Skip to content

Please consider the following formatting changes to #14214 #121

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
alibuild wants to merge 2 commits into from
Closed

Please consider the following formatting changes to #14214 #121

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
07a22ca
CUDA ORT: Must use api struct to call functions
davidrohr Apr 22, 2025
c6b07f8
Please consider the following formatting changes
alibuild Apr 22, 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
18 changes: 14 additions & 4 deletions GPU/GPUTracking/Base/cuda/GPUReconstructionCUDA.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -621,24 +621,34 @@ void GPUReconstructionCUDA::loadKernelModules(bool perKernel)
}
}

#define ORTCHK(command) \
{ \
OrtStatus* status = command; \
if (status != nullptr) { \
const char* msg = api->GetErrorMessage(status); \
GPUFatal("ONNXRuntime Error: %s", msg); \
} \
}

void GPUReconstructionCUDA::SetONNXGPUStream(Ort::SessionOptions& session_options, int32_t stream, int32_t* deviceId)
{
GPUChkErr(cudaGetDevice(deviceId));
#if !defined(__HIPCC__) && defined(ORT_CUDA_BUILD)
const OrtApi* api = OrtGetApiBase()->GetApi(ORT_API_VERSION);
OrtCUDAProviderOptionsV2* cuda_options = nullptr;
CreateCUDAProviderOptions(&cuda_options);
ORTCHK(api->CreateCUDAProviderOptions(&cuda_options));

// std::vector<const char*> keys{"device_id", "gpu_mem_limit", "arena_extend_strategy", "cudnn_conv_algo_search", "do_copy_in_default_stream", "cudnn_conv_use_max_workspace", "cudnn_conv1d_pad_to_nc1d"};
// std::vector<const char*> values{"0", "2147483648", "kSameAsRequested", "DEFAULT", "1", "1", "1"};
// UpdateCUDAProviderOptions(cuda_options, keys.data(), values.data(), keys.size());

// this implicitly sets "has_user_compute_stream"
cuda_options.has_user_compute_stream = 1;
UpdateCUDAProviderOptionsWithValue(cuda_options, "user_compute_stream", mInternals->Streams[stream]);
cuda_options->has_user_compute_stream = 1;
ORTCHK(api->UpdateCUDAProviderOptionsWithValue(cuda_options, "user_compute_stream", mInternals->Streams[stream]));
session_options.AppendExecutionProvider_CUDA_V2(cuda_options);

// Finally, don't forget to release the provider options
ReleaseCUDAProviderOptions(cuda_options);
api->ReleaseCUDAProviderOptions(cuda_options);
#elif defined(ORT_ROCM_BUILD)
// const auto& api = Ort::GetApi();
// api.GetCurrentGpuDeviceId(deviceId);
Expand Down