Skip to content

P
Paddle

csdn_franckjun / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle

体验新版 GitCode，发现更多精彩内容 >>

未验证 提交 b5aab4f0 编写于 作者: Z zlsh80826 提交者： GitHub
浏览文件
操作

[Paddle-TRT] Add noexcept on methods inherited from TensorRT (#34157) 

* add trt noexcept definition

* add trt noexcept on trt plugin

* add trt noexcept on trt int8 calibrator

* remove noexcept on base serialize

* add trt noexcept on split plugin

* add trt noexcept on elementwise plugin

* add trt noexcept on prelu plugin

* add trt noexcept on pool plugin

* add trt noexcept on swish plugin

* add trt noexcept on gelu plugin

* add trt noexcept on layer norm plugin

* add trt noexcept on instance norm plugin

* add trt noexcept on emb eltwise layernorm plugin

* add trt noexcept on qkv2context plugin

* add trt noexcept on skip layernorm plugin

* add trt noexcept on slice plugin

* add trt noexcept on hard swish plugin

* add trt noexcept on stack plugin

* add trt noexcept on special slice plugin

* add trt noexcept on anchor generator plugin

* add trt noexcept on yolobox plugin

* add trt noexcept on roi align plugin

* add trt noexcept on gather nd plugin
上级 c0133e01
隐藏空白更改
内联 并排
Showing with 1143 addition and 942 deletion
paddle/fluid/inference/tensorrt/helper.h
浏览文件 @ b5aab4f0
......@@ -39,6 +39,12 @@ namespace tensorrt {
NV_TENSORRT_MAJOR * 1000 + NV_TENSORRT_MINOR * 100 + \
NV_TENSORRT_PATCH * 10 + NV_TENSORRT_BUILD
#if IS_TRT_VERSION_GE(8000)
#define TRT_NOEXCEPT noexcept
#else
#define TRT_NOEXCEPT
#endif
namespace dy = paddle::platform::dynload;
// TensorRT data type to size
......@@ -72,7 +78,8 @@ static int GetInferLibVersion() {
// A logger for create TensorRT infer builder.
class NaiveLogger : public nvinfer1::ILogger {
public:
void log(nvinfer1::ILogger::Severity severity, const char* msg) override {
void log(nvinfer1::ILogger::Severity severity,
const char* msg) TRT_NOEXCEPT override {
switch (severity) {
case Severity::kVERBOSE:
VLOG(3) << msg;
......@@ -105,7 +112,7 @@ class NaiveProfiler : public nvinfer1::IProfiler {
typedef std::pair<std::string, float> Record;
std::vector<Record> mProfile;
virtual void reportLayerTime(const char* layerName, float ms) {
virtual void reportLayerTime(const char* layerName, float ms) TRT_NOEXCEPT {
auto record =
std::find_if(mProfile.begin(), mProfile.end(),
[&](const Record& r) { return r.first == layerName; });
......
paddle/fluid/inference/tensorrt/plugin/anchor_generator_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -110,16 +110,18 @@ AnchorGeneratorPlugin::AnchorGeneratorPlugin(const void* data, size_t length) {
PrepareParamsOnDevice();
}
const char* AnchorGeneratorPlugin::getPluginType() const {
const char* AnchorGeneratorPlugin::getPluginType() const TRT_NOEXCEPT {
return "anchor_generator_plugin";
}
const char* AnchorGeneratorPlugin::getPluginVersion() const { return "1"; }
const char* AnchorGeneratorPlugin::getPluginVersion() const TRT_NOEXCEPT {
return "1";
}
int AnchorGeneratorPlugin::getNbOutputs() const { return 2; }
int AnchorGeneratorPlugin::getNbOutputs() const TRT_NOEXCEPT { return 2; }
nvinfer1::Dims AnchorGeneratorPlugin::getOutputDimensions(
int index, const nvinfer1::Dims* inputs, int nb_input_dims) {
int index, const nvinfer1::Dims* inputs, int nb_input_dims) TRT_NOEXCEPT {
nvinfer1::Dims dims{};
dims.nbDims = 4;
dims.d[0] = height_;
......@@ -130,20 +132,21 @@ nvinfer1::Dims AnchorGeneratorPlugin::getOutputDimensions(
}
bool AnchorGeneratorPlugin::supportsFormat(
nvinfer1::DataType type, nvinfer1::TensorFormat format) const {
nvinfer1::DataType type, nvinfer1::TensorFormat format) const TRT_NOEXCEPT {
// static shape plugin can't support different type between input/out
// it may cause addition overhead in half mode
return (type == data_type_ && format == nvinfer1::TensorFormat::kLINEAR);
}
size_t AnchorGeneratorPlugin::getWorkspaceSize(int max_batch_size) const {
size_t AnchorGeneratorPlugin::getWorkspaceSize(int max_batch_size) const
TRT_NOEXCEPT {
return 0;
}
template <typename T>
int AnchorGeneratorPlugin::enqueue_impl(int batch_size,
const void* const* inputs,
void** outputs, void* workspace,
void* const* outputs, void* workspace,
cudaStream_t stream) {
const int block = 512;
const int gen_anchor_grid = (box_num_ + block - 1) / block;
......@@ -169,15 +172,15 @@ int AnchorGeneratorPlugin::enqueue(int batch_size, const void* const* inputs,
#else
void* const* outputs, void* workspace,
#endif
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
return enqueue_impl<float>(batch_size, inputs, outputs, workspace, stream);
}
int AnchorGeneratorPlugin::initialize() { return 0; }
int AnchorGeneratorPlugin::initialize() TRT_NOEXCEPT { return 0; }
void AnchorGeneratorPlugin::terminate() {}
void AnchorGeneratorPlugin::terminate() TRT_NOEXCEPT {}
size_t AnchorGeneratorPlugin::getSerializationSize() const {
size_t AnchorGeneratorPlugin::getSerializationSize() const TRT_NOEXCEPT {
size_t serialize_size = 0;
serialize_size += SerializedSize(data_type_);
serialize_size += SerializedSize(anchor_sizes_);
......@@ -192,7 +195,7 @@ size_t AnchorGeneratorPlugin::getSerializationSize() const {
return serialize_size;
}
void AnchorGeneratorPlugin::serialize(void* buffer) const {
void AnchorGeneratorPlugin::serialize(void* buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, data_type_);
SerializeValue(&buffer, anchor_sizes_);
SerializeValue(&buffer, aspect_ratios_);
......@@ -205,28 +208,31 @@ void AnchorGeneratorPlugin::serialize(void* buffer) const {
SerializeValue(&buffer, box_num_);
}
void AnchorGeneratorPlugin::destroy() {}
void AnchorGeneratorPlugin::destroy() TRT_NOEXCEPT {}
void AnchorGeneratorPlugin::setPluginNamespace(const char* lib_namespace) {
void AnchorGeneratorPlugin::setPluginNamespace(const char* lib_namespace)
TRT_NOEXCEPT {
namespace_ = std::string(lib_namespace);
}
const char* AnchorGeneratorPlugin::getPluginNamespace() const {
const char* AnchorGeneratorPlugin::getPluginNamespace() const TRT_NOEXCEPT {
return namespace_.c_str();
}
nvinfer1::DataType AnchorGeneratorPlugin::getOutputDataType(
int index, const nvinfer1::DataType* input_type, int nb_inputs) const {
int index, const nvinfer1::DataType* input_type,
int nb_inputs) const TRT_NOEXCEPT {
return input_type[0];
}
bool AnchorGeneratorPlugin::isOutputBroadcastAcrossBatch(
int output_index, const bool* input_is_broadcast, int nb_inputs) const {
int output_index, const bool* input_is_broadcast,
int nb_inputs) const TRT_NOEXCEPT {
return true;
}
bool AnchorGeneratorPlugin::canBroadcastInputAcrossBatch(
int input_index) const {
bool AnchorGeneratorPlugin::canBroadcastInputAcrossBatch(int input_index) const
TRT_NOEXCEPT {
return false;
}
......@@ -236,9 +242,9 @@ void AnchorGeneratorPlugin::configurePlugin(
const nvinfer1::DataType* input_types,
const nvinfer1::DataType* output_types, const bool* input_is_broadcast,
const bool* output_is_broadcast, nvinfer1::PluginFormat float_format,
int max_batct_size) {}
int max_batct_size) TRT_NOEXCEPT {}
nvinfer1::IPluginV2Ext* AnchorGeneratorPlugin::clone() const {
nvinfer1::IPluginV2Ext* AnchorGeneratorPlugin::clone() const TRT_NOEXCEPT {
auto plugin = new AnchorGeneratorPlugin(
data_type_, anchor_sizes_, aspect_ratios_, stride_, variances_, offset_,
height_, width_, num_anchors_, box_num_);
......@@ -246,30 +252,32 @@ nvinfer1::IPluginV2Ext* AnchorGeneratorPlugin::clone() const {
return plugin;
}
void AnchorGeneratorPluginCreator::setPluginNamespace(
const char* lib_namespace) {
void AnchorGeneratorPluginCreator::setPluginNamespace(const char* lib_namespace)
TRT_NOEXCEPT {
namespace_ = std::string(lib_namespace);
}
const char* AnchorGeneratorPluginCreator::getPluginNamespace() const {
const char* AnchorGeneratorPluginCreator::getPluginNamespace() const
TRT_NOEXCEPT {
return namespace_.c_str();
}
const char* AnchorGeneratorPluginCreator::getPluginName() const {
const char* AnchorGeneratorPluginCreator::getPluginName() const TRT_NOEXCEPT {
return "anchor_generator_plugin";
}
const char* AnchorGeneratorPluginCreator::getPluginVersion() const {
const char* AnchorGeneratorPluginCreator::getPluginVersion() const
TRT_NOEXCEPT {
return "1";
}
const nvinfer1::PluginFieldCollection*
AnchorGeneratorPluginCreator::getFieldNames() {
AnchorGeneratorPluginCreator::getFieldNames() TRT_NOEXCEPT {
return &field_collection_;
}
nvinfer1::IPluginV2Ext* AnchorGeneratorPluginCreator::createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) {
const char* name, const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT {
const nvinfer1::PluginField* fields = fc->fields;
int type_id = -1;
std::vector<float> anchor_sizes, aspect_ratios, stride, variances;
......@@ -315,7 +323,8 @@ nvinfer1::IPluginV2Ext* AnchorGeneratorPluginCreator::createPlugin(
}
nvinfer1::IPluginV2Ext* AnchorGeneratorPluginCreator::deserializePlugin(
const char* name, const void* serial_data, size_t serial_length) {
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT {
auto plugin = new AnchorGeneratorPlugin(serial_data, serial_length);
plugin->setPluginNamespace(namespace_.c_str());
return plugin;
......@@ -374,7 +383,8 @@ AnchorGeneratorPluginDynamic::AnchorGeneratorPluginDynamic(void const* data,
PrepareParamsOnDevice();
}
nvinfer1::IPluginV2DynamicExt* AnchorGeneratorPluginDynamic::clone() const {
nvinfer1::IPluginV2DynamicExt* AnchorGeneratorPluginDynamic::clone() const
TRT_NOEXCEPT {
auto plugin = new AnchorGeneratorPluginDynamic(
data_type_, anchor_sizes_, aspect_ratios_, stride_, variances_, offset_,
num_anchors_);
......@@ -384,7 +394,7 @@ nvinfer1::IPluginV2DynamicExt* AnchorGeneratorPluginDynamic::clone() const {
nvinfer1::DimsExprs AnchorGeneratorPluginDynamic::getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) {
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT {
nvinfer1::DimsExprs ret{};
ret.nbDims = 4;
ret.d[0] = inputs[0].d[2]; // feature height
......@@ -396,7 +406,7 @@ nvinfer1::DimsExprs AnchorGeneratorPluginDynamic::getOutputDimensions(
bool AnchorGeneratorPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* inOut, int nbInputs,
int nbOutputs) {
int nbOutputs) TRT_NOEXCEPT {
// input can be any, doesn't matter
// anchor generator doesn't read input raw data, only need the shape info
auto type = inOut[pos].type;
......@@ -412,11 +422,12 @@ bool AnchorGeneratorPluginDynamic::supportsFormatCombination(
void AnchorGeneratorPluginDynamic::configurePlugin(
const nvinfer1::DynamicPluginTensorDesc* in, int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) {}
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) TRT_NOEXCEPT {}
size_t AnchorGeneratorPluginDynamic::getWorkspaceSize(
const nvinfer1::PluginTensorDesc* inputs, int nbInputs,
const nvinfer1::PluginTensorDesc* outputs, int nbOutputs) const {
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const TRT_NOEXCEPT {
return 0;
}
......@@ -449,7 +460,7 @@ int AnchorGeneratorPluginDynamic::enqueue_impl(
int AnchorGeneratorPluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc, const void* const* inputs,
void* const* outputs, void* workspace, cudaStream_t stream) {
void* const* outputs, void* workspace, cudaStream_t stream) TRT_NOEXCEPT {
assert(outputDesc[0].type == nvinfer1::DataType::kFLOAT);
assert(outputDesc[1].type == nvinfer1::DataType::kFLOAT);
return enqueue_impl<float>(inputDesc, outputDesc, inputs, outputs, workspace,
......@@ -457,21 +468,24 @@ int AnchorGeneratorPluginDynamic::enqueue(
}
nvinfer1::DataType AnchorGeneratorPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType* inputTypes, int nbInputs) const {
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT {
return inputTypes[0];
}
const char* AnchorGeneratorPluginDynamic::getPluginType() const {
const char* AnchorGeneratorPluginDynamic::getPluginType() const TRT_NOEXCEPT {
return "anchor_generator_plugin_dynamic";
}
int AnchorGeneratorPluginDynamic::getNbOutputs() const { return 2; }
int AnchorGeneratorPluginDynamic::getNbOutputs() const TRT_NOEXCEPT {
return 2;
}
int AnchorGeneratorPluginDynamic::initialize() { return 0; }
int AnchorGeneratorPluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
void AnchorGeneratorPluginDynamic::terminate() {}
void AnchorGeneratorPluginDynamic::terminate() TRT_NOEXCEPT {}
size_t AnchorGeneratorPluginDynamic::getSerializationSize() const {
size_t AnchorGeneratorPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
size_t serialize_size = 0;
serialize_size += SerializedSize(data_type_);
serialize_size += SerializedSize(anchor_sizes_);
......@@ -483,7 +497,7 @@ size_t AnchorGeneratorPluginDynamic::getSerializationSize() const {
return serialize_size;
}
void AnchorGeneratorPluginDynamic::serialize(void* buffer) const {
void AnchorGeneratorPluginDynamic::serialize(void* buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, data_type_);
SerializeValue(&buffer, anchor_sizes_);
SerializeValue(&buffer, aspect_ratios_);
......@@ -493,32 +507,35 @@ void AnchorGeneratorPluginDynamic::serialize(void* buffer) const {
SerializeValue(&buffer, num_anchors_);
}
void AnchorGeneratorPluginDynamic::destroy() {}
void AnchorGeneratorPluginDynamic::destroy() TRT_NOEXCEPT {}
void AnchorGeneratorPluginDynamicCreator::setPluginNamespace(
const char* lib_namespace) {
const char* lib_namespace) TRT_NOEXCEPT {
namespace_ = std::string(lib_namespace);
}
const char* AnchorGeneratorPluginDynamicCreator::getPluginNamespace() const {
const char* AnchorGeneratorPluginDynamicCreator::getPluginNamespace() const
TRT_NOEXCEPT {
return namespace_.c_str();
}
const char* AnchorGeneratorPluginDynamicCreator::getPluginName() const {
const char* AnchorGeneratorPluginDynamicCreator::getPluginName() const
TRT_NOEXCEPT {
return "anchor_generator_plugin_dynamic";
}
const char* AnchorGeneratorPluginDynamicCreator::getPluginVersion() const {
const char* AnchorGeneratorPluginDynamicCreator::getPluginVersion() const
TRT_NOEXCEPT {
return "1";
}
const nvinfer1::PluginFieldCollection*
AnchorGeneratorPluginDynamicCreator::getFieldNames() {
AnchorGeneratorPluginDynamicCreator::getFieldNames() TRT_NOEXCEPT {
return &field_collection_;
}
nvinfer1::IPluginV2Ext* AnchorGeneratorPluginDynamicCreator::createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) {
const char* name, const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT {
const nvinfer1::PluginField* fields = fc->fields;
int type_id = -1;
std::vector<float> anchor_sizes, aspect_ratios, stride, variances;
......@@ -555,7 +572,8 @@ nvinfer1::IPluginV2Ext* AnchorGeneratorPluginDynamicCreator::createPlugin(
}
nvinfer1::IPluginV2Ext* AnchorGeneratorPluginDynamicCreator::deserializePlugin(
const char* name, const void* serial_data, size_t serial_length) {
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT {
auto plugin = new AnchorGeneratorPluginDynamic(serial_data, serial_length);
plugin->setPluginNamespace(namespace_.c_str());
return plugin;
......
paddle/fluid/inference/tensorrt/plugin/anchor_generator_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -34,34 +34,35 @@ class AnchorGeneratorPlugin : public nvinfer1::IPluginV2Ext {
const int width, const int num_anchors, const int box_num);
AnchorGeneratorPlugin(const void* data, size_t length);
~AnchorGeneratorPlugin() override;
const char* getPluginType() const override;
const char* getPluginVersion() const override;
int getNbOutputs() const override;
const char* getPluginType() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
int getNbOutputs() const TRT_NOEXCEPT override;
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nb_input_dims) override;
bool supportsFormat(nvinfer1::DataType type,
nvinfer1::TensorFormat format) const override;
size_t getWorkspaceSize(int max_batch_size) const override;
int nb_input_dims) TRT_NOEXCEPT override;
bool supportsFormat(nvinfer1::DataType type, nvinfer1::TensorFormat format)
const TRT_NOEXCEPT override;
size_t getWorkspaceSize(int max_batch_size) const TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batch_size, const void* const* inputs, void** outputs,
#else
int enqueue(int batch_size, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
int initialize() override;
void terminate() override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
void destroy() override;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* input_type,
int nb_inputs) const override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
void destroy() TRT_NOEXCEPT override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_type,
int nb_inputs) const TRT_NOEXCEPT override;
bool isOutputBroadcastAcrossBatch(int output_index,
const bool* input_is_broadcast,
int nb_inputs) const override;
bool canBroadcastInputAcrossBatch(int input_index) const override;
int nb_inputs) const TRT_NOEXCEPT override;
bool canBroadcastInputAcrossBatch(int input_index) const
TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::Dims* input_dims, int nb_inputs,
const nvinfer1::Dims* output_dims, int nb_outputs,
const nvinfer1::DataType* input_types,
......@@ -69,13 +70,13 @@ class AnchorGeneratorPlugin : public nvinfer1::IPluginV2Ext {
const bool* input_is_broadcast,
const bool* output_is_broadcast,
nvinfer1::PluginFormat float_format,
int max_batct_size) override;
nvinfer1::IPluginV2Ext* clone() const override;
int max_batct_size) TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* clone() const TRT_NOEXCEPT override;
private:
template <typename T>
int enqueue_impl(int batch_size, const void* const* inputs, void** outputs,
void* workspace, cudaStream_t stream);
int enqueue_impl(int batch_size, const void* const* inputs,
void* const* outputs, void* workspace, cudaStream_t stream);
nvinfer1::DataType data_type_;
std::vector<float> anchor_sizes_;
std::vector<float> aspect_ratios_;
......@@ -97,16 +98,17 @@ class AnchorGeneratorPluginCreator : public nvinfer1::IPluginCreator {
