p85971406
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

!47299 add shape cpu kernel 

Merge pull request !47299 from NaCN/add_shape_cpu_kernel
This commit is contained in:
i-robot 2023-02-02 06:52:25 +00:00 committed by Gitee
parent bfbdbc1627 5b4cebb7c9
commit 2952ee029a
No known key found for this signature in database
GPG Key ID: 173E9B9CA92EEF8F
4 changed files with 149 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
mindspore/ccsrc/backend/common/optimizer/dynamic_shape/dynamic_shape_helper.cc
View File

@ -74,30 +74,6 @@ void InferShapeForNopNode(const AnfNodePtr &input_node) {
}
}
bool InferShapeForDefiniteOutputNode(const CNodePtr &cnode) {
MS_EXCEPTION_IF_NULL(cnode);
if (!common::AnfAlgo::CheckPrimitiveType(cnode, prim::kPrimShape)) {
return false;
}
auto input_size = common::AnfAlgo::GetInputTensorNum(cnode);
if (input_size != 1) {
MS_LOG(EXCEPTION) << "Node only has one input: " << cnode->fullname_with_scope();
}
auto cur_shape = dynamic_cast<mindspore::abstract::Shape *>(cnode->Shape().get())->shape();
if (std::any_of(cur_shape.begin(), cur_shape.end(), [](int64_t x) { return x == kInvalidShape; })) {
return false;
}
std::vector<int64_t> output_shape = {static_cast<int64_t>(cur_shape.size())};
mindspore::abstract::BaseShapePtr shape = std::make_shared<mindspore::abstract::Shape>(output_shape);
// cppcheck-suppress unreadVariable
auto lock = AnfUtils::GetAbstractLock(cnode.get());
auto abstract = cnode->abstract();
MS_EXCEPTION_IF_NULL(abstract);
abstract->set_shape(shape);
return true;
}
TypeId GetSequenceType(const abstract::AbstractSequencePtr &seq_abs) {
auto elems = seq_abs->elements();
if (!elems[0]->isa<abstract::AbstractScalar>()) {
@ -265,10 +241,6 @@ void InferShape(const CNodePtr &cnode, std::map<uint32_t, tensor::TensorPtr> *de
MS_EXCEPTION_IF_NULL(depend_tensor_map);
MS_LOG(DEBUG) << "InferShape start, node:" << cnode->fullname_with_scope();
std::set<int64_t> depend_list = abstract::GetValueDependArgIndices(cnode);
auto ret = InferShapeForDefiniteOutputNode(cnode);
if (ret) {
return;
}
depend_tensor_map->clear();
auto &inputs = cnode->inputs();

6
mindspore/ccsrc/kernel/common_utils.h
View File

@ -460,14 +460,14 @@ BACKEND_EXPORT int64_t CalOutputTupleSize(const AnfNodePtr &node);
BACKEND_EXPORT void SetDynamicInputSizeAttr(const CNodePtr &cnode);
BACKEND_EXPORT bool IsDynamicParamKernel(const std::string &op_name);
template <typename Derived>
template <typename Derived, typename AddressType = AddressPtr>
class MatchKernelHelper {
public:
MatchKernelHelper() = default;
virtual ~MatchKernelHelper() = default;
using KernelRunFunc = std::function<bool(Derived *, const std::vector<AddressPtr> &, const std::vector<AddressPtr> &,
const std::vector<AddressPtr> &)>;
using KernelRunFunc = std::function<bool(Derived *, const std::vector<AddressType> &,
const std::vector<AddressType> &, const std::vector<AddressPtr> &)>;
virtual const std::vector<std::pair<KernelAttr, KernelRunFunc>> &GetFuncList() const = 0;
protected:

86
mindspore/ccsrc/plugin/device/cpu/kernel/shape_cpu_kernel.cc Normal file
View File

