Skip to content

[NPU] Fix importing deallocation issue #31882

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into from
+55 −0
Open

[NPU] Fix importing deallocation issue #31882

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
d156fb6
Don't destroy imported memory if already set
pereanub Aug 26, 2025
2f8488e
Add test case
pereanub Aug 26, 2025
91588a1
Fix for remote tensors
pereanub Aug 26, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 7 additions & 0 deletions src/plugins/intel_npu/src/backend/src/zero_infer_request.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -315,6 +315,13 @@ void ZeroInferRequest::set_tensor_data(const std::shared_ptr<ov::ITensor>& tenso
if (zeroUtils::memory_was_allocated_in_the_same_l0_context(_initStructs->getContext(), tensor->data())) {
_logger.debug("ZeroInferRequest::set_tensor_data - tensor was created in the same L0 context, size: %zu",
tensor->get_byte_size());

if (levelZeroTensors != nullptr && !is_remote_tensor(levelZeroTensors) &&
levelZeroTensors->data() == tensor->data()) {
_logger.debug("ZeroInferRequest::set_tensor_data - same data was provided, don't do anything");
return;
}

levelZeroTensors = tensor;

auto zero_tensor = std::dynamic_pointer_cast<ZeroTensor>(levelZeroTensors);
Expand Down
48 changes: 48 additions & 0 deletions src/plugins/intel_npu/tests/functional/behavior/infer_request_run.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1427,6 +1427,54 @@ TEST_P(SetShapeInferRunTests, checkResultsAfterStateTensorsReallocation) {

using CpuVaTensorsTests = InferRequestRunTests;

TEST_P(CpuVaTensorsTests, DontDestroyImportedMemory) {
// Skip test according to plugin specific disabledTestPatterns() (if any)
SKIP_IF_CURRENT_TEST_IS_DISABLED()

ov::InferRequest inference_request;
ov::Tensor first_output;
ov::Tensor second_output;
ov::Tensor global_input;
float* data;

OV_ASSERT_NO_THROW(compiled_model = core->compile_model(ov_model, target_device, configuration));
OV_ASSERT_NO_THROW(inference_request = compiled_model.create_infer_request());
global_input = inference_request.get_input_tensor();
memset(global_input.data(), 1, global_input.get_byte_size());
OV_ASSERT_NO_THROW(inference_request.infer());
first_output = inference_request.get_output_tensor(0);

compiled_model = {};
inference_request = {};

OV_ASSERT_NO_THROW(compiled_model = core->compile_model(ov_model, target_device, configuration));
OV_ASSERT_NO_THROW(inference_request = compiled_model.create_infer_request());
OV_ASSERT_NO_THROW(inference_request.set_input_tensor(global_input));
OV_ASSERT_NO_THROW(inference_request.infer());
second_output = inference_request.get_output_tensor(0);

EXPECT_NE(first_output.data(), second_output.data());
EXPECT_EQ(memcmp(first_output.data(), second_output.data(), second_output.get_byte_size()), 0);

const size_t byte_size = global_input.get_byte_size();
data = static_cast<float*>(::operator new(byte_size, std::align_val_t(4096)));
memset(data, 1, byte_size);

for (int i = 0; i < 10; i++) {
{
ov::Tensor input_data_tensor{ov::element::f32, global_input.get_shape(), data};
OV_ASSERT_NO_THROW(inference_request.set_input_tensor(input_data_tensor));
OV_ASSERT_NO_THROW(inference_request.infer());
}
second_output = inference_request.get_output_tensor(0);

EXPECT_NE(first_output.data(), second_output.data());
EXPECT_EQ(memcmp(first_output.data(), second_output.data(), second_output.get_byte_size()), 0);
}

::operator delete(data, std::align_val_t(4096));
}

TEST_P(CpuVaTensorsTests, SetMultiplePageAllignedTensors) {
SKIP_IF_CURRENT_TEST_IS_DISABLED()

Expand Down
Loading