public:
AnchorGeneratorPluginCreator() = default;
~AnchorGeneratorPluginCreator() override = default;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
const char* getPluginName() const override;
const char* getPluginVersion() const override;
const nvinfer1::PluginFieldCollection* getFieldNames() override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
const char* getPluginName() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override;
nvinfer1::IPluginV2Ext* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override;
const char* name,
const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override;
private:
std::string namespace_;
......@@ -127,35 +129,36 @@ class AnchorGeneratorPluginDynamic : public DynamicPluginTensorRT {
const int num_anchors);
AnchorGeneratorPluginDynamic(void const* data, size_t length);
~AnchorGeneratorPluginDynamic();
nvinfer1::IPluginV2DynamicExt* clone() const override;
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) override;
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override;
int nbOutputs) TRT_NOEXCEPT override;
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override;
int nbOutputs) const TRT_NOEXCEPT override;
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
const char* getPluginType() const override;
int getNbOutputs() const override;
int initialize() override;
void terminate() override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
void destroy() override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
const char* getPluginType() const TRT_NOEXCEPT override;
int getNbOutputs() const TRT_NOEXCEPT override;
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
void destroy() TRT_NOEXCEPT override;
private:
template <typename T>
......@@ -181,16 +184,17 @@ class AnchorGeneratorPluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
AnchorGeneratorPluginDynamicCreator() = default;
~AnchorGeneratorPluginDynamicCreator() override = default;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
const char* getPluginName() const override;
const char* getPluginVersion() const override;
const nvinfer1::PluginFieldCollection* getFieldNames() override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
const char* getPluginName() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override;
nvinfer1::IPluginV2Ext* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override;
const char* name,
const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override;
private:
std::string namespace_;
......
paddle/fluid/inference/tensorrt/plugin/elementwise_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -48,7 +48,7 @@ __global__ void elementwise_kernel(const size_t total, const T *x_data,
}
nvinfer1::Dims ElementWisePlugin::getOutputDimensions(
int index, const nvinfer1::Dims *input_dims, int num_inputs) {
int index, const nvinfer1::Dims *input_dims, int num_inputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"There is only one output in TRT elementwise "
"op plugin, but got output index: %d.",
......@@ -64,7 +64,7 @@ nvinfer1::Dims ElementWisePlugin::getOutputDimensions(
return input_dims[0];
}
int ElementWisePlugin::initialize() {
int ElementWisePlugin::initialize() TRT_NOEXCEPT {
PADDLE_ENFORCE_GT(dims_y_.nbDims, 0,
platform::errors::InvalidArgument(
"The dimension of input Y of TRT elementwise op plugin "
......@@ -120,7 +120,7 @@ int ElementWisePlugin::enqueue(int batch_size, const void *const *inputs,
#else
void *const *outputs, void *workspace,
#endif
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
const float *x = reinterpret_cast<const float *>(inputs[0]);
const float *y = reinterpret_cast<const float *>(inputs[1]);
float *out = reinterpret_cast<float *>(outputs[0]);
......@@ -147,26 +147,26 @@ int ElementWisePlugin::enqueue(int batch_size, const void *const *inputs,
// Dynamic Plugin below.
#if IS_TRT_VERSION_GE(6000)
int ElementwisePluginDynamic::initialize() { return 0; }
int ElementwisePluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
size_t ElementwisePluginDynamic::getSerializationSize() const {
size_t ElementwisePluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
return SerializedSize(type_.c_str()) + SerializedSize(axis_);
}
void ElementwisePluginDynamic::serialize(void *buffer) const {
void ElementwisePluginDynamic::serialize(void *buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, type_.c_str());
SerializeValue(&buffer, axis_);
}
nvinfer1::DimsExprs ElementwisePluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
return inputs[0];
}
bool ElementwisePluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -189,7 +189,8 @@ bool ElementwisePluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType ElementwisePluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0,
platform::errors::InvalidArgument(
"The Elementwise Plugin only has one input, so the "
......@@ -201,7 +202,7 @@ nvinfer1::DataType ElementwisePluginDynamic::getOutputDataType(
int ElementwisePluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc, const void *const *inputs,
void *const *outputs, void *workspace, cudaStream_t stream) {
void *const *outputs, void *workspace, cudaStream_t stream) TRT_NOEXCEPT {
auto x_dims = input_desc[0].dims;
auto y_dims = input_desc[1].dims;
int axis = (axis_ == -1) ? x_dims.nbDims - y_dims.nbDims : axis_;
......
paddle/fluid/inference/tensorrt/plugin/elementwise_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -48,33 +48,35 @@ class ElementWisePlugin : public PluginTensorRT {
DeserializeValue(&serial_data, &serial_length, &post_size_);
}
ElementWisePlugin* clone() const override {
ElementWisePlugin* clone() const TRT_NOEXCEPT override {
return new ElementWisePlugin(type_, dims_x_, dims_y_, axis_);
}
const char* getPluginType() const override { return "elementwise_plugin"; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "elementwise_plugin";
}
nvinfer1::Dims getOutputDimensions(int index,
const nvinfer1::Dims* input_dims,
int num_inputs) override;
int num_inputs) TRT_NOEXCEPT override;
int initialize() override;
int initialize() TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batch_size, const void* const* inputs, void** outputs,
#else
int enqueue(int batch_size, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream);
void* workspace, cudaStream_t stream) TRT_NOEXCEPT;
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize() + SerializedSize(type_.c_str()) +
SerializedSize(dims_x_) + SerializedSize(dims_y_) +
SerializedSize(axis_) + SerializedSize(prev_size_) +
SerializedSize(midd_size_) + SerializedSize(post_size_);
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, type_.c_str());
SerializeValue(&buffer, dims_x_);
......@@ -97,13 +99,15 @@ class ElementWisePlugin : public PluginTensorRT {
class ElementWisePluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "elementwise_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "elementwise_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new ElementWisePlugin(serial_data, serial_length);
}
};
......@@ -120,48 +124,49 @@ class ElementwisePluginDynamic : public DynamicPluginTensorRT {
type_ = std::string(elementwise_type);
DeserializeValue(&serialData, &serialLength, &axis_);
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new ElementwisePluginDynamic(type_, axis_);
}
const char* getPluginType() const override {
const char* getPluginType() const TRT_NOEXCEPT override {
return "elementwise_plugin_dynamic";
}
int getNbOutputs() const override { return 1; }
int initialize() override;
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
private:
std::string type_;
......@@ -171,33 +176,34 @@ class ElementwisePluginDynamic : public DynamicPluginTensorRT {
class ElementwisePluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
ElementwisePluginDynamicCreator() {}
const char* getPluginName() const override {
const char* getPluginName() const TRT_NOEXCEPT override {
return "elementwise_plugin_dynamic";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
const nvinfer1::PluginFieldCollection* getFieldNames() override {
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override {
return &field_collection_;
}
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override {
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override {
return nullptr;
}
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
auto plugin = new ElementwisePluginDynamic(serial_data, serial_length);
return plugin;
}
void setPluginNamespace(const char* lib_namespace) override {
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override {
plugin_namespace_ = lib_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/emb_eltwise_layernorm_plugin.cu
浏览文件 @ b5aab4f0
......@@ -128,7 +128,7 @@ template <typename T>
int EmbEltwiseLayernormPluginDynamicImpl<T>::enqueue(
const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc, const void *const *inputs,
void *const *outputs, void *workspace, cudaStream_t stream) {
void *const *outputs, void *workspace, cudaStream_t stream) TRT_NOEXCEPT {
auto id_dims = input_desc[0].dims;
int batch = id_dims.d[0];
int seq_len = id_dims.d[1];
......@@ -181,17 +181,19 @@ template class EmbEltwiseLayernormPluginDynamicImpl<float>;
template class EmbEltwiseLayernormPluginDynamicImpl<half>;
#endif
int EmbEltwiseLayernormPluginDynamic::initialize() {
int EmbEltwiseLayernormPluginDynamic::initialize() TRT_NOEXCEPT {
impl_->initialize();
return 0;
}
void EmbEltwiseLayernormPluginDynamic::terminate() { impl_->terminate(); }
void EmbEltwiseLayernormPluginDynamic::terminate() TRT_NOEXCEPT {
impl_->terminate();
}
nvinfer1::DimsExprs EmbEltwiseLayernormPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) { // NOLINT
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT { // NOLINT
PADDLE_ENFORCE_EQ(output_index, 0,
platform::errors::InvalidArgument(
"There is only one output of the EmbEltwiseLayernorm, "
......@@ -208,7 +210,7 @@ nvinfer1::DimsExprs EmbEltwiseLayernormPluginDynamic::getOutputDimensions(
bool EmbEltwiseLayernormPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -256,7 +258,8 @@ bool EmbEltwiseLayernormPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType EmbEltwiseLayernormPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(
index, 0, platform::errors::InvalidArgument(
"The EmbEltwiseLayernorm Plugin only has one input, so the "
......@@ -271,7 +274,7 @@ nvinfer1::DataType EmbEltwiseLayernormPluginDynamic::getOutputDataType(
int EmbEltwiseLayernormPluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc, const void *const *inputs,
void *const *outputs, void *workspace, cudaStream_t stream) {
void *const *outputs, void *workspace, cudaStream_t stream) TRT_NOEXCEPT {
impl_->enqueue(input_desc, output_desc, inputs, outputs, workspace, stream);
return cudaGetLastError() != cudaSuccess;
}
......
paddle/fluid/inference/tensorrt/plugin/emb_eltwise_layernorm_plugin.h
浏览文件 @ b5aab4f0
......@@ -68,7 +68,7 @@ class EmbEltwiseLayernormPluginDynamicImpl
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream);
cudaStream_t stream) TRT_NOEXCEPT;
void shareGPUData(const EmbEltwiseLayernormPluginDynamicImplBase* anthor);
private:
......@@ -189,7 +189,7 @@ class EmbEltwiseLayernormPluginDynamic : public DynamicPluginTensorRT {
}
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
auto ptr = new EmbEltwiseLayernormPluginDynamic(
embs_, bias_, scale_, emb_sizes_, bias_size_, scale_size_, hidden_size_,
eps_, with_fp16_);
......@@ -197,14 +197,14 @@ class EmbEltwiseLayernormPluginDynamic : public DynamicPluginTensorRT {
return ptr;
}
const char* getPluginType() const override {
const char* getPluginType() const TRT_NOEXCEPT override {
return "fused_embedding_eltwise_layernorm_plugin";
}
int getNbOutputs() const override { return 1; }
int initialize() override;
void terminate() override;
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
int sum_num = 0;
sum_num += SerializedSize(emb_sizes_);
......@@ -223,7 +223,7 @@ class EmbEltwiseLayernormPluginDynamic : public DynamicPluginTensorRT {
return sum_num;
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
SerializeValue(&buffer, emb_sizes_);
for (size_t i = 0; i < emb_sizes_.size(); i++) {
auto size = emb_sizes_[i];
......@@ -248,33 +248,34 @@ class EmbEltwiseLayernormPluginDynamic : public DynamicPluginTensorRT {
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* in_out,
int nb_inputs, int nb_outputs) override;
int nb_inputs,
int nb_outputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nb_inputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nb_outputs) override {}
int nb_outputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nb_inputs,
const nvinfer1::PluginTensorDesc* outputs,
int nb_outputs) const override {
int nb_outputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* input_types,
int nb_inputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT override;
void destroy() override {
void destroy() TRT_NOEXCEPT override {
if (own_host_buff_) {
for (auto ptr : embs_) {
delete[] ptr;
......@@ -310,32 +311,33 @@ class EmbEltwiseLayernormPluginDynamicCreator
: public nvinfer1::IPluginCreator {
public:
EmbEltwiseLayernormPluginDynamicCreator() {}
const char* getPluginName() const override {
const char* getPluginName() const TRT_NOEXCEPT override {
return "fused_embedding_eltwise_layernorm_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
const nvinfer1::PluginFieldCollection* getFieldNames() override {
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override {
return &field_collection_;
}
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override {
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override {
return nullptr;
}
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new EmbEltwiseLayernormPluginDynamic(serial_data, serial_length);
}
void setPluginNamespace(const char* lib_namespace) override {
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override {
plugin_namespace_ = lib_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/gather_nd_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -58,19 +58,19 @@ __global__ void GatherNdCUDAKernel(const T* input, const int32_t* input_dims,
}
}
int GatherNdPluginDynamic::initialize() { return 0; }
int GatherNdPluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
size_t GatherNdPluginDynamic::getSerializationSize() const {
size_t GatherNdPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
return SerializedSize(with_fp16_);
}
void GatherNdPluginDynamic::serialize(void* buffer) const {
void GatherNdPluginDynamic::serialize(void* buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, with_fp16_);
}
nvinfer1::DimsExprs GatherNdPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) {
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(
nb_inputs, 2,
platform::errors::InvalidArgument(
......@@ -100,7 +100,7 @@ nvinfer1::DimsExprs GatherNdPluginDynamic::getOutputDimensions(
bool GatherNdPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of gather_nd plugin should not be nullptr."));
......@@ -134,14 +134,15 @@ bool GatherNdPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType GatherNdPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType* input_types, int nb_inputs) const {
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT {
return input_types[0];
}
int GatherNdPluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc, const void* const* inputs,
void* const* outputs, void* workspace, cudaStream_t stream) {
void* const* outputs, void* workspace, cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
auto index_dims = input_desc[1].dims;
auto input_dims_size = input_dims.nbDims;
......
paddle/fluid/inference/tensorrt/plugin/gather_nd_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -35,46 +35,49 @@ class GatherNdPluginDynamic : public DynamicPluginTensorRT {
DeserializeValue(&serial_data, &serial_length, &with_fp16_);
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new GatherNdPluginDynamic(with_fp16_);
}
const char* getPluginType() const override { return "gather_nd_plugin"; }
int getNbOutputs() const override { return 1; }
int initialize() override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "gather_nd_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) override;
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override {
void destroy() TRT_NOEXCEPT override {
if (input_dims_data_) {
cudaFree(input_dims_data_);
}
......@@ -88,31 +91,34 @@ class GatherNdPluginDynamic : public DynamicPluginTensorRT {
class GatherNdPluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
GatherNdPluginDynamicCreator() {}
const char* getPluginName() const override { return "gather_nd_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "gather_nd_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
const nvinfer1::PluginFieldCollection* getFieldNames() override {
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override {
return &field_collection_;
}
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override {
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override {
return nullptr;
}
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
auto plugin = new GatherNdPluginDynamic(serial_data, serial_length);
return plugin;
}
void setPluginNamespace(const char* lib_namespace) override {
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override {
plugin_namespace_ = lib_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/gelu_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -30,8 +30,8 @@ static const float kAT = 0.5;
static const float kBT = 0.7978845608028654; // sqrt(2.0/M_PI)
static const float kCT = 0.035677408136300125; // 0.044715 * sqrt(2.0/M_PI)
