add_kernel.h"
#include "paddle/phi/kernels/elementwise_divide_kernel.h"
#include "paddle/phi/kernels/elementwise_multiply_kernel.h"
#include "paddle/phi/kernels/elementwise_subtract_kernel.h"
#include "paddle/phi/backends/onednn/onednn_reuse.h"
#include "paddle/phi/core/kernel_registry.h"
namespace phi {
template <typename T, dnnl::algorithm BINARY_OP>
void ElementwiseKernel(const OneDNNContext& dev_ctx,
const DenseTensor& x,
const DenseTensor& y,
int axis,
DenseTensor* out) {
const auto& onednn_engine = dev_ctx.GetEngine();
float scale_x = dev_ctx.HasDnnAttr("Scale_x")
? PADDLE_GET_CONST(float, dev_ctx.GetDnnAttr("Scale_x"))
: 1;
float scale_y = dev_ctx.HasDnnAttr("Scale_y")
? PADDLE_GET_CONST(float, dev_ctx.GetDnnAttr("Scale_y"))
: 1;
float scale_out =
dev_ctx.HasDnnAttr("Scale_out")
? PADDLE_GET_CONST(float, dev_ctx.GetDnnAttr("Scale_out"))
: 1;
dnnl::post_ops post_operations;
funcs::AppendActivation(dev_ctx, post_operations);
auto* non_const_x = &x;
auto* non_const_y = &y;
funcs::BinaryOneDNNHandler<T> handler(BINARY_OP,
axis,
onednn_engine,
dev_ctx.GetPlace(),
non_const_x,
non_const_y,
out,
scale_x,
scale_y,
scale_out,
true,
post_operations);
// oneDNN's binary is optimized for broadcasting y into x, so in other case
// we have to swap tensors to achieve optimal performance
if (x.numel() < y.numel()) {
std::swap(non_const_x, non_const_y);
}
const auto src_x_memory = handler.AcquireSrcMemory(non_const_x);
const auto src_y_memory = handler.AcquireSecondSrcMemory(non_const_y);
// (jczaja) For Inplace src and dst should be the same memory object.
// So x should share buffer with z. But UT mechanics is testing inplace
// execution for this op not checking that x can be bradcasted to match in
// shape y tensor.
// This is wrong as when x is to be broadcasted then z(out) will match the
// shape of y which is bigger than x. Hence if x is smaller in shape than z
// and they share a buffer (of
// shape x) then this buffer is not big enough to hold result of elementwise
// operation.
const bool reuse_x_memory = non_const_x->numel() == out->numel() &&
non_const_x->IsSharedBufferWith(*out);
std::shared_ptr<dnnl::memory> dst_memory;
if (reuse_x_memory) {
dst_memory = src_x_memory;
// NOTE(chenfeiyu): when the output reuses memory from other tensor rather
// than allocate its own, it's still need to take care of its data type.
// Unfortunately, paddle's operator only infers the output' shape, but not
// the data type. Alloc<T> takes care of allocation and data type
// normally, but if the memory is already allocated and there is no need
// to re-allocate, it just set the data type. So this it added there to
// get the right data type.
dev_ctx.template Alloc<T>(out);
} else {
dst_memory = handler.AcquireDstMemory(out);
}
const auto binary_prim = handler.AcquireForwardPrimitive();
auto& astream = OneDNNContext::tls().get_stream();
const std::unordered_map<int, dnnl::memory> args = {
{DNNL_ARG_SRC_0, *src_x_memory},
{DNNL_ARG_SRC_1, *src_y_memory},
{DNNL_ARG_DST, *dst_memory}};
binary_prim->execute(astream, args);
astream.wait();
if (handler.use_broadcasting_hack == false) {
out->set_mem_desc(dst_memory->get_desc());
} else {
auto dims = dst_memory->get_desc().dims();
dims.insert(dims.begin(), non_const_x->dims()[0]);
dims[1] /= dims[0];
out->set_mem_desc(dst_memory->get_desc().reshape(dims));
}
}
#define DEFINE_ONEDNN_ELEMENTWISE_KERNEL(name, algorithm) \
template <typename T, typename Context> \
void name##RawKernel(const Context& dev_ctx, \
const DenseTensor& x, \
const DenseTensor& y, \
int axis, \
DenseTensor* out) { \
ElementwiseKernel<T, algorithm>(dev_ctx, x, y, axis, out); \
}
DEFINE_ONEDNN_ELEMENTWISE_KERNEL(Divide, dnnl::algorithm::binary_div)
} // namespace phi
PD_REGISTER_KERNEL(divide_raw,
OneDNN,
ONEDNN,
phi::DivideRawKernel,
float,
phi::dtype::bfloat16) {}
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册