Skip to content

Commit

Permalink
Update code according to the review
Browse files Browse the repository at this point in the history 
Signed-off-by: Bogdan Pereanu <[email protected]>
  • Loading branch information
@pereanub
pereanub committed Jan 10, 2025
1 parent 6e1d738 commit 3a38ec7
Show file tree
Hide file tree
Showing 4 changed files with 29 additions and 28 deletions.
4 changes: 2 additions & 2 deletions src/plugins/intel_npu/src/backend/include/zero_pipeline.hpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -21,8 +21,8 @@ struct Pipeline {
zeroProfiling::ProfilingPool& profiling_pool,
zeroProfiling::ProfilingQuery& profiling_query,
const std::shared_ptr<zeroProfiling::NpuInferProfiling>& npu_profiling,
const std::vector<std::vector<std::shared_ptr<ov::ITensor>>>& input_tensors_data,
const std::vector<std::shared_ptr<ov::ITensor>>& output_tensors_data,
const std::vector<std::vector<std::shared_ptr<ov::ITensor>>>& input_tensors,
const std::vector<std::shared_ptr<ov::ITensor>>& output_tensors,
uint32_t group_ordinal);

Pipeline(const Pipeline&) = delete;
Expand Down
6 changes: 3 additions & 3 deletions src/plugins/intel_npu/src/backend/include/zero_tensor.hpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -17,9 +17,9 @@ namespace intel_npu {

/**
* @brief Constructs Tensor using element type and shape. Allocate internal host storage using custom allocator.
* @details The implementation is simillar with the AllocatedTensor class from OV repository.
* @note Set_shape method throw an error in case re-allocation is needed but this is not supported by the driver.
* There are two extra methods to notify the consumer if memorey changed or not and to reset the flag.
* @details The implementation is simillar to the AllocatedTensor class from OV namespace.
* @note Set_shape method throws an error in case re-allocation is needed but this is not supported by the driver.
* There are two extra methods to notify the consumer if memory changed or not and to reset the flag.
*/
class ZeroTensor final : public ov::ITensor {
public:
Expand Down
44 changes: 22 additions & 22 deletions src/plugins/intel_npu/src/backend/src/zero_pipeline.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -38,8 +38,8 @@ Pipeline::Pipeline(const Config& config,
zeroProfiling::ProfilingPool& profiling_pool,
zeroProfiling::ProfilingQuery& profiling_query,
const std::shared_ptr<zeroProfiling::NpuInferProfiling>& npu_profiling,
const std::vector<std::vector<std::shared_ptr<ov::ITensor>>>& input_tensors_data,
const std::vector<std::shared_ptr<ov::ITensor>>& output_tensors_data,
const std::vector<std::vector<std::shared_ptr<ov::ITensor>>>& input_tensors,
const std::vector<std::shared_ptr<ov::ITensor>>& output_tensors,
uint32_t group_ordinal)
: _graph(graph),
_config(config),
Expand Down Expand Up @@ -74,11 +74,11 @@ Pipeline::Pipeline(const Config& config,
for (size_t i = 0; i < _number_of_command_lists; i++) {
size_t io_index = 0;
for (const auto& desc : graph->get_input_descriptors()) {
if (input_tensors_data.at(io_index).size() > 1) {
if (input_tensors.at(io_index).size() > 1) {
void* data = nullptr;
auto remote_tensor = std::dynamic_pointer_cast<ZeroRemoteTensor>(input_tensors_data.at(io_index).at(i));
auto remote_tensor = std::dynamic_pointer_cast<ZeroRemoteTensor>(input_tensors.at(io_index).at(i));
if (remote_tensor == nullptr) {
data = input_tensors_data.at(io_index).at(i)->data();
data = input_tensors.at(io_index).at(i)->data();
} else {
data = extract_object(remote_tensor->get_properties(), ov::intel_npu::mem_handle);
}
Expand All @@ -90,35 +90,35 @@ Pipeline::Pipeline(const Config& config,
}

void* data = nullptr;
auto remote_tensor = std::dynamic_pointer_cast<ZeroRemoteTensor>(input_tensors_data.at(io_index).at(0));
auto remote_tensor = std::dynamic_pointer_cast<ZeroRemoteTensor>(input_tensors.at(io_index).at(0));
if (remote_tensor == nullptr) {
data = input_tensors_data.at(io_index).at(0)->data();
data = input_tensors.at(io_index).at(0)->data();
} else {
data = extract_object(remote_tensor->get_properties(), ov::intel_npu::mem_handle);
}

graph->set_argument_value(
desc.idx,
static_cast<unsigned char*>(data) +
(i * input_tensors_data.at(io_index).at(0)->get_byte_size()) / _number_of_command_lists);
(i * input_tensors.at(io_index).at(0)->get_byte_size()) / _number_of_command_lists);