bool GeluPlugin::supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const {
bool GeluPlugin::supportsFormat(
nvinfer1::DataType type, nvinfer1::PluginFormat format) const TRT_NOEXCEPT {
if (with_fp16_) {
return ((type == nvinfer1::DataType::kFLOAT ||
type == nvinfer1::DataType::kHALF) &&
......@@ -44,7 +44,7 @@ bool GeluPlugin::supportsFormat(nvinfer1::DataType type,
nvinfer1::Dims GeluPlugin::getOutputDimensions(int index,
const nvinfer1::Dims* in_dims,
int nb_inputs) {
int nb_inputs) TRT_NOEXCEPT {
assert(nb_inputs == 1);
assert(index < this->getNbOutputs());
nvinfer1::Dims const& input_dims = in_dims[0];
......@@ -96,7 +96,8 @@ int GeluPlugin::enqueue(int batch_size, const void* const* inputs,
#if IS_TRT_VERSION_LT(8000)
void** outputs, void*, cudaStream_t stream) {
#else
void* const* outputs, void*, cudaStream_t stream) {
void* const* outputs, void*,
cudaStream_t stream) TRT_NOEXCEPT {
#endif
const auto& input_dims = this->getInputDims(0);
int num = batch_size;
......@@ -132,13 +133,13 @@ int GeluPlugin::enqueue(int batch_size, const void* const* inputs,
nvinfer1::DimsExprs GeluPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) {
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT {
return inputs[0];
}
bool GeluPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -167,7 +168,8 @@ bool GeluPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType GeluPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType* input_types, int nb_inputs) const {
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"The Gelu Plugin only has one input, so the "
"index value should be 0, but get %d.",
......@@ -178,7 +180,8 @@ nvinfer1::DataType GeluPluginDynamic::getOutputDataType(
int GeluPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc,
const void* const* inputs, void* const* outputs,
void* workspace, cudaStream_t stream) {
void* workspace,
cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
size_t num = ProductDim(input_dims);
const int block_size = 256;
......
paddle/fluid/inference/tensorrt/plugin/gelu_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -35,40 +35,48 @@ class GeluPlugin : public PluginTensorRT {
}
~GeluPlugin() {}
GeluPlugin* clone() const override { return new GeluPlugin(with_fp16_); }
GeluPlugin* clone() const TRT_NOEXCEPT override {
return new GeluPlugin(with_fp16_);
}
const char* getPluginType() const override { return "gelu_plugin"; }
int getNbOutputs() const override { return 1; }
int initialize() override { return 0; }
bool supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "gelu_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override { return 0; }
bool supportsFormat(nvinfer1::DataType type, nvinfer1::PluginFormat format)
const TRT_NOEXCEPT override;
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nb_input_dims) override;
int nb_input_dims) TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batch_size, const void* const* inputs, void** outputs,
#else
int enqueue(int batch_size, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize();
}
// TRT will call this func to serialize the configuration of TRT
// It should not be called by users.
void serialize(void* buffer) const override { serializeBase(buffer); }
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
}
};
class GeluPluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "gelu_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "gelu_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new GeluPlugin(serial_data, serial_length);
}
};
......@@ -83,61 +91,66 @@ class GeluPluginDynamic : public DynamicPluginTensorRT {
}
~GeluPluginDynamic() {}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new GeluPluginDynamic(with_fp16_);
}
const char* getPluginType() const override { return "gelu_plugin_dynamic"; }
int getNbOutputs() const override { return 1; }
int initialize() override { return 0; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "gelu_plugin_dynamic";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override { return 0; }
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return SerializedSize(with_fp16_);
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
SerializeValue(&buffer, with_fp16_);
}
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* in_out,
int nb_inputs, int nb_outputs) override;
int nb_inputs,
int nb_outputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nb_inputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nb_outputs) override {}
int nb_outputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nb_inputs,
const nvinfer1::PluginTensorDesc* outputs,
int nb_outputs) const override {
int nb_outputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* input_types,
int nb_inputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
};
class GeluPluginDynamicCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "gelu_plugin_dynamic"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "gelu_plugin_dynamic";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
auto plugin = new GeluPluginDynamic(serial_data, serial_length);
return plugin;
}
......
paddle/fluid/inference/tensorrt/plugin/hard_swish_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -22,7 +22,7 @@ namespace tensorrt {
namespace plugin {
nvinfer1::Dims HardSwishPlugin::getOutputDimensions(
int index, const nvinfer1::Dims* in_dims, int nb_inputs) {
int index, const nvinfer1::Dims* in_dims, int nb_inputs) TRT_NOEXCEPT {
assert(nb_inputs == 1);
assert(index < this->getNbOutputs());
nvinfer1::Dims const& input_dims = in_dims[0];
......@@ -54,7 +54,8 @@ int HardSwishPlugin::enqueue(int batch_size, const void* const* inputs,
#if IS_TRT_VERSION_LT(8000)
void** outputs, void*, cudaStream_t stream) {
#else
void* const* outputs, void*, cudaStream_t stream) {
void* const* outputs, void*,
cudaStream_t stream) TRT_NOEXCEPT {
#endif
const auto& input_dims = this->getInputDims(0);
int num = batch_size;
......
paddle/fluid/inference/tensorrt/plugin/hard_swish_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -40,30 +40,32 @@ class HardSwishPlugin : public PluginTensorRT {
}
~HardSwishPlugin() {}
HardSwishPlugin* clone() const override {
HardSwishPlugin* clone() const TRT_NOEXCEPT override {
return new HardSwishPlugin(threshold_, scale_, offset_);
}
const char* getPluginType() const override { return "hard_swish_plugin"; }
int getNbOutputs() const override { return 1; }
int initialize() override { return 0; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "hard_swish_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override { return 0; }
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nbInputDims) override;
int nbInputDims) TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batchSize, const void* const* inputs, void** outputs,
#else
int enqueue(int batchSize, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize() + SerializedSize(threshold_) +
SerializedSize(scale_) + SerializedSize(offset_);
}
// TRT will call this func to serialize the configuration of TRT
// It should not be called by users.
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, threshold_);
SerializeValue(&buffer, scale_);
......@@ -78,13 +80,15 @@ class HardSwishPlugin : public PluginTensorRT {
class HardSwishPluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "hard_swish_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "hard_swish_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new HardSwishPlugin(serial_data, serial_length);
}
};
......
paddle/fluid/inference/tensorrt/plugin/instance_norm_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -39,10 +39,10 @@ cudnnStatus_t convert_trt2cudnn_dtype(nvinfer1::DataType trt_dtype,
return CUDNN_STATUS_SUCCESS;
}
int InstanceNormPlugin::initialize() { return 0; }
int InstanceNormPlugin::initialize() TRT_NOEXCEPT { return 0; }
nvinfer1::Dims InstanceNormPlugin::getOutputDimensions(
int index, const nvinfer1::Dims *inputDims, int nbInputs) {
int index, const nvinfer1::Dims *inputDims, int nbInputs) TRT_NOEXCEPT {
assert(nbInputs == 1);
assert(index < this->getNbOutputs());
nvinfer1::Dims const &input_dims = inputDims[0];
......@@ -50,8 +50,8 @@ nvinfer1::Dims InstanceNormPlugin::getOutputDimensions(
return output_dims;
}
bool InstanceNormPlugin::supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const {
bool InstanceNormPlugin::supportsFormat(
nvinfer1::DataType type, nvinfer1::PluginFormat format) const TRT_NOEXCEPT {
return ((type == nvinfer1::DataType::kFLOAT ||
type == nvinfer1::DataType::kHALF) &&
(format == nvinfer1::PluginFormat::kLINEAR));
......@@ -63,7 +63,7 @@ int InstanceNormPlugin::enqueue(int batch_size, const void *const *inputs,
#else
void *const *outputs, void *workspace,
#endif
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
const auto &input_dims = this->getInputDims(0);
PADDLE_ENFORCE_EQ(input_dims.nbDims, 3,
......
paddle/fluid/inference/tensorrt/plugin/instance_norm_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -39,7 +39,7 @@ class InstanceNormPlugin : public PluginTensorRT {
cudnnTensorDescriptor_t x_desc_, y_desc_, b_desc_;
public:
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize() + SerializedSize(eps_) +
SerializedSize(scale_) + SerializedSize(bias_);
}
......@@ -47,7 +47,7 @@ class InstanceNormPlugin : public PluginTensorRT {
// TRT will call this func when we need to serialize the configuration of
// tensorrt.
// It should not be called by users.
void serialize(void *buffer) const override {
void serialize(void *buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, eps_);
SerializeValue(&buffer, scale_);
......@@ -89,37 +89,41 @@ class InstanceNormPlugin : public PluginTensorRT {
platform::dynload::cudnnDestroyTensorDescriptor(b_desc_);
}
int initialize() override;
int initialize() TRT_NOEXCEPT override;
InstanceNormPlugin *clone() const override {
InstanceNormPlugin *clone() const TRT_NOEXCEPT override {
return new InstanceNormPlugin(eps_, scale_, bias_);
}
const char *getPluginType() const override { return "instance_norm_plugin"; }
int getNbOutputs() const override { return 1; }
const char *getPluginType() const TRT_NOEXCEPT override {
return "instance_norm_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims *inputs,
int nbInputDims) override;
int nbInputDims) TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batchSize, const void *const *inputs, void **outputs,
#else
int enqueue(int batchSize, const void *const *inputs, void *const *outputs,
#endif
void *workspace, cudaStream_t stream) override;
void *workspace, cudaStream_t stream) TRT_NOEXCEPT override;
bool supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const override;
bool supportsFormat(nvinfer1::DataType type, nvinfer1::PluginFormat format)
const TRT_NOEXCEPT override;
};
class InstanceNormPluginCreator : public TensorRTPluginCreator {
public:
const char *getPluginName() const override { return "instance_norm_plugin"; }
const char *getPluginName() const TRT_NOEXCEPT override {
return "instance_norm_plugin";
}
const char *getPluginVersion() const override { return "1"; }
const char *getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2 *deserializePlugin(const char *name,
const void *serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2 *deserializePlugin(
const char *name, const void *serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new InstanceNormPlugin(serial_data, serial_length);
}
};
......
paddle/fluid/inference/tensorrt/plugin/layer_norm_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -24,10 +24,10 @@ namespace inference {
namespace tensorrt {
namespace plugin {
int LayerNormPlugin::initialize() { return 0; }
int LayerNormPlugin::initialize() TRT_NOEXCEPT { return 0; }
nvinfer1::Dims LayerNormPlugin::getOutputDimensions(
int index, const nvinfer1::Dims *inputDims, int nbInputs) {
int index, const nvinfer1::Dims *inputDims, int nbInputs) TRT_NOEXCEPT {
assert(nbInputs == 1);
assert(index < this->getNbOutputs());
nvinfer1::Dims const &input_dims = inputDims[0];
......@@ -41,10 +41,10 @@ int LayerNormPlugin::enqueue(int batch_size, const void *const *inputs,
#else
void *const *outputs, void *workspace,
#endif
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
const auto &input_dims = this->getInputDims(0);
const float *input = reinterpret_cast<const float *>(inputs[0]);
float *output = reinterpret_cast<float **>(outputs)[0];
float *output = reinterpret_cast<float *const *>(outputs)[0];
int begin_norm_axis = begin_norm_axis_;
float eps = eps_;
......@@ -91,13 +91,13 @@ int LayerNormPlugin::enqueue(int batch_size, const void *const *inputs,
nvinfer1::DimsExprs LayerNormPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputDims, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
return inputDims[0];
}
bool LayerNormPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of layernorm plugin shoule not be nullptr."));
......@@ -118,7 +118,8 @@ bool LayerNormPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType LayerNormPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0,
platform::errors::InvalidArgument(
"The LayerNormPlugin only has one input, so the "
......@@ -130,7 +131,7 @@ nvinfer1::DataType LayerNormPluginDynamic::getOutputDataType(
int LayerNormPluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc, const void *const *inputs,
void *const *outputs, void *workspace, cudaStream_t stream) {
void *const *outputs, void *workspace, cudaStream_t stream) TRT_NOEXCEPT {
const auto &input_dims = input_desc[0].dims;
int begin_norm_axis = begin_norm_axis_;
float eps = eps_;
......
paddle/fluid/inference/tensorrt/plugin/layer_norm_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -40,7 +40,7 @@ class LayerNormPlugin : public PluginTensorRT {
std::vector<int64_t> variance_shape_;
public:
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize() + SerializedSize(bias_) +
SerializedSize(scale_) + SerializedSize(begin_norm_axis_) +
SerializedSize(eps_) + SerializedSize(mean_shape_) +
......@@ -50,7 +50,7 @@ class LayerNormPlugin : public PluginTensorRT {
// TRT will call this func when we need to serialize the configuration of
// tensorrt.
// It should not be called by users.
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, bias_);
SerializeValue(&buffer, scale_);
......@@ -86,35 +86,39 @@ class LayerNormPlugin : public PluginTensorRT {
DeserializeValue(&serialData, &serialLength, &variance_shape_);
}
~LayerNormPlugin() {}
int initialize() override;
int initialize() TRT_NOEXCEPT override;
LayerNormPlugin* clone() const override {
LayerNormPlugin* clone() const TRT_NOEXCEPT override {
return new LayerNormPlugin(bias_.data(), bias_.size(), scale_.data(),
scale_.size(), begin_norm_axis_, eps_,
mean_shape_, variance_shape_);
}
const char* getPluginType() const override { return "layernorm_plugin"; }
int getNbOutputs() const override { return 1; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "layernorm_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nbInputDims) override;
int nbInputDims) TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batchSize, const void* const* inputs, void** outputs,
#else
int enqueue(int batchSize, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
};
class LayerNormPluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "layernorm_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "layernorm_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new LayerNormPlugin(serial_data, serial_length);
}
};
......@@ -145,25 +149,25 @@ class LayerNormPluginDynamic : public DynamicPluginTensorRT {
DeserializeValue(&serialData, &serialLength, &mean_shape_);
DeserializeValue(&serialData, &serialLength, &variance_shape_);
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new LayerNormPluginDynamic(bias_.data(), bias_.size(), scale_.data(),
scale_.size(), begin_norm_axis_, eps_,
mean_shape_, variance_shape_);
}
const char* getPluginType() const override {
const char* getPluginType() const TRT_NOEXCEPT override {
return "layernorm_plugin_dynamic";
}
int getNbOutputs() const override { return 1; }
int initialize() override { return 0; }
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override { return 0; }
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return SerializedSize(bias_) + SerializedSize(scale_) +
SerializedSize(begin_norm_axis_) + SerializedSize(eps_) +
SerializedSize(mean_shape_) + SerializedSize(variance_shape_);
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
SerializeValue(&buffer, bias_);
SerializeValue(&buffer, scale_);
SerializeValue(&buffer, begin_norm_axis_);
......@@ -174,33 +178,34 @@ class LayerNormPluginDynamic : public DynamicPluginTensorRT {
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
private:
std::vector<float> bias_;
......@@ -217,15 +222,15 @@ class LayerNormPluginDynamic : public DynamicPluginTensorRT {
class LayerNormPluginDynamicCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override {
const char* getPluginName() const TRT_NOEXCEPT override {
return "layernorm_plugin_dynamic";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new LayerNormPluginDynamic(serial_data, serial_length);
}
};
......
paddle/fluid/inference/tensorrt/plugin/pool_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -22,7 +22,7 @@ namespace plugin {
nvinfer1::Dims PoolPlugin::getOutputDimensions(int index,
const nvinfer1::Dims *inputDims,
int nbInputs) {
int nbInputs) TRT_NOEXCEPT {
assert(nbInputs == 1);
assert(index == 0);
assert(inputDims[0].nbDims == 3);
......@@ -37,15 +37,16 @@ nvinfer1::Dims PoolPlugin::getOutputDimensions(int index,
int PoolPlugin::enqueue(int batchSize, const void *const *inputs,
#if IS_TRT_VERSION_LT(8000)
void **outputs, void *workspace, cudaStream_t stream) {
void **outputs, void *workspace,
cudaStream_t stream) TRT_NOEXCEPT {
#else
