Skip to content
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

Jump to bottom

DFT on DirectML #12710

Merged
merged 6 commits into from Sep 1, 2022
Merged

DFT on DirectML #12710

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
09dcadc
DFT on DirectML
smk2007 Aug 24, 2022
ed027de
feedback
smk2007 Aug 26, 2022
620a96d
fix misc build issues
smk2007 Aug 26, 2022
b92236f
fixes
smk2007 Aug 30, 2022
0150eac
fix constant cpu inputs and optional tensors for external operators
smk2007 Aug 31, 2022
e055c09
disable dft tests on 'pure' dml
smk2007 Aug 31, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
8 changes: 4 additions & 4 deletions cmake/onnxruntime_providers.cmake
View file
Open in desktop
Expand Up @@ -621,7 +621,7 @@ if (onnxruntime_USE_TENSORRT)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-unused-parameter -Wno-missing-field-initializers")
endif()
set(CXX_VERSION_DEFINED TRUE)

if (onnxruntime_USE_TENSORRT_BUILTIN_PARSER)
# Add TensorRT library
find_path(TENSORRT_INCLUDE_DIR NvInfer.h
Expand Down Expand Up @@ -658,9 +658,9 @@ if (onnxruntime_USE_TENSORRT)
include_directories(${TENSORRT_INCLUDE_DIR})
set(onnxparser_link_libs nvonnxparser_static)
endif()

set(trt_link_libs cudnn ${CMAKE_DL_LIBS} ${TENSORRT_LIBRARY})

file(GLOB_RECURSE onnxruntime_providers_tensorrt_cc_srcs CONFIGURE_DEPENDS
"${ONNXRUNTIME_ROOT}/core/providers/tensorrt/*.h"
"${ONNXRUNTIME_ROOT}/core/providers/tensorrt/*.cc"
Expand Down Expand Up @@ -1157,7 +1157,7 @@ if (onnxruntime_USE_DML)
if (GDK_PLATFORM STREQUAL Scarlett)
target_link_libraries(onnxruntime_providers_dml PRIVATE ${gdk_dx_libs})
else()
target_link_libraries(onnxruntime_providers_dml PRIVATE d3d12.lib dxgi.lib)
target_link_libraries(onnxruntime_providers_dml PRIVATE dxguid.lib d3d12.lib dxgi.lib)
endif()

target_link_libraries(onnxruntime_providers_dml PRIVATE delayimp.lib)
Expand Down
3 changes: 1 addition & 2 deletions onnxruntime/core/providers/dml/DmlExecutionProvider/src/AbiCustomRegistry.cpp
View file
Open in desktop
Expand Up @@ -544,8 +544,7 @@ HRESULT STDMETHODCALLTYPE AbiCustomRegistry::RegisterOperatorKernel(
// Currently unsupported for external operators
if (canAliasFirstInput ||
supportsGraph ||
requiredInputCountForGraph ||
requiredConstantCpuInputs)
requiredInputCountForGraph)
{
ORT_THROW_HR(E_INVALIDARG);
}
Expand Down
33 changes: 21 additions & 12 deletions onnxruntime/core/providers/dml/DmlExecutionProvider/src/MLOperatorAuthorImpl.cpp
View file
Open in desktop
Expand Up @@ -1408,11 +1408,14 @@ namespace Windows::AI::MachineLearning::Adapter
ComPtr<IMLOperatorTensor> tensor;
ORT_THROW_IF_FAILED(GetInputTensor(i, tensor.GetAddressOf()));

ComPtr<IUnknown> resource;
tensor->GetDataInterface(resource.GetAddressOf());
if (resource)
if (tensor)
{
resourcesToTransition.push_back(resource.Get());
ComPtr<IUnknown> resource;
tensor->GetDataInterface(resource.GetAddressOf());
if (resource)
{
resourcesToTransition.push_back(resource.Get());
}
}
}

Expand Down Expand Up @@ -1525,21 +1528,27 @@ namespace Windows::AI::MachineLearning::Adapter

ML_CHECK_BOOL(inputIndex < m_inputTensors.size());

auto opKernelContextWrapper = const_cast<OpKernelContextWrapper*>(this);
if (m_inputTensors[inputIndex]->GetInterface() == nullptr)
{
auto inputTensor = m_impl->Input<onnxruntime::Tensor>(inputIndex);
if (inputTensor != nullptr)
{
ComPtr<TensorWrapper> tensorWrapper = wil::MakeOrThrow<TensorWrapper>(
const_cast<onnxruntime::Tensor*>(inputTensor),
IsAllocationInterface(inputTensor->Location()),
m_winmlProvider.Get(),
m_internalOperator);

ComPtr<TensorWrapper> tensorWrapper = wil::MakeOrThrow<TensorWrapper>(
const_cast<onnxruntime::Tensor*>(inputTensor),
IsAllocationInterface(inputTensor->Location()),
m_winmlProvider.Get(),
m_internalOperator);

const_cast<OpKernelContextWrapper*>(this)->m_inputTensors[inputIndex] = tensorWrapper;
opKernelContextWrapper->m_inputTensors[inputIndex] = tensorWrapper;
}
}

const_cast<OpKernelContextWrapper*>(this)->m_inputTensors[inputIndex].CopyTo(tensor);

if (opKernelContextWrapper->m_inputTensors[inputIndex] != nullptr)
{
opKernelContextWrapper->m_inputTensors[inputIndex].CopyTo(tensor);
}
return S_OK;
}
ORT_CATCH_RETURN
Expand Down