++io_index;
}

io_index = 0;
for (const auto& desc : graph->get_output_descriptors()) {
void* data = nullptr;
auto remote_tensor = std::dynamic_pointer_cast<ZeroRemoteTensor>(output_tensors_data.at(io_index));
auto remote_tensor = std::dynamic_pointer_cast<ZeroRemoteTensor>(output_tensors.at(io_index));
if (remote_tensor == nullptr) {
data = output_tensors_data.at(io_index)->data();
data = output_tensors.at(io_index)->data();
} else {
data = extract_object(remote_tensor->get_properties(), ov::intel_npu::mem_handle);
}

graph->set_argument_value(
desc.idx,
static_cast<unsigned char*>(data) +
(i * output_tensors_data.at(io_index)->get_byte_size()) / _number_of_command_lists);
(i * output_tensors.at(io_index)->get_byte_size()) / _number_of_command_lists);
++io_index;
}

Expand Down Expand Up @@ -226,22 +226,22 @@ void Pipeline::updateCommandList(uint32_t arg_index, const void* arg_data, size_
OV_ITT_TASK_CHAIN(ZERO_EXECUTOR_IP_UMCL, itt::domains::LevelZeroBackend, "Pipeline", "updateCommandList");
_logger.debug("Pipeline - updateCommandList");

const size_t _number_of_command_lists = _command_lists.size();
const size_t number_of_command_lists = _command_lists.size();

for (size_t i = 0; i < _number_of_command_lists; i++) {
for (size_t i = 0; i < number_of_command_lists; i++) {
_command_lists.at(i)->updateMutableCommandList(
arg_index,
static_cast<const unsigned char*>(arg_data) + (i * byte_size) / _number_of_command_lists);
static_cast<const unsigned char*>(arg_data) + (i * byte_size) / number_of_command_lists);
}
};

void Pipeline::closeCommandList() {
OV_ITT_TASK_CHAIN(ZERO_EXECUTOR_IP_UMCL, itt::domains::LevelZeroBackend, "Pipeline", "closeCommandList");
_logger.debug("Pipeline - closeCommandList");

const size_t _number_of_command_lists = _command_lists.size();
const size_t number_of_command_lists = _command_lists.size();

for (size_t i = 0; i < _number_of_command_lists; i++) {
for (size_t i = 0; i < number_of_command_lists; i++) {
_command_lists.at(i)->close();
}
};
Expand All @@ -250,10 +250,10 @@ void Pipeline::updateCommandListIndex(uint32_t arg_index, const void* arg_data,
OV_ITT_TASK_CHAIN(ZERO_EXECUTOR_IP_UMCL, itt::domains::LevelZeroBackend, "Pipeline", "updateCommandListIndex");
_logger.debug("Pipeline - updateCommandListIndex");

const size_t _number_of_command_lists = _command_lists.size();
const size_t number_of_command_lists = _command_lists.size();

OPENVINO_ASSERT(command_list_index < _number_of_command_lists,
"Command list index is higgher than the number of Command lists ",
OPENVINO_ASSERT(command_list_index < number_of_command_lists,
"Command list index is higher than the number of Command lists ",
command_list_index);

_command_lists.at(command_list_index)->updateMutableCommandList(arg_index, arg_data);
Expand All @@ -263,10 +263,10 @@ void Pipeline::closeCommandListIndex(size_t command_list_index) {
OV_ITT_TASK_CHAIN(ZERO_EXECUTOR_IP_UMCL, itt::domains::LevelZeroBackend, "Pipeline", "closeCommandListIndex");
_logger.debug("Pipeline - closeCommandListIndex");

const size_t _number_of_command_lists = _command_lists.size();
const size_t number_of_command_lists = _command_lists.size();

OPENVINO_ASSERT(command_list_index < _number_of_command_lists,
"Command list index is higgher than the number of Command lists ",
OPENVINO_ASSERT(command_list_index < number_of_command_lists,
"Command list index is higher than the number of Command lists ",
command_list_index);

_command_lists.at(command_list_index)->close();
Expand Down
3 changes: 2 additions & 1 deletion src/plugins/intel_npu/src/common/src/remote_tensor.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -40,8 +40,9 @@ const ov::AnyMap& RemoteTensor::get_properties() const {
}

void RemoteTensor::set_shape(ov::Shape new_shape) {
if (_shape == new_shape)
if (_shape == new_shape) {
return;
}

_shape = std::move(new_shape);

Expand Down

0 comments on commit 3a38ec7

Please sign in to comment.