void *const *outputs, void *workspace,
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
#endif
auto const &input_dims = this->getInputDims(0);
int input_size = 0;
float const *idata = reinterpret_cast<float const *>(inputs[0]);
float **odatas = reinterpret_cast<float **>(outputs);
float *const *odatas = reinterpret_cast<float *const *>(outputs);
std::vector<int> input_shape = input_shape_;
std::vector<int> output_shape = output_shape_;
......@@ -87,14 +88,14 @@ PoolPluginDynamic::PoolPluginDynamic(void const *serialData,
DeserializeValue(&serialData, &serialLength, &is_global_);
}
size_t PoolPluginDynamic::getSerializationSize() const {
size_t PoolPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
return SerializedSize(ceil_mode_) + SerializedSize(pool_type_.c_str()) +
SerializedSize(adaptive_) + SerializedSize(ksize_) +
SerializedSize(strides_) + SerializedSize(paddings_) +
SerializedSize(is_global_);
}
void PoolPluginDynamic::serialize(void *buffer) const {
void PoolPluginDynamic::serialize(void *buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, ceil_mode_);
SerializeValue(&buffer, pool_type_.c_str());
SerializeValue(&buffer, adaptive_);
......@@ -106,7 +107,7 @@ void PoolPluginDynamic::serialize(void *buffer) const {
nvinfer1::DimsExprs PoolPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(nb_inputs, 1,
platform::errors::InvalidArgument(
"The Split plugin should be only one input."));
......@@ -181,7 +182,7 @@ nvinfer1::DimsExprs PoolPluginDynamic::getOutputDimensions(
bool PoolPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -198,7 +199,8 @@ bool PoolPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType PoolPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"The Pool Plugin only has one input, so the "
"index value should be 0, but get %d.",
......@@ -212,7 +214,8 @@ nvinfer1::DataType PoolPluginDynamic::getOutputDataType(
int PoolPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc,
const void *const *inputs, void *const *outputs,
void *workspace, cudaStream_t stream) {
void *workspace,
cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
int n = input_dims.d[0];
int c = input_dims.d[1];
......
paddle/fluid/inference/tensorrt/plugin/pool_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -57,7 +57,7 @@ static std::vector<int> CalcOutputSize(const std::vector<int>& input_shape,
class PoolPlugin : public PluginTensorRT {
public:
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize() + SerializedSize(ceil_mode_) +
SerializedSize(pool_type_) + SerializedSize(adaptive_) +
SerializedSize(ksize_) + SerializedSize(strides_) +
......@@ -67,7 +67,7 @@ class PoolPlugin : public PluginTensorRT {
// TRT will call this func when we need to serialize the configuration of
// tensorrt.
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, ceil_mode_);
SerializeValue(&buffer, pool_type_);
......@@ -116,22 +116,24 @@ class PoolPlugin : public PluginTensorRT {
DeserializeValue(&serialData, &serialLength, &output_shape_);
}
PoolPlugin* clone() const override {
PoolPlugin* clone() const TRT_NOEXCEPT override {
return new PoolPlugin(ceil_mode_, pool_type_, adaptive_, ksize_, strides_,
paddings_, input_shape_);
}
const char* getPluginType() const override { return "pool_plugin"; }
int getNbOutputs() const override { return 1; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "pool_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nbInputDims) override;
int initialize() override { return 0; }
int nbInputDims) TRT_NOEXCEPT override;
int initialize() TRT_NOEXCEPT override { return 0; }
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batchSize, const void* const* inputs, void** outputs,
#else
int enqueue(int batchSize, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
private:
bool ceil_mode_;
......@@ -146,13 +148,15 @@ class PoolPlugin : public PluginTensorRT {
class PoolPluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "pool_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "pool_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new PoolPlugin(serial_data, serial_length);
}
};
......@@ -176,47 +180,50 @@ class PoolPluginDynamic : public DynamicPluginTensorRT {
PoolPluginDynamic(void const* serialData, size_t serialLength);
~PoolPluginDynamic() {}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new PoolPluginDynamic(ceil_mode_, pool_type_, adaptive_, ksize_,
strides_, paddings_, is_global_);
}
const char* getPluginType() const override { return "pool_plugin_dynamic"; }
int getNbOutputs() const override { return 1; }
int initialize() override { return 0; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "pool_plugin_dynamic";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override { return 0; }
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
private:
bool ceil_mode_;
......@@ -230,13 +237,15 @@ class PoolPluginDynamic : public DynamicPluginTensorRT {
class PoolPluginDynamicCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "pool_plugin_dynamic"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "pool_plugin_dynamic";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new PoolPluginDynamic(serial_data, serial_length);
}
};
......
paddle/fluid/inference/tensorrt/plugin/prelu_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -26,14 +26,14 @@ namespace inference {
namespace tensorrt {
namespace plugin {
int PReluPlugin::initialize() {
int PReluPlugin::initialize() TRT_NOEXCEPT {
cudaMalloc(&p_gpu_weight_, sizeof(float) * weight_.size());
cudaMemcpy(p_gpu_weight_, weight_.data(), weight_.size() * sizeof(float),
cudaMemcpyHostToDevice);
return 0;
}
void PReluPlugin::terminate() {
void PReluPlugin::terminate() TRT_NOEXCEPT {
if (p_gpu_weight_) {
cudaFree(p_gpu_weight_);
p_gpu_weight_ = nullptr;
......@@ -42,7 +42,7 @@ void PReluPlugin::terminate() {
nvinfer1::Dims PReluPlugin::getOutputDimensions(int index,
const nvinfer1::Dims *inputDims,
int nbInputs) {
int nbInputs) TRT_NOEXCEPT {
assert(nbInputs == 1);
assert(index < this->getNbOutputs());
nvinfer1::Dims const &input_dims = inputDims[0];
......@@ -55,14 +55,14 @@ int PReluPlugin::enqueue(int batch_size, const void *const *inputs,
void **outputs, void *workspace, cudaStream_t stream) {
#else
void *const *outputs, void *workspace,
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
#endif
// input dims is CHW.
const auto &input_dims = this->getInputDims(0);
const float *input = reinterpret_cast<const float *>(inputs[0]);
// const float *alpha = reinterpret_cast<const float *>(alpha_.get().values);
const float *alpha = p_gpu_weight_;
float *output = reinterpret_cast<float **>(outputs)[0];
float *const output = reinterpret_cast<float *const *>(outputs)[0];
int numel = 1;
for (int i = 0; i < input_dims.nbDims; i++) {
numel *= input_dims.d[i];
......@@ -86,13 +86,13 @@ int PReluPlugin::enqueue(int batch_size, const void *const *inputs,
#if IS_TRT_VERSION_GE(6000)
void PReluPluginDynamic::terminate() {
void PReluPluginDynamic::terminate() TRT_NOEXCEPT {
if (p_gpu_weight_) {
cudaFree(p_gpu_weight_);
}
}
int PReluPluginDynamic::initialize() {
int PReluPluginDynamic::initialize() TRT_NOEXCEPT {
cudaMalloc(&p_gpu_weight_, sizeof(float) * weight_.size());
cudaMemcpy(p_gpu_weight_, weight_.data(), weight_.size() * sizeof(float),
cudaMemcpyHostToDevice);
......@@ -107,24 +107,24 @@ PReluPluginDynamic::PReluPluginDynamic(void const *serialData,
mode_ = std::string(prelu_mode);
}
size_t PReluPluginDynamic::getSerializationSize() const {
size_t PReluPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
return SerializedSize(mode_.c_str()) + SerializedSize(weight_);
}
void PReluPluginDynamic::serialize(void *buffer) const {
void PReluPluginDynamic::serialize(void *buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, weight_);
SerializeValue(&buffer, mode_.c_str());
}
nvinfer1::DimsExprs PReluPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
return inputs[0];
}
bool PReluPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -141,7 +141,8 @@ bool PReluPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType PReluPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"The PRelu Plugin only has one input, so the "
"index value should be 0, but get %d.",
......@@ -155,7 +156,8 @@ nvinfer1::DataType PReluPluginDynamic::getOutputDataType(
int PReluPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc,
const void *const *inputs, void *const *outputs,
void *workspace, cudaStream_t stream) {
void *workspace,
cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
const float *alpha = p_gpu_weight_;
const float *input = static_cast<const float *>(inputs[0]);
......
paddle/fluid/inference/tensorrt/plugin/prelu_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -34,7 +34,7 @@ class PReluPlugin : public PluginTensorRT {
std::string mode_;
public:
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize() + SerializedSize(mode_.c_str()) +
SerializedSize(weight_);
}
......@@ -42,7 +42,7 @@ class PReluPlugin : public PluginTensorRT {
// TRT will call this func when we need to serialize the configuration of
// tensorrt.
// It should not be called by users.
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, weight_);
SerializeValue(&buffer, mode_.c_str());
......@@ -65,36 +65,40 @@ class PReluPlugin : public PluginTensorRT {
mode_ = std::string(prelu_mode);
}
~PReluPlugin() {}
int initialize() override;
void terminate() override;
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
PReluPlugin* clone() const override {
PReluPlugin* clone() const TRT_NOEXCEPT override {
auto* ptr = new PReluPlugin(weight_.data(), weight_.size(), mode_);
ptr->p_gpu_weight_ = p_gpu_weight_;
return ptr;
}
const char* getPluginType() const override { return "prelu_plugin"; }
int getNbOutputs() const override { return 1; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "prelu_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nbInputDims) override;
int nbInputDims) TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batchSize, const void* const* inputs, void** outputs,
#else
int enqueue(int batchSize, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
};
class PReluPluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "prelu_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "prelu_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new PReluPlugin(serial_data, serial_length);
}
};
......@@ -112,49 +116,52 @@ class PReluPluginDynamic : public DynamicPluginTensorRT {
PReluPluginDynamic(void const* serialData, size_t serialLength);
~PReluPluginDynamic() {}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
auto ptr = new PReluPluginDynamic(weight_.data(), weight_.size(), mode_);
ptr->p_gpu_weight_ = p_gpu_weight_;
return ptr;
}
const char* getPluginType() const override { return "prelu_plugin_dynamic"; }
int getNbOutputs() const override { return 1; }
int initialize() override;
void terminate() override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "prelu_plugin_dynamic";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
private:
std::vector<float> weight_;
......@@ -165,13 +172,15 @@ class PReluPluginDynamic : public DynamicPluginTensorRT {
class PReluPluginDynamicCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "prelu_plugin_dynamic"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "prelu_plugin_dynamic";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new PReluPluginDynamic(serial_data, serial_length);
}
};
......
paddle/fluid/inference/tensorrt/plugin/qkv_to_context_plugin.cu
浏览文件 @ b5aab4f0
......@@ -147,11 +147,11 @@ inline void TransposeQKV(const int batch, const int seq_len,
}
}
int QkvToContextPluginDynamic::initialize() { return 0; }
int QkvToContextPluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
nvinfer1::DimsExprs QkvToContextPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
// input[0], (B, S, 3 * N * H, 1, 1)
// input[1], (B, head_num, seq_len, seq_len)
// output, (B, seq_len, hidden)
......@@ -177,7 +177,7 @@ nvinfer1::DimsExprs QkvToContextPluginDynamic::getOutputDimensions(
bool QkvToContextPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -215,7 +215,8 @@ bool QkvToContextPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType QkvToContextPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(
index, 0, platform::errors::InvalidArgument(
"The EmbEltwiseLayernorm Plugin only has one input, so the "
......@@ -235,7 +236,7 @@ __global__ void apply_scale(T *data, T scale, int n) {
int QkvToContextPluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc, const void *const *inputs,
void *const *outputs, void *workspace, cudaStream_t stream) {
void *const *outputs, void *workspace, cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
int input_num = ProductDim(input_dims);
// input[0], (B, S, 3 * N * H, 1, 1)
......
paddle/fluid/inference/tensorrt/plugin/qkv_to_context_plugin.h
浏览文件 @ b5aab4f0
......@@ -59,21 +59,23 @@ class QkvToContextPluginDynamic : public DynamicPluginTensorRT {
DeserializeValue(&serial_data, &serial_length, &scale_);
DeserializeValue(&serial_data, &serial_length, &with_fp16_);
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new QkvToContextPluginDynamic(hidden_, head_number_, head_size_,
scale_, with_fp16_);
}
const char* getPluginType() const override { return "qkv_to_context_plugin"; }
int getNbOutputs() const override { return 1; }
int initialize() override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "qkv_to_context_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return SerializedSize(hidden_) + SerializedSize(head_number_) +
SerializedSize(head_size_) + SerializedSize(scale_) +
SerializedSize(with_fp16_);
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
SerializeValue(&buffer, hidden_);
SerializeValue(&buffer, head_number_);
SerializeValue(&buffer, head_size_);
......@@ -83,33 +85,34 @@ class QkvToContextPluginDynamic : public DynamicPluginTensorRT {
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* in_out,
int nb_inputs, int nb_outputs) override;
int nb_inputs,
int nb_outputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nb_inputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nb_outputs) override {}
int nb_outputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nb_inputs,
const nvinfer1::PluginTensorDesc* outputs,
int nb_outputs) const override {
int nb_outputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* input_types,
int nb_inputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
private:
int hidden_;
......@@ -121,31 +124,34 @@ class QkvToContextPluginDynamic : public DynamicPluginTensorRT {
class QkvToContextPluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
QkvToContextPluginDynamicCreator() {}
const char* getPluginName() const override { return "qkv_to_context_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "qkv_to_context_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
const nvinfer1::PluginFieldCollection* getFieldNames() override {
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override {
return &field_collection_;
}
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override {
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override {
return nullptr;
}
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
auto plugin = new QkvToContextPluginDynamic(serial_data, serial_length);
return plugin;
}
void setPluginNamespace(const char* lib_namespace) override {
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override {
plugin_namespace_ = lib_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/roi_align_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -200,7 +200,8 @@ RoiAlignPluginDynamic::RoiAlignPluginDynamic(void const* data, size_t length) {
smem_per_block_ = smem_per_block;
}
nvinfer1::IPluginV2DynamicExt* RoiAlignPluginDynamic::clone() const {
nvinfer1::IPluginV2DynamicExt* RoiAlignPluginDynamic::clone() const
TRT_NOEXCEPT {
auto* plugin =
new RoiAlignPluginDynamic(data_type_, pooled_height_, pooled_width_,
spatial_scale_, sampling_ratio_);
......@@ -210,7 +211,7 @@ nvinfer1::IPluginV2DynamicExt* RoiAlignPluginDynamic::clone() const {
nvinfer1::DimsExprs RoiAlignPluginDynamic::getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) {
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT {
nvinfer1::DimsExprs ret{};
ret.nbDims = 4;
ret.d[0] = inputs[1].d[0]; // roi
......@@ -222,7 +223,7 @@ nvinfer1::DimsExprs RoiAlignPluginDynamic::getOutputDimensions(
bool RoiAlignPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* inOut, int nbInputs,
int nbOutputs) {
int nbOutputs) TRT_NOEXCEPT {
if (inOut[pos].format != nvinfer1::TensorFormat::kLINEAR) {
return false;
}
......@@ -234,11 +235,12 @@ bool RoiAlignPluginDynamic::supportsFormatCombination(
void RoiAlignPluginDynamic::configurePlugin(
const nvinfer1::DynamicPluginTensorDesc* in, int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) {}
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) TRT_NOEXCEPT {}
size_t RoiAlignPluginDynamic::getWorkspaceSize(
const nvinfer1::PluginTensorDesc* inputs, int nbInputs,
const nvinfer1::PluginTensorDesc* outputs, int nbOutputs) const {
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const TRT_NOEXCEPT {
return 0;
}
......@@ -287,7 +289,7 @@ int RoiAlignPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs,
void* const* outputs, void* workspace,
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(outputDesc[0].type, data_type_,
platform::errors::InvalidArgument(
"TRT RoiAlignPluginDynamic expects outputDesc[0].type "
......@@ -302,21 +304,22 @@ int RoiAlignPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
}
nvinfer1::DataType RoiAlignPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType* inputTypes, int nbInputs) const {
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT {
