Skip to content

GPU ONNX: Fix compiler warning and simplify code #14210

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

Merged
davidrohr merged 1 commit into from
Apr 22, 2025
Merged

GPU ONNX: Fix compiler warning and simplify code #14210

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
53 changes: 24 additions & 29 deletions GPU/GPUTracking/Base/cuda/GPUReconstructionCUDA.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -621,21 +621,10 @@ void GPUReconstructionCUDA::loadKernelModules(bool perKernel)
}
}

#ifndef __HIPCC__ // CUDA
void GPUReconstructionCUDA::startGPUProfiling()
{
GPUChkErr(cudaProfilerStart());
}

void GPUReconstructionCUDA::endGPUProfiling()
{
GPUChkErr(cudaProfilerStop());
}

void GPUReconstructionCUDA::SetONNXGPUStream(Ort::SessionOptions& session_options, int32_t stream, int32_t* deviceId)
{
#ifdef ORT_CUDA_BUILD
cudaGetDevice(deviceId);
GPUChkErr(cudaGetDevice(deviceId));
#if !defined(__HIPCC__) && defined(ORT_CUDA_BUILD)
OrtCUDAProviderOptionsV2* cuda_options = nullptr;
CreateCUDAProviderOptions(&cuda_options);

Expand All @@ -650,22 +639,7 @@ void GPUReconstructionCUDA::SetONNXGPUStream(Ort::SessionOptions& session_option

// Finally, don't forget to release the provider options
ReleaseCUDAProviderOptions(cuda_options);
#endif // ORT_CUDA_BUILD
}

#else // HIP
void* GPUReconstructionHIP::getGPUPointer(void* ptr)
{
void* retVal = nullptr;
GPUChkErr(hipHostGetDevicePointer(&retVal, ptr, 0));
return retVal;
}

void GPUReconstructionHIP::SetONNXGPUStream(Ort::SessionOptions& session_options, int32_t stream, int32_t* deviceId)
{
#ifdef ORT_ROCM_BUILD
// Create ROCm provider options
cudaGetDevice(deviceId);
#elif defined(ORT_ROCM_BUILD)
// const auto& api = Ort::GetApi();
// api.GetCurrentGpuDeviceId(deviceId);
OrtROCMProviderOptions rocm_options;
Expand All @@ -676,4 +650,25 @@ void GPUReconstructionHIP::SetONNXGPUStream(Ort::SessionOptions& session_options
session_options.AppendExecutionProvider_ROCM(rocm_options);
#endif // ORT_ROCM_BUILD
}

#ifndef __HIPCC__ // CUDA

void GPUReconstructionCUDA::startGPUProfiling()
{
GPUChkErr(cudaProfilerStart());
}

void GPUReconstructionCUDA::endGPUProfiling()
{
GPUChkErr(cudaProfilerStop());
}

#else // HIP
void* GPUReconstructionHIP::getGPUPointer(void* ptr)
{
void* retVal = nullptr;
GPUChkErr(hipHostGetDevicePointer(&retVal, ptr, 0));
return retVal;
}

#endif // __HIPCC__