Skip to content

re-enable debugging for dml #1348

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

Closed
aciddelgado wants to merge 2 commits into from
Closed

re-enable debugging for dml #1348

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
14ae470
re-enable debugging for dml
aciddelgado Mar 25, 2025
a692a7a
clean it up like ryan wants it
aciddelgado Mar 27, 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
40 changes: 26 additions & 14 deletions src/models/debugging.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,22 +83,34 @@ void DumpTensor(const Model& model, std::ostream& stream, OrtValue* value, bool
stream << SGR::Fg_Green << " Location: " << SGR::Reset;

const auto& memory_info = value->GetTensorMemoryInfo();
switch (memory_info.GetDeviceType()) {
case OrtMemoryInfoDeviceType_CPU:
stream << "CPU\r\n";
if (memory_info.GetDeviceType() == OrtMemoryInfoDeviceType_CPU) {
stream << "CPU\r\n";
if (dump_value) {
DumpValues(stream, type_info->GetElementType(), value->GetTensorRawData(), element_count);
break;
case OrtMemoryInfoDeviceType_GPU: {
stream << "GPU\r\n";
auto type = type_info->GetElementType();
auto tensor_span = std::span<uint8_t>{const_cast<OrtValue*>(value)->GetTensorMutableData<uint8_t>(), SizeOf(type) * element_count};
auto device_span = model.p_device_->WrapMemory<uint8_t>(tensor_span);
DumpValues(stream, type, device_span.CopyDeviceToCpu().data(), element_count);
break;
}
default:
stream << "Unhandled device type: " << static_cast<int>(memory_info.GetDeviceType()) << "\r\n";
break;
// Internally there are 5 device types defined in onnxruntime but only 3 are exposed in the public API
// https://github.com/microsoft/onnxruntime/blob/9dbfee91ca9c2ba2074d19805bb6dedccedbcfe3/include/onnxruntime/core/framework/ortdevice.h#L15
} else if (memory_info.GetDeviceType() < 5) {
switch (model.p_device_->GetType()) {
case DeviceType::CUDA:
Copy link
Collaborator

@baijumeswani baijumeswani Mar 27, 2025 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Shall we try to define a model.p_device_->GetType<std::string>() for each device type? We can avoid the switch statement here then.

stream << model.p_device_->GetType<std::string>() << "\r\n";

Copy link
Contributor

@RyanUnderhill RyanUnderhill Mar 27, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we can fairly safely say that if the tensor is not on CPU, it's on the model.p_device_

Copy link
Contributor

@RyanUnderhill RyanUnderhill Mar 27, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh, I see. We shouldn't use a switch here, use this:

std::string to_string(DeviceType device_type);

stream << "CUDA\r\n";
break;
case DeviceType::DML:
stream << "DML\r\n";
break;
case DeviceType::QNN:
stream << "QNN\r\n";
break;
default:
stream << "Unknown\r\n";
break;
}
auto type = type_info->GetElementType();
auto tensor_span = std::span<uint8_t>{const_cast<OrtValue*>(value)->GetTensorMutableData<uint8_t>(), SizeOf(type) * element_count};
auto device_span = model.p_device_->WrapMemory<uint8_t>(tensor_span);
DumpValues(stream, type, device_span.CopyDeviceToCpu().data(), element_count);
} else {
stream << "Unhandled device type: " << static_cast<int>(memory_info.GetDeviceType()) << "\r\n";
}
}

Expand Down
Loading