return inputTypes[0];
}
const char* RoiAlignPluginDynamic::getPluginType() const {
const char* RoiAlignPluginDynamic::getPluginType() const TRT_NOEXCEPT {
return "roi_align_plugin_dynamic";
}
int RoiAlignPluginDynamic::getNbOutputs() const { return 1; }
int RoiAlignPluginDynamic::getNbOutputs() const TRT_NOEXCEPT { return 1; }
int RoiAlignPluginDynamic::initialize() { return 0; }
int RoiAlignPluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
void RoiAlignPluginDynamic::terminate() {}
void RoiAlignPluginDynamic::terminate() TRT_NOEXCEPT {}
size_t RoiAlignPluginDynamic::getSerializationSize() const {
size_t RoiAlignPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
size_t serialize_size = 0;
serialize_size += SerializedSize(data_type_);
serialize_size += SerializedSize(pooled_height_);
......@@ -326,7 +329,7 @@ size_t RoiAlignPluginDynamic::getSerializationSize() const {
return serialize_size;
}
void RoiAlignPluginDynamic::serialize(void* buffer) const {
void RoiAlignPluginDynamic::serialize(void* buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, data_type_);
SerializeValue(&buffer, pooled_height_);
SerializeValue(&buffer, pooled_width_);
......@@ -334,40 +337,43 @@ void RoiAlignPluginDynamic::serialize(void* buffer) const {
SerializeValue(&buffer, sampling_ratio_);
}
void RoiAlignPluginDynamic::destroy() {}
void RoiAlignPluginDynamic::destroy() TRT_NOEXCEPT {}
RoiAlignPluginDynamicCreator::RoiAlignPluginDynamicCreator() {}
void RoiAlignPluginDynamicCreator::setPluginNamespace(
const char* lib_namespace) {
void RoiAlignPluginDynamicCreator::setPluginNamespace(const char* lib_namespace)
TRT_NOEXCEPT {
namespace_ = std::string(lib_namespace);
}
const char* RoiAlignPluginDynamicCreator::getPluginNamespace() const {
const char* RoiAlignPluginDynamicCreator::getPluginNamespace() const
TRT_NOEXCEPT {
return namespace_.c_str();
}
const char* RoiAlignPluginDynamicCreator::getPluginName() const {
const char* RoiAlignPluginDynamicCreator::getPluginName() const TRT_NOEXCEPT {
return "roi_align_plugin_dynamic";
}
const char* RoiAlignPluginDynamicCreator::getPluginVersion() const {
const char* RoiAlignPluginDynamicCreator::getPluginVersion() const
TRT_NOEXCEPT {
return "1";
}
const nvinfer1::PluginFieldCollection*
RoiAlignPluginDynamicCreator::getFieldNames() {
RoiAlignPluginDynamicCreator::getFieldNames() TRT_NOEXCEPT {
return &field_collection_;
}
nvinfer1::IPluginV2Ext* RoiAlignPluginDynamicCreator::createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) {
const char* name, const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT {
const nvinfer1::PluginField* fields = fc->fields;
return nullptr;
}
nvinfer1::IPluginV2Ext* RoiAlignPluginDynamicCreator::deserializePlugin(
const char* name, const void* serial_data, size_t serial_length) {
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT {
auto plugin = new RoiAlignPluginDynamic(serial_data, serial_length);
plugin->setPluginNamespace(namespace_.c_str());
return plugin;
......
paddle/fluid/inference/tensorrt/plugin/roi_align_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -34,37 +34,38 @@ class RoiAlignPluginDynamic : public DynamicPluginTensorRT {
int sampling_ratio);
RoiAlignPluginDynamic(void const* data, size_t length);
~RoiAlignPluginDynamic() = default;
nvinfer1::IPluginV2DynamicExt* clone() const override;
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) override;
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override;
int nbOutputs) TRT_NOEXCEPT override;
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override;
int nbOutputs) const TRT_NOEXCEPT override;
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
const char* getPluginType() const override;
int getNbOutputs() const override;
int initialize() override;
void terminate() override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
void destroy() override;
const char* getPluginType() const TRT_NOEXCEPT override;
int getNbOutputs() const TRT_NOEXCEPT override;
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
void destroy() TRT_NOEXCEPT override;
private:
template <typename T, typename OutT>
......@@ -87,17 +88,18 @@ class RoiAlignPluginDynamicCreator : public nvinfer1::IPluginCreator {
RoiAlignPluginDynamicCreator();
~RoiAlignPluginDynamicCreator() override = default;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
const char* getPluginName() const override;
const char* getPluginVersion() const override;
const nvinfer1::PluginFieldCollection* getFieldNames() override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
const char* getPluginName() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override;
nvinfer1::IPluginV2Ext* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override;
const char* name,
const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override;
private:
std::string namespace_;
......
paddle/fluid/inference/tensorrt/plugin/skip_layernorm_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -29,7 +29,7 @@ namespace plugin {
// Dynamic Plugin below.
#if IS_TRT_VERSION_GE(6000)
int SkipLayerNormPluginDynamic::initialize() {
int SkipLayerNormPluginDynamic::initialize() TRT_NOEXCEPT {
cudaMalloc(&bias_gpu_, sizeof(float) * bias_size_);
cudaMemcpy(bias_gpu_, bias_.data(), bias_size_ * sizeof(float),
cudaMemcpyHostToDevice);
......@@ -39,7 +39,7 @@ int SkipLayerNormPluginDynamic::initialize() {
return 0;
}
void SkipLayerNormPluginDynamic::terminate() {
void SkipLayerNormPluginDynamic::terminate() TRT_NOEXCEPT {
if (bias_gpu_) {
cudaFree(bias_gpu_);
bias_gpu_ = nullptr;
......@@ -52,13 +52,13 @@ void SkipLayerNormPluginDynamic::terminate() {
nvinfer1::DimsExprs SkipLayerNormPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
return inputs[0];
}
bool SkipLayerNormPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -96,7 +96,8 @@ bool SkipLayerNormPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType SkipLayerNormPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0,
platform::errors::InvalidArgument(
"The SkipLayerNorm Plugin only has one input, so the "
......@@ -112,7 +113,7 @@ nvinfer1::DataType SkipLayerNormPluginDynamic::getOutputDataType(
int SkipLayerNormPluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc, const void *const *inputs,
void *const *outputs, void *workspace, cudaStream_t stream) {
void *const *outputs, void *workspace, cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
size_t num = ProductDim(input_dims);
int hidden = input_dims.d[2];
......
paddle/fluid/inference/tensorrt/plugin/skip_layernorm_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -39,6 +39,7 @@ class SkipLayerNormPluginDynamic : public DynamicPluginTensorRT {
std::copy(bias, bias + bias_size, bias_.data());
std::copy(scale, scale + scale_size, scale_.data());
}
SkipLayerNormPluginDynamic(void const* serial_data, size_t serial_length) {
DeserializeValue(&serial_data, &serial_length, &bias_);
DeserializeValue(&serial_data, &serial_length, &scale_);
......@@ -48,7 +49,7 @@ class SkipLayerNormPluginDynamic : public DynamicPluginTensorRT {
DeserializeValue(&serial_data, &serial_length, &with_fp16_);
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
auto ptr = new SkipLayerNormPluginDynamic(
bias_.data(), scale_.data(), bias_size_, scale_size_, eps_, with_fp16_);
ptr->bias_gpu_ = bias_gpu_;
......@@ -56,17 +57,19 @@ class SkipLayerNormPluginDynamic : public DynamicPluginTensorRT {
return ptr;
}
const char* getPluginType() const override { return "skip_layernorm_plugin"; }
int getNbOutputs() const override { return 1; }
int initialize() override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "skip_layernorm_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
size_t ser_size = SerializedSize(bias_) + SerializedSize(scale_) +
SerializedSize(bias_size_) + SerializedSize(scale_size_) +
SerializedSize(eps_) + SerializedSize(with_fp16_);
return ser_size;
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
SerializeValue(&buffer, bias_);
SerializeValue(&buffer, scale_);
SerializeValue(&buffer, bias_size_);
......@@ -77,34 +80,35 @@ class SkipLayerNormPluginDynamic : public DynamicPluginTensorRT {
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* in_out,
int nb_inputs, int nb_outputs) override;
int nb_inputs,
int nb_outputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nb_inputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nb_outputs) override {}
int nb_outputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nb_inputs,
const nvinfer1::PluginTensorDesc* outputs,
int nb_outputs) const override {
int nb_outputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* input_types,
int nb_inputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void terminate() override;
void destroy() TRT_NOEXCEPT override { delete this; }
void terminate() TRT_NOEXCEPT override;
private:
std::vector<float> bias_;
......@@ -122,31 +126,34 @@ class SkipLayerNormPluginDynamic : public DynamicPluginTensorRT {
class SkipLayerNormPluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
SkipLayerNormPluginDynamicCreator() {}
const char* getPluginName() const override { return "skip_layernorm_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "skip_layernorm_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
const nvinfer1::PluginFieldCollection* getFieldNames() override {
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override {
return &field_collection_;
}
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override {
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override {
return nullptr;
}
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
auto plugin = new SkipLayerNormPluginDynamic(serial_data, serial_length);
return plugin;
}
void setPluginNamespace(const char* lib_namespace) override {
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override {
plugin_namespace_ = lib_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/slice_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -75,12 +75,12 @@ SlicePlugin::~SlicePlugin() {
cudaFree(offset_temp_data_);
}
SlicePlugin *SlicePlugin::clone() const {
SlicePlugin *SlicePlugin::clone() const TRT_NOEXCEPT {
return new SlicePlugin(starts_, ends_, axes_, with_fp16_);
}
bool SlicePlugin::supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const {
bool SlicePlugin::supportsFormat(
nvinfer1::DataType type, nvinfer1::PluginFormat format) const TRT_NOEXCEPT {
if (with_fp16_) {
return ((type == nvinfer1::DataType::kFLOAT ||
type == nvinfer1::DataType::kHALF) &&
......@@ -91,9 +91,8 @@ bool SlicePlugin::supportsFormat(nvinfer1::DataType type,
}
}
nvinfer1::Dims SlicePlugin::getOutputDimensions(int index,
const nvinfer1::Dims *inputs,
int nb_input_dims) {
nvinfer1::Dims SlicePlugin::getOutputDimensions(
int index, const nvinfer1::Dims *inputs, int nb_input_dims) TRT_NOEXCEPT {
auto in_dims = inputs[0];
nvinfer1::Dims out_dims = in_dims;
for (size_t i = 0; i < axes_.size(); i++) {
......@@ -109,7 +108,7 @@ int SlicePlugin::enqueue(int batch_size, const void *const *inputs,
void **outputs, void *workspace, cudaStream_t stream) {
#else
void *const *outputs, void *workspace,
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
#endif
auto input_dims = getInputDims(0);
......@@ -187,13 +186,13 @@ int SlicePlugin::enqueue(int batch_size, const void *const *inputs,
return cudaGetLastError() != cudaSuccess;
}
size_t SlicePlugin::getSerializationSize() const {
size_t SlicePlugin::getSerializationSize() const TRT_NOEXCEPT {
return getBaseSerializationSize() + SerializedSize(getPluginType()) +
SerializedSize(starts_) + SerializedSize(ends_) +
SerializedSize(axes_);
}
void SlicePlugin::serialize(void *buffer) const {
void SlicePlugin::serialize(void *buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, getPluginType());
serializeBase(buffer);
SerializeValue(&buffer, starts_);
......@@ -222,23 +221,23 @@ SlicePluginDynamic::SlicePluginDynamic(void const *serialData,
cudaStreamCreate(&copy_stream_);
}
void SlicePluginDynamic::destroy() {
void SlicePluginDynamic::destroy() TRT_NOEXCEPT {
cudaStreamDestroy(copy_stream_);
cudaEventDestroy(copy_event_);
cudaFree(offset_temp_data_);
delete this;
}
int SlicePluginDynamic::initialize() { return 0; }
int SlicePluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
size_t SlicePluginDynamic::getSerializationSize() const {
size_t SlicePluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
size_t size = SerializedSize(starts_) + SerializedSize(ends_) +
SerializedSize(axes_) + SerializedSize(with_fp16_);
return size;
}
void SlicePluginDynamic::serialize(void *buffer) const {
void SlicePluginDynamic::serialize(void *buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, starts_);
SerializeValue(&buffer, ends_);
SerializeValue(&buffer, axes_);
......@@ -247,7 +246,7 @@ void SlicePluginDynamic::serialize(void *buffer) const {
nvinfer1::DimsExprs SlicePluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
auto in_dims = inputs[0];
nvinfer1::DimsExprs ret = in_dims;
// start, ends should greater 0
......@@ -261,7 +260,7 @@ nvinfer1::DimsExprs SlicePluginDynamic::getOutputDimensions(
bool SlicePluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -289,7 +288,8 @@ bool SlicePluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType SlicePluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"The Slice Plugin only has one input, so the "
"index value should be 0, but get %d.",
......@@ -304,7 +304,8 @@ nvinfer1::DataType SlicePluginDynamic::getOutputDataType(
int SlicePluginDynamic::enqueue(const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc,
const void *const *inputs, void *const *outputs,
void *workspace, cudaStream_t stream) {
void *workspace,
cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
auto out_dims = output_desc[0].dims;
auto num_dims = input_dims.nbDims;
......
paddle/fluid/inference/tensorrt/plugin/slice_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -35,27 +35,29 @@ class SlicePlugin : public PluginTensorRT {
// It should not be called by users.
SlicePlugin(void const* serial_data, size_t serial_length);
~SlicePlugin();
SlicePlugin* clone() const override;
SlicePlugin* clone() const TRT_NOEXCEPT override;
const char* getPluginType() const override { return "slice_plugin"; }
int getNbOutputs() const override { return 1; }
int initialize() override { return 0; }
bool supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "slice_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override { return 0; }
bool supportsFormat(nvinfer1::DataType type, nvinfer1::PluginFormat format)
const TRT_NOEXCEPT override;
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nb_input_dims) override;
int nb_input_dims) TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batch_size, const void* const* inputs, void** outputs,
#else
int enqueue(int batch_size, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
size_t getSerializationSize() const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
// TRT will call this func to serialize the configuration of TRT
// It should not be called by users.
void serialize(void* buffer) const override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
private:
std::vector<int> starts_;
......@@ -68,13 +70,15 @@ class SlicePlugin : public PluginTensorRT {
class SlicePluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "slice_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "slice_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new SlicePlugin(serial_data, serial_length);
}
};
......@@ -86,48 +90,51 @@ class SlicePluginDynamic : public DynamicPluginTensorRT {
explicit SlicePluginDynamic(std::vector<int> starts, std::vector<int> ends,
std::vector<int> axes, bool with_fp16);
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new SlicePluginDynamic(starts_, ends_, axes_, with_fp16_);
}
SlicePluginDynamic(void const* serialData, size_t serialLength);
const char* getPluginType() const override { return "slice_plugin_dynamic"; }
int getNbOutputs() const override { return 1; }
int initialize() override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "slice_plugin_dynamic";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override;
void destroy() TRT_NOEXCEPT override;
private:
std::vector<int> starts_;
......@@ -140,13 +147,15 @@ class SlicePluginDynamic : public DynamicPluginTensorRT {
class SlicePluginDynamicCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "slice_plugin_dynamic"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "slice_plugin_dynamic";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serialData,
size_t serialLength) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serialData,
size_t serialLength) TRT_NOEXCEPT override {
return new SlicePluginDynamic(serialData, serialLength);
}
};
......
paddle/fluid/inference/tensorrt/plugin/special_slice_plugin.cu
浏览文件 @ b5aab4f0
......@@ -30,28 +30,29 @@ SpecialSlicePluginDynamic::SpecialSlicePluginDynamic(void const* serial_data,
SpecialSlicePluginDynamic::~SpecialSlicePluginDynamic() {}
nvinfer1::IPluginV2DynamicExt* SpecialSlicePluginDynamic::clone() const {
nvinfer1::IPluginV2DynamicExt* SpecialSlicePluginDynamic::clone() const
TRT_NOEXCEPT {
return new SpecialSlicePluginDynamic();
}
const char* SpecialSlicePluginDynamic::getPluginType() const {
const char* SpecialSlicePluginDynamic::getPluginType() const TRT_NOEXCEPT {
return "special_slice_plugin";
}
int SpecialSlicePluginDynamic::getNbOutputs() const { return 1; }
int SpecialSlicePluginDynamic::getNbOutputs() const TRT_NOEXCEPT { return 1; }
int SpecialSlicePluginDynamic::initialize() { return 0; }
int SpecialSlicePluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
size_t SpecialSlicePluginDynamic::getSerializationSize() const {