@ -0,0 +1,86 @@
/**
* Copyright 2019-2022 Huawei Technologies Co., Ltd
*
* 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 "plugin/device/cpu/kernel/shape_cpu_kernel.h"
#include "plugin/device/cpu/hal/device/cpu_device_address.h"
namespace mindspore {
namespace kernel {
namespace {
constexpr size_t kShapeInputsNum = 1;
constexpr size_t kShapeOutputsNum = 1;
} // namespace
bool ShapeCpuKernelMod::Init(const BaseOperatorPtr &base_operator, const std::vector<KernelTensorPtr> &inputs,
const std::vector<KernelTensorPtr> &outputs) {
MS_EXCEPTION_IF_NULL(base_operator);
kernel_name_ = base_operator->name();
CHECK_KERNEL_INPUTS_NUM(inputs.size(), kShapeInputsNum, kernel_name_);
CHECK_KERNEL_OUTPUTS_NUM(outputs.size(), kShapeOutputsNum, kernel_name_);
return MatchKernelFunc(base_operator, inputs, outputs);
}
int ShapeCpuKernelMod::Resize(const BaseOperatorPtr &base_operator, const std::vector<KernelTensorPtr> &inputs,
const std::vector<KernelTensorPtr> &outputs,
const std::map<uint32_t, tensor::TensorPtr> &) {
if (auto ret = KernelMod::Resize(base_operator, inputs, outputs); ret != KRET_OK) {
return ret;
}
input_shape_ = inputs.at(kIndex0)->GetShapeVector();
output_shape_ = outputs.at(kIndex0)->GetShapeVector();
if (output_shape_.size() != 1) {
MS_LOG(EXCEPTION) << "For '" << kernel_name_
<< "', the dimension of output must be 1-D, but got: " << output_shape_.size();
}
if (output_shape_[0] != SizeToLong(input_shape_.size())) {
MS_LOG(EXCEPTION) << "For '" << kernel_name_
<< "', 'output_shape[0]' must be equal to the dimension of input, but got 'output_shape[0]': "
<< output_shape_[0] << " and the dimension of input: " << input_shape_.size();
}
return KRET_OK;
}
bool ShapeCpuKernelMod::LaunchKernel(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &,
const std::vector<AddressPtr> &outputs) {
auto output_addr = reinterpret_cast<int64_t *>(outputs[0]->addr);
for (size_t i = 0; i < LongToSize(output_shape_[0]); ++i) {
output_addr[i] = input_shape_[i];
}
return true;
}
const std::vector<std::pair<KernelAttr, ShapeCpuKernelMod::KernelRunFunc>> &ShapeCpuKernelMod::GetFuncList() const {
static const std::vector<std::pair<KernelAttr, ShapeCpuKernelMod::KernelRunFunc>> func_list = {
{KernelAttr().AddInputAttr(kNumberTypeFloat16).AddOutputAttr(kObjectTypeTuple, kNumberTypeInt64),
&ShapeCpuKernelMod::LaunchKernel},
{KernelAttr().AddInputAttr(kNumberTypeFloat32).AddOutputAttr(kObjectTypeTuple, kNumberTypeInt64),
&ShapeCpuKernelMod::LaunchKernel},
{KernelAttr().AddInputAttr(kNumberTypeFloat64).AddOutputAttr(kObjectTypeTuple, kNumberTypeInt64),
&ShapeCpuKernelMod::LaunchKernel},
{KernelAttr().AddInputAttr(kNumberTypeInt16).AddOutputAttr(kObjectTypeTuple, kNumberTypeInt64),
&ShapeCpuKernelMod::LaunchKernel},
{KernelAttr().AddInputAttr(kNumberTypeInt32).AddOutputAttr(kObjectTypeTuple, kNumberTypeInt64),
&ShapeCpuKernelMod::LaunchKernel},
{KernelAttr().AddInputAttr(kNumberTypeInt64).AddOutputAttr(kObjectTypeTuple, kNumberTypeInt64),
&ShapeCpuKernelMod::LaunchKernel},
{KernelAttr().AddInputAttr(kNumberTypeBool).AddOutputAttr(kObjectTypeTuple, kNumberTypeInt64),
&ShapeCpuKernelMod::LaunchKernel}};
return func_list;
}
MS_KERNEL_FACTORY_REG(NativeCpuKernelMod, Shape, ShapeCpuKernelMod);
} // namespace kernel
} // namespace mindspore

60
mindspore/ccsrc/plugin/device/cpu/kernel/shape_cpu_kernel.h Normal file
View File

@ -0,0 +1,60 @@
/**
* Copyright 2019-2022 Huawei Technologies Co., Ltd
*
* 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.
*/
#ifndef MINDSPORE_CCSRC_PLUGIN_DEVICE_CPU_KERNEL_SHAPE_CPU_KERNEL_H_
#define MINDSPORE_CCSRC_PLUGIN_DEVICE_CPU_KERNEL_SHAPE_CPU_KERNEL_H_
#include <vector>
#include <memory>
#include <utility>
#include <map>
#include "plugin/device/cpu/kernel/cpu_kernel.h"
#include "plugin/factory/ms_factory.h"
namespace mindspore {
namespace kernel {
class ShapeCpuKernelMod : public NativeCpuKernelMod, public MatchKernelHelper<ShapeCpuKernelMod> {
public:
ShapeCpuKernelMod() = default;
~ShapeCpuKernelMod() override = default;
bool Launch(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs) override {
MS_EXCEPTION_IF_NULL(kernel_func_);
return kernel_func_(this, inputs, workspace, outputs);
}
int Resize(const BaseOperatorPtr &base_operator, const std::vector<KernelTensorPtr> &inputs,
const std::vector<KernelTensorPtr> &outputs, const std::map<uint32_t, tensor::TensorPtr> &) override;
bool Init(const BaseOperatorPtr &base_operator, const std::vector<KernelTensorPtr> &inputs,
const std::vector<KernelTensorPtr> &outputs) override;
const std::vector<std::pair<KernelAttr, KernelRunFunc>> &GetFuncList() const override;
protected:
std::vector<KernelAttr> GetOpSupport() override { return OpSupport(); }
bool LaunchKernel(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &workspace,
const std::vector<AddressPtr> &outputs);
private:
ShapeVector input_shape_;
ShapeVector output_shape_;
};
} // namespace kernel
} // namespace mindspore
#endif // MINDSPORE_CCSRC_PLUGIN_DEVICE_CPU_KERNEL_SHAPE_CPU_KERNEL_H_