size_t SpecialSlicePluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
size_t serialize_size = 0;
return serialize_size;
}
void SpecialSlicePluginDynamic::serialize(void* buffer) const {}
void SpecialSlicePluginDynamic::serialize(void* buffer) const TRT_NOEXCEPT {}
nvinfer1::DimsExprs SpecialSlicePluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) {
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT {
nvinfer1::DimsExprs output(inputs[0]);
output.nbDims++;
for (int i = output.nbDims - 1; i > 1; i--) {
......@@ -69,21 +70,22 @@ nvinfer1::DimsExprs SpecialSlicePluginDynamic::getOutputDimensions(
void SpecialSlicePluginDynamic::configurePlugin(
const nvinfer1::DynamicPluginTensorDesc* in, int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) {}
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) TRT_NOEXCEPT {}
size_t SpecialSlicePluginDynamic::getWorkspaceSize(
const nvinfer1::PluginTensorDesc* inputs, int nbInputs,
const nvinfer1::PluginTensorDesc* outputs, int nbOutputs) const {
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const TRT_NOEXCEPT {
return 0;
}
void SpecialSlicePluginDynamic::destroy() { delete this; }
void SpecialSlicePluginDynamic::destroy() TRT_NOEXCEPT { delete this; }
void SpecialSlicePluginDynamic::terminate() {}
void SpecialSlicePluginDynamic::terminate() TRT_NOEXCEPT {}
bool SpecialSlicePluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* desc, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
if (pos == 0) // slice tensor
return (desc[pos].type == nvinfer1::DataType::kHALF &&
desc[pos].format ==
......@@ -101,7 +103,8 @@ bool SpecialSlicePluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType SpecialSlicePluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType* input_types, int nb_inputs) const {
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"The index should be equal to 0"));
return input_types[0];
......@@ -120,7 +123,7 @@ __global__ void SpecialSliceKernel(const T* slice_input,
int SpecialSlicePluginDynamic::enqueue(
const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc, const void* const* inputs,
void* const* outputs, void* workspace, cudaStream_t stream) {
void* const* outputs, void* workspace, cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims; // (sum(S), 768, 1, 1)
auto out_dims = output_desc[0].dims; // (batch, 768, 1, 1)
......@@ -142,36 +145,40 @@ int SpecialSlicePluginDynamic::enqueue(
SpecialSlicePluginDynamicCreator::SpecialSlicePluginDynamicCreator() {}
const char* SpecialSlicePluginDynamicCreator::getPluginName() const {
const char* SpecialSlicePluginDynamicCreator::getPluginName() const
TRT_NOEXCEPT {
return "special_slice_plugin";
}
const char* SpecialSlicePluginDynamicCreator::getPluginVersion() const {
const char* SpecialSlicePluginDynamicCreator::getPluginVersion() const
TRT_NOEXCEPT {
return "1";
}
const nvinfer1::PluginFieldCollection*
SpecialSlicePluginDynamicCreator::getFieldNames() {
SpecialSlicePluginDynamicCreator::getFieldNames() TRT_NOEXCEPT {
return &field_collection_;
}
nvinfer1::IPluginV2* SpecialSlicePluginDynamicCreator::createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) {
const char* name, const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT {
return new SpecialSlicePluginDynamic();
}
nvinfer1::IPluginV2* SpecialSlicePluginDynamicCreator::deserializePlugin(
const char* name, const void* serial_data, size_t serial_length) {
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT {
auto plugin = new SpecialSlicePluginDynamic(serial_data, serial_length);
return plugin;
}
void SpecialSlicePluginDynamicCreator::setPluginNamespace(
const char* lib_namespace) {
const char* lib_namespace) TRT_NOEXCEPT {
plugin_namespace_ = lib_namespace;
}
const char* SpecialSlicePluginDynamicCreator::getPluginNamespace() const {
const char* SpecialSlicePluginDynamicCreator::getPluginNamespace() const
TRT_NOEXCEPT {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/special_slice_plugin.h
浏览文件 @ b5aab4f0
......@@ -31,37 +31,38 @@ class SpecialSlicePluginDynamic : public DynamicPluginTensorRT {
SpecialSlicePluginDynamic();
SpecialSlicePluginDynamic(void const* serial_data, size_t serial_length);
~SpecialSlicePluginDynamic();
nvinfer1::IPluginV2DynamicExt* clone() const override;
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) override;
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override;
int nbOutputs) TRT_NOEXCEPT override;
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override;
int nbOutputs) const TRT_NOEXCEPT override;
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
const char* getPluginType() const override;
int getNbOutputs() const override;
int initialize() override;
void terminate() override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
void destroy() override;
const char* getPluginType() const TRT_NOEXCEPT override;
int getNbOutputs() const TRT_NOEXCEPT override;
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
void destroy() TRT_NOEXCEPT override;
private:
int axis_;
......@@ -71,16 +72,17 @@ class SpecialSlicePluginDynamic : public DynamicPluginTensorRT {
class SpecialSlicePluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
SpecialSlicePluginDynamicCreator();
const char* getPluginName() const override;
const char* getPluginVersion() const override;
const nvinfer1::PluginFieldCollection* getFieldNames() override;
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override;
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
const char* getPluginName() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override;
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override;
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
private:
std::string plugin_namespace_;
......
paddle/fluid/inference/tensorrt/plugin/split_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -38,7 +38,7 @@ __device__ int upper_bound(T const* vals, int n, T const& key) {
}
nvinfer1::Dims SplitPlugin::getOutputDimensions(
int index, const nvinfer1::Dims* input_dims, int num_inputs) {
int index, const nvinfer1::Dims* input_dims, int num_inputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(num_inputs, 1,
platform::errors::InvalidArgument(
"Invalid number of inputs of split TRT plugin. "
......@@ -66,7 +66,7 @@ void SplitPlugin::shareData(const SplitPlugin* another) {
d_output_ptrs_.resize(another->d_output_ptrs_.size(), nullptr);
}
int SplitPlugin::initialize() {
int SplitPlugin::initialize() TRT_NOEXCEPT {
PADDLE_ENFORCE_LE(axis_, nvinfer1::Dims::MAX_DIMS,
platform::errors::InvalidArgument(
"Axis dimension exceeds max dimension in TensorRT. "
......@@ -98,7 +98,7 @@ int SplitPlugin::initialize() {
}
// nothing to release according to initialize
void SplitPlugin::terminate() {}
void SplitPlugin::terminate() TRT_NOEXCEPT {}
// The following part of the code refers to onnx-tensorrt
// https://github.com/onnx/onnx-tensorrt/blob/master/Split.cu
......@@ -129,7 +129,7 @@ int SplitPlugin::enqueue(int batchSize, const void* const* inputs,
void** outputs, void* workspace, cudaStream_t stream) {
#else
void* const* outputs, void* workspace,
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
#endif
const int* d_segment_offsets_ptr =
thrust::raw_pointer_cast(&d_segment_offsets_[0]);
......@@ -155,14 +155,14 @@ int SplitPlugin::enqueue(int batchSize, const void* const* inputs,
// Dynamic Plugin below.
#if IS_TRT_VERSION_GE(6000)
int SplitPluginDynamic::initialize() { return 0; }
int SplitPluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
size_t SplitPluginDynamic::getSerializationSize() const {
size_t SplitPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
return SerializedSize(axis_) + SerializedSize(output_length_) +
SerializedSize(with_fp16_);
}
void SplitPluginDynamic::serialize(void* buffer) const {
void SplitPluginDynamic::serialize(void* buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, axis_);
SerializeValue(&buffer, output_length_);
SerializeValue(&buffer, with_fp16_);
......@@ -170,7 +170,7 @@ void SplitPluginDynamic::serialize(void* buffer) const {
nvinfer1::DimsExprs SplitPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) {
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(nb_inputs, 1,
platform::errors::InvalidArgument(
"The Split plugin should be only one input."));
......@@ -188,7 +188,7 @@ nvinfer1::DimsExprs SplitPluginDynamic::getOutputDimensions(
bool SplitPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of split plugin should not be nullptr."));
......@@ -217,14 +217,16 @@ bool SplitPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType SplitPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType* input_types, int nb_inputs) const {
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT {
return input_types[0];
}
int SplitPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc,
const void* const* inputs, void* const* outputs,
void* workspace, cudaStream_t stream) {
void* workspace,
cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
int outer_rows = 1;
int inner_cols = 1;
......
paddle/fluid/inference/tensorrt/plugin/split_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -39,43 +39,47 @@ class SplitPlugin : public PluginTensorRTV2Ext {
DeserializeValue(&serial_data, &serial_length, &output_length_);
}
nvinfer1::IPluginV2Ext* clone() const override {
nvinfer1::IPluginV2Ext* clone() const TRT_NOEXCEPT override {
SplitPlugin* ptr = new SplitPlugin(axis_, output_length_, with_fp16_);
ptr->setPluginNamespace(this->getPluginNamespace());
ptr->shareData(this);
return ptr;
}
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* input_types,
int nb_inputs) const override {
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT override {
return input_types[0];
}
const char* getPluginType() const override { return "split_plugin_v2ext"; }
int getNbOutputs() const override { return output_length_.size(); }
const char* getPluginType() const TRT_NOEXCEPT override {
return "split_plugin_v2ext";
}
int getNbOutputs() const TRT_NOEXCEPT override {
return output_length_.size();
}
nvinfer1::Dims getOutputDimensions(int index,
const nvinfer1::Dims* input_dims,
int num_inputs) override;
int num_inputs) TRT_NOEXCEPT override;
int initialize() override;
void terminate() override;
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batch_size, const void* const* inputs, void** outputs,
#else
int enqueue(int batch_size, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
protected:
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return SerializedSize(axis_) + SerializedSize(output_length_) +
getBaseSerializationSize();
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, axis_);
SerializeValue(&buffer, output_length_);
......@@ -98,32 +102,35 @@ class SplitPlugin : public PluginTensorRTV2Ext {
class SplitPluginCreator : public nvinfer1::IPluginCreator {
public:
SplitPluginCreator() {}
const char* getPluginName() const override { return "split_plugin_v2ext"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "split_plugin_v2ext";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
const nvinfer1::PluginFieldCollection* getFieldNames() override {
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override {
return &field_collection_;
}
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override {
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override {
// not implemented
return nullptr;
}
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
auto plugin = new SplitPlugin(serial_data, serial_length);
return plugin;
}
void setPluginNamespace(const char* lib_namespace) override {
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override {
plugin_namespace_ = lib_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return plugin_namespace_.c_str();
}
......@@ -151,46 +158,51 @@ class SplitPluginDynamic : public DynamicPluginTensorRT {
DeserializeValue(&serial_data, &serial_length, &with_fp16_);
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new SplitPluginDynamic(axis_, output_length_, with_fp16_);
}
const char* getPluginType() const override { return "split_plugin"; }
int getNbOutputs() const override { return output_length_.size(); }
int initialize() override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "split_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override {
return output_length_.size();
}
int initialize() TRT_NOEXCEPT override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) override;
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
private:
int axis_;
......@@ -200,31 +212,34 @@ class SplitPluginDynamic : public DynamicPluginTensorRT {
class SplitPluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
SplitPluginDynamicCreator() {}
const char* getPluginName() const override { return "split_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "split_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
const nvinfer1::PluginFieldCollection* getFieldNames() override {
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override {
return &field_collection_;
}
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override {
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override {
return nullptr;
}
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
auto plugin = new SplitPluginDynamic(serial_data, serial_length);
return plugin;
}
void setPluginNamespace(const char* lib_namespace) override {
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override {
plugin_namespace_ = lib_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/stack_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -37,17 +37,19 @@ StackPluginDynamic::StackPluginDynamic(void const* serial_data,
StackPluginDynamic::~StackPluginDynamic() {}
nvinfer1::IPluginV2DynamicExt* StackPluginDynamic::clone() const {
nvinfer1::IPluginV2DynamicExt* StackPluginDynamic::clone() const TRT_NOEXCEPT {
return new StackPluginDynamic(axis_, num_stack_, with_fp16_);
}
const char* StackPluginDynamic::getPluginType() const { return "stack_plugin"; }
const char* StackPluginDynamic::getPluginType() const TRT_NOEXCEPT {
return "stack_plugin";
}
int StackPluginDynamic::getNbOutputs() const { return 1; }
int StackPluginDynamic::getNbOutputs() const TRT_NOEXCEPT { return 1; }
int StackPluginDynamic::initialize() { return 0; }
int StackPluginDynamic::initialize() TRT_NOEXCEPT { return 0; }
size_t StackPluginDynamic::getSerializationSize() const {
size_t StackPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
size_t serialize_size = 0;
serialize_size += SerializedSize(axis_);
serialize_size += SerializedSize(num_stack_);
......@@ -55,7 +57,7 @@ size_t StackPluginDynamic::getSerializationSize() const {
return serialize_size;
}
void StackPluginDynamic::serialize(void* buffer) const {
void StackPluginDynamic::serialize(void* buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, axis_);
SerializeValue(&buffer, num_stack_);
SerializeValue(&buffer, with_fp16_);
......@@ -63,7 +65,7 @@ void StackPluginDynamic::serialize(void* buffer) const {
nvinfer1::DimsExprs StackPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) {
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT {
nvinfer1::DimsExprs output(inputs[0]);
output.nbDims = inputs[0].nbDims + 1;
......@@ -76,21 +78,22 @@ nvinfer1::DimsExprs StackPluginDynamic::getOutputDimensions(
void StackPluginDynamic::configurePlugin(
const nvinfer1::DynamicPluginTensorDesc* in, int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) {}
const nvinfer1::DynamicPluginTensorDesc* out, int nbOutputs) TRT_NOEXCEPT {}
size_t StackPluginDynamic::getWorkspaceSize(
const nvinfer1::PluginTensorDesc* inputs, int nbInputs,
const nvinfer1::PluginTensorDesc* outputs, int nbOutputs) const {
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const TRT_NOEXCEPT {
return num_stack_ * sizeof(uintptr_t);
}
void StackPluginDynamic::destroy() { delete this; }
void StackPluginDynamic::destroy() TRT_NOEXCEPT { delete this; }
void StackPluginDynamic::terminate() {}
void StackPluginDynamic::terminate() TRT_NOEXCEPT {}
bool StackPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of stack plugin should not be nullptr."));
......@@ -118,7 +121,8 @@ bool StackPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType StackPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType* input_types, int nb_inputs) const {
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"The index should be equal to 0"));
return input_types[0];
......@@ -139,7 +143,8 @@ __global__ void StackKernel(const T* const* input, T* output, int num_stack,
int StackPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc,
const void* const* inputs, void* const* outputs,
void* workspace, cudaStream_t stream) {
void* workspace,
cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims; // (batch, seq, seq)
auto out_dims = output_desc[0].dims; // (batch, num_head, seq, seq)
auto out_num_dims = out_dims.nbDims;
......@@ -195,19 +200,21 @@ int StackPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc* input_desc,
StackPluginDynamicCreator::StackPluginDynamicCreator() {}
const char* StackPluginDynamicCreator::getPluginName() const {
const char* StackPluginDynamicCreator::getPluginName() const TRT_NOEXCEPT {
return "stack_plugin";
}
const char* StackPluginDynamicCreator::getPluginVersion() const { return "1"; }
const char* StackPluginDynamicCreator::getPluginVersion() const TRT_NOEXCEPT {
return "1";
}
const nvinfer1::PluginFieldCollection*
StackPluginDynamicCreator::getFieldNames() {
StackPluginDynamicCreator::getFieldNames() TRT_NOEXCEPT {
return &field_collection_;
}
nvinfer1::IPluginV2* StackPluginDynamicCreator::createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) {
const char* name, const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT {
int axis = -1;
int num_stack = -1;
bool with_fp16 = false;
......@@ -230,16 +237,18 @@ nvinfer1::IPluginV2* StackPluginDynamicCreator::createPlugin(
}
nvinfer1::IPluginV2* StackPluginDynamicCreator::deserializePlugin(
const char* name, const void* serial_data, size_t serial_length) {
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT {
auto plugin = new StackPluginDynamic(serial_data, serial_length);
return plugin;
}
void StackPluginDynamicCreator::setPluginNamespace(const char* lib_namespace) {
void StackPluginDynamicCreator::setPluginNamespace(const char* lib_namespace)
TRT_NOEXCEPT {
plugin_namespace_ = lib_namespace;
}
const char* StackPluginDynamicCreator::getPluginNamespace() const {
const char* StackPluginDynamicCreator::getPluginNamespace() const TRT_NOEXCEPT {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/stack_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -31,37 +31,36 @@ class StackPluginDynamic : public DynamicPluginTensorRT {
explicit StackPluginDynamic(int axis, int num_stack, bool with_fp16);
StackPluginDynamic(void const* serial_data, size_t serial_length);
~StackPluginDynamic();
nvinfer1::IPluginV2DynamicExt* clone() const override;
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int outputIndex, const nvinfer1::DimsExprs* inputs, int nbInputs,
nvinfer1::IExprBuilder& exprBuilder) override;
nvinfer1::IExprBuilder& exprBuilder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override;
int nbOutputs) TRT_NOEXCEPT override;
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override;
int nbOutputs) const TRT_NOEXCEPT override;
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
const char* getPluginType() const override;
int getNbOutputs() const override;
int initialize() override;
void terminate() override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
void destroy() override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
const char* getPluginType() const TRT_NOEXCEPT override;
int getNbOutputs() const TRT_NOEXCEPT override;
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
void destroy() TRT_NOEXCEPT override;
private:
int axis_;
......@@ -71,16 +70,17 @@ class StackPluginDynamic : public DynamicPluginTensorRT {
class StackPluginDynamicCreator : public nvinfer1::IPluginCreator {
public:
StackPluginDynamicCreator();
const char* getPluginName() const override;
const char* getPluginVersion() const override;
const nvinfer1::PluginFieldCollection* getFieldNames() override;
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override;
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
const char* getPluginName() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override;
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override;
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
private:
std::string plugin_namespace_;
......
paddle/fluid/inference/tensorrt/plugin/swish_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -23,11 +23,11 @@ namespace inference {
namespace tensorrt {
namespace plugin {
int SwishPlugin::initialize() { return 0; }
int SwishPlugin::initialize() TRT_NOEXCEPT { return 0; }
nvinfer1::Dims SwishPlugin::getOutputDimensions(int index,
const nvinfer1::Dims *inputDims,
int nbInputs) {
int nbInputs) TRT_NOEXCEPT {
assert(nbInputs == 1);
assert(index < this->getNbOutputs());
nvinfer1::Dims const &input_dims = inputDims[0];
......@@ -83,12 +83,12 @@ int SwishPlugin::enqueue(int batch_size, const void *const *inputs,
void **outputs, void *workspace, cudaStream_t stream) {
#else
void *const *outputs, void *workspace,
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
#endif
// input dims is CHW.
const auto &input_dims = this->getInputDims(0);
const float *input = reinterpret_cast<const float *>(inputs[0]);
float *output = reinterpret_cast<float **>(outputs)[0];
float *output = reinterpret_cast<float *const *>(outputs)[0];
int num = batch_size;
for (int i = 0; i < input_dims.nbDims; i++) {
num *= input_dims.d[i];
......@@ -103,29 +103,29 @@ int SwishPlugin::enqueue(int batch_size, const void *const *inputs,
// Dynamic Plugin below.
#if IS_TRT_VERSION_GE(6000)
int SwishPluginDynamic::initialize() {
int SwishPluginDynamic::initialize() TRT_NOEXCEPT {
getPluginNamespace();
return 0;
}
size_t SwishPluginDynamic::getSerializationSize() const {
size_t SwishPluginDynamic::getSerializationSize() const TRT_NOEXCEPT {
return SerializedSize(beta_) + SerializedSize(with_fp16_);
}
void SwishPluginDynamic::serialize(void *buffer) const {
void SwishPluginDynamic::serialize(void *buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, beta_);
SerializeValue(&buffer, with_fp16_);
}
nvinfer1::DimsExprs SwishPluginDynamic::getOutputDimensions(
int output_index, const nvinfer1::DimsExprs *inputs, int nb_inputs,
nvinfer1::IExprBuilder &expr_builder) {
nvinfer1::IExprBuilder &expr_builder) TRT_NOEXCEPT {
return inputs[0];
}
bool SwishPluginDynamic::supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc *in_out, int nb_inputs,
int nb_outputs) {
int nb_outputs) TRT_NOEXCEPT {
PADDLE_ENFORCE_NOT_NULL(
in_out, platform::errors::InvalidArgument(
"The input of swish plugin shoule not be nullptr."));
......@@ -154,7 +154,8 @@ bool SwishPluginDynamic::supportsFormatCombination(
}
nvinfer1::DataType SwishPluginDynamic::getOutputDataType(
int index, const nvinfer1::DataType *input_types, int nb_inputs) const {
int index, const nvinfer1::DataType *input_types,
int nb_inputs) const TRT_NOEXCEPT {
PADDLE_ENFORCE_EQ(index, 0, platform::errors::InvalidArgument(
"The Swish Plugin only has one input, so the "
"index value should be 0, but get %d.",
......@@ -165,7 +166,8 @@ nvinfer1::DataType SwishPluginDynamic::getOutputDataType(
int SwishPluginDynamic::enqueue(const nvinfer1::PluginTensorDesc *input_desc,
const nvinfer1::PluginTensorDesc *output_desc,
const void *const *inputs, void *const *outputs,
void *workspace, cudaStream_t stream) {
void *workspace,
cudaStream_t stream) TRT_NOEXCEPT {
auto input_dims = input_desc[0].dims;
size_t num = ProductDim(input_dims);
int threads = 1024;
......
paddle/fluid/inference/tensorrt/plugin/swish_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -31,11 +31,11 @@ class SwishPlugin : public PluginTensorRT {
float beta_;
public:
size_t getSerializationSize() const override {
size_t getSerializationSize() const TRT_NOEXCEPT override {
return getBaseSerializationSize() + SerializedSize(beta_);
}
void serialize(void* buffer) const override {
void serialize(void* buffer) const TRT_NOEXCEPT override {
serializeBase(buffer);
SerializeValue(&buffer, beta_);
}
......@@ -53,33 +53,37 @@ class SwishPlugin : public PluginTensorRT {
~SwishPlugin() {}
int initialize() override;
int initialize() TRT_NOEXCEPT override;
SwishPlugin* clone() const override {
SwishPlugin* clone() const TRT_NOEXCEPT override {
return new SwishPlugin(beta_, with_fp16_);
}
const char* getPluginType() const override { return "swish_plugin"; }
int getNbOutputs() const override { return 1; }
const char* getPluginType() const TRT_NOEXCEPT override {
return "swish_plugin";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nbInputDims) override;
int nbInputDims) TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batchSize, const void* const* inputs, void** outputs,
#else
int enqueue(int batchSize, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
};
class SwishPluginCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "swish_plugin"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "swish_plugin";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new SwishPlugin(serial_data, serial_length);
}
};
......@@ -96,46 +100,49 @@ class SwishPluginDynamic : public DynamicPluginTensorRT {
DeserializeValue(&serialData, &serialLength, &beta_);
DeserializeValue(&serialData, &serialLength, &with_fp16_);
}
nvinfer1::IPluginV2DynamicExt* clone() const override {
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT override {
return new SwishPluginDynamic(beta_, with_fp16_);
}
const char* getPluginType() const override { return "swish_plugin_dynamic"; }
int getNbOutputs() const override { return 1; }
int initialize() override;
const char* getPluginType() const TRT_NOEXCEPT override {
return "swish_plugin_dynamic";
}
int getNbOutputs() const TRT_NOEXCEPT override { return 1; }
int initialize() TRT_NOEXCEPT override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
int nbInputs,
int nbOutputs) TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
int nbOutputs) TRT_NOEXCEPT override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
int nbOutputs) const TRT_NOEXCEPT override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
cudaStream_t stream) TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* inputTypes,
int nbInputs) const TRT_NOEXCEPT override;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
private:
float beta_;
......@@ -143,13 +150,15 @@ class SwishPluginDynamic : public DynamicPluginTensorRT {
class SwishPluginDynamicCreator : public TensorRTPluginCreator {
public:
const char* getPluginName() const override { return "swish_plugin_dynamic"; }
const char* getPluginName() const TRT_NOEXCEPT override {
return "swish_plugin_dynamic";
}
const char* getPluginVersion() const override { return "1"; }
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override {
nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override {
return new SwishPluginDynamic(serial_data, serial_length);
}
};
......
paddle/fluid/inference/tensorrt/plugin/trt_plugin.cc
浏览文件 @ b5aab4f0
......@@ -60,8 +60,8 @@ size_t PluginTensorRT::getBaseSerializationSize() const {
return SeriaSize(input_dims_, data_type_, data_format_, with_fp16_);
}
bool PluginTensorRT::supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const {
bool PluginTensorRT::supportsFormat(
nvinfer1::DataType type, nvinfer1::PluginFormat format) const TRT_NOEXCEPT {
return ((type == nvinfer1::DataType::kFLOAT) &&
(format == nvinfer1::PluginFormat::kLINEAR));
}
......@@ -69,7 +69,7 @@ bool PluginTensorRT::supportsFormat(nvinfer1::DataType type,
void PluginTensorRT::configureWithFormat(
const nvinfer1::Dims* input_dims, int num_inputs,
const nvinfer1::Dims* output_dims, int num_outputs, nvinfer1::DataType type,
nvinfer1::PluginFormat format, int max_batch_size) {
nvinfer1::PluginFormat format, int max_batch_size) TRT_NOEXCEPT {
data_type_ = type;
data_format_ = format;
input_dims_.assign(input_dims, input_dims + num_inputs);
......@@ -95,26 +95,28 @@ void PluginTensorRTV2Ext::configurePlugin(
const nvinfer1::DataType* input_types,
const nvinfer1::DataType* output_types, const bool* input_is_broadcast,
const bool* output_is_broadcast, nvinfer1::PluginFormat float_format,
int32_t max_batch_size) {
int32_t max_batch_size) TRT_NOEXCEPT {
input_dims_.assign(input_dims, input_dims + nb_inputs);
data_format_ = float_format;
data_type_ = input_types[0];
}
const nvinfer1::PluginFieldCollection* TensorRTPluginCreator::getFieldNames() {
const nvinfer1::PluginFieldCollection* TensorRTPluginCreator::getFieldNames()
TRT_NOEXCEPT {
return &field_collection_;
}
nvinfer1::IPluginV2* TensorRTPluginCreator::createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) {
const char* name, const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT {
return nullptr;
}
void TensorRTPluginCreator::setPluginNamespace(const char* lib_namespace) {
void TensorRTPluginCreator::setPluginNamespace(const char* lib_namespace)
TRT_NOEXCEPT {
plugin_namespace_ = lib_namespace;
}
const char* TensorRTPluginCreator::getPluginNamespace() const {
const char* TensorRTPluginCreator::getPluginNamespace() const TRT_NOEXCEPT {
return plugin_namespace_.c_str();
}
......
paddle/fluid/inference/tensorrt/plugin/trt_plugin.h
浏览文件 @ b5aab4f0
......@@ -64,35 +64,35 @@ class PluginTensorRT : public nvinfer1::IPluginV2 {
nvinfer1::PluginFormat getDataFormat() const { return data_format_; }
// IPluginV2
virtual const char* getPluginType() const = 0;
virtual const char* getPluginType() const TRT_NOEXCEPT = 0;
virtual const char* getPluginVersion() const { return "1"; }
virtual const char* getPluginVersion() const TRT_NOEXCEPT { return "1"; }
int getNbOutputs() const { return 1; }
int getNbOutputs() const TRT_NOEXCEPT { return 1; }
virtual nvinfer1::Dims getOutputDimensions(int index,
const nvinfer1::Dims* input_dims,
int num_inputs) = 0;
int num_inputs) TRT_NOEXCEPT = 0;
// Check format support. The default is FLOAT32 and kLINEAR.
bool supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const override;
bool supportsFormat(nvinfer1::DataType type, nvinfer1::PluginFormat format)
const TRT_NOEXCEPT override;
// Configure the layer
void configureWithFormat(const nvinfer1::Dims* input_dims, int num_inputs,
const nvinfer1::Dims* output_dims, int num_outputs,
nvinfer1::DataType type,
nvinfer1::PluginFormat format,
int max_batch_size) override;
int max_batch_size) TRT_NOEXCEPT override;
// Initialize the layer for execution.
int initialize() override { return 0; }
int initialize() TRT_NOEXCEPT override { return 0; }
// Shutdown the layer. This is called when the engine is destroyed
void terminate() override {}
void terminate() TRT_NOEXCEPT override {}
// Find the workspace size required by the layer
size_t getWorkspaceSize(int) const override { return 0; }
size_t getWorkspaceSize(int) const TRT_NOEXCEPT override { return 0; }
// Execute the layer
#if IS_TRT_VERSION_LT(8000)
......@@ -101,25 +101,27 @@ class PluginTensorRT : public nvinfer1::IPluginV2 {
virtual int enqueue(int batch_size, const void* const* inputs,
void* const* outputs,
#endif
void* workspace, cudaStream_t stream) = 0;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT = 0;
// Find the size of the serialization buffer required
virtual size_t getSerializationSize() const = 0;
virtual size_t getSerializationSize() const TRT_NOEXCEPT = 0;
// Serialize the layer config to buffer.
// TensorRT will call this func to serialize the configuration of TensorRT
// engine. It should not be called by users.
virtual void serialize(void* buffer) const = 0;
virtual void serialize(void* buffer) const TRT_NOEXCEPT = 0;
void destroy() override { delete this; }
void destroy() TRT_NOEXCEPT override { delete this; }
virtual nvinfer1::IPluginV2* clone() const = 0;
virtual nvinfer1::IPluginV2* clone() const TRT_NOEXCEPT = 0;
void setPluginNamespace(const char* plugin_namespace) override {
void setPluginNamespace(const char* plugin_namespace) TRT_NOEXCEPT override {
namespace_ = plugin_namespace;
}
const char* getPluginNamespace() const override { return namespace_.c_str(); }
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return namespace_.c_str();
}
protected:
// Deserialize input_dims, max_batch_size, data_type, data_format
......@@ -155,15 +157,16 @@ class PluginTensorRTV2Ext : public nvinfer1::IPluginV2Ext {
// The Func in IPluginV2Ext
virtual nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const = 0;
int nb_inputs) const TRT_NOEXCEPT = 0;
virtual bool isOutputBroadcastAcrossBatch(int32_t output_index,
const bool* input_is_broadcasted,
int32_t nb_inputs) const {
virtual bool isOutputBroadcastAcrossBatch(
int32_t output_index, const bool* input_is_broadcasted,
int32_t nb_inputs) const TRT_NOEXCEPT {
return false;
}
virtual bool canBroadcastInputAcrossBatch(int32_t input_index) const {
virtual bool canBroadcastInputAcrossBatch(int32_t input_index) const
TRT_NOEXCEPT {
return false;
}
......@@ -174,37 +177,37 @@ class PluginTensorRTV2Ext : public nvinfer1::IPluginV2Ext {
const bool* input_is_broadcast,
const bool* output_is_broadcast,
nvinfer1::PluginFormat float_format,
int32_t max_batch_size) override;
int32_t max_batch_size) TRT_NOEXCEPT override;
virtual IPluginV2Ext* clone() const = 0;
virtual IPluginV2Ext* clone() const TRT_NOEXCEPT = 0;
void attachToContext(cudnnContext*, cublasContext*,
nvinfer1::IGpuAllocator*) override {}
nvinfer1::IGpuAllocator*) TRT_NOEXCEPT override {}
void detachFromContext() override {}
void detachFromContext() TRT_NOEXCEPT override {}
// The Func in IPluginV2
virtual const char* getPluginType() const = 0;
const char* getPluginVersion() const override { return "1"; }
virtual int32_t getNbOutputs() const { return 1; }
virtual const char* getPluginType() const TRT_NOEXCEPT = 0;
const char* getPluginVersion() const TRT_NOEXCEPT override { return "1"; }
virtual int32_t getNbOutputs() const TRT_NOEXCEPT { return 1; }
virtual nvinfer1::Dims getOutputDimensions(int32_t index,
const nvinfer1::Dims* inputs,
int32_t nb_input) = 0;
int32_t nb_input) TRT_NOEXCEPT = 0;
// Check format support. The default is FLOAT32 and NCHW.
bool supportsFormat(nvinfer1::DataType type,
nvinfer1::PluginFormat format) const override {
bool supportsFormat(nvinfer1::DataType type, nvinfer1::PluginFormat format)
const TRT_NOEXCEPT override {
return ((type == nvinfer1::DataType::kFLOAT) &&
(format == nvinfer1::PluginFormat::kLINEAR));
}
// Initialize the layer for execution.
// This is called when the engine is created.
int initialize() override { return 0; }
int initialize() TRT_NOEXCEPT override { return 0; }
// Shutdown the layer. This is called when the engine is destroyed
void terminate() override {}
void terminate() TRT_NOEXCEPT override {}
// Find the workspace size required by the layer
size_t getWorkspaceSize(int) const override { return 0; }
size_t getWorkspaceSize(int) const TRT_NOEXCEPT override { return 0; }
// Execute the layer
#if IS_TRT_VERSION_LT(8000)
......@@ -213,23 +216,23 @@ class PluginTensorRTV2Ext : public nvinfer1::IPluginV2Ext {
virtual int enqueue(int batch_size, const void* const* inputs,
void* const* outputs,
#endif
void* workspace, cudaStream_t stream) = 0;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT = 0;
// Find the size of the serialization buffer required
virtual size_t getSerializationSize() const = 0;
virtual size_t getSerializationSize() const TRT_NOEXCEPT = 0;
// Serialize the layer config to buffer.
// TensorRT will call this func to serialize the configuration of TensorRT
// engine. It should not be called by users.
virtual void serialize(void* buffer) const = 0;
virtual void serialize(void* buffer) const TRT_NOEXCEPT = 0;
virtual void destroy() = 0;
virtual void destroy() TRT_NOEXCEPT = 0;
void setPluginNamespace(const char* plugin_namespace) override {
void setPluginNamespace(const char* plugin_namespace) TRT_NOEXCEPT override {
name_space_ = plugin_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return name_space_.c_str();
}
......@@ -256,52 +259,52 @@ class DynamicPluginTensorRT : public nvinfer1::IPluginV2DynamicExt {
DynamicPluginTensorRT(const void* serialized_data, size_t length) {}
// The Func in IPluginExt or IpluginExtV2
virtual const char* getPluginVersion() const { return "1"; }
virtual const char* getPluginType() const = 0;
int getNbOutputs() const { return 1; }
int initialize() override { return 0; }
void terminate() override{};
virtual const char* getPluginVersion() const TRT_NOEXCEPT { return "1"; }
virtual const char* getPluginType() const TRT_NOEXCEPT = 0;
int getNbOutputs() const TRT_NOEXCEPT { return 1; }
int initialize() TRT_NOEXCEPT override { return 0; }
void terminate() TRT_NOEXCEPT override{};
virtual size_t getSerializationSize() const = 0;
virtual void serialize(void* buffer) const = 0;
virtual size_t getSerializationSize() const TRT_NOEXCEPT = 0;
virtual void serialize(void* buffer) const TRT_NOEXCEPT = 0;
// The Func in IPluginV2
nvinfer1::IPluginV2DynamicExt* clone() const = 0;
nvinfer1::IPluginV2DynamicExt* clone() const TRT_NOEXCEPT = 0;
virtual nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) = 0; // NOLINT
nvinfer1::IExprBuilder& expr_builder) TRT_NOEXCEPT = 0; // NOLINT
virtual bool supportsFormatCombination(
int pos, const nvinfer1::PluginTensorDesc* in_out, int nb_inputs,
int nb_outputs) = 0;
int nb_outputs) TRT_NOEXCEPT = 0;
virtual void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nb_inputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nb_outputs) = 0;
int nb_outputs) TRT_NOEXCEPT = 0;
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nb_inputs,
const nvinfer1::PluginTensorDesc* outputs,
int nb_outputs) const override {
int nb_outputs) const TRT_NOEXCEPT override {
return 0;
}
virtual int enqueue(const nvinfer1::PluginTensorDesc* input_desc,
const nvinfer1::PluginTensorDesc* output_desc,
const void* const* inputs, void* const* outputs,
void* workspace, cudaStream_t stream) = 0;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT = 0;
virtual nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_types,
int nb_inputs) const = 0;
void setPluginNamespace(const char* plugin_namespace) override {
int nb_inputs) const TRT_NOEXCEPT = 0;
void setPluginNamespace(const char* plugin_namespace) TRT_NOEXCEPT override {
name_space_ = plugin_namespace;
}
const char* getPluginNamespace() const override {
const char* getPluginNamespace() const TRT_NOEXCEPT override {
return name_space_.c_str();
}
virtual void destroy() = 0;
virtual void destroy() TRT_NOEXCEPT = 0;
protected:
void deserializeBase(void const*& serial_data, // NOLINT
......@@ -320,22 +323,23 @@ class TensorRTPluginCreator : public nvinfer1::IPluginCreator {
public:
TensorRTPluginCreator() = default;
virtual const char* getPluginName() const = 0;
virtual const char* getPluginName() const TRT_NOEXCEPT = 0;
virtual const char* getPluginVersion() const = 0;
virtual const char* getPluginVersion() const TRT_NOEXCEPT = 0;
const nvinfer1::PluginFieldCollection* getFieldNames() override;
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override;
nvinfer1::IPluginV2* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override;
nvinfer1::IPluginV2* createPlugin(const char* name,
const nvinfer1::PluginFieldCollection* fc)
TRT_NOEXCEPT override;
virtual nvinfer1::IPluginV2* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) = 0;
virtual nvinfer1::IPluginV2* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT = 0;
void setPluginNamespace(const char* lib_namespace) override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
private:
std::string plugin_namespace_;
......
paddle/fluid/inference/tensorrt/plugin/yolo_box_op_plugin.cu
浏览文件 @ b5aab4f0
......@@ -70,15 +70,16 @@ YoloBoxPlugin::~YoloBoxPlugin() {
}
}
const char* YoloBoxPlugin::getPluginType() const { return "yolo_box_plugin"; }
const char* YoloBoxPlugin::getPluginType() const TRT_NOEXCEPT {
return "yolo_box_plugin";
}
const char* YoloBoxPlugin::getPluginVersion() const { return "1"; }
const char* YoloBoxPlugin::getPluginVersion() const TRT_NOEXCEPT { return "1"; }
int YoloBoxPlugin::getNbOutputs() const { return 2; }
int YoloBoxPlugin::getNbOutputs() const TRT_NOEXCEPT { return 2; }
nvinfer1::Dims YoloBoxPlugin::getOutputDimensions(int index,
const nvinfer1::Dims* inputs,
int nb_input_dims) {
nvinfer1::Dims YoloBoxPlugin::getOutputDimensions(
int index, const nvinfer1::Dims* inputs, int nb_input_dims) TRT_NOEXCEPT {
const int anchor_num = anchors_.size() / 2;
const int box_num = inputs[0].d[1] * inputs[0].d[2] * anchor_num;
......@@ -90,13 +91,15 @@ nvinfer1::Dims YoloBoxPlugin::getOutputDimensions(int index,
return nvinfer1::Dims2(box_num, class_num_);
}
bool YoloBoxPlugin::supportsFormat(nvinfer1::DataType type,
nvinfer1::TensorFormat format) const {
bool YoloBoxPlugin::supportsFormat(
nvinfer1::DataType type, nvinfer1::TensorFormat format) const TRT_NOEXCEPT {
return ((type == data_type_ || type == nvinfer1::DataType::kINT32) &&
format == nvinfer1::TensorFormat::kLINEAR);
}
size_t YoloBoxPlugin::getWorkspaceSize(int max_batch_size) const { return 0; }
size_t YoloBoxPlugin::getWorkspaceSize(int max_batch_size) const TRT_NOEXCEPT {
return 0;
}
template <typename T>
__device__ inline T sigmoid(T x) {
......@@ -219,7 +222,7 @@ __global__ void KeYoloBoxFw(const T* const input, const int* const imgsize,
template <typename T>
int YoloBoxPlugin::enqueue_impl(int batch_size, const void* const* inputs,
void** outputs, void* workspace,
void* const* outputs, void* workspace,
cudaStream_t stream) {
const int n = batch_size;
const int h = input_h_;
......@@ -247,7 +250,7 @@ int YoloBoxPlugin::enqueue(int batch_size, const void* const* inputs,
#else
void* const* outputs, void* workspace,
#endif
cudaStream_t stream) {
cudaStream_t stream) TRT_NOEXCEPT {
if (data_type_ == nvinfer1::DataType::kFLOAT) {
return enqueue_impl<float>(batch_size, inputs, outputs, workspace, stream);
} else if (data_type_ == nvinfer1::DataType::kHALF) {
......@@ -256,11 +259,11 @@ int YoloBoxPlugin::enqueue(int batch_size, const void* const* inputs,
assert("unsupported type.");
}
int YoloBoxPlugin::initialize() { return 0; }
int YoloBoxPlugin::initialize() TRT_NOEXCEPT { return 0; }
void YoloBoxPlugin::terminate() {}
void YoloBoxPlugin::terminate() TRT_NOEXCEPT {}
size_t YoloBoxPlugin::getSerializationSize() const {
size_t YoloBoxPlugin::getSerializationSize() const TRT_NOEXCEPT {
size_t serialize_size = 0;
serialize_size += SerializedSize(data_type_);
serialize_size += SerializedSize(anchors_);
......@@ -274,7 +277,7 @@ size_t YoloBoxPlugin::getSerializationSize() const {
return serialize_size;
}
void YoloBoxPlugin::serialize(void* buffer) const {
void YoloBoxPlugin::serialize(void* buffer) const TRT_NOEXCEPT {
SerializeValue(&buffer, data_type_);
SerializeValue(&buffer, anchors_);
SerializeValue(&buffer, class_num_);
......@@ -286,28 +289,30 @@ void YoloBoxPlugin::serialize(void* buffer) const {
SerializeValue(&buffer, input_w_);
}
void YoloBoxPlugin::destroy() {}
void YoloBoxPlugin::destroy() TRT_NOEXCEPT {}
void YoloBoxPlugin::setPluginNamespace(const char* lib_namespace) {
void YoloBoxPlugin::setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT {
namespace_ = std::string(lib_namespace);
}
const char* YoloBoxPlugin::getPluginNamespace() const {
const char* YoloBoxPlugin::getPluginNamespace() const TRT_NOEXCEPT {
return namespace_.c_str();
}
nvinfer1::DataType YoloBoxPlugin::getOutputDataType(
int index, const nvinfer1::DataType* input_type, int nb_inputs) const {
int index, const nvinfer1::DataType* input_type,
int nb_inputs) const TRT_NOEXCEPT {
return input_type[0];
}
bool YoloBoxPlugin::isOutputBroadcastAcrossBatch(int output_index,
const bool* input_is_broadcast,
int nb_inputs) const {
bool YoloBoxPlugin::isOutputBroadcastAcrossBatch(
int output_index, const bool* input_is_broadcast,
int nb_inputs) const TRT_NOEXCEPT {
return false;
}
bool YoloBoxPlugin::canBroadcastInputAcrossBatch(int input_index) const {
bool YoloBoxPlugin::canBroadcastInputAcrossBatch(int input_index) const
TRT_NOEXCEPT {
return false;
}
......@@ -317,9 +322,9 @@ void YoloBoxPlugin::configurePlugin(
const nvinfer1::DataType* input_types,
const nvinfer1::DataType* output_types, const bool* input_is_broadcast,
const bool* output_is_broadcast, nvinfer1::PluginFormat float_format,
int max_batct_size) {}
int max_batct_size) TRT_NOEXCEPT {}
nvinfer1::IPluginV2Ext* YoloBoxPlugin::clone() const {
nvinfer1::IPluginV2Ext* YoloBoxPlugin::clone() const TRT_NOEXCEPT {
return new YoloBoxPlugin(data_type_, anchors_, class_num_, conf_thresh_,
downsample_ratio_, clip_bbox_, scale_x_y_, input_h_,
input_w_);
......@@ -327,26 +332,30 @@ nvinfer1::IPluginV2Ext* YoloBoxPlugin::clone() const {
YoloBoxPluginCreator::YoloBoxPluginCreator() {}
void YoloBoxPluginCreator::setPluginNamespace(const char* lib_namespace) {
void YoloBoxPluginCreator::setPluginNamespace(const char* lib_namespace)
TRT_NOEXCEPT {
namespace_ = std::string(lib_namespace);
}
const char* YoloBoxPluginCreator::getPluginNamespace() const {
const char* YoloBoxPluginCreator::getPluginNamespace() const TRT_NOEXCEPT {
return namespace_.c_str();
}
const char* YoloBoxPluginCreator::getPluginName() const {
const char* YoloBoxPluginCreator::getPluginName() const TRT_NOEXCEPT {
return "yolo_box_plugin";
}
const char* YoloBoxPluginCreator::getPluginVersion() const { return "1"; }
const char* YoloBoxPluginCreator::getPluginVersion() const TRT_NOEXCEPT {
return "1";
}
const nvinfer1::PluginFieldCollection* YoloBoxPluginCreator::getFieldNames() {
const nvinfer1::PluginFieldCollection* YoloBoxPluginCreator::getFieldNames()
TRT_NOEXCEPT {
return &field_collection_;
}
nvinfer1::IPluginV2Ext* YoloBoxPluginCreator::createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) {
const char* name, const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT {
const nvinfer1::PluginField* fields = fc->fields;
int type_id = -1;
......@@ -392,7 +401,8 @@ nvinfer1::IPluginV2Ext* YoloBoxPluginCreator::createPlugin(
}
nvinfer1::IPluginV2Ext* YoloBoxPluginCreator::deserializePlugin(
const char* name, const void* serial_data, size_t serial_length) {
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT {
auto plugin = new YoloBoxPlugin(serial_data, serial_length);
plugin->setPluginNamespace(namespace_.c_str());
return plugin;
......
paddle/fluid/inference/tensorrt/plugin/yolo_box_op_plugin.h
浏览文件 @ b5aab4f0
......@@ -35,38 +35,39 @@ class YoloBoxPlugin : public nvinfer1::IPluginV2Ext {
YoloBoxPlugin(const void* data, size_t length);
~YoloBoxPlugin() override;
const char* getPluginType() const override;
const char* getPluginVersion() const override;
int getNbOutputs() const override;
const char* getPluginType() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
int getNbOutputs() const TRT_NOEXCEPT override;
nvinfer1::Dims getOutputDimensions(int index, const nvinfer1::Dims* inputs,
int nb_input_dims) override;
bool supportsFormat(nvinfer1::DataType type,
nvinfer1::TensorFormat format) const override;
size_t getWorkspaceSize(int max_batch_size) const override;
int nb_input_dims) TRT_NOEXCEPT override;
bool supportsFormat(nvinfer1::DataType type, nvinfer1::TensorFormat format)
const TRT_NOEXCEPT override;
size_t getWorkspaceSize(int max_batch_size) const TRT_NOEXCEPT override;
#if IS_TRT_VERSION_LT(8000)
int enqueue(int batch_size, const void* const* inputs, void** outputs,
#else
int enqueue(int batch_size, const void* const* inputs, void* const* outputs,
#endif
void* workspace, cudaStream_t stream) override;
void* workspace, cudaStream_t stream) TRT_NOEXCEPT override;
template <typename T>
int enqueue_impl(int batch_size, const void* const* inputs, void** outputs,
void* workspace, cudaStream_t stream);
int initialize() override;
void terminate() override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
void destroy() override;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
int enqueue_impl(int batch_size, const void* const* inputs,
void* const* outputs, void* workspace, cudaStream_t stream);
int initialize() TRT_NOEXCEPT override;
void terminate() TRT_NOEXCEPT override;
size_t getSerializationSize() const TRT_NOEXCEPT override;
void serialize(void* buffer) const TRT_NOEXCEPT override;
void destroy() TRT_NOEXCEPT override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* input_type,
int nb_inputs) const override;
nvinfer1::DataType getOutputDataType(
int index, const nvinfer1::DataType* input_type,
int nb_inputs) const TRT_NOEXCEPT override;
bool isOutputBroadcastAcrossBatch(int output_index,
const bool* input_is_broadcast,
int nb_inputs) const override;
bool canBroadcastInputAcrossBatch(int input_index) const override;
int nb_inputs) const TRT_NOEXCEPT override;
bool canBroadcastInputAcrossBatch(int input_index) const
TRT_NOEXCEPT override;
void configurePlugin(const nvinfer1::Dims* input_dims, int nb_inputs,
const nvinfer1::Dims* output_dims, int nb_outputs,
const nvinfer1::DataType* input_types,
......@@ -74,8 +75,8 @@ class YoloBoxPlugin : public nvinfer1::IPluginV2Ext {
const bool* input_is_broadcast,
const bool* output_is_broadcast,
nvinfer1::PluginFormat float_format,
int max_batct_size) override;
nvinfer1::IPluginV2Ext* clone() const override;
int max_batct_size) TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* clone() const TRT_NOEXCEPT override;
private:
nvinfer1::DataType data_type_;
......@@ -96,17 +97,18 @@ class YoloBoxPluginCreator : public nvinfer1::IPluginCreator {
YoloBoxPluginCreator();
~YoloBoxPluginCreator() override = default;
void setPluginNamespace(const char* lib_namespace) override;
const char* getPluginNamespace() const override;
const char* getPluginName() const override;
const char* getPluginVersion() const override;
const nvinfer1::PluginFieldCollection* getFieldNames() override;
void setPluginNamespace(const char* lib_namespace) TRT_NOEXCEPT override;
const char* getPluginNamespace() const TRT_NOEXCEPT override;
const char* getPluginName() const TRT_NOEXCEPT override;
const char* getPluginVersion() const TRT_NOEXCEPT override;
const nvinfer1::PluginFieldCollection* getFieldNames() TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* createPlugin(
const char* name, const nvinfer1::PluginFieldCollection* fc) override;
nvinfer1::IPluginV2Ext* deserializePlugin(const char* name,
const void* serial_data,
size_t serial_length) override;
const char* name,
const nvinfer1::PluginFieldCollection* fc) TRT_NOEXCEPT override;
nvinfer1::IPluginV2Ext* deserializePlugin(
const char* name, const void* serial_data,
size_t serial_length) TRT_NOEXCEPT override;
private:
std::string namespace_;
......
paddle/fluid/inference/tensorrt/trt_int8_calibrator.cc
浏览文件 @ b5aab4f0
......@@ -22,7 +22,7 @@ namespace inference {
namespace tensorrt {
// set the batch size before constructing the thread to execute engine
int TRTInt8Calibrator::getBatchSize() const { return batch_size_; }
int TRTInt8Calibrator::getBatchSize() const TRT_NOEXCEPT { return batch_size_; }
TRTInt8Calibrator::TRTInt8Calibrator(
const std::unordered_map<std::string, size_t>& buffers, int batch_size,
......@@ -95,7 +95,7 @@ bool TRTInt8Calibrator::setBatch(
}
bool TRTInt8Calibrator::getBatch(void** bindings, const char** names,
int num_bindings) {
int num_bindings) TRT_NOEXCEPT {
VLOG(4) << "get batch: " << engine_name_;
std::unique_lock<std::mutex> lk(mut_);
// The consumer has just finished processing a data.
......@@ -131,14 +131,15 @@ void TRTInt8Calibrator::setDone() {
cond_.notify_all();
}
const void* TRTInt8Calibrator::readCalibrationCache(size_t& length) {
const void* TRTInt8Calibrator::readCalibrationCache(size_t& length)
TRT_NOEXCEPT {
if (calibration_table_.empty()) return nullptr;
length = calibration_table_.size();
return calibration_table_.data();
}
void TRTInt8Calibrator::writeCalibrationCache(const void* ptr,
std::size_t length) {
std::size_t length) TRT_NOEXCEPT {
calibration_table_ = std::string((const char*)ptr, length);
VLOG(4) << "Got calibration data for " << engine_name_ << " " << ptr
<< " length=" << length;
......
paddle/fluid/inference/tensorrt/trt_int8_calibrator.h
浏览文件 @ b5aab4f0
......@@ -43,17 +43,18 @@ struct TRTInt8Calibrator : public nvinfer1::IInt8EntropyCalibrator2 {
explicit TRTInt8Calibrator(const std::string& calibration_data);
~TRTInt8Calibrator();
int getBatchSize() const override;
int getBatchSize() const TRT_NOEXCEPT override;
bool getBatch(void* bindings[], const char* names[],
int num_bindings) override;
int num_bindings) TRT_NOEXCEPT override;
bool setBatch(const std::unordered_map<std::string, void*>& data);
void setDone();
void waitAndSetDone();
const void* readCalibrationCache(std::size_t& length) override;
void writeCalibrationCache(const void* ptr, std::size_t length) override;
const void* readCalibrationCache(std::size_t& length) TRT_NOEXCEPT override;
void writeCalibrationCache(const void* ptr,
std::size_t length) TRT_NOEXCEPT override;
const std::string& getCalibrationTableAsString() {
return calibration_table_;
}
......
Markdown is supported
0% 或 .
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 或